Erratum

Surg Endosc (2000) 14: 411 DOI: 10.1007/s004640030001 Springer-Verlag New York Inc. 2000

The following four articles were inadvertently printed in Surgical Endoscopy, vol. 14, no. 3, March 2000, with incorrect DOI numbers. Their correct DOI numbers are: F.L. Greene, The impact of laparoscopy on cancer management (pages 217218), DOI: 10.1007/s004640030045; Vitale et al., Endoscopic treatment of distal bile duct stricture from chronic pancreatitis (pages 227231), DOI: 10.1007/s004640030046; Ishida et al., Pneumoperitoneum with carbon dioxide enhances liver metastases of cancer cells implanted into the portal vein in rabbits (pages 239242), DOI: 10.1007/ s004640030047; and Lichtenbaum et al., Preoperative abdominal ultrasound may be misleading in risk stratification for presence of common bile duct abnormalities (pages 254257), DOI: 10.1007/s004640030049.

Erratum
Surg Endosc (2000) 14: 411 DOI: 10.1007/s004640030001 Springer-Verlag New York Inc. 2000

The following four articles were inadvertently printed in Surgical Endoscopy, vol. 14, no. 3, March 2000, with incorrect DOI numbers. Their correct DOI numbers are: F.L. Greene, The impact of laparoscopy on cancer management (pages 217218), DOI: 10.1007/s004640030045; Vitale et al., Endoscopic treatment of distal bile duct stricture from chronic pancreatitis (pages 227231), DOI: 10.1007/s004640030046; Ishida et al., Pneumoperitoneum with carbon dioxide enhances liver metastases of cancer cells implanted into the portal vein in rabbits (pages 239242), DOI: 10.1007/ s004640030047; and Lichtenbaum et al., Preoperative abdominal ultrasound may be misleading in risk stratification for presence of common bile duct abnormalities (pages 254257), DOI: 10.1007/s004640030049.

Surg Endosc (2000) 14: 349353 DOI: 10.1007/s004640020062

Springer-Verlag New York Inc. 2000

Minimally invasive surgery for posterior gastric stromal tumors

C. C. Hepworth, D. Menzies, R. W. Motson
Colchester General Hospital, Turner Road, Colchester, Essex CO4 5JL, UK Received: 30 April 1999/Accepted: 12 July 1999

Abstract Background: Because involvement is extremely rare, surgery for gastric stromal tumors consists of local excision with clear resection margins. The aim of this study was to report the results of a consecutive series of nine patients with posterior gastric stromal tumors that were excised using a minimally invasive method. Methods: Patients received a general anesthetic before placement of three laparoscopic ports a 10-mm (umbilical) port for the telescope and two working ports, a 12-mm port (left upper quadrant) and a 10-mm port (right upper quadrant). Grasping forceps were placed through an anteriorly placed gastrotomy to deliver the tumor through the gastrotomy into the abdominal cavity, thus allowing an endoscopic linear cutter to excise the tumor with a cuff of normal gastric tissue. Results: Nine consecutive patients with a median age of 73 years (range, 4783) were treated. In seven patients, laparoscopic removal of the tumor was achieved. Two patients required conversion to an open operation because the tumor could not be delivered into the abdominal cavity. The median length of postoperative stay for the seven patients in whom the procedure was completed laparoscopically was 3 days (range, 26). Conclusions: Posterior gastric stromal tumors can be removed safely using this minimally invasive method. Delivery of the tumor through the gastrotomy is essential for success. Key words: Gastric stromal tumors Minimally invasive surgery Stomach

Smooth muscle tumors (leiomyomas, leiomyosarcomas, and leiomyoblastomas) arising in gastric tissue are now called gastric stromal tumors. They were formerly distinguished as leiomyomas and leiomyosarcomas when composed of spindle cells and as either benign leiomyoblastomas (epiCorrespondence to: R. W. Motson

thelioid leiomyomas) or malignant leiomyoblastomas (epithelioid leiomyosarcoma) when composed of epithelioid cells [8]. They are uncommon, and the majority are small (<2 cm) and clinically insignificant [20, 23, 38]. However, symptoms may occur due to their size, location, or as a result of bleeding. Many of these tumors are discovered incidentally at endoscopy. Hemorrhage and anemia are the most common presentations [6, 20, 33]; they are caused by ulceration of the overlying gastric mucosa and may require interventional endoscopy or surgical resection. The stomach and small intestine [4, 7] are the most frequent gastrointestinal sites of origin for stromal tumors. Endoscopy is the usual mode of investigation used to identify submucosal tumors, of which gastric stromal tumors are a common subtype. The smooth submucosal tumor has overlying mucosa that tents normally over it when biopsied. It may be ulcerated. Schindlers folds [34] may be seen on either side of the tumor. Endoscopic ultrasound (EUS) [36, 40] is used to reveal the anatomical layer of the stomach wall from which the lesion arises; it is capable of determining local invasion and nodal status but does not provide a histological diagnosis. EUS is currently the most accurate method of visualizing submucosal tumors [32]. Histological confirmation can be difficult to obtain preoperatively because standard endoscopic biopsies are not likely to reach the layer where the tumor originates. Because of difficulties in determining histology and because small biopsy samples may not be adequate to assess malignancy, the clinician often faces a dilemma when deciding whether to refer a patient for a surgical procedure or to watch and wait. Pathologists are aware of the difficulties in differentiating benign from malignant gastric stromal tumors [20, 22]. They are classified [17] as benign, borderline, or malignant depending upon mitotic activity, size, [1], and differentiation [20, 22]. Although mitotic activity remains the most important prognostic factor, tumors have been known to recur locally and to metastasize even when mitotic figures are rare or entirely absent [23, 31, 38]. The standard treatment for gastric stromal tumors (apart from those associated with Carneys triad) is complete excision with clear resection margins and without tumor rupture [7, 27]. Only then can the histological nature of the

350

lesion be determined with absolute certainty. If the tumor is subsequently found to be malignant, no further treatment [41] is required as long as the resection margins are clear, as these tumors are usually radio-resistant and not responsive to chemotherapy [6]. Both wedge resection and gastrectomy have been performed [6, 7, 12, 35] for malignant tumors, but neither method appears to confer a survival advantage [12, 15]. This may be due in part to the fact that nodal metastasis is rare [7, 20, 41] and there does not appear to be a need for regional lymph node dissection. If malignancy is suspected or confirmed on subsequent histological grounds, a careful surveillance program for local recurrence and hepatic or peritoneal deposits may be required. Surveillance is necessary because following complete resection of the malignant tumor, only 10% of patients remain free of the disease [28]. One study showed that the median interval to recurrence was 18 months (range 3232) and that recurrences arose mainly in the peritoneum (78%) and/or liver (70%) [28]. Resection of recurrences is associated with a prolonged median survival [27, 28, 30]. Extraabdominal metastases (lung and bone are most common sites) have a poor prognosis. Carneys triad [3, 21] occurs mainly in young women; the average age of presentation is 16 years. This syndrome includes epithelioid cell gastric sarcomas, pulmonary chondromas, and extraadrenal paragangliomas (which cause hypertension). Intraadrenal pheochromocytomas have been reported. In Carneys triad, the gastric sarcomas are usually multifocal; if locally excised, they recur in the gastric remnant. If this syndrome is suspected, a total gastrectomy may be indicated. In this paper, we report the results of a consecutive series of nine patients with posterior gastric stromal tumors that were excised using a minimally invasive method. Materials and methods Patients
Nine consecutive patients (median age, 73 years; range, 4783) with gastric stromal tumors identified at endoscopy on the posterior wall of the stomach underwent minimally invasive surgery for their removal.

stomach (Fig. 2) and then back through the anterior abdominal wall. Traction on the stay sutures lifted up the anterior wall of the stomach (Fig. 3) so that the anterior wall of the stomach lay in a vertical plane. The tumor was located by the gastroscope on the posterior part of the stomach. The 30 telescope visualized the anterior part of the stomach. By turning the laparoscopic light off and rotating the gastroscope to face the anterior wall after it has located the tumor, transillumination of the anterior wall of the stomach can be achieved. A gastrotomy was made using cutting diathermy at the point of greatest transillumination, allowing access directly onto the tumor. The gastroscope was withdrawn, and the 30 telescope and grasping forceps (Storz, Slough, England) were introduced through the gastrotomy into the gastric cavity. The grasping forceps are placed through the gastrotomy before the 30 telescope. The tumor can then be grasped and pulled through the gastrotomy by withdrawing the grasping forceps. To aid delivery, the stay sutures are relaxed so that the stomach again rests in its normal anatomical position. Delivery of the tumor everts the posterior wall of the stomach through the gastrotomy (Fig. 3), forming a pedicle of normal gastric tissue. If the tumor is too large to be delivered or cannot be delivered with a pedicle of normal gastric tissue, conversion to an open operation is required. This problem may arise if the tumor has invaded surrounding organs. If the tumor is on the anterior wall [13] of the stomach, transillumination by the gastroscope (viewed by the 30 telescope) will demarcate the tumor clearly. The tumor appears as a dark area surrounded by a halo of light. For posterior tumors, once the tumor is delivered with its cuff of normal gastric tissue, sequential application of the endoscopic linear cutter across the pedicle both transects and staples the posterior wall of the stomach (Fig. 4). The tumour is then placed into an endoscopic retrieval bag (Vernon Carus, Preston, Lancashire, England) and removed through the 12-mm trocar site. Thereafter, tension on the stay sutures lifts the anterior wall of the stomach into a vertical plane and allows laparoscopic closure of the gastrotomy with a continuous single Vicryl (Ethicon) suture (Fig. 5). The stay sutures are then cut and the stomach resumes its original position in the abdomen. Laparoscopic ports are removed with closure of the port sites. The patient is allowed a normal diet the following day and monitored carefully for evidence of upper gastrointestinal bleeding, vomiting, or peritonism.

Results Nine consecutive patients (six female and three male) with a median age of 73 years (range, 4783) were found at gastroscopy to have probable gastric stromal tumors (Table 1). Seven of these nine presented acutely because of melena with a median hemoglobin of 9.1 g/dl (range, 7.513.7). Another patient was being investigated for upper abdominal pain (hemoglobin 14.0 g/dl) due to cholecystitis and had a gastric stromal tumor that was found incidentally at gastroscopy. The remaining patient was being investigated for iron deficiency anemia (hemoglobin 8.1 g/dl). None of the patients required either emergency interventional endoscopy or surgery. All of them underwent gastroscopy. Although the tumor was visualized in all cases, in only one case did an endoscopic biopsy yield a specific diagnosis of gastric stromal tumor. The endoscopic biopsy suggested a malignant gastric stromal tumor, but surgical histology identified the lesion as a borderline malignancy. At operation, seven of our nine patients had a laparoscopic procedure to remove the tumor. Two of nine patients required conversion to an open operation because a pedicle of gastric tissue could not be formed to allow transection with the linear stapler. One of these two patients underwent a partial gastrectomy; surgical histology of the specimen identified the lesion as a malignant tumor. Twenty-seven months later, this patient developed a recurrence and required total gastrectomy.

Technique
After receiving a general anesthetic, the patient was placed in a supine position with the legs abducted in Allen supports (CLS Medical Ltd., Newcastle-upon-Tyne, England). Using a direct puncture technique, a 10mm laparoscopic port (Eurosurgical Ltd., Guildford, Surrey, England) was introduced at the umbilicus [24]. Carbon dioxide was insufflated (Olympus, model A5845; Keymed, Southend-on-Sea, England) through this port to create a pneumoperitoneum with an insufflation pressure of 1215 mmHg. A 30 telescope (Olympus model A5255), was then placed through the 10-mm laparoscopic port to visualize the abdominal contents and to allow placement of the remaining two ports under direct vision. A 12-mm port (Fig. 1) was placed in the left upper quadrant to allow the use of the endoscopic linear cutter (Ethicon, Bracknell, Berkshire, England). Another 10-mm port was placed in the right upper quadrant. The camera was held by an assistant standing on the patients left while the surgeon stood between the patients legs. An IT 20 gastroscope (Olympus; Keymed) was inserted orally, and the tumor was visualized in the stomach. Two stay sutures (8 cm apart) were placed through the anterior abdominal wall into the peritoneal cavity. They were passed through the seromuscular layer of the anterior aspect of the

351 Fig. 1. Abdominal placement of laparascopic ports. , 10-mm laparoscopic port; q, 12-mm laparoscopic port. Fig. 2. Two stay sutures are placed through the anterior abdominal wall into the peritoneal cavity and then through the seromuscular layer of the anterior aspect of the stomach. Fig. 3. Traction on the two stay sutures lifts up the anterior wall of the stomach. Fig. 4. Sequential application of the endoscopic linear cutter transects and closes (with staples) the everted posterior wall of the stomach (pedicle) after delivery of the stromal tumor. Fig. 5. Closure of the gastrotomy with a single continuous suture.

The remaining eight patients histology revealed benign tumors in three cases and a borderline malignancy in five cases. The most common location was the fundus. An open operation was required for two tumors that were larger (11 7 cm, 12 12 cm) than those removed laparoscopically (median, 6.5 4.5 cm; range, 6 39 6). All seven patients who were treated laparoscopically were eating, drinking, and ambulant the following day.

There was no postoperative bleeding or leakage, and none of the patients died. The overall median length of postoperative stay was 3 days (range, 210). For patients in whom the procedure was completed laparoscopically, the median length of stay was 3 days (range, 26). The two patients who required conversion to an open operation remained in hospital for 6 and 10 days, respectively.

352 Table 1. Features of gastric stromal tumors Feature Tumor locationa antrum bodyb fundusc cardia Histological diagnosis benign borderlineb malignantc Mode of presentation melena anemia incidental finding
a b

No. of patients 1 2 4 1 3 5 1 7 1 1

Tumor location is shown for only eight of the nine patients One patient required conversion to an open procedure for a borderline malignant tumor c One patient required conversion to an open procedure. Two years later, the patient developed a recurrence and underwent a total gastrectomy

adjacent structures or the size of the tumor. Preoperative endoscopic ultrasound may determine which posterior gastric tumors would be suitable for this type of removal. Use of the linear cutter across the pedicle does raise the theoretical possibility that other tissue could be incorporated. There was no clinical evidence of this problem in this study. If unwanted tissue were incorporated, one would expect to encounter some difficulty in everting the tumor to form the gastric pedicle. Had the tumor been located on the anterior wall of the stomach, the use of the stay sutures and transillumination from the gastroscope would have identified the outline of the tumor, facilitating its removal [13]. There are several methods (endoscopic and laparoscopic) for resecting gastric stromal tumors. Preoperative investigations that best determine the position and size of the tumor should allow the most suitable method to be chosen. For patients in whom minimally invasive therapy can be performed, the well-known advantages of a shorter recovery period, reduced pain, and shorter hospital stay may be attained. However, a randomized prospective controlled study is needed to verify this possibility. References
1. Amin MB, Ma CK, Linden MD, Kubus JJ, Zarbo RJ (1993) Prognostic value of proliferating cell nuclear antigen index in gastric stromal tumors: correlation with mitotic count and clinical outcome. Am J Clin Pathol 100: 428432 2. Basso N, Silecchia G, Pizzuto G, Surgo D, Picconi T, Materia A (1996) Laparoscopic excision of posterior gastric wall leiomyoma. Surg Laparosc Endosc 6: 6567 3. Blei E, Gonzalez-Crussi F (1992) The intriguing nature of gastric tumours in Carneys triad. Cancer 69: 292300 4. Bruce AW, Stalker AL (1958) Smooth muscle tumors of the alimentary tract. Br J Surg 46: 629633 5. Clancy TV, Moore PM, Ramshaw DG, Kays CR (1994) Laparoscopic excision of a benign gastric tumor. J Laparoendosc Surg 4: 277280 6. Dasarthy S, Pandey GK, Bhargava DK, Gupta SS, Gupta SD (1995) Tropical Gastroenterol 16: 4346 7. Dougherty MJ, Compton CC, Talbert M, Wood WC (1991) Sarcomas of the gastrointestinal tract: Separation into favorable and unfavorable prognostic groups by mitotic count. Ann Surg 214: 569574 8. Franquemont DW (1995) Differentiation and risk assessment of gastrointestinal stromal tumours. Am J Clin Pathol 103: 4147 9. Fujisaki J, Mine T, Akimoto K, Yoshida S, Hasegawa Y, Ogata E (1988) Enucleation of a gastric leiomyoma by a combined laser and snare electrocutting technique. Gastrointest Endosc 34: 128130 10. Goh P, Kum CK (1993) Laparoscopic Billroth II gastrectomy: a review. Surg Oncol 2 (Suppl 1): 1318 11. Gorbuz AT, Peetz ME (1997) Resection of a gastric leiomyoma using combined laparoscopic and gastroscopic approach. Surg Endosc 11: 285286 12. Grant CS, Kim CH, Farrugia G, Zinsmeister A, Goellner JR (1991) Gastric leiomyosarcoma: prognostic factors and surgical management. Arch Surg 126: 985990 13. Gurbuz AT, Peetz ME (1997) Resection of a gastric leiomyoma using combined laparoscopic and gastroscopic approach. Surg Endosc 11: 285286 14. Ibrahim IM, Silvestri F, Zingler B (1997) Laparoscopic resection of posterior gastric leiomyoma. Surg Endosc 11: 277279 15. Katai H, Sasako M, Sano T, Maruyama K (1997) Wedge resection of the stomach for gastric leiomyosarcoma. Br J Surg 84: 560561 16. Kuramata H, Etoh S, Miyamoto S (1976) On the transendoscopic extirpation of gastric submucosal tumor. Stomach Intestine 11: 1475 1484 17. Lewin KJ, Appelman HD (1995) Tumors of the esophagus and stomach. In: Atlas of tumor pathology. 3rd se. Armed Forces Institute of Pathology, Washington, DC.

Discussion Our review of the literature revealed case reports of endoscopic resection [9, 16, 36], laparoscopic resection [2, 39], laparoscopic intragastric resection [37], and endoscopically assisted laparoscopic resections [5, 11, 13, 14, 25, 29]. This paper describes nine consecutive patients with posterior gastric stromal tumors that were excised using a minimally invasive approach. Due to difficulties in biopsying subcutaneous tumors at endoscopy, a clear distinction between benign and malignant tumours is rarely made before surgery. In this series, endoscopic biopsies were diagnostic on only one occasion. Surgical histology showed borderline malignancy in five cases and unequivocal malignancy in one case. Given that only three of our nine patients had definitely benign histology, it is arguable that most gastric stromal tumors should be removed surgically. Placement of stay sutures through the anterior abdominal wall and the anterior aspect of the stomach facilitated both gastrotomy formation and its subsequent closure. Grasping forceps had to be placed through the gastrotomy before the telescope. If the telescope is placed first, it is difficult to locate the gastrotomy with the forceps. Everting the gastric stromal tumor produces a pedicle of normal gastric tissue. The laparoscopic linear stapling device facilitated simultaneous division and closure of the tissues, reducing both the spillage of gastric contents and the amount of bleeding. If the tumor was too large to be everted or a pedicle of normal gastric tissue could not be formed to allow application of the endoscopic linear cutter, the procedure was converted to an open operation. We encountered this problem in two cases; both were converted to an open procedure. Laparoscopic resection of the stomach [10, 19] has been reported, but it was not performed in this study. At present, the laparoscopic resection of malignant tumors is controversial due to the possibility of port site recurrence [26]. The reason for failure to form a pedicle is not clear. The source of the problem might be fixation of the tumor to

353 18. Llorente J (1994) Laparoscopic gastric resection for gastric leiomyoma. Surg Endosc 8: 887889 19. Lointier P, Leroux S, Ferrier C, Dapoigny M (1993) A technique of laparoscopic gastrectomy and Billroth II gastrojejunosotmy. J Laparoendosc Surg 3: 353364 20. Ludwig DJ, Traverso LW (1997) Gut stromal tumors and their clinical behavior. Am J Surg 173: 390394 21. Margulies KB, Sheps SG (1988) Carneys triad: guidelines for management. Mayo Clin Proc 63: 496502 22. Mazur MT, Clark HB (1983) Gastric stromal tumors: reappraisal of histogenesis. Am J Surg Pathol 7: 507519 23. Morgan BK, Compton C, Talbert M, Gallagher W, Wood W (1990) Benign smooth muscle tumors of the gastrointestinal tract. Ann Surg 211: 6366 24. Motson RW (1994) Direct puncture technique for laparoscopy. Ann R Coll Surg Engl 76: 346347 25. Motson RW, Fisher PW, Dawson JW (1995) Laparoscopic resection of a benign intragastric stromal tumour. Br J Surg 82: 1670 26. Neuhaus SJ, Texler M, Hewett PJ, Watson DI (1998) Port-site metastases following laparoscopic surgery. Br J Surg 85: 735741 27. Ng EH, Pollock RE, Munsell MF, Atkinson EN, Romsdahl MM (1992) Prognostic factors influencing survival in gastrointestinal leiomyosarcomas: implications for surgical management and staging. Ann Surg 215: 6877 28. Ng EH, Pollock RE, Romsdahl MM (1992) Prognostic implications of patterns of failure for gastrointestinal leiomyosarcomas. Cancer 69: 13341341 29. Payne WG, Murphy CG, Grossbard LJ (1995) Combined laparoscopic and endoscopic approach to resection of gastric leiomyoma. J Laparoendosc Surg 5: 119122 30. Persson S, Kindblom L-G, Angervall L, Tisell L-E (1992) Metastasizing gastric epithelioid leiomyosarcomas (leiomyoblastomas) in young individuals with long-term survival. Cancer 70: 721732 31. Ranchod M, Kempson RL (1977) Smooth muscle tumors of the gastrointestinal tract and retroperitoneum. Cancer 39: 255261 32. Rosch T (1995) Endoscopic ultrasonography in upper gastrointestinal submucosal tumors: a literature review. Gastrointest Endosc Clin North Am 5: 609614 33. Sanders L, Silverman M, Rossi R, Braasch J, Munson L (1996) Gastric smooth muscle tumors: diagnostic dilemmas and factors affecting outcome. World J Surg 20: 992995 34. Schindler R (1937) The endoscopic study of gastric pathology. In: Gastroscopy. Univ. of Chicago Press, Chicago, pp 233234 35. Shiu MH, Farr GH, Papachristou DN, Hajdu SI (1983) Myosarcomas of the stomach: natural history, prognostic fators and management. Cancer 49: 177187 36. Spinelli P, Cerrai FG, Cambareri AR, Meroni E, Pizzetti P (1993) Two-step endoscopic resection of gastric leiomyomas. Surg Endosc 7: 9092 37. Taniguchi E, Kamiike W, Yamanishi H, Ito T, Nezu R, Nishida T, Momiyama T, Ohashi S, Okada T, Matsuda H (1997) Laparoscopic intragastric surgery for gastric leiomyoma. Surg Endosc 11: 287289 38. Welch JP (1975) Smooth muscle tumors of the stomach. Am J Surg 130: 279285 39. Yamashita Y, Bekki F, Kakegawa T, Umetani H, Yatsuka K (1995) Two laparoscopic techniques for resection of leiomyoma in the stomach. Surg Laparosc Endosc 5: 3842 40. Yasuda K, Nakajima M, Yoshida S, Kiyota K, Kawai K (1989) The diagnosis of submucosal tumors of the stomach by endoscopic ultrasonography. Gastrointest Endosc 35: 1015 41. Yoshida M, Otani Y, Ohgami M, Kubota T, Kumai K, Mukai M, Kitajima M (1997) Surgical management of gastric leiomyosarcoma: evaluation of the propriety of laparoscopic wedge resection. World J Surg 21: 440443

News and notices

Surg Endosc (2000) 14: 412415 DOI: 10.1007/s004640000181 Springer-Verlag New York Inc. 2000

The Association of Endoscopic Surgeons of Great Britain and Ireland

The AESGBI office is moving from the Oaks Hospital Colchester. All correspondence should be addressed to: Association of Endoscopic Surgeons of Great Britain and Ireland The Royal College of Surgeons 35/43 Lincolns Inn Fields London WC2A 3PN, UK Tel: 0044 (0)171 973 0305 Fax: 0044 (0)171 430 9235 E-mail: jrabone@asgbi.org.uk

mens as well as in vivo animal models. Course fees depend on length and complexity of instructions. In addition, attendees may observe live surgery and attend other educational activities at the George Washington University. For further details and brochure please contact: Debbie Moser Washington Institute of Surgical Endoscopy George Washington University Department of Surgery 2150 Pennsylvania Avenue, N.W. Suite 6-B Washington, DC 20037, USA Tel: 202-994-5441 Fax: 202-944-0567

Volunteer Surgeons Needed Northwestern Nicaragua Laparoscopic Surgery Teaching Program, Leon, Nicaragua
Volunteer surgeons are needed to tutor laparoscopic cholecystectomy for this non-profit collaboration between the Nicaraguan Ministry of Health, the National Autonomous University of Nicaragua, and Medical Training Worldwide. The program consists of tutoring general surgeons who have already undergone a basic laparoscopic cholecystectomy course. Medical Training Worldwide will provide donated equipment and supplies when needed. For further information please contact: Medical Training Worldwide Ramon Berguer, MD, Chairman Tel: 707-423-5192 Fax: 707-423-7578 E-mail: berguer.r@martinez.va.gov

Essentials of Laparoscopic Surgery Surgical Skills Unit University of Dundee Scotland, UK

Under the direction of Sir A. Cuschieri the Surgical Skills Unit is offering a three-day practical course designed for surgeons who wish to undertake the procedures such as laparoscopic cholecystectomy. This intensely practical program develops the necessary operating skills, emphasizes safe practice, and highlights the common pitfalls and difficulties encountered when starting out. Each workshop has a maximum of 18 participants who will learn both camera and instrument-manipulation skills in a purposebuilt skills laboratory. During the course there is a live demonstration of a laparoscopic cholecystectomy. The unit has a large library of operative videos edited by Sir Cuschieri, and the latest books on endoscopic surgery are on display in our Resource area. Course fee including lunch and course materials is $860. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY, Scotland, UK Tel: +44 382 645857 Fax: +44 382 646042

Courses at the Washington Institute of Surgical Endoscopy

We are delighted to offer a variety of regular courses in laparoscopic techniques year round. In addition, special arrangements can be made for individual tuition in all minimally invasive disciplines. CME credit is available and course fees depend on the instruction offered. For further information please contact: Debbie Moser Washington Institute of Surgical Endoscopy 2150 Pennsylvania Avenue, N.W. Suite 6-B Washington, DC 20037, USA Tel: 202-994-8425, or 1-888-8WISEDOC Fax: 202-994-0567

Minimal Access Therapy Training Courses Surgical Skills Unit University of Dundee, Scotland, UK Unit Director: Sir Alfred Cuschieri
The Surgical Skills unit offers practical courses in minimal access therapies. Each program is organized to have maximum hands on practice sessions in the purpose-built laboratory. Subject experts are available for tuition and personal feedback. Live operative demonstrations on specific procedures form part of each course; these are backed up by video illustrations. Each course has a workbook with essential information, clear explanations, and guidelines. Informal discussion sessions with subject experts form part of each program. The courses are run regularly throughout the year. Essentials of Laparoscopic Surgery, a 3-day course covering the essential skills and safety procedures for basic laparoscopic operations. Foundations of Laparoscopy Surgery, a 5-day course presenting an extensive and detailed introduction to laparoscopic surgery. Practical Training in Pediatric Endoscopic Surgery, a 5-day course for surgeons who wish to perform or refine endoscopic surgical techniques in children.

Courses at the Washington Institute of Surgical Endoscopy

Courses are available in difficult cholecystectomy and common duct exploration, laparoscopic antireflux surgery, laparoscopic colorectal surgery, laparoscopic solid organ surgery, transanal endoscopic microsurgery, and advanced techniques. Additionally, tailor-made courses are available for single surgeons or groups of surgeons to fit their requirements. Courses consist of didactic construction, review of videos, and experience in our superbly-equipped laboratory. We utilize training boxes, phantom abdo-

413 The Advanced Endoscopic Surgical Courses of the Royal College of Surgeons of Edinburgh are held in the Surgical Skills Unit. These are: Advanced Endoscopic Skills Course, a 5-day endoscopic skills training in advanced surgical tasks common to all specialities, e.g., intracorporeal and extracorporeal knot tying, internal continuous and interrupted suturing, anastomosis of organs, stapling, use of sophisticated equipment such as ultrasound and harmonic dissections and Specialist Procedure-related courses, which are 23 days long and concentrate on specific clinical situations particularly relevant to a speciality and focused on specific endoscopic operation(s). Ductal Calculi: The Laparoscopic Approach Practical Aspects of Laparoscopic Fundoplication Laparoscopic Colorectal Course Thoracoscopic Sympathectomy Other specialist courses held in the unit include courses in anaesthetic techniques, arthroscopy, colonoscopy, endoscopy, gynecology, and otolaryngology. For further information you can visit our web site or contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY, Scotland, UK Tel: +44 1382 645857 Fax: +44 1382 646042 E-mail: j.e.a.struthers@dundee.ac.uk Web page: http://www.dundee.ac.uk/surgicalskills/ For further information please contact: Wanda Toy, Program Administrator Microsurgery & Operative Endoscopy Training (MOET) Institute 153 States Street San Francisco, CA 94114, USA Tel: (415) 626-3400 Fax: (415) 626-3444

Fellowships in Minimally Invasive Thoracic and General Surgery University of Pittsburgh Medical Center Pittsburgh, PA, USA
One-year fellowships in advanced Minimally Invasive Surgery in General and Thoracic Surgery are being offered at the University of Pittsburgh Medical Center. Requirements include completion of residence training programs in General or Thoracic Surgery. The fellowships will involve extensive clinical exposure as well as clinical and basic science research. These positions include a very competitive salary and travel allowance. Interested candidates are invited to send a letter of inquiry with a curriculum vitae to: Philip R. Schauer, M.D. Assistant Professor of Surgery and Co-Director for General Surgery James Luketich, M.D. Assistant Professor of Surgery and Co-Director for Thoracic Surgery The University of Pittsburgh Medical Center C-800 Presbyterian University Hospital 200 Lothrop Street Pittsburgh, PA 15213-3221, USA

Courses at the Royal Adelaide Centre for Endoscopic Surgery

Basic and Advanced Laparoscopic Skills Courses are conducted by the Royal Adelaide Centre for Endoscopic Surgery on a regular basis. The courses are limited to six places to maximize skill development and tuition. Basic courses are conducted over two days for trainees and surgeons seeking an introduction to laparoscopic cholecystectomy. Animal viscera in simulators is used to develop practical skills. Advanced courses are conducted over four days for surgeons already experienced in laparoscopic cholecystectomy who wish to undertake more advanced procedures. A wide range of procedures are included, although practical sessions can be tailored to one or two procedures at the participants request. Practical skills are developed using training simulators and anaesthetised pigs. Course fees: $A300 ($US225) for the basic course and $A1,600 ($US1,200) for the advanced course. For further details and a brochure please contact: Dr. D. I. Watson or Professor G. G. Jamieson The Royal Adelaide Centre for Endoscopic Surgery Department of Surgery Royal Adelaide Hospital Adelaide, SA 5000, Australia Tel: +61 8 224 5516 Fax: +61 8 232 3471

Visiting Scholar Opportunities in Minimally Invasive Surgery Ohio State University Medical Center Columbus, OH, USA
A minimum experience of one month and a maximum of six months is being offered to Visiting Scholars at the Ohio State University Medical Center starting immediately. Applicants must have completed a recognized training program in general surgery. The program is designed to expose trained surgeons to advanced applications of new techniques in minimally invasive surgery, prepare individuals to train others in the application of this new technology, to include this individual in the development of the new application of technology, and to provide the individual with the guidance and the resources to develop independent investigational projects evaluating treatment modalities in minimally invasive surgery. International applicants are encouraged to apply. Interested candidates are invited to send a letter of inquiry with a curriculum vitae to: W. Scott Melvin, M.D. Assistant Professor of Surgery and Director Center for Minimally Invasive Surgery The Ohio State University Department of Surgery N737 Doan Hall 410 West Tenth Avenue Columbus, OH 43210, USA

Advanced Laparoscopic Suturing and Surgical Skills Courses MOET Institute San Francisco, CA, USA
Courses are offered year-round by individual arrangement. The MOET Institute is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide continuing medical education for physicians and designates these CME activities for 2040 credit hours in Category 1 of the Physicians Recognition Award of the American Medical Association. These programs are also endorsed by the Society of Gastrointestinal Endoscopic Surgeons (SAGES).

Fellowships in Minimally Invasive Thoracic and General Surgery Ohio State University Medical Center Columbus, OH, USA
One-year research or clinical fellowships in Minimally Invasive Surgery are being offered at the Ohio State University Medical Center starting July 1, 2000. Requirements include completion of residence training programs in General or Thoracic Surgery. The clinical fellowship will include extensive clinical exposure to advanced techniques in laparoscopic surgery. The research fellowship will involve both clinical and basic science research. The positions offer a competitive salary and travel allowance. Interested candidates are invited to send a letter of inquiry with a curriculum vitae to:

414 W. Scott Melvin, M.D. Assistant Professor of Surgery and Director Center for Minimally Invasive Surgery Robert E. Michler, M.D. Professor of Surgery and Chief Cardio Thoracic Surgery The Ohio State University Department of Surgery N737 Doan Hall 410 West Tenth Avenue Columbus, OH 43210, USA For further information please contact: Judy Quesenberry Program Service Coordinator 704-355-4823

The Royal Surrey County Hospital Minimal Access Therapy Training Unit CME accredited Laparoscopic Courses 2000
Directed by Professor Michael Bailey. The Minimal Access Therapy Training Unit offers a full range of Laparoscopy Courses. All courses include live demonstrations, interactive presentations and hands-on skills training. Courses on offer include: An Introduction to Laparoscopic Surgery; Basic Surgical Skills; Laparoscopic Cholecystectomy [RCS Certified]; Advanced Laparoscopic Workshop; Extra Peritoneal Hernia repair; Nissen Fundoplication; Upper Gastrointestinal Laparoscopy; Laparoscopic Colorectal Surgery; EAES, AESGBI approved. For further details contact: Alison Snook MATTU The Royal Surrey County Hospital Egerton Road, Guildford Surrey GU2 5XX, UK Tel: 00 44 [0] 1483 303053 Fax: 00 44 [0] 1483 406636 E-mail alison@mattu.org.uk www.mattu.org.uk- online registration.

European Course on Laparoscopic Surgery under the auspices of E.A.E.S. April 47, 2000 November 2124, 2000 University Hospital Saint-Pierre (U.L.B.) Brussels, Belgium
Course Director: G. B. Cadiere, M.D. Universit Libre de Bruxelles (U.L.B.) Department of G.I. Surgery University Hospital Saint-Pierre Live demonstrations. Interactive dialogue with the operating surgeons; Video forum: videotaped discussions, technical details, pitfalls: Topics: Functional gastric surgery (Nissen-Toupet-Gastroplasty); Colon (colectomy-rectopexy); Hernia (trans-/pre-peritoneal approach, balloon); Retroperitoneoscopy; Splenectomy; Needle surgery; Biliary surgery; New technologies. Surgeons: J. Bruyns, G. B. Cadiere, J. Himpens, J. Leroy, M. Vertruyen Official languages: EnglishFrench. Internet site: http://www.LAP-SURGERY.com For any further information: Scientific Information C.H.U. Saint-Pierre Service de Chirurgie Digestive Rue Haute 322 B-1000 Bruxelles, Belgium Tel: 322 535 41 15 Fax: 322 535 31 66 E-mail: coelio@resulb.ulb.ac.be or Administrative Secretariat Conference Services S.A. Avenue de lObservatoire 3, bte 17 B-1180 Bruxelles, Belgium Tel: 322 375 16 48 Fax: 322 375 32 99 E-mail: conference.services@skynet.be

Carolinas Laparoscopic and Advanced Surgery Program (CLASP) A Division of Carolinas Medical Center Charlotte, NC, USA Courses offered in 2000
Laparoscopic Ventral and Incisional Herniorrhaphy Workshop: The purpose of this course is to teach laparoscopic techniques for ventral hernia repair, discuss the different prosthetic materials, their application, and associated healing process, and describing the techniques for minimally invasive access in the multiply operated abdomen. Dates offered: April 7, June 9, and September 22, 2000. Course consists of 12-day didactic and 12-day hands-on workshops. Course fee: $250.00. Laparoscopic Solid Organ-Adrenal, Spleen, and Liver Workshop: The purpose of this course is to identify the appropriate work-up for adrenal pathology and determine those that require resection, as well as the techniques of laparoscopic adrenalectomy. One should be able to determine the appropriate pathology that necessitates splenectomy, and describe techniques of laparoscopic splenectomy. Additionally, there will be a detailed description of laparoscopic cryoablation of hepatic metastasis. Course consists of 12-day didactic and 12-day hands-on workshops. Course fee: $450.00; workshop date: May 5, 2000. Hand Assisted Laparoscopic Surgery for General and Colorectal Surgeons: The purpose of this course is to discuss the use and indications for laparoscopic hand-assisted surgery in regard to splenectomy, morbid obesity surgery, and colon resection. Surgeons will perform splenectomies, nephrectomies, and a bowel resection during the laboratory session. Course date: September 23, 2000. Fundamentals in Critical Care: Course date: April 21, 2000. Course Directors: B. Todd Heniford, M.D. Chief of Minimal Access Surgery CodirectorCarolinas Laparoscopic and Advanced Surgery Program Carolinas Medical Center Frederick L. Greene, M.D. Chairman, Department of General Surgery CodirectorCarolinas Laparoscopic and Advanced Surgery Program Carolinas Medical Center

III International Czech-Polish-Slovak Symposium on Videosurgery September 2021, 2001 Tr inec, Czech Republic
The main memory of the traditional surgery (competition miniinvasive vs traditional surgery in the Year 2001). Stanislav Czudek, M.D., Ph.D. Center for Miniinvasive Surgery Hospital Podles 73961 Tr inec Czech Republic

1st World Congress of Pediatric Thoracic Disciplines April 2022, 2000 Izmir, Turkey URL: http://www.med.ege.edu.tr/pedsurg/congress.htm
Contact: Prof. Dr. Oktay Mutaf Department of Pediatric Surgery Ege University Faculty of Medicine zmir, Turkey Bornova 35100 I Fax: +90 232 375 12 88 E-mail: omutaf@med.ege.edu.tr

415

Surgical Skills Unit May 1519, 2000 Dundee, Scotland, UK

Practical Training in Pediatric Endoscopic Surgery Information: Angela Duncan, Registration Secretary, Surgical Skills Unit, Ninewells Hospital and Medical School, Dundee, Scotland, UK Tel: +44 1382 645857 Fax: +44 1382 646042 E-mail: a.z.duncan@dundee.ac.uk http://www.dundee.ac.uk/surgicalskills/

7th Southeast European Symposium of Pediatric Surgery Intestinal Motility Disorders June 23, 2000 University of Graz, Austria/Europe
Correspondence to: Professor Gu nther Schimpl, M.D. Department of Pediatric Surgery Auenbruggerplatz 34 A-8036 Graz, Austria/Europe Tel.: +43/316/385-3762 Fax: +43/316/385-3775 E-mail: kinderchirurgie@kfunigraz.ac.at

7th World Congress of Endoscopic Surgery June 14, 2000 Singapore International Convention and Exhibition Centre Singapore
The World Congress of Endoscopic Surgery coupled with the World Expo of Surgical Technologies is one of the most important and largest Endoscopic Surgery Congress and Exhibition in the world. This 7th World Congress includes a full 12-day Live Transmission from the National University Hospital and 3 days of intensive lectures by the worlds distinguished pioneers and leaders in this field. The world of surgeons and members from all the familial organizations under the International Federation of Societies of Endoscopic Surgeons, will come together in Singapore to welcome in the new century and a new millennium. The Scientific Programmes for the main meeting has a motto: A New Millennium, A New Vision. Participants will get a glimpse into the future and an idea of how surgery will evolve in the next century. Plenary Symposium covers the hottest, newest, and the most controversial topics covering a total of 34 topics. These topics include state-of-the-art lectures, endoscopy: new technologies, and new surgical technologies in the 21st century. The Education Programme will also be conducted and consists of 12 preand postcongress workshops covering such topics as the Advance Workshop in Needlescopic Surgery; GI Therapeutic Endoscopy Workshop, which includes a 12-day live transmission and 2 days of symposium; Workshop in Pediatric Laparoscopic Urology, plus six subspecialty seminars, and a nurses seminar. For the postgraduates, there will be a one-day workshop on Common Laparoscopic Procedures from Revolution to Standard of Care. There will be ample time for questions and answers, and an exchange of views with all the experts during the panel discussions. Abstract submission is extended to March 15, 2000. Delegates can now submit their abstracts via the web at www.worldendo2000.com. Registration for all courses, and bookings of hotels can also be done via the Internet. All forms can be downloaded from the Congress website. For more information on scientific programme, please contact: Professor Peter Goh Congress President Minimally Invasive Surgical Centre National University Hospital Tel: (65) 772 5264 Fax: (65) 775 4007 For other information and requisition of brochure, please contact: Congress Secretariat 7th World Congress of Endoscopic Surgery c/o Ace:Daytons Direct (International) Pte Ltd 2 Leng Kee Road, #04-02 Thye Hong Centre Singapore 159086 Tel: (65) 475 9377 Fax: (65) 475 3041/472 3177 Email: directnt@pacific.net.sg

Joint Meeting of the American Pancreatic Association and the International Association of Pancreatology November 15, 2000 Ambassador West Hotel Chicago, IL, USA
Symposia, posters, scientific sessions, Pancreatology at the Millenium. For further information, contact Peter A. Banks, M.D., Fax: (617) 5660338, or Howard A. Reber, M.D., Fax: (310) 206-2472.

Association of Endoscopic Surgeons of Great Britain and Ireland Autumn Meeting November 1617, 2000
The Autumn Meeting will be held at Colchester General Hospital, Colchester, Essex, UK. Further information from the AESGBI at The Royal College of Surgeons of England, 3543 Lincolns Inn Fields, London WC2A 3PN. Telephone: 0171 973 0305; Fax: 0171 930 9235; Email: jrabone@ASGBI.org.uk.

Fifteenth International Workshop on Therapeutic Endoscopy December 57, 2000 Hong Kong
The Chinese University of Hong Kong and the Hong Kong Society of Digestive Endoscopy will hold the 15th International Workshop on Therapeutic Endoscopy on December 57, 2000. The workshop, intended for experienced endoscopists interested in endoscopic therapy, consists of 3 days of live demonstration of advanced techniques of therapeutic endoscopy. There will also be poster presentations by the participants. This year we have invited Rikiya Fujita (Japan), Mitsuhiro Kida (Japan), Joseph Leung (USA), Horst Neuhaus (Germany), Jeffrey Ponsky (USA), Chan-Sup Shim (Korea) and Paul Swain (UK) to be our international faculty. For further information, please write to: Prof. Sydney Chung Endoscopy Centre Prince of Wales Hospital The Chinese University of Hong Kong Shatin, N.T., Hong Kong Tel: (852) 2632 2233 Fax: (852) 2635 0075 E-mail: info@hksde.org

Surg Endosc (2000) 14: 340344 DOI: 10.1007/s004640020014

Springer-Verlag New York Inc. 2000

Pain after microlaparoscopic cholecystectomy

A randomized double-blind controlled study
T. Bisgaard, B. Klarskov, R. Trap, H. Kehlet, J. Rosenberg
Department of Surgical Gastroenterology 435, University of Copenhagen, Hvidovre Hospital, DK-2650, Hvidovre, Denmark Received: 12 July 1999/Accepted: 21 October 1999

Abstract Background: Laparoscopic cholecystectomy (LC) is traditionally performed with two 10-mm and two 5-mm trocars. The effect of smaller port incisions on pain has not been established in controlled studies. Methods: In a double-blind controlled study, patients were randomized to LC or cholecystectomy with three 2-mm trocars and one 10-mm trocar (micro-LC). All patients received a multimodal analgesic regimen, including incisional local anesthetics at the beginning of surgery, NSAID, and paracetamol. Pain was registered preoperatively, for the first 3 h postoperatively, and daily for the 1st week. Results: The study was discontinued after inclusion of 26 patients because five of the 13 patients (38%) randomized to micro-LC were converted to LC. In the remaining 21 patients, overall pain and incisional pain intensity during the first 3 h postoperatively increased in the LC group (n 13) compared with preoperative pain levels (p < 0.01), whereas pain did not increase in the micro-LC group (n 8). Conclusions: Micro-LC in combination with a prophylactic multimodal analgesic regimen reduced postoperative pain for the first 3 h postoperatively. However, the micro-LC led to an unacceptable rate of conversion to LC (38%). The micro-LC instruments therefore need further technical development before this surgical technique can be used on a routine basis for laparoscopic cholecystectomy. Key words: Gallbladder Microlaparoscopic cholecystectomy Pain Randomized controlled trial

be reduced even further [6]. In some recent uncontrolled series, laparoscopic cholecystectomy was performed with three 2-mm [5, 10] or 3-mm trocars [3, 8, 9], assisted by one 10-mm trocar for gallbladder retraction. Compared with earlier reports of patients undergoing LC, postoperative analgesic requirements were reduced by 70% when 2-mm instruments were used [5]. However, the effect of 2-mm or 3-mm port incisions (micro-LC) on pain after laparoscopic cholecystectomy has not been established in controlled studies. We therefore decided to investigate the effect of 2-mm micro-LC on postoperative pain in a randomized double-blind controlled study. Patients and methods Patients
We originally planned to randomize 50 patients. The study started October 1998 and was discontinued in January 1999. The criteria for exclusion were ASA physical status III or IV, estimated abdominal wall layer of >10 cm, and patients having papillotomy by endocopic retrograde cholangiopancreatography within 1 month before operation. Moreover, patients were excluded if they had chronic pain diseases other than gallstone disease, if they received opioids or tranquilizers (treatment for >1 week up to the cholecystectomy), or if they had a history of alcohol or drug abuse. Patients were also excluded if the operation was converted from micro-LC to LC or to an open operation, or from LC to an open operation.

Surgical technique
The same two experienced surgeons carried out all operations. The microLC learning curve [13] was overcome prior to the beginning of the study. In both surgical groups, laparoscopic cholecystectomy was performed by the standard French technique with identical placement of trocars [4]. The micro-LC was conducted as follows: Pneumoperitoneum was established with a Veress cannula. A nondisposable 10-mm trocar with a pyramidal tip was inserted (Karl Storz Endoscope system) above the umbilicus. Three 2-mm MiniSite disposable introducers and trocars (United States Surgical Corporation, Norwalk, CT, USA), a 2-mm laparoscope (USSC), a 10-mm laparoscope (Olympus), two grasping instruments (MiniSite Endo Grasp, Single Action 2-mm; USSC), an electro-scissor

Laparoscopic cholecystectomy (LC) is traditionally performed with two 10-mm and two 5-mm trocars [4], but most patients still suffer from early incisional pain [1, 11]. Some authors have therefore suggested that port incisions should

Correspondence to: T. Bisgaard

341 (MiniSite MiniShears 2-mm; USSC), and a 10-mm endoscopic rotating multiple clip applier (10-mm ER320; Ethicon ERCA) was used for the operation. Dissection was performed via the upper left epigastric port visualized by the 10-mm laparoscope through the supraumbilical port, as in the French technique. The 10-mm clip applier was inserted via the supraumbilical 10-mm port to secure the cystic duct and artery. Thus, it was necessary to shift the camera to the 2-mm laparoscope during clipping and retraction of the gallbladder. The gallbladder was retracted via the supraumbilical port. Micro-LC was converted to LC in case of lack of progress in the operation, technical problems, or if safety was compromised. LC was performed using two 10-mm and two 5-mm trocars with pyramidal tips (Karl Storz Endoscope system). If necessary, a lateral umbilical fascial incision (0.51 cm) was made in patients of both surgical groups to ease retraction of the gallbladder. Intraoperative cholangiography was not performed. All patients received gentamicin (160 mg) at the beginning of surgery. During laparoscopy, intraabdominal pressure was maintained at 12 mmHg. The CO2 was carefully evacuated at the end of surgery by manual compression of the abdomen with open trocars. In both surgical groups, only the fascia and the skin of the 10-mm supra-umbilical incision was sutured. The 5-mm and 2-mm incisions were closed at the skin level with sterilized strips. ings (5 5 cm, DuoDerm, Mini, ConvaTec). Prior to the operation, the patients were instructed to keep the dressings for the 1st postoperative week.

Statistics and ethics

Data from 100 consecutive patients who underwent uncomplicated laparoscopic cholecystectomy in our department have shown that 50% had moderate or severe incisional pain for the first 2 postoperative days [unpublished data]. We decided that a clinically relevant treatment should reduce the incidence of severe or moderate overall pain from 50% to 10% (Minimal Relevant Difference, MIREDIF 40%). With a type I error of 0.05 and a type II error of 0.20, the necessary sample size would be 40 patients (20 patients in each group). We chose to study a minimum of 50 patients. For comparison between groups, repeated comparisons at different time points were avoided by adding together the pain scores (total pain scores, TPS). Median differences between groups and 95% confidence limits of median differences, as well as the Friedman, Mann-Whitney U, and Fishers exact tests were used when appropriate. All values are presented as median (range) and percentages when appropriate. The local ethics committee approved the study, and all patients gave their written informed consent to participate in the study.

Anesthesia and analgesia

The same general anesthetic technique was followed in all cases. No premedication was used. General anesthesia was induced with intravenous fentanyl (0.003 mg/kg) and propofol (2.02.5 mg/kg). Orotracheal intubation was facilitated by cisatracurium (0.15 mg/kg). Anesthesia was maintained with oxygen in air (1:2), propofol (10 mg/kg/h), and supplemental doses of alfentanil (0.51.0 mg) if required, given at the discretion of the anesthesiologist. Minute ventilation was adjusted to keep end-tidal PCO2 at 4.55.5 kPa. For analgesia, both surgical groups received incisional local anesthetics in all port sites using a total of 140 mg of bupivacaine (0.5% bupivacaine 10 ml in the supraumbilical incision and 6 ml in the other three incisions). The infiltration technique has been described in detail elsewhere [1]. Furthermore, all patients had intravenous ketorolac (30 mg) 20 min before the end of surgery. In the recovery room, immediately after surgery, a single dose of 2 g paracetamol was given as suppositories. At 3 h after surgery, all patients commenced oral treatment with ibuprofen (600 mg at 8-h intervals for 4 days). If patients requested supplementary analgesic treatment, opioids were administered (5 mg intravenously in the recovery room, 30 mg orally or 510 mg IV in the ward).

Results This study was discontinued after the inclusion of 26 patients (13 patients in each group) because we became convinced that the 2-mm micro-LC conversion rate was unacceptable: Five of 13 patients (38%) in the micro-LC group were converted to LC. All five patients had gross anatomic signs of chronic cholecystitis and/or dense adhesions in the gallbladder region, whereas only one patient among the remaining eight patients who underwent successful micro-LC had chronic cholecystitis/dense adhesions (p 0.0093) (Table 1). All patients in both study groups were preoperatively regarded as simple elective cases suitable for laparoscopy. Intention-to-treat analysis showed that the micro-LC group had longer operations (Table 1). Furthermore, only the patients in the micro-LC group needed suspension sutures anchored in the gallbladder to facilitate dissection [2] (Table 1), since it was often difficult to grasp the gallbladder with the very tiny 2-mm instruments. Micro-LC in combination with a prophylactic multimodal analgesic regimen eliminated postoperative pain during the first 3 h postoperatively (nonsignificant Friedman test). In contrast, the LC group experienced a significant increase in pain intensity compared with preoperative values, as shown by a statistically significant Friedman test (Fig. 1a, b and Table 2). Comparison between groups showed that the micro-LC group experienced significantly less overall pain during mobilization (as measured by VRS) and significantly less incisional pain at rest and during mobilization (as measured by VRS) (Table 2). However, statistical comparisons of VAS pain registrations did not reach statistical significance (Table 2). Nevertheless, comparison between groups (median differences and 95% confidence intervals of median differences) showed a tendency toward a decrease in pain intensity in the micro-LC group compared with the LC group in the first 3 h postoperatively (Table 2). One patient in the LC group did not receive alfentanil during anesthetic maintenance (Table 1). This patient received no more or no less fentanyl than the other patients

Pain assessment
Before the operation, patients were instructed to use a 100-mm visual analogue scale (VAS) (VAS: end points labelled no pain and worst possible pain) and a verbal rating scale (VRS) (VRS: no pain 0, light pain 1, moderate pain 2, severe pain 3). Overall pain and incisional pain was registered the morning before the operation and hourly for the first 3 h postoperatively using VAS and VRS at rest (supine) and during mobilization (supine to sitting). Incisional pain was defined as a superficial pain, wound pain, or pain located in the abdominal wall [1]. During the same period, morphine requirements were recorded. In addition, patients registered overall pain and incisional pain in a structured diary from the day before operation and daily throughout the 1st week at 8 P.M. as average VAS pain during the preceding 24 hs. Average incisional pain was rated by VRS. Information sheets, contact telephone numbers, and structured pain questionnaires for the daily pain diary were given to all patients on discharge.

Randomization and blinding

The surgeon randomized patients to micro-LC or LC by the envelope method after anesthesia was induced and did not attend to the patient or the two observers who recorded pain at any time after the operation. The starting time of the operation was blinded to the two observers. At the end of the operation, the incisions were blinded with waterproof standard dress-

342 Table 1. Clinical, operative, and general anesthetic data for 26 patients randomized to either microlaparoscopic cholecystectomy (micro-LC) or traditional laparoscopic cholecystectomy (LC) micro-LC (n 13) Clinical data sex ratio (M:F) age (yr) body mass index (kg/m2) ASA physical class (I:II) Operative data duration of surgery (min) gallbladder suspension suture (no. of patients) gross anatomic findings: chronic cholecystitis/dense adhesions (no. of patients) normal anatomic findings (no. of patients) fascial incision for gallbladder retraction (no. of patients) General anesthesia fentanyl (mg) propofol (mg) cisatracurium (mg) alfentanil (mg) Values given as median (range) 3:10 46 (2571) 25 (1832) 10:3 85 (45155) 5 6 (5 converted to LC) 7 (0 converted to LC) 9 0.3 (0.20.4) 1240 (10202490) 11.0 (7.515.0) 2.0 (0.55.0) LC (n 13) 9:4 53 (2468) 26 (2333) 10:3 55 (30180) 0 8 5 11 0.2 (0.20.5) 1120 (4582,175) 10.0 (7.014.0) 1.5 (04.0) p 0.047 0.857 0.573 1 0.016 0.039 0.695 0.695 1 0.511 0.045 0.360 0.264

and had no pain in the first 3 h after the operation (TPS 0). One patient in the micro-LC group, as compared with two patients in the LC group, required morphine for postoperative analgesia (5 mg and 10 mg, respectively) (p 1). During the 1st week, there was no difference in TPS between the groups (Fig. 1c and Table 2).

Discussion Although it was terminated prematurely, our study basically showed that the use of micro-LC reduced pain during the first 3 h postop, whereas the LC group had significant postoperative pain in spite of prophylactic multimodal analgesic treatment. The study was discontinued halfway through because we became convinced that 2-mm micro-LC was technically inferior to LC because of an unacceptable conversion rate (micro-LC to LC) of 38%. Gross anatomic findings of chronic cholecystitis/dense adhesions indicated conversion of micro-LC to LC because of problems with grasping and dissection, bending of the thin instruments, and a narrow vision angle with the 2-mm laparoscope. Thus, the area of contact between the jaws of the 2-mm grasping instrument and the tissue was too small to achieve adequate traction and manipulation in case of dense adhesions. However, with normal anatomy, it was quite easy to perform a successful laparoscopic cholecystectomy using these instruments. Therefore, further improvement of the surgical equipment for the 2-mm technique is required before its routine use for laparoscopic cholecystectomy. These improvements should include the development of stronger instruments for dissection in dense tissue; better grasping ability, probably with a traumatic grip; and the development of a 2-mm laparoscope, which can provide a larger and better picture.

We mainly used a 10-mm laparoscope for micro-LC. Micro-LC operations prior to commencement of the study convinced us that the 2-mm laparoscope during dissection provided inadequate picture quality for use during the entire operation. Unfortunately, none of the micro-LC series in the literature have provided sufficient information on perioperative gross anatomic findings and thus possible difficulty of the operation. Earlier studies have used the periumbilical 10-mm port for dissection of the gallbladder, and for clipping and division of the structures with standard LC instruments [9, 12, 13]. These studies comprised a total of 84 patients, and the conversion rate (micro-LC to LC) ranged from zero to 24%. Durations of operation were comparable to those of our 2-mm micro-LC operations. In one prospective study of 60 elective selected patients, Gagner and Garcia-Ruz [5] performed 2-mm micro-LC using a technique similar to the one in our study. The operations lasted a median of 98 min (range, 40150), but only 5% were converted to LC. However, very few perioperative details were provided. Davides et al. [3] recently reported a series of 25 selected elective patients who underwent successful micro-LC using three 3-mm and one 10-mm trocar. The median operation time was somewhat shorter (75 min; range, 45180), and there were no conversions to LC. As in our study, the micro-laparoscope was only used for the application of clips and retraction of the gallbladder from the abdomen. In the present study, which was closed prematurely, only a limited number of patients were available for data analysis. However, overall pain and incisional pain intensity were eliminated in the micro-LC group during the first 3 h postop, as opposed to patients undergoing LC. Furthermore, median differences and their confidence limits of pain scores during the first 3 h postop indicated that the smaller incisions were associated with less early pain after laparo-

343

Fig. 1. The effects of microlaparoscopic cholecystectomy (micro-LC) compared with conventional laparoscopic cholecystectomy (LC) on overall pain (a) and incisional pain (b) at rest and during mobilization using visual analogue scale (VAS) preoperatively and at 1, 2, and 3 h after operation. Average daily overall pain (VAS) during the 1st week is shown in (c). Values are median. For p values, see Table 2.

scopic. Some of the tests of pain scores between the groups did not reach statistical significance (p < 0.05). However, the finding of median differences and confidence limits that indicated a tendency toward a reduction in pain with smaller instruments may simply reflect a type II error, due to the inclusion of too few patients in the study. Generally, pain is most intense the first 4 h following traditional laparoscopic cholecystectomy [1, 7]. Therefore, treatment that provides analgesia for even a short duration is clinically important. In the present study, both incisional pain and overall pain were reduced, which suggests that incisional pain is a significant component of overall pain

after laparoscopic cholecystectomy. These findings are in accordance with those of a previous study of patients who had traditional laparoscopic cholecystectomy using anesthesia and analgesia similar to the present study [1]. In this study, incisional pain dominated over other pain localizations [1] throughout the 1st postoperative week. The effect on pain of reducing port incisions has only been evaluated in one randomized controlled study that replaced one epigastric 10-mm port with a 5-mm port [6], but there was no effect on pain intensity. We conclude that 2-mm micro-LC in combination with a prophylactic multimodal analgesic regimen eliminated

344 Table 2. Overall pain and incisional pain scores Pain variation over time (Friedman test), p-valuesa micro-LC (n 8) 03 h overall pain VAS at rest VAS during mobilization VRS at rest VRS during mobilization 03 h incisional pain VAS at rest VAS during mobilization VRS at rest VRS during mobilization daily for 1st week overall pain (VAS) incisional pain (VRS) 0.199 0.345 0.172 0.295 0.290 0.392 0.392 0.392 <0.001 0.003 LC (n 13) 0.003 0.001 0.001 <0.001 0.003 0.003 0.002 0.001 <0.001 <0.001 Comparison of TPS (between groups) Median difference (95% CI) 17 23 1 2 9 16 1 2 (80 to 25) (69 to 10) (4 to 1) (4 to 0) (46 to 0) (59 to 0) (4 to 0) (3 to 0) pb 0.215 0.109 0.183 0.039 0.104 0.109 0.044 0.031 0.491 0.510

16 (81 to 71) 1 (5 to 2)

VAS, visual analogue scale; VRS, verbal rating scale a p values after Friedmans test for pain variation over time within same surgical group b p values after Mann-Whitney test for comparison of total pain scores (TPS) between surgical groups

early postoperative pain for the first 3 h but carried an unacceptable conversion rate to LC. Thus, the micro-LC (2-mm) instruments need further development before they can be recommended for routine use.
Acknowledgments. This work was supported by grants from the University of Copenhagen and the Danish Medical Research Council (journal no. 9601607) and the King Christian X Foundation.

References
1. Bisgaard T, Klarskov B, Kristiansen VB, Callesen T, Schulze S, Kehlet H, Rosenberg J (1999) Multi-regional local anesthetic infiltration during laparoscopic cholecystectomy in patients receiving prophylactic multi-modal analgesia: a randomized, double-blind, placebocontrolled study. Anesth Analg 89: 10171024 2. Bresadola F, Pasqualucci A, Donini A, Chiarandini P, Anania G, Terrosu G, Sistu MA, Pasetto A (1999) Elective transumbilical compared with standard laparoscopic cholecystectomy. Eur J Surg 165: 2934 3. Davides D, Dexter SP, Vezakis A, Larvin M, Moran P, McMahon MJ (1999) Micropuncture laparoscopic cholecystectomy. Surg Endosc 13: 236238 4. Dubois F, Icard P, Berthelot G, Levard H (1990) Coelioscopic cholecystectomy: preliminary report of 36 cases. Ann Surg 211: 6062

5. Gagner M, Garcia-Ruiz A (1998) Technical aspects of minimally invasive abdominal surgery performed with needlescopic instruments. Surg Laparosc Endosc 8: 171179 6. Golder M, Rhodes M (1998) Prospective randomized trial of 5- and 10-mm epigastric ports in laparoscopic cholecystectomy. Br J Surg 85: 10661067 7. Joris J, Thiry E, Paris P, Weerts J, Lamy M (1995) Pain after laparoscopic cholecystectomy: characteristics and effect of intraperitoneal bupivacaine. Anesth Analg 81: 379384 8. Kimura T, Sakuramachi S, Yoshida M, Kobayashi T, Takeuchi Y (1998) Laparoscopic cholecystectomy using fine-caliber instruments. Surg Endosc 12: 283286 9. Reardon PR, Kamelgard JI, Applebaum B, Rossman L, Brunicardi FC (1999) Feasibility of laparoscopic cholecystectomy with miniaturized instrumentation in 50 consecutive cases. World J Surg 23: 128131 10. Tanaka J, Andoh H, Koyama K (1998) Minimally invasive needlescopic cholecystectomy. Surg Today 28: 111113 11. Ure BM, Troidl H, Spangenberger W, Dietrich A, Lefering R, Neugebauer E (1994) Pain after laparoscopic cholecystectomy: intensity and localization of pain and analysis of predictors in preoperative symptoms and intraoperative events. Surg Endosc 8: 9096 12. Watanabe Y, Sato M, Ueda S, Abe Y, Horiuchi A, Doi T, Kawachi K (1998) Microlaparoscopic cholecystectomythe first 20 cases: is it an alternative to conventional LC? Eur J Surg 164: 623625 13. Yuan RH, Lee WJ, Yu SC (1997) Mini-laparoscopic cholecystectomy: a cosmetically better, almost scarless procedure. J Laparoendosc Adv Surg Tech 7: 205211

Surg Endosc (2000) 14: 358361 DOI: 10.1007/s004640020088

Springer-Verlag New York Inc. 2000

Laparoscopic cholecystectomy in acute cholecystitis

A prospective comparative study in patients with acute vs chronic cholecystitis
P. Pessaux, J. J. Tuech, C. Rouge, R. Duplessis, C. Cervi, J. P. Arnaud
Department of Visceral Surgery, 4 rue Larrey, Angers 49100, France Received: 13 May 1999/Accepted: 7 October 1999

Abstract Background: The aim of this prospective study was to compare the outcome of laparoscopic cholecystectomy (LC) in patients with acute cholecystitis versus those with chronic cholecystitis and to determine the optimal timing for LC in patients with acute cholecystitis. Methods: From January 1991 to July 1998, 796 patients (542 women and 254 men) underwent LC. In 132 patients (67 women and 65 men), acute cholecystitis was confirmed via histopathological examination. These patients were divided into two groups. Group 1 (n 85) had an LC prior to 3 days after the onset of the symptoms of acute cholecystitis, and group 2 (n 47) had an LC after 3 days. Results: There were no mortalities. The conversion rates were 38.6% in acute cholecystitis and 9.6% in chronic cholecystitis ( p < 108). Length of surgery (150.3 min vs 107.8 min; p < 109), postoperative morbidity (15% vs 6.6%; p 0.001), and postoperative length of stay (7.9 days vs 5 days; p < 109) were significantly different between LC for acute cholecystitis and elective LC. For acute cholecystitis, we found a statistical difference between the successful group and the conversion group in terms of length of surgery and postoperative stay. The conversion rates in patients operated on before and after 3 days following the onset of symptoms were 27% and 59.5%, respectively ( p 0.0002). There was no statistical difference between early and delayed surgery in terms of operative time and postoperative complications. However, total hospital stay was significantly shorter for group 1. Conclusions: LC for acute cholecystitis is a safe procedure with a shorter postoperative stay, lower morbidity, and less mortality than open surgery. LC should be carried out as soon as the diagnosis of acute cholecystitis is established and preferably before 3 days following the onset of symptoms. Early laparoscopic cholecystectomy can reduce both the conversion rate and the total hospital stay as medical and economic benefits.

Key words: Acute cholecystitis Gallbladder Laparoscopic cholecystectomy Optimal timing

Since its introduction in 1987, laparoscopic cholecystectomy (LC) has increasingly been accepted as the procedure of choice for the treatment of symptomatic gallstones and chronic cholecystitis [22]. A successful LC is associated with a less painful postoperative course, a lower analgesic requirement, a shorter hospital stay, and less cosmetic disfigurement [15, 25]. Acute cholecystitis, which is generally found in 20% of all admissions for gallbladder disease [24], is no longer considered a contraindication for LC [4, 22]. However, the role and timing of LC in the management of acute cholecystitis remains controversial. In the prelaparoscopic era, prospective randomized studies [11, 18] demonstrated that the outcome for patients undergoing early open cholecystectomy within 7 days of the onset of symptoms was superior to delayed interval surgery. The aim of this prospective study was to compare the laparoscopic treatment of patients with acute cholecystitis to those with chronic cholecystitis. Our assessment of the results of attempted LC for acute cholecystitis paid particular attention to the interval from the onset of symptoms to the time of operation.

Materials and methods Patients

From January 1991 to July 1998, a total of 796 patients underwent laparoscopic cholecystectomy (LC). During that period, 132 patients (16.6%) had acute cholecystitis (68 phlegmonous and 64 gangrenous). They were admitted on an emergency basis with a diagnosis of acute cholecystitis based on the following symptoms: (a) acute upper abdominal pain with tenderness under the right costal margin, (b) fever >37.8C and/or leukocytosis >10 109/L (normal, <10 109/L), and (c) ultrasonographic evidence (thickened gallbladder wall, edematous gallbladder wall, distended gallbladder, presence of gallstones, ultrasonographic Murphys sign, and

Correspondence to: J. P. Arnaud

359 Table 1. Details of patients with acute cholecystitis (n 132) Symptom or finding Upper abdominal pain Right upper quadrant tenderness Leukocytosis >10 109/L Fever (>37.8C) Ultrasonographic evidence Histopathology n 116 108 91 65 118 132 % 87.8 81.8 68.9 49.2 89.4 100

pericholecystic fluid collection) [7, 19]. Histopathological examination of the excised gallbladder confirmed the presence of acute inflammation (Table I). All preoperative, intraoperative, and postoperative data were collected on standardized forms. We then analyzed the ultrasound findings, time of operation from the onset of symptoms, length of surgery, histologic gallbladder features, conversion rate to open laparotomy, postoperative mortality and morbidity, postoperative length of stay, and rate of early reintervention. The patients with acute cholecystitis were divided into two groups. Group 1 consisted of those who had an LC prior to 3 days after the onset of symptoms of acute cholecystitis; group 2 consisted of those who had an LC after 3 days following the onset of symptoms.

the procedure was converted to an open laparotomy. The average length of postoperative hospital stay for those who underwent successful LC was 6.1 days (range, 230). For patients in whom the procedure was converted to an open operation, the average length of stay was 10.5 days (range, 327). There was no statistical difference between the successful and the conversion groups in terms of morbidity (Table 5). When conversion to laparotomy in patients with acute cholecystitis was required, 11 of 51 patients (21.5%) developed postoperative complications, as compared to nine of 81 (11.1%) who had a successful LC. In particular, chest infections developed more frequently after conversion to open cholecystectomy (Table 6).

Timing of surgery in acute cholecystitis In 85 (64.4%) of 132 patients, the LC was performed prior to 3 days after the onset of symptoms (group 1), whereas 47 (35.6%) underwent the operation after 3 days (group 2). There were 23 patients (27%) in whom the surgical procedures was converted to an open operation in group 1 and 28 (59.5%) in group 2 (p 0.002). There was no statistical difference between early and delayed surgery in terms of operative time and postoperative complications. However, the total hospital stay was significantly shorter in group 1 (Table 7). Six patients (12.7%) in the delayed group either failed conservative treatment or developed a recurrent attack of acute cholecystitis, requiring emergency LC.

Statistical analysis
Statistical analysis was performed by the chi-square test, exact Fishers test, or students t-test, as appropriate. Statistical significance was set at p < 0.05.

Results Elective LC vs LC for acute cholecystitis The patient population was comprised of 542 women and 254 men with a mean age of 54.5 years (range, 1495). In our series of patients with acute cholecystitis, there were 67 women and 65 men with a mean age of 58.7 years (range, 1490). The average age of these patients was significantly different from that of the 664 patients (475 women and 189 men) who underwent elective cholecystectomy (mean age, 52.7 years; p < 0.001; range, 1995). Fifty-one patients (38.6%) with acute cholecystitis and 64 (9.6%) who had elective LC were converted to open surgery (p < 108) (Table 2). The length of the surgery (150.3 vs 107.8 min), postoperative length of stay (7.9 vs 5 days), and morbidity (15% vs 6.6%) were significantly different between acute cholecystitis and elective LC. There were no mortalities in either group (Table 3). The most common reason for conversion in patients with acute cholecystitis (n 30) was an inability to complete the procedure due to difficulties with exposure and dissection associated with the inflammatory reaction. The most common reason for conversion in patients with elective LC was adhesions (n 29). The other indications for conversion to an open procedure are summarized in Table 4. Acute cholecystitis The average length of the operative procedures was 141 min (range, 65380) for 81 patients who underwent successful LC and 170 min (range, 45350) for 51 patients in whom

Discussion Laparoscopic cholecystectomy is the treatment of choice for most patients with symptomatic cholelithiasis. Initially, acute cholecystitis was a contraindication for this procedure [4, 22]. The laparoscopic management of acute cholecystitis is a logical progression from elective LC. LC patients with acute cholecystitis have significantly higher rates of conversion to open cholecystectomy, longer operating times, longer hospital stays, and an increased morbidity rate compared with elective LC. The conversion rate reported in the literature varies from 7 to 60% [8, 10, 20, 22]. The total conversion rate in our series was high (38.6%). Our study was performed over a long period, and the existence of a learning curve was clearly demonstrated by the fact that the conversion rate decreased from 80% to 20% from the beginning to the end of the study (Table 2). An inability to define the anatomy of the cystic duct due to inflammation was the most common reason for conversion in all reported series (58.8% in our series). Conversion may be required if the operation progresses poorly, if there are pathologic conditions best dealt with by open surgery, or if complications arise. Conversion to an open procedure after an adequate attempt by an experienced laparoscopic surgeon should not be regarded as a complication or an operative failure but as a means of preventing complications. Our complication rate of 15% compares favorably to the 17.439.9% morbidity rate reported for traditional open cholecystectomy in patients with acute cholecystitis [2, 13, 21]. There were no mortalities. In cases where the operation

360 Table 2. Conversion rate during the study in laparoscopic cholecystectomy for acute cholecystitis 1991 LC for acute cholecystitis Conversions (n) Conversion rate (%) 4 3 75 1992 7 6 85 1993 11 7 63.6 1994 19 8 42.1 1995 21 8 38.1 1996 27 9 33.3 1997 29 7 24.1 1998 14 3 21.4

Table 3. Results for patients undergoing elective LC and patients with acute cholecystitis Elective LC (n 664) Mean age (yr) Operative time (min) Conversions Mortality Morbidity Early reinterventions Postoperative stay (days) 52.7 107.8 64 (9.6%) 0 44 (6.6%) 4 (0.6%) 5 Acute cholecystitis (n 132) 58.7 150.3 51 (38.6%) 0 20 (15%) 3 (2.2%) 7.9

Table 6. Postoperative complications Successful group (n 81) Conversion group (n 51) 4 1 2 1 0 3 0 11 (21.5%) NS

p <0.001 <109 <108 NS 0.001 NS <109

Chest infection Subhepatic collection Biliary fistula Hernia Bile duct injury Retained CBDS Others Total

0 1 2 1 1 0 4 9 (11.1%)

CBDS, common bile duct stones

Table 4. Reasons for conversion to laparotomy Elective LC (n 664) Inflammatory reaction Technical difficulty Adhesions Common bile duct stones Biliary injury Bleeding Cholecystoduodenal fistula Bilio-biliary fistula Intolerance to pneumoperitoneum Others Total 0 15 29 8 1 4 0 1 3 3 64 (23.5%) (46%) (12.5%) (1.5%) (6%) (1.5%) (4.5%) (4.5%) (9.6%) Acute cholecystitis (n 132) 30 0 6 9 2 1 1 0 0 2 51 (58.8%) (11.7%) (17.5%) (4%) (2%) (2%) (4%) (38.6%)

Table 7. Results of early versus delayed LC for acute cholecystitis Early surgery (group 1) (n 85) Operative time (min) Conversion Mortality Morbidity Total hospital stay (days) 149.7 23 (27%) 0 12 (14%) 7.6 Delayed surgery (group 2) (n 47) 158.4 28 (59.5%) 0 6 (12.7%) 11.4

p NS 0.0002 NS NS p < 0.001

Table 5. Comparison of successful and conversion groups in patients with acute cholecystitis Successful group (n 81) Mean age (yr) Operative time (min) Postoperative stay (days) Morbidity 57.7 141 6.1 9 (11.1%) Conversion group (n 51) 60.4 170 10.5 11 (21.5%)

p NS <0.02 <107 NS

was performed successfully, the patients enjoyed a shorter surgical time and a briefer postoperative hospital stay (6.1 vs 10.5 days; p < 107). Many modifications in the surgical technique have been described, including the use of additional cannulas, more versatile angled or side-viewing laparoscopes, sterile specimen bags to retrieve lost stones or extract infected tissue, decompression of the gallbladder, routine intraoperative cholangiography, and liberal use of sutures to control the cystic duct and artery [26]. Most of them were developed in an attempt to facilitate exposure of the biliary anatomy and diminish the incidence of gallstone or bile spillage.

The results of earlier studies aimed at defining preoperative factors that would predict conversion to open cholecystectomy have been contradictory. Lo et al. [16] found that the only factors predictive of conversion were advanced patient age, larger gallstones in the gallbladder, and severe adhesions. Although several authors claimed that male sex does not affect conversion rate [6, 16, 23], others found a positive correlation [1,9]. Other authors reported that severity of the inflammation was an important prognostic factor for a successful laparoscopic approach in acute cholecystitis [1, 3, 10] and that the intraoperative finding of empyema of gallbladder or gangrenous cholecystitis increased the odds for conversion. Perhaps the most important predictor of the success of attempted LC in patients with acute cholecystitis was the timing of surgery. Previous reports [11, 18] on patients undergoing conventional open cholecystectomies suggested that cholecystectomy should be performed early in the course of the disease, preferably in the first 72 h after the onset of symptoms. Is there an optimal timing for the performance of laparoscopic cholecystectomy? The conversion rates in patients operated on before and after 3 days after the onset of symptoms were 27% and 59.5%, respectively, and were significantly different (p 0.0002). Other laparoscopic studies [12, 15, 20] have also found that patients with prolonged preoperative illnesses had higher conversion rates. The most common reason for conversion was the severity of inflammation, along with the existence of severe

361

adhesions. In the early phase of acute inflammation, adhesions are easily separated, and there is usually an edematous plane around the gallbladder that facilitates dissection. After a period of conservative treatment, the inflammation and edema are replaced by fibrotic adhesions between the gallbladder and surrounding structures, which occasionally render laparoscopic dissection extremely difficult. Postoperative complications were similar in both groups. The absence of major complications, such as bile duct injury, suggests that LC can be performed safely in both early or delayed settings. The major advantage of early LC is the reduction of the total hospital stay. One of its main benefits is the potential for an earlier return to work; however, the relative recuperation periods were not compared in this study. Early LC offers a definitive treatment during the same admission and avoids the problems associated with failed conservative management (12.7% in our series). Two recent prospective randomized studies [14, 17] showed that initial conservative treatment followed by delayed interval surgery cannot reduce the morbidity and conversion rates of laparoscopic cholecystectomy for acute cholecystitis. But early operation before 3 days of admission yields both medical and socioeconomic benefits, in terms of a briefer hospital stay and an accelerated recuperation period. In summary, laparoscopic intervention appears to be a safe and beneficial option in the management of selected patients with acute cholecystitis. Higher rates of conversion and morbidity were observed, as compared with those for elective laparoscopic cholecystectomy. The complication rate compares favorably with that of open cholecystectomy. Laparoscopic cholecystectomy should be carried out soon as the diagnosis of acute cholecystitis is established and preferably within 3 days after the onset of symptoms. Early laparoscopic cholecystectomy allows a reduction in both the conversion rate and the total hospital stay as significant medical and economic benefits. References
1. Bickel A, Rappaport A, Kanievski V, Vaksman I, Haj M, Geron N, Eitan A (1996) Laparoscopic management of acute cholecystitis: prognostic factors for success. Surg Endosc 10: 10451049 2. Cox MR, Gunn IF Eastman MC, et al. (1992) Open cholecystectomy: a control group for comparison with laparoscopic cholecystectomy. Aust N Z J Surg 62: 795801 3. Cox MR, Wilson TG, Luck AJ, Jeans PL, Padbury RTA, Tooulis J (1993) Laparoscopic cholecystectomy for acute inflammation of the gallbladder. Ann Surg 218: 630634 4. Cuscheiri A, Berci G, McSherry CK (1990) Laparoscopic cholecystectomy [Editorial]. Am J Surg 159: 273 5. El Madani A, Badawy A, Henry C, Nicolet J, Vons Smadja C, Franco

6. 7.

8. 9.

10. 11. 12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25. 26.

D (1999) Cholecystectomie laparoscopique dans les cholecystites aigue s Chirurgie 124: 171176 Fabre JM, Fagot H Domergue J, et al. (1994) Laparoscopic cholecystectomy in complicated cholelithiasis. Surg Endosc 8: 11981201 Fink Bennett D, Freitas JE, Ripley SD, Bree RL (1985) The sensitivity of hepatobiliary imaging and real-time ultrasonography in the detection of acute cholecystitis. Arch Surg 120: 904906 Graves Jr MA, Ballinger WF, Anderson WJ (1991) Appraisal of laparoscopic cholecystectomy. Ann Surg 213: 655664 Hutchinson CH, Traverso LW, Lee FT (1994) Laparoscopic cholecystectomy: do preoperative factors predict the need to convert to open? Surg Endosc 8: 875878 Jacobs M, Verdeja JC, Goldstein HS (1991) Laparoscopic cholecystectomy in acute cholecystitis. J Laparoendosc Surg 1: 175177 Jarvinen HJ, Hastbacka J (1980) Early cholecystectomy for acute cholecystitis: a prospective randomized study. Ann Surg 501505 Kenny P, Richard C (1996) Laparoscopic cholecystectomy in acute cholecystitis: what is the optimal timing for operation? Arch Surg 131: 540545 Kiviluoto T, Siren J, Luukkonen P, Kivilaakso E (1998) Randomised trial of laparoscopic versus open cholecystectomy for acute and gangrenous cholecystitis. Lancet 351: 321325 Lai PBS, Kwong KH, Leung KL, Kwok SPY, Chan ACW, Chung SCS, Lau WY (1998) Randomized trial of early versus delayed laparoscopic cholecystectomy for acute cholecystitis. Br J Surg 85: 764 767 Lo CM, Liu CL, Lai ECS, Fan ST, Wong J (1996) Early versus delayed laparoscopic cholecystectomy for treatment of acute cholecystitis. Ann Surg 223: 3742 Lo CM, Fan ST, Liu CL, Lai ECS, Wong J (1997) Early decision for conversion of laparoscopic to open cholecystectomy for treatment of acute cholecystitis. Am J Surg 173: 513517 Lo CM, Liu CL, Fan ST, Lai ECS, Wong J (1998) Prospective randomized study of early versus delayed laparoscopic cholecystectomy for acute cholecystitis. Ann Surg 227: 461467 Norrby S, Herlin P, Holmin T, Sjodahl R, Tagesson C (1983) Early or delayed cholecystectomy in acute cholecystitis? A clinical trial. Br J Surg 70: 163165 Ralls PN, Colletti PM, Lapin SA, et al. (1985) Real-time sonography in suspected acute cholecystitis: prospective evaluation of primary and secondary signs. Radiology 155: 767771 Rattner DW, Ferguson C, Warshaw AL (1993) Factors associated with successful laparoscopic cholecystectomy for acute cholecystitis. Ann Surg 217: 233236 Roslyn JJ, Binns GS, Hughes EFX, et al. (1993) Open cholecystectomy: a contemporary analysis of 42,474 patients. Ann Surg 218: 129 137 Schirmer BD, Edge BS, Dix J, et al. (1991) Laparoscopic cholecystectomy: treatment of choice for symptomatic cholelithiasis. Ann Surg 213: 665676 Schrenk P, Woisetschlager R, Wayand NU (1995) Laparoscopic cholecystectomy: cause in conversion in 1,300 patients and analysis of risk factors. Surg Endosc 9: 2528 Sharp KW (1998) Acute cholecystitis. Surg Clin North Am 68: 269 279 Southern Surgeons Club (1991) A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 334: 10731078 Zucker KA, Flowers JL, Bailey RW, Graham SM, Buell J, Imbembo AL (1993) Laparoscopic management of acute cholecystitis. Am J Surg 165: 508514

Original articles
Surg Endosc (2000) 14: 326329 DOI: 10.1007/s004640020013 Springer-Verlag New York Inc. 2000

Laparoscopic intraperitoneal polytetrafluoroethylene (PTFE) prosthetic patch repair of ventral hernia

Prospective comparison to open prefascial polypropylene mesh repair
E. J. DeMaria, J. M. Moss, H. J. Sugerman
Department of Surgery, Section of General and Endoscopic Surgery, MCV Station Box 980519, Medical College of Virginia of Virginia Commonwealth University, Richmond, VA 23298, USA Received: 5 June 1998/Accepted: 15 October 1999

Abstract Background: The purpose of this study was to determine whether laparoscopic intraperitoneal polytetrafluoroethylene (PTFE) prosthetic patch (LIPP) repair of a ventral hernia is superior to open prefascial polypropylene mesh (OPPM) repair in a tertiary care university hospital in an urban environment. Methods: Data on 39 consecutive patients undergoing either LIPP repair (n 21) or OPPM repair (n 18) were compared. Results: Findings showed that LIPP repair is characterized by less painful recovery and shorter hospital stay, with 90% of patients treated successfully as outpatients as compared with 7% in the OPPM group. The total facility costs for the LIPP repair ($8,273 $2,950) was significantly lower than for the OPPM repair ($12,461 $5,987) (p < 0.05). Two serious delayed complications in the LIPP group were treated by reoperation (colocutaneous fistula, mesh infection), but the higher readmission costs in this group did not negate the overall cost advantage for LIPP repair. In the follow-up evaluation, 1 hernia recurrence was found in the LIPP repair group, and none in the OPPM group. Conclusions: Initial experience suggests that LIPP repair has advantages over OPPM repair in terms of decreased hospitalization, postoperative pain, and disability. Refinements in the technique to reduce complications may make LIPP repair the procedure of choice for repair of ventral hernias. Key words: Laparoscopic intraperitoneal polytetrafluoroethylene Laparoscopic intraperitoneal prosthetic patch

Correspondence to: E. J. DeMaria

Ventral hernias remain a major cause of morbidity and the need for reoperation in patients undergoing abdominal surgery. Furthermore, subsequent hernia recurrence as well as significant incisional pain and postoperative disability often accompany ventral hernia repair performed using traditional surgical techniques. Recently, laparoscopic techniques for prosthetic mesh repair of ventral abdominal wall hernias have been described [5, 6, 8, 10]. Although these preliminary reports suggest superior results in terms of postoperative disability and pain as compared with traditional open techniques for repair, at this writing, no prospective comparison with traditional repair techniques has been reported. Furthermore, in this era when the economic impact of health care interventions is increasingly scrutinized, it has been shown that minimally invasive surgical techniques do not necessarily provide a cost advantage over traditional surgical techniques. Such is the case for laparoscopic cholecystectomy and groin hernia repair, in which decrease in the length of hospitalization has not translated into a decrease in the cost for the procedure. However, in the case of ventral hernia, for which a significant period of hospitalization is required for recovery from the traditional surgical procedure, it is conceivable that decreasing the postoperative hospitalization through laparoscopic techniques may, in fact, decrease the cost of the procedure. The current study was undertaken to compare the results of laparoscopic ventral hernia repair directly with a traditional open surgical technique in which repair of the hernia defect is reinforced by polypropylene mesh anchored to the anterior abdominal fascia in the prefascial position. We have previously used this open technique in a large number of patients, demonstrating a low recurrence risk in using the procedure for ventral hernia [11, 12]. The laparoscopic procedure used was a modification of the technique described by Park et al. [8], in which the fascial defect was repaired by intraperitoneal placement of a patch of expanded polytetra-

327 Table 1. Characteristics of 21 patients undergoing laparoscopic intraperitoneal polytetrafluoroethylene patch repair (LIPP) compared with those of 18 patients treated by open prefascial polypropylene repair (OPPM) as the primary procedure for repair of ventral hernia LIPP Number of patients Males:females Age (range) in years History of previous ventral hernia repair (%) Previous gastric bypass for obesity (%) Large hernia defect for repair (%) 21 8:13 46.2 (2676) 11 (54) 10 (48) 14 (68) OPPM 18 7:11 46.7 (2770) 3 (17) 8 (44) 8 (43) Mean hospital days (range) Successful outpatient surgery (%) Postoperative parenteral narcotic therapy (%) Subjective pain Assessment Not available (%) At 6 hours after operation Mild (%) Moderate (%) Severe (%) At 24 hours after operation None (%) Mild (%) Moderate (%) Severe (%) Cost of primary procedure Table 2. Outcome data for 21 patients undergoing laparoscopic intraperitoneal polytetrafluoroethylene patch repair (LIPP) compared with data for 18 patients treated by open prefascial polypropylene repair (OPPM) as the primary procedure for repair of ventral hernia LIPP (n 21) 0.8 (0.53.0) 19 (90) 2 (10) 4 (19) 9 (52) 7 (41) 1 (6) 1 (6) 10 (59) 6 (35) 0 $8,273 $2,950 OPPM (n 18) 4.4 (0.514) 1 (7) 11 (79) 4 (22) 3 (21) 9 (64) 2 (14) 0 8 (57) 4 (29) 2 (14) $12,461 $5,987

fluoroethylene (PTFE) anchored to the fascia by means of a commercially available tacking device (Origin Medsystems, Inc., Menlo Park, CA). Materials and methods
All patients undergoing ventral hernia repair, performed by the authors as the primary surgical procedure at Medical College of Virginia Hospitals between January 1, 1996 and June 1997, were studied. After a full informed consent discussion, the surgeon chose to perform either a traditional open surgical placement of polypropylene mesh (OPPM) as a prefascial reinforcement [11, 12] or the laparoscopic intraperitoneal placement of a polytetrafluoroethylene (PTFE) patch (LIPP). The open surgical procedure was performed in a standard fashion using a technique previously published by our group and found to have a 1% risk of hernia recurrence during the follow-up period [11, 12]. The laparoscopic repair was performed using a modification of the technique previously described by Park et al. [8]. Laparoscopic access to the abdominal cavity was attained using an open insertion technique away from previous areas of incision followed by insertion of two or three additional 5- or 10-mm trocars under direct vision. Adhesions were lysed sharply to expose the hernia defect, and no attempt was made to excise the hernia sac. An appropriate size of 1-mm-thick PTFE mesh (Dualmesh, W. L. Gore and Associates, Inc., Phoenix, AZ, USA) was chosen to cover the fascial defect, with a 2-cm overlap of the fascial edges. The mesh was placed through a 10-mm trocar after four-quadrant-long sutures of nonabsorbable material were placed in the four quadrants of the material to facilitate positioning the patch by passage of a grasping needle device (Endoclose, United States Surgical, Norwalk, CT, USA) through the abdominal wall in the appropriate positions and pulling the sutures through to position the patch against the fascia covering the defect. The patch then was anchored to the fascia using a commercially available 5-mm tacking device (Origin Medsystems). Data are presented as mean standard error. Data between groups were compared utilizing students t-test, Fishers exact test, and by the Wilcoxan rank sum test as appropriate. A p value of less than 0.05 was required for statistical significance.

Results Baseline preoperative data in the two groups are presented in Table 1. Patients undergoing LIPP repair did not differ from those undergoing OPPM in terms of age or gender. Previously, 46% of the patient population had been treated by proximal gastric bypass surgery to manage morbid obesity, and there was no difference between the two groups in the frequency of previous bariatric surgery. The LIPP group had a significantly greater proportion of patients with recurrent ventral hernias. Similarly, more patients in the LIPP group had large hernia defects (>6 8 cm), although this difference between the groups did not reach statistical significance.

Data on the outcome of hernia surgery in the two groups examined are presented in Table 2. Despite the increased frequency of large hernia defects and more recurrent hernias, LIPP repair led to the successful performance of surgery on an outpatient basis in 90% of patients as compared with only 7% success in OPPM repair patients (p < 0.05). Similarly, length of postoperative stay was significantly shorter for LIPP repair patients than for OPPM repair patients (0.8 vs 4.4 days; p < 0.05). One reason behind the increased success of outpatient surgery in the LIPP group appeared to be a significant decrease in the need for parenteral narcotic therapy in these patients as compared with the OPPM repair group (10% vs 79%; p < 0.05). Although parenteral narcotics were used more often in the OPPM repair group, data on the subjective assessment of pain by the patients suggested that patients in the OPPM repair group still experienced more pain than those in the LIPP repair group. The shorter hospital stay led to a significantly lower cost for the ventral hernia repair procedure in patients undergoing LIPP repair. Complications, hospital readmission data, and the impact of readmission on hospital costs are presented in Table 3. The overall incidence of complications did not differ between the two groups, with the exception of an increase in the incidence of wound seromas in the LIPP repair group (43% vs 22%). In most cases, seromas resolved completely after treatment by needle aspiration on an outpatient basis on one or two occasions. There were two major complications that required treatment by reoperation, both in patients undergoing initial LIPP repair. One was a patient with a large midline abdominal hernia defect from a laparotomy for trauma who developed a seroma, which was treated by aspiration in the early postop period. He subsequently did well and then presented 4 months after surgery with an abdominal wall abscess penetrating through the skin overlying the hernia defect. The infection recurred despite incision and drainage, wound packing, and several courses of antibiotic therapy. He ultimately was admitted 8 months after the original procedure

328 Table 3. Complications, readmissions to the hospital, and the cost impact of readmissions on overall procedure cost in 21 patients undergoing laparoscopic intraperitoneal polytetrafluoroethylene patch repair (LIPP) compared with those in 18 patients treated by open prefascial polypropylene repair (OPPM) as the primary procedure for repair of ventral hernia LIPP (n 21) Complications Paralytic ileus Bowel obstruction Wound infections Major (readmission) Minor (local outpatient care) Seroma (%) Other majora Hernia recurrence (%) Readmission after discharge (%) Days of readmission stay (range) Cost of readmission stay Reoperation required (%) Overall cost (initial plus readmission)
a

OPPM (n 18) 0 2 2 4 4 (22) 1 0 5 (28) 4.6 (115) $4,100 + $7,007 0 $13,600 + $8,720

1 0 1 1 9 (43) 1 1 (5%) 3 (14) 16.7 (336) $19,173 + $16,312 2 (10) $11,013 + $8,393

demonstrated a significantly lower overall cost than the OPPM repair group. At this writing, only one patient in either group has experienced hernia recurrence. This patient had previous surgery for morbid obesity and a large incisional hernia developed encompassing the entire midline incision in the upper abdomen. For this patient, LIPP repair was successfully accomplished on an outpatient basis, but the patient did require readmission for abdominal distension from paralytic ileus. A small area of hernia recurrence was noted on follow-up assessment 18 months after the initial repair. A further factor in this case was recurrent obesity, with significant regained weight, documented by contrast studies as resulting from disruption of the staple line as a complication of the previous gastric bypass. At this writing, reoperation has not yet been completed, but the recurrent defect appears to be in the area of the umbilicus at the inferior margin of the patch.

Discussion The current data suggest that laparoscopic repair of ventral hernia may offer significant advantages over traditional open ventral hernia repair techniques. The data document a dramatic shift toward successful outpatient management of ventral hernia repair using the laparoscopic technique as well as a decrease in the pain experienced by patients after the procedure, a decrease in the requirement for parenteral narcotic therapy, and a significant decrease in the cost of hospitalization for ventral hernia surgery using the laparoscopic techniques. Furthermore, the recurrence risk associated with laparoscopic ventral hernia repair appears to be acceptable because only one recurrence was documented over the 12- to 24-month follow-up period in this study. The current study does suffer from a lack of randomization of patients into the two groups studied. It might therefore be assumed that a selection bias on the part of the surgeons led to improved results in terms of postoperative pain, control, or the length of hospitalization in the laparoscopic group. Conversely, however, the data suggest that the selection bias for the surgeons involved triaging more difficult ventral hernia cases to the laparoscopic group as demonstrated by the higher frequency of patients with prior ventral hernia surgery or large (>6 8 cm) hernia defects in that group. This was, in fact, an intentional bias on the part of the authors: A number of patients with hernia recurrence after anterior open repair with mesh using various techniques were approached subsequently using the laparoscopic intraperitoneal patch technique to avoid the very extensive dissection required after previous mesh repair. Therefore, it might be suggested that the current results are even more meaningful in light of this selection bias toward managing more complex cases laparoscopically. Erosion and fistulization occurred in one patient in this study. This is the major feared complication of intraperitoneal mesh placement for hernia repair. In fact, the expanded PTFE patch (Dualmesh) used in this trial was chosen specifically in an attempt to minimize the risk of erosion and fistulization. This PTFE patch is characterized by two different surfaces, one that promotes fibrous ingrowth into the mesh and another that is relatively resistant to adhesion

LIPP colocutaneous fistula, OPPM deep venous thrombosis

for surgical excision of the PTFE patch, which was accomplished without additional complication. The cost of subsequent readmission for treatment in this case was $21,337 for an 11-day hospitalization. The other major complication requiring operative therapy occurred in a patient who developed a seroma, which was drained by needle aspiration on postoperative day 7, only to be followed by signs of wound infection 3 days later. She claimed that the sudden abdominal pain developed the day after the needle aspiration procedure when she was struck in the abdomen by a patient in a wheelchair at the chronic nursing care facility where she worked. Feculent drainage developed from the wound after local incision of the skin to treat infection, and she underwent subsequent exploration and diversion for a colocutaneous fistula. She recovered and underwent colonic reanastomosis and repair of the hernia as a subsequent procedure 4 months later. The cost of readmission for treatment on two occasions, including the two reoperations in this case, was $34,300 for a total of 36 days of additional inpatient treatment. The cause of the colonic fistulization presumably was erosion of the colon from an adjacent surgical tack used to anchor the patch to the fascia. Several tacks were seen adjacent to the colonic erosion at the time of surgical exploration, and no adhesions to the patch itself were identified, although it is possible that the colon adhered to the patch and was torn when the patients abdomen was traumatized. Although more patients were readmitted to the hospital for treatment of complications after OPPM repair (28% vs 14%), the length of hospitalization and cost for readmission treatment was much greater in patients undergoing initial LIPP repair, primarily because of the need for reoperative treatment in two patients in this group. However, when the costs for readmission treatment were added to the costs for initial hernia surgery, including the cost for management of complications in each group, the LIPP repair group still

329

formation and placed adjacent to the abdominal viscera. Others [5, 10] have used polypropylene mesh in the peritoneal position to accomplish laparoscopic hernia repair. This mesh promotes fibrous ingrowth and adhesion formation to the abdominal viscera. The PTFE material used in this series has been studied both scientifically and clinically, with documentation of less adhesion formation [1, 2, 3, 4, 7, 9]. In our experience with the two reoperations in this series, as well as a third case not included in this report because multiple simultaneous procedures were performed in addition to ventral hernia repair, few adhesions to the intraperitoneal patch itself have been found. However, adhesions to the tacks placed to hold the mesh to the abdominal fascia have been found. In the case of erosion with fistulization reported here, it apparently was a surgical tack in the location of colonic injury that participated in the process of erosion. Sutures, staples, or tacks used to hold the mesh in place likely create adhesions no matter what type of mesh material is used to prevent their formation. In the development of laparoscopic ventral hernia repair, erosion and fistulization has not been reported [5, 6, 8, 10]. However, with intraperitoneal mesh placement, the use of mesh materials such as polypropylene, which promotes fibrous ingrowth and adhesions, will lead almost certainly to this complication in some patients. Despite the data suggesting that the PTFE material used in this study is the best available material currently available for intraperitoneal hernia repair [1, 2, 3, 4, 7, 9], a case of erosion and fistulization did occur. Technical advances in the procedure likely will be required to eliminate this risk. However, on the basis of our experience, this problem probably can be minimized if the surgical tacks are anchored deeply in the patch and fascia and do not protrude. Any tacks that protrude should be removed. Seroma formation occurred in nearly half of the laparoscopically treated patients in this study. These patients usually presented with a firm mass in the abdominal wall in the area of the prior hernia. Confirmation of seroma was made by needle aspiration under sterile technique using a smallbore 25-gauge needle. It is important to exclude the possibility of early hernia recurrence in some of these cases before inserting a larger needle. This could be accomplished easily when necessary using ultrasound. In the current study, seromas responded to a single needle aspiration, or to three at the most, before complete resolution, and were not a long-term problem. In the current case of mesh infection requiring removal, it appears likely that inadequate sterile technique during aspiration of a seroma contributed to the subsequent mesh infection. Mesh infection requiring mesh excision is uncommon, yet remains a risk with any mesh herniorrhaphy, whether laparoscopic or traditional techniques are used. Therefore, it ap-

pears unlikely that the laparoscopic technique contributed to the development of this complication. No difference in the incidence of complications were identified between groups in the current study. However, clearly the three readmissions in the laparoscopic group, as described earlier, represented complex management situations and resulted in long hospitalizations, reoperations, and expensive hospital bills. However, even when the increased readmission costs in the laparoscopic group were factored into the overall cost for the procedures in both groups, there remained a cost advantage for the laparoscopic repair. Hopefully, refinements in the technique for laparoscopic intraperitoneal prosthetic mesh repair will decrease the complication rate that we experienced as part of our learning curve for the procedure and further increase the benefits of this technique. It is clear from the current study that laparoscopic ventral hernia repair offers advantages over traditional techniques, and that this approach may emerge as the procedure of choice for many ventral hernia patients. References
1. (American Fertility Society (1992) Prophylaxis of pelvic sidewall adhesions with Gore-Tex surgical membrane: a multicenter investigation. Fertil Steril 57: 921923 2. Boyers SP (1994) Gore-Tex surgical membrane. Infertil Reprod Med Clin North Am 5 3. Grow DR, Seltman HJ, Coddington CC, Hodgen GD (1994) The reduction of postoperative adhesions by two different barrier methods versus control in cynomolgus monkeys: a prospective, randomized, crossover study. Fertil Steril 61: 11411146 4. Haney AF, Hesla J, Hurst BS, Kettel LM, Murphy AA, Rock JA, Rowe G, Schlaff WD (1995) Expanded polytetrafluoroethylene (GoreTex surgical membrane) is superior to oxidized regenerated cellulose (Interceed TC7) in preventing adhesions. Fertil Steril 63: 10211026 5. Holzman MD, Purut CM, Reintgen K, Eubanks S, Pappas TN (1997) Laparoscopic ventral and incisional hernioplasty. Surg Endosc 11: 3235 6. LeBlanc KA, Booth WV (1993) Laparoscopic repair of incisional abdominal hernias using expanded polytetrafluoroethylene: preliminary findings. Surg Laparosc Endosc 3: 3941 7. March CM, Boyers S, Franklin R, Haney AF, Hurst B, Lotze E, Rock JA, Rowe G, Schalaff W (1993) Prevention of adhesion formation/ reformation with the gore-tex surgical membrane. Gynecol Surg Adhes Prev 253259 8. Park A, Gagner M, Pomp A (1996) Laparoscopic repair of large incisional hernias. Surg Laparosc Endosc 6: 123128 9. Rowe G, Rubio PA, Jansen D (1993) Clinical use of the Gore-tex surgical membrane for preventing pelvic-peritoneal adhesions. Houston Med 9: 8688 10. Saiz AA, Willis IH, Paul DK, Sivina M (1996) Laparoscopic ventral hernia repair: a community hospital experience. Am Surg 5: 336338 11. Sugerman HJ, Kellum Jr JM, Reines HD, DeMaria EJ, Newsome HH, Lowry JW (1996) Greater risk of incisional hernia with morbidly obese than steroid-dependent patients and low recurrence with prefascial polypropylene mesh. Am J Surg 171: 8084 12. Wagman LD, Barnhart GR, Sugerman HJ (1985) Recurrent midline hernia repair. Surg Gynecol Obstet 161: 181

Surg Endosc (2000) 14: 367371 DOI: 10.1007/s004640000156

Springer-Verlag New York Inc. 2000

Effect of pressure and gas type on intraabdominal, subcutaneous, and blood pH in laparoscopy
C. Kuntz,1 A. Wunsch,2 C. Bo deker,1 F. Bay,1 R. Rosch,1 J. Windeler,3 C. Herfarth1
1 2

Department of Surgery, University of Heidelberg, Im Neuenheimer Feld 110, D-69120 Heidelberg, Germany Department of Transplant Surgery, University of Mainz, Laugenbeckstrasse 1, D55101 Mainz, Germany 3 Institute of Medical Biometry, University of Heidelberg, Im Neuenheimer Feld 305, D 69120 Heidelberg, Germany Received: 11 June 1998/Accepted: 12 February 1999

Abstract Background: According to the literature, the number of port-site metastases in laparoscopic surgery varies considerably depending on the type of gas used for the pneumoperitoneum. In order to investigate this observation we studied the changes in blood, subcutaneous, and intra-abdominal pH during laparoscopy with helium, CO2 and room air in a rat model. In addition, we looked at the influence of intraabdominal pressure and duration of pneumoperitoneum on the pH during the laparoscopy. Methods: pH was measured by tonometry, intra-abdominally and subcutaneously. A pH electrode was additionally placed into the subcutaneous tissue and the results compared to those measured by tonometry. Blood samples were taken from a catheter in the carotid artery. The intraabdominal pressure was 0, 3, 6, 9 mmHg for 30 min in each case. We investigated the effect of pneumoperitoneum with CO2, helium and air in randomized groups of 5 rats. In an additional series the pressure was held constant at 3 mmHg and the pH was measured every 30 min. Results: Due to the different absorption capacity of the peritoneum, laparoscopy with CO2 decreases the subcutaneous pH from 7.35 to 6.81. Blood pH is reduced from 7.37 to 7.17 and the intra-abdominal pH from 7.35 to 6.24. Other, less absorbable gases induce smaller changes of blood and subcutaneous pH (only 10% of CO2). In a variance analysis the p value is less than 0.001. The influence of duration of laparoscopy (30 min vs 90 min) on the subcutaneous pH is less compared to the influence of intra-abdominal pressure (0, 3, 6, 9 mmHg). Conclusions: Depending on the type of gas (CO2, air, helium) used for laparoscopy blood, subcutaneous and intraabdominal pH are influenced differently. Because lower pH is known to impair local defense mechanisms, these results may be one explanation for the higher incidence of port-site

metastasis in laparoscopy with CO2 than with other gases, as reported in the literature. Key words: Pneumoperitoneum CO2 Helium Blood pH Intra-abdominal pH Subcutaneous pH Laparoscopy

Correspondence to: C. Kuntz

Since its introduction, laparoscopic surgery and especially laparoscopic cholecystectomy have been rapidly accepted for clinical use. This minimal invasive procedure generally requires a pneumoperitoneum for adequate visualization and exposure of the structures to be operated upon [6, 11, 16, 22, 25, 27, 35]. Carbon dioxide (CO2) has been used for insufflation because it is low in cost and nonflammable. Its diffusion capacity is high, with subsequent rapid absorption and excretion [11, 12]. Side effects of a pneumoperitoneum with CO2 are hypercapnia, acidosis and pulmonary hypertension [9, 22, 23, 30]. Therefore, other gases, such as helium, argon, N2O and gasless laparoscopy were investigated [5, 6, 16, 24, 25, 32, 35]. One problem of laparoscopy in oncologic surgery is the occurrence of port-site metastases after operating on malignancies [13, 15, 21, 28, 35]. A correlation between the incidence of port-site metastases and the type of gas used during laparoscopy has been suggested in certain animal and in-vitro studies [15, 17, 28]. In some animal studies more port-site metastases were observed when using CO2 than with air or helium [15]. Consequently, it may be assumed that changes in pH play a role in the development of these metastases [15, 35]. Therefore we investigated the influence of different gases (CO2, air, helium) on the pH measured in blood, subcutaneous fat tissue and in the abdominal cavity in rats in a prospective randomized study. Furthermore, we examined the influence of the respective intra-abdominal pressure (0, 3, 6, 9 mmHg) on the pH (blood, subcutaneous fat tissue, intra-abdominal) with each gas tested.

368

Materials and methods Animals

We used male Wistar rats with a body weight of 350380 g. Rats were bred under specific pathogen-free conditions to avoid pulmonary infections. The animals were kept under standardized conditions: temperature between 22C and 24C relative humidity 5060%, 12 h of light followed by 12 h of darkness. They were fed a standard labaratory diet with free access to food and water. Food was withdrawn 12 h before surgery.

each rat and each gas. In a second experiment in 5 rats, the intra-abdominal pressure of CO2 was 3 mmHg for a period of 90 min to investigate the influence of duration and weight of the intra-abdominal pressure.

Statistics
Statistical analysis was performed by an analysis of variance with repeated measures to check whether there were any differences between the groups over the time course (laparoscopic or open procedure); p < 0.01 was defined as significant.

Anesthesia
Each operation was performed on spontaneously breathing rats which had anesthesia with ketamine (100 mg/kg body weight) i.m. and Xylazine (8 mg/kg body weight) i.m. Postoperative analgesia was obtained by tramadol added to each water bottle over 24 h (40 mg Tramadol/100 ml of water).

Results The effects of laparoscopy with CO2 on the pH are significantly different in the three tested compartments (blood, subcutaneous fat tissue, intra-abdominal). As long as the intra-abdominal pressure is 0 mmHg, the pH is around 7.38 in all compartments. The blood pH decreases to 7.28 at 6 mmHg and to 7.17 at 9 mmHg when using CO2 for pneumoperitoneum. The subcutaneous pH decreases to 7.03 at 6 mmHg and to 6.81 at 9 mmHg, while the intra-abdominal pH decreases to 6.18 when using CO2 for pneumoperitoneum (Fig. 1). Using helium as the gas, the subcutaneous pH decreases to 7.29 at 9 mmHg intra-abdominal pressure, whereas the intra-abdominal pH increases to 7.49 (Fig. 2). Using air as the gas, the blood pH decreases to 7.29 at 9 mmHg intra-abdominal pressure. The subcutaneous pH decreases to 7.24, while the intra-abdominal pH remains at 7.32 (Fig. 3). The influence of the different gases on the pH of blood is shown in Fig. 4. The considerable decrease of the blood pH using CO2 (from 7.37 to 7.17) is remarkable compared to the changes of blood pH using air or helium (7.37 to 7.29). In an analysis of variance these changes (blood pH alteration due to different gases and intra-abdominal pressures) are significant (p 0.0001). The influence of the different gases on the subcutaneous pH is shown in Fig. 5. While the influences of air and helium are similar, CO2 leads to an obvious decrease of pH, from 7.35 to 6.81. In an analysis of variance the p value is 0.0001 for tonometry as well as pH-metry. The near-concordance between these two methods is demonstrated (Figs. 13). The p value in the analysis of variance is 0.9885. This shows that these curves are most probably identical. In Fig. 1 the effect of CO2 on the different compartments is shown. The strong decrease of the intra-abdominal pH is demonstrated. Furthermore, the considerable decrease of pH in the subcutaneous fat tissue is shown when using CO2 for laparoscopy (7.35 to 6.81). Using helium, the decrease of the subcutaneous pH is less than 10% of what occurred with CO2 (Fig. 2). Using air, the decrease of the subcutaneous pH is 10% of what occurred with CO2 (Figs. 1 and 3). The influence of the weight of the intra-abdominal pressure (0, 3, 6, 9 mmHg CO2 over 30 min) compared to the influence of duration of the pneumoperitoneum (3 mmHg over 90 min) is plotted in Fig. 6. The influence of duration is significantly less important than the influence of pressure (p 0.015).

Operative procedure
At first a catheter was inserted into the left carotid artery for blood sampling. Then the rats were secured to the operating table in a supine position. A laparoscope (2.7 mm trocar with a 30 optic) was placed into the abdomen in the midline just above the penis vein. Through a second small incision in the right lower abdomen a tonometry catheter (TRIP NGF) was placed into the abdomen. Then 10 ml of saline fluid (NaCl 0.9%, pH 7.0 without buffer capacity) were instilled into the abdominal cavity in order to measure the abdominal pH in the installed saline fluid between the bowels and the abdominal wall in the supine position. The initial correct position of the catheter tip was confirmed by camera vision. A second tonometry catheter (TRIP NGF) was inserted through another small incision into the subcutaneous space in the left upper abdomen. A pH-meter probe was placed in another subcutaneous space in the right upper abdomen. Finally, a temperature sensor was placed subcutaneously. The abdominal cavity was connected to a large airtight box to equalize pressure fluctuations during the procedure. We used a fluid manometer (range 012 mmHg) to control the pressure indicated by the insufflator.

Tonometry
The TRIP NGF catheter (Tonometrics Division, Instrumentarium Corp., Helsinki, Finland) is used clinically for measuring the regional pCO2 and thereby the intramucosal pH of gastric mucosal tissue. Instead of the pCO2 of the gastric mucosa, the pCO2 of any other surrounding tissue can also be measured. After insertion of the catheter, it was filled with 1 ml of saline fluid (NaCl 0.9%). After equilibration (30 min) 0.3 ml of the fluid were removed and discharged (dead space of the catheter). In the remaining 0.7 ml the pCO2 was measured with the help of a blood gas analyzer (Fa Radometer). At the same time a sample of arterial blood was drawn from the carotid artery and also analyzed in a blood gas analyzer. The pH was determined according to the following formula: pH = 6.1 + log10
HCO3arterial

pCO2 EF 0.031

EF = equilibration factor

pH-metry
pH-metry was performed with a pH probe (range pH 014, isopotential 68 pH, pH temperature compensation 0100C with a co-working temperature probe) in the subcutaneous fat tissue (Fa Microelectrodes, Bedford, MA).

Gas/pressure
The pneumoperitoneum was created with CO2, air, or helium (helium insufflator, Fa Wolf, Knittlingen, Germany) using 5 rats for each gas. The intra-abdominal pressure was 0, 3, 6, 9 mmHg for 30 min per pressure for

369

Fig. 1. pH changes using CO2. Fig. 2. pH changes using helium. Fig. 3. pH changes using air. Fig. 4. pH changes in blood. Fig. 5. pH changes in the subcutaneous fat tissue (tonometrie). Fig. 6. Subcutaneous pH in correlation to pressure and duration of CO2 insufflation.

Discussion In the last years, studies of systemic effects of laparoscopy increasingly outranged studies on the technical feasibility of laparoscopic procedures. Changes of the cardiorespiratory

system as well as alterations in the immune system by laparoscopy with CO2 or other gases were demonstrated [5, 6, 8, 11, 1416, 18, 20, 2224]. CO2 increases arterial blood pressure and cardiac output, leads to respiratory acidosis and alters intracranial pressure [3, 6, 11, 21, 26, 28, 30]. An

370

impairment of pulmonary function as induced in an experimental study by intravenous injection of dextran microspheres [4] caused no adverse hemodynamic effects during laparoscopic nephrectomy using CO2. Alterations of the humoral and the cellular immune system do occur but are less profound after laparoscopy than after laparotomy [1, 8, 20, 33]. This may be a reason to prefer laparoscopy in oncologic operations. However, one of the most important problems in laparoscopic surgery for malignancies is port-site metastases. They occur more often than abdominal wall metastases in the incision area in conventional surgery. Hughes et al. [13] report a 0.78% metastasis rate in the laparotomy region in 1,603 patients after conventional colorectal surgery. Portsite metastases are reported in 0% to 21% of laparoscopically operated patients [27]. Reasons for port-site metastases may be the chimney effect [10, 19, 28], the pH and the type of gas used for laparoscopy [10, 13, 15, 28], or the pneumoperitoneum alone [18]. Jacobi et al. [15, 16] observed the strong positive influence of CO2 on tumor cell growth, whereas helium reduced tumor cell growth significantly. These in-vitro results were confirmed by animal experiments [1517]. One reason for this effect may be the changes of the pH [15]. Therefore we measured the pH in the subcutaneous fat tissue, where most port-site metastases occur [15, 17]. The decrease of the subcutaneous pH during laparoscopy with CO2 is considerable (see Figs. 1 and 5). This effect is due to the absorption of CO2. Furthermore, the tension of the abdominal wall may lead to reduced blood flow and therefore reduced blood gas exchange. The reduced removal of acidic cell products (H+ excretion from every cell, excretion of intracellularly formed acids such as lactate, keton bodies, etc.) and the diffusing CO2 result in an increase of acidosis in the subcutaneous fat tissue. Comparable to the literature [27], we found a continuous decrease of the pH when raising the pressure. In the case of CO2 this is a direct effect of the pressure, as shown in Fig. 6. The pH value in blood (CO2) decreases from 7.37 to 7.17, which is significantly less than the pH value decrease in the subcutaneous fat tissue (Figs. 1, 4, and 5). In pigs, Ho et al. [10, 11] and Volz et al. [35] found that a CO2 pneumoperitoneum resulted in systemic CO2 absorption across the peritoneum. The absorption capacity of the peritoneum is about 200 ml/h/kg for CO2 [10]. Therefore the CO2 absorption in a 70-kg man approximates to 14 l/h. To eliminate this CO2 volume, the minute ventilation must be increased by about 75% [10]. In reality, however, the mechanical minute ventilation (in humans) is increased by only about 2030% at the most during laparoscopy. The blood pH then is more or less in an acceptable range but the subcutaneous pH is still decreased. The use of spontaneously breathing rats results in an improved minute ventilation due to the decreased pH. Therefore, spontaneously breathing rats accentuate the differences between the different gases, because mechanical ventilation of rats would be too susceptible to changes and would not produce better results. In contrast to the high absorption capacity of the peritoneum for CO2 (see above), the low aqueous solubility of helium minimizes its tissue absorption [23]. Dissolved helium is rapidly and totally excreted from pulmonary arterial blood into the alveoli, with essentially no helium retained in systemic arterial blood [23].

Whereas Ho et al. [11] report that the absorption of CO2 through the peritoneum alone (and not the intra-abdominal pressure) leads to acidemia, hypercapnia, and depressed hemodynamics, Shuto et al. [32] assume the contrary: they found similar effects using CO2 and helium, whereas the hemodynamic effects changed more with increasing pressure (CO2 or helium) [32]. The influence of the intraabdominal pressure on the decrease of subcutaneous pH is shown when using air as the gas (Fig. 3). Here the changes of the subcutaneous pH are caused by the restricted circulation due to the distended abdominal wall. Helium leads to an increase of the pH, which is confirmed through the results in the literature [9]. Laparoscopy with air or helium leads to a decrease of the pH in the subcutaneous fat tissue from 7.37 to 7.28. In the case of CO2, the pH value decreases to 6.88. The decrease of the subcutaneous pH is influenced by both the type of gas and the intra-abdominal pressure. The O2 diffusion from erythrocytes to the tissue is influenced by the pH (Bohr effect). The pH corresponds to the CO2 as shown above (see tonometry). The tension of the abdominal wall and therefore the decreased blood flow in the abdominal wall prevent a correction of the local pH in the subcutaneous fat tissue. Furthermore, the subcutaneous fat tissue does not have a high blood circulation [7]. Alterations of the splanchnic microcirculation during laparoscopy with CO2 were described by Schilling et al. [29], who partially measured the mucosal pH of the stomach with tonometry. Other authors describe mechanical effects as reasons for port-site metastases [27] as well as disseminated tumor cells in the smoke of laparoscopy caused by electrosurgery [2, 10, 12, 27]. Whether the subcutaneous pH alone or other factors [2, 9, 26, 27], or all of them combined, are the promotors for port-site metastases is unclear. In the literature there are publications suggesting a connection between pH, cell function, and the immune system [31, 34]. Our data confirm significant changes in the pHlocally (subcutaneously, intra-abdominally) and systemically (blood)during laparoscopy, as postulated by other authors [14, 15, 31]. However, these results were obtained through an experimental animal model and the demonstrated effects may be less marked in humans because ventilation can be controlled and therefore CO2 partially exhaled.

References
1. Allendorf JD, Bessler M, Wheelan RL, Trokel M, Laird DA, Terry MB, Treat MR (1997) Postoperative immune function varies with the degree of surgical trauma in a murine model. Surg Endosc 11: 427 430 2. Champault G, Taffinder N, Ziol M, Riskalla H, Catheline JMC (1997) Cells are present in the smoke created during laparoscopy surgery. Br J Surg 84: 993995 3. McDermott JP, Regan MC, Page R, Stoke MA, Berry K, Moriarty DC, Caushaj PF, Fitzpatrick JM, Gorey TF (1995) Cardiorespiratory effects of laparoscopy with and without gas insufflation. Arch Surg 130: 984988 4. Do rsam J, Bucuras CV, Mieck U, Brcovic D, Motsch J, Staehler G (1997) Effects of carbon dioxide pneumoperitoneum in pigs with impaired pulmonary function. J Endourol 11: 185189 5. Eisenhauer DM, Saunders CJ, Ho HS, Wolfe BM (1994) Hemodynamic effects of argon pneumoperitoneum. Surg Endosc 8: 315321 6. Fitzgerald SD, Andrus CH, Baudendistel LJ, Dahms TE, Kaminski DL

371 (1992) Hypercarbia during carbon dioxide pneumoperitoneum. Am J Surg 163: 186190 Glick Z, Wickler SJ, Stern JS, Horwitz BA (1984) Regional blood flow in rats after a single low protein, high carbohydrate test meal. Am J Physiol (1984) 247: R160R166 Griffith JP, Everitt NJ, Lancaster F, Boylston A, Richards SJ, Scaott CS, Benson EA, Sue-Ling HM, McMahon MJ (1995) Influence of laparoscopic and conventional cholecystectomy upon cell-mediated immunity. Br J Surg 82: 677680 Hewett PJ, Thomas WM, King G, Eaton M (1996) Intraperitoneal cell movement during abdominal carbon dioxide insufflation and laparoscopy. Dis Col Rect 39: S62S66 Ho HS, Saunders CJ, Gunther RA, Wolfe BM (1995) Effector of hemodynamics during laparoscopy: CO 2 absorption or intraabdominal pressure? J Surg Res 59: 497503 Ho HS, Gunther RA, Wolfe BM (1992) Intraperitoneal carbon dioxide insufflation and cardiopulmonary functions. Arch Surg 127: 928932 Hubens G, Pauwels M, Hubens A, Vermeulen P, van Marck E, Eyskens E (1996) The influence of a pneumoperitoneum on the peritoneal implantation of free intraperitoneal colon cancer cells. Surg Endosc 10: 809812 Hughes ESR, McDermott FT, Polglase AL, Johnson WR (1983) Tumour recurrence in the abdominal wall scar tissue after large-bowel cancer surgery. Dis Col Rect 26: 571572 Hunter JG, Staheli J, Oddsdottir M, Trus T (1995) Nitrous oxid pneumoperitoneum revisted. Surg Endosc 9: 501504 Jacobi CA, Sabat R, Bo hm B, Zieren HU, Volk HD, Muller JM (1997) Pneumoperitoneum with carbon dioxide stimulates growth of malignant colonic cells. Surgery 121: 7278 Jacobi CA, Bo hm B, Volk HD, Zieren HU, Mu ller JM (1997) The influence of laparotomy and laparoscopy with CO2 or helium on the bacterial translocation in a rat model. Langenbecks Arch Chir 247251 Jacobi CA, Ordemann J, Bo hm B, Zieren H, Liebenthal C, Volk H, Mu ller JM (1997) The influence of laparotomy and laparoscopy on tumor growth in a rat model. Surg Endosc 11: 618621 Jones DB, Guo LW, Reinhard MK, Soper NJ, Philpott GW, Connett J, Fleshman JW (1995) Impact of pneumoperitoneum on trocar site implantation of colon cancer in hamster model. Dis Col Rect 38: 1182 1188 Kazemier G, Bonjer HJ, Berends FJ, Lange FJ (1995) Port site metastases after laparoscopic colorectal surgery for cure of malignancy. Br J Surg 82: 11411142 Kuntz C, Wunsch A, Bay F, Windeler J, Glaser F, Herfarth C (1998) Prospective randomised study of stress and immune response after laparoscopic versus conventional colonic resection. Surg Endosc 12: 963967 21. McLaughlin JG, Scheeres DE, Dean RJ, Bonnell BW (1995) The adverse hemodynamic effects of laparoscopic cholecystectomy. Surg Endosc 9: 121124 22. Leighton TA, Pianim N, Liu SY, Kono M, Klein S, Bongard FS (1992) Effectors of hypercarbia during experimental pneumoperitoneum. Am Surg 58: 717721 23. Leighton TA, Liu SY, Bongard FS (1993) Comparative cardiopulmonary effects of carbon dioxide versus helium pneumoperitoneum. Surgery 113: 527531 24. McMahon AJ, Baxter JN, Murray W, Imrie CW, Kenny G, ODywer PJ (1994) Helium pneumoperitoneum for laparoscopic cholecystectomy: ventilatory and blood gas changes. Br J Surg 81: 10331036 25. Obeid F, Saba A, Fath J, Guslits B, Chung R, Sorensen V, Buck J, Horst M (1995) Increases in intra-abdominal pressure affect pulmonary compliance. Arch Surg 130: 544547 26. Rademaker BMP, Meyer DW, Bannenberg JJ, Klopper PJ, Kalkman CJ (1995) Laparoscopy without pneumoperitoneum. Surg Endosc 9: 797801 27. Reymond MA, Schneider C, Hohenberger W, Ko ckerling F (1997) Pathogenese von Impfmetastasen nach Laparoskopie. Zentralbl Chir 122: 387394 28. Safran DB, Orlando R (1994) Physiologic effects of pneumoperitoneum. Am J Surg 167: 281286 29. Schilling MK, Redaelli C, Kra henbu hl L, Signer C, Bu chler MW (1996) Splanchnic microcirculatory changes during CO2 laparoscopy. J Am Coll Surg 184: 378382 30. Scho b OM, Allen DC, Benzel E, Curet MJ, Adams MS, Baldwin NG, Largiader F, Zucker KA (1996) A comparison of the pathophysiologic effects of carbon dioxide, nitrous oxide and helium pneumoperitoneum on intracranial pressure. Am J Surg 172: 248252 31. Severin T, Mu ller B, Giese G, Uhl B, Wolf B, Hauschild S, Kreutz W (1994) pH dependent LAK cell cytotoxicity. Tumor Biol 15: 304310 32. Shuto K, Kitano S, Yoshida T, Bandoh T, Mitarai Y, Kobayashi M (1995). Hemodynamic and arterial blood gas changes during carbon dioxide and helium pneumoperitoneum in pigs. Surg Endosc 9: 1173 1178 33. Trokel MJ, Bessler M, Treat MR, Whelan RL, Nowygrod R (1994) Preservation of immune response after laparoscopy. Surg Endosc 8: 13851388 34. Turner GA (1979) Increased release of tumor cells by collagenase at acid pH: a possible mechanism for metastasis. Experientia 15: 1657 1658 35. Volz J, Ko ster S, Weis M, Schmidt R, Urbaschek R, Melchert F, Albrecht M (1996) Pathophysiologic features of a pneumoperitoneum at laparoscopy: a swine model. Am J Ostet Gynecol 174: 132140

7. 8.

9. 10. 11. 12.

13. 14. 15. 16. 17. 18.

19. 20.

Surg Endosc (2000) 14: 354357 DOI: 10.1007/s004640000154

Springer-Verlag New York Inc. 2000

More than two structures in Calots triangle

A postmortem study
R. Bergamaschi,1 D. Ignjatovic2
1 2

Department of Surgery, University of Bergen, SSSF Hospital, Forde, Norway KBC Dr Dragisa Misovic University Hospital, Belgrade, Yugoslavia

Received: 17 December 1998/Accepted: 26 March 1999

Abstract Background: Large laparoscopic cholecystectomy series often fail to report the rate at which a third structure is encountered in Calots triangle. Methods: During a 6-month period, the liver and hepatoduodenal ligament of 90 consecutive human cadavers underwent corrosion casting (n 50), postmortem arteriography (n 20), and postmortem cholangiography (n 20). Results: Third structures within Calots triangle were arteries (0.65.7 mm diameter) in 36.2% (early division of the right hepatic artery, 8.6%; caterpillar hump right hepatic artery, 12.9%; liver branch of the cystic artery, 10%; double cystic arteries, 5.7%), bile ducts (0.31.6 mm diameter) in 5.7% (small-caliber sectoral ducts, 1.4%; right posterior hepatic ducts, 4.3%), and veins (0.91.6 mm diameter) merging with the portal vein in 4% of the specimens. Conclusion: Knowledge of the aforementioned anatomy is critical to surgeons facing more than two structures within Calots triangle during laparoscopic cholecystectomy. Key words: Anatomy Complications Laparoscopic cholecystectomy

triangle are based on the surgeons interpretation of laparoscopic anatomy and on intraoperative cholangiography, respectively. The aim of this postmortem study was to provide anatomic data critical to the correct handling of more than two structures within Calots triangle during laparoscopic cholecystectomy. Materials and methods
In Belgrade, 90 consecutive fresh (<24-h-old) human autopsy specimens from patients (47 men and 43 women) with a mean age of 59 years (range 2683 years) were obtained from January to September 1997. Nine cadavers with macroscopic pathologic changes in the liver (metastatic disease [n 5], cirrhosis [n 1], echinococcus cyst [n 1], trauma [n 2] were excluded from the study. The liver and hepatoduodenal ligament were removed en bloc through a midline incision of the anterior abdominal wall. The inferior vena cava was divided cranially to the mobilized liver. The duodenum was dissected free from the retroperitoneum, and the hepatic artery proper, common bile duct, and portal vein were transected close to the head of the pancreas. The inferior vena cava then was divided caudally to the liver. Specimens then were placed into a 0.9% NaCl normotonic solution (37C). Of these specimens, 50 underwent corrosion casting, 20 postmortem arteriography, and 20 postmortem cholangiography. The hepatic artery proper was identified, and a 10-French polyethylene catheter was placed in it and secured with a suture. The artery then was irrigated with 0.9% saline solution to wash out all blood clots and to identify collateral vessels. Branches of the artery were identified and ligated to prevent leakage of cold polymerizing methyl acrylate during injection. The common bile duct was prepared in a similar manner. After placement of the 10-French catheter, irrigation was performed to remove sludge from the bile ducts, particularly the cystic duct. Ligation of bile duct branches was not performed. All blood clots were removed with forceps from the portal vein lumen, which then was thoroughly irrigated through a polyethylene catheter until clear saline solution was derived from the inferior vena cava. The identification and ligation of collateral vessels (right gastric vein, umbilical veins, and others) were performed during portal vein irrigation. Corrosion casting was carried out by injecting cold polymerizing methyl acrylate (dyed with different colors) through the catheters. During injection, specimens were immersed in water to regain their original shape. Acrylate was first injected into the hepatic artery proper and then into the portal vein. Once the solidification occurred in the arteries and veins, the

Large series of patients surgically treated with laparoscopic cholecystectomy [3, 7, 13, 14, 16] often fail to report the rate of anatomic variations encountered in Calots triangle that might have played a role in the occurrence of intra- and postoperative complications. On the other hand, case reports of late complications caused by right hepatic artery ligation [17] and unrecognized sectoral bile duct [1, 2] have appeared in the literature. Available data on anatomic variations of arteries [6, 15] and bile ducts [4, 8, 9, 19] in Calots
Correspondence to: R. Bergamaschi, National Cancer for Advanced Laparoscopic Surgery, Trondheim, Norway

355 cystic duct was identified and ligated at the gallbladder neck to prevent filling of the gallbladder, which otherwise would have broken the specimen (because of the weight of the gallbladder cast). The bile ducts then were injected. Corrosion was performed in a heated 35% potassium hydroxide solution to accelerate saponification. The casts were rinsed in water until all remnants of organic tissue disappeared, then mounted on stands. Postmortem cholangiography was performed without ligature of the cystic duct. Once the 10-French polyethylene catheter was placed and sutured to the common bile duct, irrigation was carried out. Postmortem arteriography was performed after the placement of the 10-French catheter and ligation of collateral vessels. In both procedures, barium sulfate suspension was injected through the catheter and x-rays were taken. A metallic ethalon was used during radiography for comparative measurements. Photographs and measurements of corrotion casts and x-ray films were made at the Department of Surgery, SSSF Hospital, Forde. The size and length of arteries, veins, and bile ducts were measured by a nonius scalable ruler and flexible copper wire, respectively. If the location of a structure was deep, surrounding vessels were shaved off to allow measuring. In the case of double cystic artery, the artery with the smallest diameter or an unusual course was named second cystic artery. Additional bile ducts that arose from the liver and merged with the cystic or common hepatic duct were referred to as sectoral ducts [20]. External diameters of arteries, veins, and bile ducts were assumed to correspond to a 20% increase of the inner values because of the methodology used.

A branch of the cystic artery (Fig. 4) was identified in 7 of 70 cadavers (10%). These branches ran underneath the lateral peritoneum of Calots triangle and entered the anterior segment of the right liver lobe. Their mean diameter and length were 0.8 (range 0.61.9 mm) and 11.4 (range 522 mm), respectively. Anastomotic communications with intrahepatic liver arteries were not found. Bile ducts Data on bile ducts were drawn from 50 casts and 20 cholangiographies. Additional bile ducts passing through Calots triangle were identified in 4 of 70 cadavers (5.7%) One sectoral duct arose from the anterior segment of the right liver lobe, ran through Calots triangle parallel to the common hepatic duct, and entered the cystic duct 7 mm from the junction to the common bile duct. The length and inner diameter of this duct were 37 mm and 0.3 mm, respectively. Three other sectoral bile ducts arose from the posterior segment of the right liver lobe, ran through the most cranial part of Calots triangle, and entered the common hepatic duct (at an acute angle) 6 mm, 11 mm, and 9 mm from the hepatic duct junction. Two of these ducts drained the posterior liver segment in part (length and diameter 14 mm and 12 mm; and 0.7 mm and 0.9 mm, respectively), whereas one was a right posterior hepatic duct draining the entire posterior live segment (length and diameter 16 mm and 1.6 mm, respectively) (Fig. 5). Veins Data on gallbladder veins were drawn from 50 casts. Gallbladder veins were identified in the Calots triangle of 50 cadavers. These veins were smaller than 0.3 mm in diameter, ran underneath the medial peritoneum of the caudal part of Calots triangle, and entered the peribiliary venous plexus in 48 of 50 specimens (96%). Any attempt to measure their inner diameter and length failed because of limitations of the scalable ruler and because of their plexus branching pattern, respectively. One vein and two veins were found in 37 of 48 casts (77%) and 11/48 casts (23%), respectively. A gallbladder vein larger than 0.3 mm in diameter was identified in the Calots triangle in 2 of 50 cadavers (4%). Both veins arose from the gallbladder neck, whereas their inlets were either on the trunc or on the right branch of the portal vein, extrahepatically. Both veins ran underneath the lateral peritoneum of the caudal part of Calots triangle, laterally and posteriorly to both the cystic duct and artery (Fig. 6). Their length and inner diameter were 14 mm and 18 mm; 0.9 mm and 1.6 mm, respectively. Discussion The right hepatic artery can be ligated inadvertently during laparoscopic cholecystectomy when it runs through Calots triangle quite close to the juncture of the gallbladder neck and the cystic duct. This is likely to occur in the case of caterpillar hump deformity and early division of the artery. The former has been reported to occur in 4% to 16% of

Results Arteries Data on arteries were drawn from 50 casts and 20 arteriographies. The right hepatic artery was identified in 70 cadavers (mean length 44.0 mm, range 2078 mm; mean diameter 4.0 mm, range 3.05.7 mm). The division of this artery into two branches (anterior segmental artery: mean diameter 2.7, range 1.93.4 mm; and posterior segmental artery: mean diameter 2.6, range 1.43.2 mm) was intrahepatic in 64 of 70 specimens (91.4%). The right hepatic artery gave early rise to two similar branches within the Calots triangle of 6 of 70 cadavers (8.6%) (Fig. 1). The mean length of these six right hepatic arteries was 24.1 mm (range 2031 mm). The mean diameter of the early divided anterior and posterior branches was 2.5 (range 1.83.1) mm and 2.4 (range 1.3 2.9) mm, respectively. The right hepatic artery ran parallel to the cystic duct, made a caterpillar-like loop (Fig. 2) close to the gallbladder neck, and gave origin to a short cystic artery in 9 of 70 specimens (12.9%). Cystic arteries were identified in 70 cadavers. A single artery was found in 59 of 70 specimens (84.3%), which ran through the Calots triangle in all but one case. The mean diameter and length of single cystic arteries were 1.4 mm and 19.7 mm (range 454 mm), respectively. A double cystic artery was recorded in 10 of 70 (cadavers) (14.3%). In addition to the first artery, a second cystic artery ran through the Calots triangle of 4 of 10 specimens (40%). This was not the case with the remaining 6 of 10 cadavers. Second cystic arteries originating from the right (n 2) and left (n 2) hepatic artery ran only through Calots triangle (Table 1). The mean diameter and length of second cystic arteries were 0.9 mm and 9.3 mm (range 214 mm), respectively. Five of 10 second cystic arteries were shorter than 5 mm. One of 70 specimens (1.4%) had three cystic arteries that all ran through Calots triangle (Fig. 3). This triple artery arose from the right hepatic (length 14 mm, diameter 0.9 mm), the left hepatic (length 17 mm, diameter 0.5 mm) and the hepatic artery proper (length 16 mm, diameter 0.6 mm).

356

Fig. 1. Early division of the right hepatic artery at postmortem arteriography. Fig. 2. Caterpillar-like loop of the right hepatic artery gives origin to a short cystic artery. A Cystic artery. B Caterpillar-like loop. C Cystic duct. Fig. 3. Triple cystic artery at postmortem arteriography. The arrow the three arteries passing through Calots triangle. Fig. 4. A branch of the cystic artery enters the liver. A Branch of the cystic artery. B Cystic artery. Fig. 5. Sectoral bile duct. A Posterior segmental duct. B Common bile duct. C Left hepatic duct. D Right hepatic duct. Fig. 6. Single cystic vein. A Cystic vein. B Right hepatic artery. C Common bile duct. D Right branch of portal vein.

cases [15], which is in accordance with the 13% rate in the current study. The occurrence of the latter was 9% in this study and otherwise is not mentioned in the literature. A branch of the cystic artery (running through Calots triangle to enter the liver) [12] can at times be approximately the same caliber as segmental hepatic vessels result-

ing from early division of the right hepatic artery, as shown by the current results. The outcome of ligating one of these arteries inadvertently may range from an uneventful course [5] to liver infarction, either necessitating [10] or not necessitating [17] liver resection. A double cystic artery has been reported to occur in 11%

357 Table 1. Arteries giving rise to a second cystic artery in 70 specimens n Gastroduodenal artery Left hepatic artery Right hepatic artery Segmental branch of the right hepatic artery 1 2 2 5 1. Albasini JLA, Aledo VS, Dexter SPL, Marton J, Martin IG, McMahon MJ (1995) Bile leakage following laparoscopic cholecystectomy. Surg Endosc 9: 12741278 2. Arian MC (1998) Two unusual cases of postcholecystectomy pain. Surg Endosc 12: 5759 3. Cuschieri A, Dubois F, Mouile J, Mouret P, Becker H, Buess G, Trede M, Troidl H (1991) The European experience with laparoscopic cholecystectomy. Am J Surg 161: 385387 4. Flowers JL, Zucker AK, Graham SM, Scovill WA, Imbembo AL, Bailey RW (1992) Laparoscopic cholangiography. Ann Surg 215: 209216 5. Halasaz NA (1991) Cholecystectomy and hepatic artery injuries. Arch Surg 126: 127128 6. Hugh BT, Kelly DM (1992) Laparoscopic anatomy of the cystic artery. Am J Surg 163: 593595 7. Ihasz M, Hung C, Regoly-Merei J, Fazekas T, Batrofi J, Balint A, Zaborszky A, Posfai G (1997) Complications of laparoscopic cholecystectomy in Hungary: a multicentre study of 13,833 patients. Eur J Surg 163: 267274 8. Khalili TM, Phillips EH, Berci G, Carroll BJ, Gabbay J, Hiatt JR (1997) Final score in laparoscopic cholecystectomy. Surg Endosc 11: 10971098 9. Kullman E, Borch K, Lindstrom E, Svanvik J, Anderberg B (1996) Value of routine intraoperative cholangiography in detecting aberrant bile ducts and bile duct injuries during laparoscopic cholecystectomy. Br J Surg 83: 171175 10. Madariaga RJ, Dodson SF, Seldby R, Todo S, Iwatzuki S, Starzl ET (1994) Corrective treatment and anatomic considerations for laparoscopic cholecystectomy injuries. J Am Coll Surg 179: 321325 11. Michels N (1951) The hepatic, cystic, and retroduodenal arteries and their relations to the biliary ducts. Ann Surg 133: 503524 12. Michels N (1966) Newer anatomy of the liver and its variant blood supply and collateral circulation. Am J Surg 112: 337347 13. Orlando R, Russel J, Lynch J, Mattie A (1993) Laparoscopic cholecystectomy: a statewide experience. Arch Surg 128: 494499 14. Peters JH, Krailadsiri W, Incarbone R, Bremner CG, Froes E, Ireland AP, Crookes P, Ortega AE, Anthone GA, Stain SA (1994) Reasons for conversion from laparoscopic to open cholecystectomy in an urban teaching hospital. Am J Surg 168: 555559 15. Scott-Conner CEH, Hall TJ (1992) Variant arterial anatomy in laparoscopic cholecystectomy. Am J Surg 163: 590592 16. The Southern Surgeons Club (1991) A prospective analysis of 1,518 laparoscopic cholecystectomies. N Engl J Med 324: 10731078 17. Wachsberg RH, Cho KC, Raina S (1994) Liver infarction following unrecognized right hepatic artery ligation at laparoscopic cholecystectomy. Abdom Imaging 19: 5354 18. Williams RL, Warwick R, Dyson M (1995) Grays anatomy (38th ed). Churchill Livingstone, Edinburg 19. Wright KD, Wellwood JM (1997) Bile duct injury during laparoscopic cholecystectomy without operative cholangiography. Br J Surg 85: 191194 20. Yoshida J, Chijiiwa K, Yamaguchi K, Yokohata K, Tanaka M (1996) Practical classification of the branching types of the biliary tree: an analysis of 1,094 consecutive direct cholangiograms. J Am Coll Surg 182: 3740.

References

to 25% of subjects [11], which accords with our 14% rate. However, whether both cystic arteries are running through Calots triangle is not reported [6]. The current data show that both arteries were found within the triangle in only 40% of cadavers with double arteries, which is a lower rate than a previously published 58% [11]. An additional hazard is the length of the second cystic artery, which was less than 5 mm in half of the current specimens. Data on venous drainage of the gallbladder are scarce [18]. A cystic vein the same caliber as that of additional arteries or bile ducts running through Calots triangle was noted in 4% of the current specimens. Such a vein may challenge the surgeon because it constitutes the third structure within Calots triangle in addition to the cystic duct and artery. These vessels, when inadequately electrocoagulated, may cause low-pressure bleeding leading to secondary laparotomy [3, 7] if they merge with the portal vein. Sectoral bile ducts were observed in 14% of cadavers in a previous postmortem study [11]. Dissimilarly, data from clinical studies based on intraoperative cholangiography have shown rates ranging from 0.3% to 8.4% [4, 79, 14, 19]. Our 5.7% rate of additional bile ducts passing through Calots triangle was based on both corrotion casts and postmortem cholangiography. Sectoral ducts may be the only drainage path for that particular part of hepatic tissue having no apparent interductal communication within the liver [9, 19]. They merge with either the cystic or the common hepatic duct [9, 10, 20]. However, the bottom line is not their inlet but rather their size [9, 10]. In fact, additional ducts may be either sectoral ducts draining a liver segment only in part or right posterior hepatic ducts draining the whole posterior liver segment [10]. This in turn makes it likely that ligation of small-caliber sectoral ducts or right posterior hepatic ducts will result in an uneventful outcome [2, 19] and bile leakage [10], respectively. On the other hand, inadvertent lesion of the former may result in a spontaneously recovering bile leak [1], whereas major reconstructive surgery often is required when a lesion occurs in the latter [10]. Knowledge of the aforementioned anatomy is critical to surgeons facing more than two structures within Calots triangle during laparoscopic cholecystectomy.

Technique
Surg Endosc (2000) 14: 400405 DOI: 10.1007/s004640000067 Springer-Verlag New York Inc. 2000

Laparoscopic radiofrequency ablation of primary and metastatic liver tumors

Technical considerations
A. Siperstein,1 A. Garland,2 K. Engle,2 S. Rogers,2 E. Berber,1 A. String,2 A. Foroutani,2 T. Ryan2
1 2

Department of General Surgery, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA Department of Surgery, University of California, San Francisco, Mount Zion Medical Center, 1600 Divisadero Street, San Francisco, CA 94143-1674, USA Received: 1 April 1999/Accepted: 23 July 1999

Abstract Background: Radiofrequency thermal ablation is a new technology for the local destruction of liver tumors. Since we first described laparoscopic radiofrequency ablation (LRFA) for the treatment of liver tumors, much has been learned about patient selection, laparoscopic ultrasound (LU) guided placement of the ablation catheter, monitoring of the ablation process, and patient follow-up. Methods: Since January 1996 we have performed LRFA of 250 tumors in 67 patients including 85 adenocarcinomas, 107 neuroendocrine tumors, 34 sarcomas, 1 melanoma, and 11 hepatomas. We used LU to guide placement of the ablation catheter and to monitor the ablation process. Most of the patients had two trocars (camera and laparoscopic ultrasound) with the 15-gauge ablation catheter (RITA Medical Systems, Mountain View, CA, USA) placed percutaneously. Results: The LRFA procedure was completed successfully in all patients, with 1 to 14 lesions per patient, ranging in size from 0.5 to 10 cm in diameter. The entire liver could be examined by LU via right subcostal ports. Criteria for successful ablation were 5-min ablation times at 100C with 1-min cool-down temperatures of 60 to 70C. Outgassing of dissolved nitrogen, monitored by ultrasound, was useful in confirming the zone of ablation. Intralesional color-flow Doppler, seen before ablation, was eliminated after ablation. Placement of the grounding pad closer to the lesion on the back rather than the thigh resulted in more efficient energy delivery to the tumor. Lesions larger than 3 cm in diameter required overlapping ablations to achieve a 1-cm margin of normal liver. Most patients required overnight hospitalization, with no coagulopathy or electrolyte disturbances noted.

Conclusions: The LRFA procedure is a novel, minimally invasive technique for treatment of liver tumors that have failed conventional therapy. This study documents the technical aspects of targeting lesions and performing reproducible zones of ablation. Familiarity with these techniques should lead to more widespread application. Key words: Ablation Laparoscopic Liver Radiofrequency Tumors

Presented at the annual meeting of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES), San Antonio, Texas, USA, 2427 March 1999 Correspondence to: A. Siperstein

Radiofrequency thermal ablation is becoming increasingly recognized as a new method for the local control of primary and metastatic liver tumors. Resection remains the gold standard for the treatment of patients with hepatic malignancies. Unfortunately, however, few patients are candidates for resection because of the number, location, or size of their hepatic tumors, or the presence of extrahepatic disease [1, 3]. Multiple treatment methods exist for the treatment of hepatic tumors. These may be classified broadly as systemic therapies, including conventional chemotherapy [5, 15], octreotide for selected neuroendocrine tumors [4], and newer modalities such as gene therapy or immunotherapy. For patients with liver predominant disease, a variety of regional therapies traditionally have been used including resection [2, 10], chemoembolization [18], ethanol injection [6], cryosurgery [11, 12], and more recently, radiofrequency thermal ablation [17]. Radiofrequency electrical energy refers to an alternating electric current at a frequency of 400 MHz (a frequency similar to that of an AM radio), whereas conventional household electricity is alternating current at a frequency of 60 Hz. Such high-frequency electricity is used in conventional electrosurgical or Bovie machines, which are used routinely in operating rooms [9] (Fig. 1). Conventional electrocautery units work in a monopolar fashion, with the current delivered via an electrosurgical pencil and the current returned to the unit via a grounding pad. When cautery of

401 107 neuroendocrine tumors, 34 sarcomas, 1 melanoma and 11 hepatomas (Fig. 2). The patients treated had unresectable liver tumors and had failed best conventional therapy. They had liver predominant disease, with an expected survival of more than 6 months. Laparoscopic thermal ablation was performed on all identified lesions, and did not preclude any subsequent chemotherapy or other treatment methods. Within 1 week before surgery, helical computed tomography (CT) scans of the abdomen were performed, noncontrast and with contrast in the arterial and portal venous phases, using a 7-mm slice thickness. Preoperative laboratory tests included a complete blood count, electrolytes, liver function tests, coagulation studies, and specific tumor markers as indicated. Postoperatively, patients were managed in a manner similar to that of patients undergoing laparoscopic cholecystectomy, with the exception that the preoperative laboratory studies were repeated on postoperative day 1. Patient follow-up studies included CT scan and laboratory studies at 1 week and every 3 months postoperatively. The CT scans were analyzed for local tumor control and the development of new intra- or extrahepatic disease. All ablations were performed with an RITA Medical Systems Model 30 thermal ablation catheter (Fig. 3). The catheter consists of a 15-gauge needle 25 cm in length. Once the catheter tip is positioned at the center of the intended ablation zone, four curved prongs are deployed that deliver radiofrequency electrical energy to the tissues and affix the catheter in place. Thermocouples mounted in the tips of the deployed prongs are used to monitor tissue temperature as the ablation proceeds, and these readings also are used to regulate the amount of energy delivered via the catheter.

Fig. 1. Principles of radiofrequency energy delivery. a Convention operating room electrical surgical units rely on radiofrequency electrical current to cause rapid tissue heating over a focal area. b Placement of a conventional electrosurgical tip into tissue would result in the formation of an elliptical zone of ablation surrounding the exposed metal. Where current density is highest immediately adjacent to the handpiece, resistive heating in the tissue occurs, resulting in local tissue necrosis. c The thermal ablation catheter uses a deployable array of prongs to deliver electrical energy and thus heating over a 3.5- to 4-cm diameter sphere of tissue.

small blood vessels is performed in the operating room, the tip of the electrical pencil is applied lightly to the tissue surface. This results in a very high local current density, with rapid burning and charring of the tissues. If the tip of the electrosurgical pencil were to be plunged into the tissue, there would be no immediate charring. However, current would continue to flow between the handpiece and the grounding pad. The high-frequency electrical energy causes the electrons in the tissue to vibrate in a back-and-forth manner. Where the current density is highest, near the tip of the electrosurgical pencil resistive heating is caused in the tissues. When the tissue temperature rises above 45 to 50C, there is protein coagulation necrosis with resultant cell death. The zone of ablation appears as an elliptical area immediately surrounding the uninsulated portion of the electrosurgical handpiece [7, 8]. The thermal ablation catheter used in this study deploys four curved prongs into the tissue to deliver the electrical energy to a spherical volume of tissue 3.5 to 4 cm in diameter. The metal of the ablation catheter itself is only 20 to 30C hotter than the surrounding tissues. Unlike cryosurgery, heat is not delivered by thermal conduction. Instead, the thermal ablation catheter acts more as an antenna to deliver the electrical energy to the surrounding tissues. Radiofrequency thermal ablation has been applied in open surgical procedures as well as percutaneously in selected patients [13, 14]. In 1997, we described studies in the porcine model that led to the development of techniques for performing laparoscopic ultrasound guided radiofrequency thermal ablation in humans [17]. The purpose of this report is to describe in detail the technical considerations for performing this procedure in a safe and reproducible manner. Materials and methods
Since January 1996, we have performed laparoscopic radiofrequency thermal ablation of 250 tumors in 67 patients, including 85 adenocarcinomas,

Results Of the 67 patients treated, laparoscopic radiofrequency thermal ablation was completed successfully in all patient. Between 1 and 14 lesions ranging in diameter from 0.5 to 10 cm were treated per patient. The maximum volume of tumor treated per patient was determined more by practical consideration of the length of the surgery required to perform multiple ablations than by a toxicity limit. The cooling effect of blood flow from adjacent arterial or venous structures prevented vessels larger than approximately 3 mm in diameter from thrombosing. Thermal energy has the potential to cause damage and stricture to the larger bile ducts, so tumors immediately adjacent to the bifurcation of the common duct were not deemed safe to treat. All the patients were treated under general endotracheal anesthesia. Most were positioned in the supine position. However, selected patients with disease limited to the right lobe of the liver, especially if there was a predominance of disease in the posterior segments, were treated in the left decubitis position with the table slightly flexed. Most of the patients could be treated with the placement of two right subcostal ports (Fig. 4). The abdominal cavity was entered in all patients using an Ethicon Optivue trocar. Most of the patients were treated using a 12- and 11-mm trocar. However, selected patients were treated with a 12and 5-mm port. To mobilize the liver adequately in patients with extensive intra-abdominal adhesions, additional trocars occasionally were necessary. Placement of an umbilical trocar was not useful because the camera or the laparoscopic ultrasound transducer could not reach the dome of the liver from such a location. The anterior and lateral aspects of the liver were mobilized. It was not deemed necessary to take down the falciform ligament or triangular ligaments. Viscera within 2 cm of an intended ablation zone required mobilization from the liver if they were adherent because of prior surgical proce-

402

Fig. 2. Number of patients and tumors treated by laparoscopic radiofrequency thermal ablation. A variety of tumor histologies have been successfully treated. Patients with neuroendocrine metastases tend to have more numerous metastases. Fig. 3. End of the thermal ablation catheter with the prongs deployed. Small thermocouples embedded near the tips of the prongs allow monitoring of temperature during and after the ablation process.

Fig. 4. Intraoperative photograph demonstrating placement of subcostal trocars for the laparoscope and laparoscopic ultrasound transducer. The radiofrequency ablation catheter is placed through a separate percutaneous puncture.

dures. Similarly, if lesions encroached on the gall bladder fossa, laparoscopic cholecystectomy was warranted to avoid thermal ablation of the wall of the gall bladder with delayed bile leakage. Once the liver was adequately mobilized, laparoscopic ultrasound examination of the entire liver parenchyma was performed using a 7.5-MHz rigid ultrasound transducer with an Aloka 2000 ultrasound machine. Interestingly, in all patients, an ultrasound of the entire liver parenchyma from a right subcostal port was found possible. The portal and hepatic vein branches could be traced easily to assign a liver segment to each of the lesions treated. Occasional patients required the use of a flexible ultrasound transducer to image the most cephalad portions of segments 4, 7, and 8. To coordinate the movement of the laparoscopic ultrasound transducer with the laparoscopic image, it was found most convenient to use a picture in picture box to superimpose a quarter-size laparoscopic image over the full-size ultrasound image. This made it much easier to coordinate the movement of the instruments and to document the procedure on a single videotape. Adhesions permitting, the re-

mainder of the abdominal cavity was examined by both laparoscope and ultrasound to provide maximal tumor staging information. Once all the hepatic lesions were identified and their size measured, color-flow Doppler was performed to assess tumor vascularity. With some tumors, discreet feeding vessels could be identified and ablated first to facilitate subsequent ablations. Under ultrasound guidance, 18-gauge core biopsies were performed, using a freehand technique, with a spring loaded biopsy gun. Tissue samples were taken of representative, but not all, tumors in a given patient. The placement of the needle under ultrasound guidance was found to be one of the most technically challenging aspects of the procedure. Unlike percutaneous and open surgical procedures, biopsy guides are not currently commercially available. The biopsy was performed, first using laparoscopic ultrasound to find the shortest path from the tumor to the liver surface. The biopsy needle was then placed into the abdominal cavity through a percutaneous approach and did not require the placement of an additional trocar. The needle was placed so that it would parallel as much as possible the plane of the ultrasound so the entire path of the needle could be seen on the ultrasound image as it traversed the liver tissue. The tip of the biopsy needle then was positioned at the periphery of the tumor and the trigger fired. Care was taken to study the 2-cm throw length of the biopsy device before firing to ensure that major hepatic vasculature or extrahepatic structures were not injured. Performing the biopsy with a smaller 18-gauge needle often was useful to establish the optimum angle of approach for subsequent placement of the larger thermal ablation catheter. The ablation catheter then was passed using an identical technique. For tumors smaller than 3 cm in diameter, a single cycle of ablation could be performed, achieving an adequate tissue margin. Great care was required to position the tip of the ablation catheter precisely in the geometric center of the tumor so that the ablated zone completely encompassed the tumor. For tumors larger than 3 cm in diameter, at least six overlapping thermal ablation zones were required to ensure an adequate volume of ablation.

403

In patients with multiple lesions, it often is useful to use a second catheter to target a remote lesion during the ablation of a prior lesion. Care must be taken, however, to not place two catheters in close proximity because if the prongs should touch they would function as a much larger array with unpredictable results. The RITA Model 500 generator was run in a temperature-controlled mode with an average target temperature of 105C and a maximum power of 50 W. With activation of the device, power would be increased automatically to 50 W until an average thermocouple temperature of 105C was obtained. Using the thermocouple temperature, the power applied would be regulated automatically to maintain an average thermocouple temperature of 105C. This would be maintained for a period of 5 min, with the overall time of an ablation cycle taking 7 to 10 min, depending on the time required to achieve target temperatures. Once the delivery of power was stopped, the monitoring of thermocouple temperatures was continued. These were seen to drop rapidly over the first 10 to 20 s as the temperature of the metal prongs decreased by 20 to 30C to equilibrate with the surrounding tissue. The temperatures then decreased at a slower rate as heat was dissipated from the zone of ablation (Fig. 5). Thermocouple temperatures in the 60 to 70C range 1 min after the ablation process had ceased indicated that a successful ablation had been performed. If there were any questions as to the technical adequacy of the ablation, the prongs could be pulled back into the catheter, the catheter rotated 45, and the prongs then redeployed into what had previously been determined in animal studies to be the coolest point of the ablated zone. If thermocouple temperatures at this location were above 60C, this was interpreted as an adequate ablation. If not, an additional cycle would be performed with the catheter in this position. Occasional tumors were highly vascular, and despite applying 50 W of power, target temperatures could not be obtained. If a plateauing of the temperatures was observed at 1 to 2 min, the prongs could be withdrawn to a diameter of approximately 2 cm. This allowed power to be delivered to a much smaller volume of tissue, resulting in vascular thrombosis and ablation of this volume. The prongs could then be extended to their full diameter of 3 cm and the ablation cycle repeated. It was found that the position of the grounding pad was a variable crucial to the efficient delivery of power. With the grounding pad placed in the conventional location, over the thigh or buttocks, the circuit impedance was seen to measure approximately 75 ohms. With the grounding pad located over the patients back, the impedance dropped to 45. This resulted in less power being dissipated to the extrahepatic tissues, so that a much higher proportion of the energy was delivered to the tumor itself. As a result target temperatures were achieved 1 to 2 min, eliminating the need to use the techniques of partial catheter deployment or catheter rotation to achieve successful ablations. This phenomenon had been identified in the porcine model, and several patients were evaluated with grounding pads placed on both the hip and back. To ensure that this phenomenon was not the result of individual patient or tumor characteristics, ablation could be started using the hip

grounding pad with an impedance of 75 ohms and tissue temperatures plateauing in a suboptimal range. By simply switching the grounding pad to the back location, the impedance was seen to drop and thermocouple temperatures to rise rapidly (Fig. 6). Because of a concern that this unconventional grounding pad placement might result in grounding pad thermal injuries, a thermocouple was sandwiched between the grounding pad and the patients skin. Despite multiple repeated ablations, thermocouple pad temperatures did not rise above 40C, and no grounding pad thermal injuries were observed. Assessment of the ablation process in the operating room was performed in three ways. The first and most crucial was monitoring of thermocouple temperatures during and after the ablation process. The second was the observed phenomenon of dissolved nitrogen outgassing into the heating tissues. As the tissues are heated, the solubility of dissolved nitrogen decreases, resulting in microbubble formation in the tissue. This appears as an echogenic blush, seen in most but not all tumors, that enlarges to encompass the zone of ablation (Fig. 7). In the tumors that demonstrated color-flow Doppler before ablation, this was measured after ablation to make sure that no blood flow was present in the ablated zones. For lesions that required multiple overlapping ablations, the zones farthest from the ultrasound transducer were ablated first so that the outgassing of dissolved nitrogen did not obscure needle placement for subsequent ablations. After approximately 10 min, the dissolved nitrogen would be resorbed and the normal echogenicity of the tumors would reappear. After ablation, there usually was no dramatic change in the echogenicity of the tumor or normal liver. Some patients demonstrated a slight hypoechoic characteristic in the ablated areas, and needle tracks or small amounts of gas often could be observed in the ablated zones. On withdrawal of the needle, bleeding from the needle track was rarely a problem. In some patients, particularly those with cirrhosis, application of 20 to 30 W of power as the needle was withdrawn through the tissues would serve to coagulate the needle track and minimize bleeding. Postoperatively, patients required only routine care. Most patients were managed without narcotics, and the intravenous tubes were removed once the patients resumed oral intake. Most patients were discharged on the first postoperative day after an average stay of 1.3 days. Discussion Through experience gained in the treatment of 250 hepatic tumors with laparoscopic radiofrequency thermal ablation, we have gained an increased understanding of the optimal technical considerations in the performance of this procedure. The procedure can safely be performed laparoscopically with reproducible results and, as demonstrated previously, a local recurrence rate of 12% [16]. Radiofrequency thermal ablation has been performed percutaneously and by open and laparoscopic surgical techniques. Although the percutaneous approach has the advantage of not requiring a surgical procedure, there are limitations. Many hepatic tumors, because of their size or location, are not well imaged by transabdominal ultrasound, which is the usual means for tumor location and targeting in

404

Fig. 5. Temperature profiles during the ablation process. a Typical ablation profile with an increase in thermocouple temperature over 1 to 2 min to average 105C. This is held for 5 min to complete the ablation cycle. Once power is turned off, temperature in the thermocouples cools rapidly to equilibrate with the surrounding tissues. Tissue temperatures of 60 to 70C 1 min after tablation indicate that lethal tissue temperatures have been achieved. b Temperature profile with one electrode adjacent to a blood vessel. This thermocouple is at a lower temperature because of the cooling effect of the adjacent blood flow. c Temperature curve for an overlapping ablation. Two of the thermocouples in previously ablated tissue are seen to achieve temperatures more rapidly than the adjacent thermocouples in unablated tissue. Fig. 6. Effect of changing grounding pad position on tissue impedance and temperature. With movement of the grounding pad from the hip to the back position, impedance is seen to decrease, with a resultant improvement in the efficiency of energy delivery and temperature increases. Fig. 7. A working view of the operating room monitors during an ablation. A picture in picture box is used to superimpose an image seen through the laparoscope over the ultrasound image. a Deployment of the ablation catheter in the tumor. b Outgassing of dissolved nitrogen during the ablation process encompasses the zone of ablation.

such procedures. The treatment of tumors on the liver surface has the potential to ablate the adjacent chest wall or, worse, adjacent viscera, with the potential for major postprocedure complications. Initial reports also have suggested that the local recurrence rate for percutaneous procedures may be higher than that seen with surgical techniques, pos-

sibly because of greater difficulty in targeting the lesions precisely. An open surgical approach carries with it the morbidity and recovery from a laparotomy. Most patients who are candidates for radiofrequency thermal ablation have been through prior laparotomies, as well as other therapies, and

405

have fairly advanced disease. In this group of patients, quality of life is an important consideration, and so many of them are reluctant to undergo laparotomy. We have had occasion to treat several patients who required laparotomy for simultaneous procedures. The technique actually is somewhat more cumbersome to perform by laparotomy rather than laparoscopy because the working space between the liver and the surrounding structures seems more confined. The laparoscopic approach offers the advantage of a minimally invasive procedure with ability to perform intraoperative high-frequency ultrasound examination of the liver for better tumor detection and more accurate targeting. We also know that because of the pneumoperitoneum and upward movement of the diaphragm, there is in fact minimal movement of the liver during a laparoscopic procedure, further facilitating needle placement. Movement of the grounding pad from the thigh to the back was seen to have a significant effect in lowering impedance of the circuit and markedly improving the efficiency of energy delivery to the target tissue. The bodily tissues serve as a resistor, so that as the path length between the ablation catheter and the grounding pad increases, so does the overall resistance. Because energy is dissipated where there is resistance in the circuit, as the path length increases, more energy is lost to the surrounding bodily tissues and less is delivered to the tumor itself. Stated in another way, tissue heating is related to current density. At a given power, as impedance decreases, current flow increases according to the following formula: power current2 impedance. Radiofrequency thermal ablation is a new technology for the local destruction of hepatic tumors. Currently, The technology is limited in that only 3.5 to 4 cm diameter volumes of tissue may be ablated at a single cycle. This makes it tedious to ablate larger tumors, and overlapping ablations, if not performed with great precision, may lead to local treatment failures. Despite these limitations, firstgeneration thermal ablation technology has been able to safely and reliably ablate considerable volumes of tissue using minimally invasive approaches. With time, it is expected that larger catheter rays may be designed, with other advances to shorten the ablation time. This technology may allow for palliation of symptoms or prolongation of life in patients with tumors refractory to treatment by conventional means.

References
1. Bengmark S, Hafstrom L (1969) The natural history of primary and secondary malignant tumors of the liver. Cancer 23: 198202 2. Butler J, Attiyeh FF, Daly JM (1986) Hepatic resection for metastases of the colon and rectum. Surg Gynecol Obstet 162: 109113 3. Cady B, Stone MD (1991) The role of surgical resection of liver metastases in colorectal carcinoma. Semin Oncol 18: 399406 4. Di Bartolomeo M, Bajetta E, Buzzoni R, Mariani L, Carnaghi C, Somma L, Zilembo N, di Leo A (1996) Clinical efficacy of octreotide in the treatment of metastatic neuroendocrine tumors: a study by the Italian trials in medical oncology group. Cancer 77: 402408 5. Faiss S, Schereubl H, Rieken E, Wiedenmann B (1996) Drug therapy in metastatic endocrine tumors of the gastrointeropancreatic system. Recent Results Cancer Res 142: 193207 6. Livraghi T, Bolondi L, Lazzaronis P, Marin G, Morabito A, Rapaccini G, Salmi A, Torzilli G (1992) Percutaneous ethanol injection in the treatment of hepatocellular carcinoma in cirrhosis: a study on 207 patients. Cancer 69: 925929 7. Lounsberry W, Goldschmidt V, Linke C (1961) The early histologic changes following electrocoagulation. J Urol 86: 321329 8. McGahan J, Brock J, Tesluk H (1992) Hepatic ablation with use of radio frequency electrocautery in the animal model. J Vasc Interv Radiol 3: 291297 9. Organ L (1976) Electrophysiologic principles of radiofrequency lesion making. Appl Neurophysiol 39: 6976 10. Que FG, Nagorney DM, Batts KP, Linz LJ, Kvols LK (1995) Hepatic resection for metastatic neuroendocrine carcinomas. Am J Surg 169: 3643 11. Ravikumar TS (1996) Interstitial therapies for liver tumors: management options in primary and secondary liver cancer. Surg Oncol Clin N Am 5: 365376 12. Ravikumar TS, Kane R, Cady B, Jenkins R, Clouse M, Steele G Jr (1991) A 5-year study of cryosurgery in the treatment of liver tumors. Arch Surg 126: 15201524 13. Rossi S, Di Stasi MD, Buscarini E (1996) Percutaneous RF interstitial thermal ablation in the treatment of hepatic cancer. Am J Radiol 167: 759768 14. Rossi S, Di Stasi M, Buscarini E, Cavanna L, Quaretti P, Squassante E, Garbagnati F, Buscarini L (1995) Percutaneous radiofrequency interstitial thermal ablation in the treatment of small hepatocellular carcinima. Cancer J 1: 7381 15. Silverberg E, Boring CC, Squires TS (1990) Cancer statistics 1990 Cancer Clin 40: 926 16. Siperstein AE, Garland A, Engle K, Rogers SJ, Berber E, Foroutani A, String A, Ryan T, Ituarte P (2000) Local recurrence after laparoscopic radiofrequency thermal ablation of hepatic tumors. Annals Surg Onc 7: 106113 17. Siperstein AE, Rogers SJ, Hansen PD, Gitomirsky A (1997) Laparoscopic thermal ablation of hepatic neuroendocrine tumor metastases. Surgery 122: 11471155 18. Tanaka K, Nakamura S, Numarata K, Okazaki H, Endo O, Inoue S, Takamura Y, Sugiyama M, Ohaki Y (1992) Hepatocellular carcinoma: treatment with percutaneous ethanol injection and transcatheter arterial embolization. Radiology 185: 457460

Review article
Surg Endosc (2000) 14: 318325 DOI: 10.1007/s004640020012 Springer-Verlag New York Inc. 2000

Indications for endo-organ gastric excision

S. K. Mittal, C. J. Filipi
Department of Surgery, Creighton University School of Medicine, 601 North 30th Street, Omaha, NE 68131-2197, USA Received: 7 April 1999/Accepted: 18 August 1999

Abstract. Intragastric surgery for benign and malignant conditions is a new form of minimally invasive surgery, to which the term endo-organ gastric surgery has been applied. This procedure may provide improved results for patients, but reported studies are small, and follow-up evaluation is limited. The indications for endo-organ surgery are evolving as technology and operative expertise begin to meet the need for continued advancements in miniaturized surgery. This new approach is applied primarily to the removal of gastric neoplasms poorly positioned or too large for standard transoral endoscopic excision. Gastric polyps, benign gastric wall tumors such as leiomyomas and carcinoids, and low-grade as well as high-grade malignancies can be removed. The history of endo-organ surgery, the background technology, and surgical experience are reviewed. In addition, current indications for endo-organ surgery and the rationale for algorithms are included. Intraluminal gastric surgery is not widely performed or studied, therefore a further understanding of its role is provided. Key words: Endo-organ Intragastric Minimally invasive Surgery

Introduction Since the introduction of laparoscopic cholecystectomy in 1987 by Phillipe Mouret, there has been a progression of minimally invasive surgery, beginning with laparoscopic Nissen fundoplication progressing to laparoscopic adrenalectomy, colectomy, splenectomy, and laparoscopic herniorrhaphy. Currently, thoracoscopic, endovascular, laparoscopic retroperitoneal, and mini-invasive neck and breast procedures are under investigation. Minimally invasive surgery has significantly decreased operative time, blood loss, postoperative discomfort, and length of hospital stay. A variety of access devices, including a modified percutaneous endoscopic gastrostomy (PEG), are now available, allowing di-

rect intraluminal gastric access for laparoscopic instruments, which is called endo-organ surgery. The widespread use of endoscopy has led to the earlier and more frequent diagnosis of gastric polyps. These lesions carry a definite risk for focal carcinoma. The risk varies with size and histology of the polyp. Up to 45% of gastric wall tumors may be malignant, and improved survival has been shown with early excision. Hence, it is imperative that a definite tissue diagnosis be made and appropriate treatment instituted. Endoscopic excision usually is limited to small lesions (<2 cm) located in the antrum and the body of the stomach. Laparoscopic gastric wedge resections are performed most easily on anterior lesions. Intraluminal surgery allows partial- and full-thickness excision near the gastroesophageal junction (GEJ), pylorus, and posterior body. Intraluminal excision is suited especially for complete (well-defined lateral margins) removal of lesions adjacent to the pylorus and gastroesophageal junction. Endo-organ surgery may be limited for lesions located on the greater curvature and incisura because the operative port-to-lesion distance is reduced. In this circumstance, laparoscopically assisted placement may provide the required distance from the operative port to the lesion. With further experience, a larger number of excisional procedures will be performed using the intraluminal approach. In this article, we outline the current indications for endo-organ excisional therapy.

Historical background Sedillot [50] performed the first gastrostomy in 1849. Later, Bilroth made landmark contributions that form the basis of modern-day gastric surgery. With the advent of endoscopy, direct visualization of the lesions became possible [22, 25]. Subsequently, technical advancements increased the use and indications for endoscopy. However, the procedure still is limited in its therapeutic capabilities. Maneuverability can be difficult, and endoscopes are unable to cut and excise large pieces of tissue with precision. Endoscopic excision currently requires electrocautery or photocoagulation, which distorts the cells at the periphery, making lateral and

Correspondence to: C. J. Filipi

319 Table 1. Incidence of focal cancer in gastric polyps based on histology Hyperplastic polyps Cancer/ polyps Orlowska [42] Daibo [13] Laxen [30] Kamiya [26] Tomasulo [54] Ming [34] 10/483 10/477 2/123 2/93 4/74 0/76 28/1326 Focal cancer (%) 2.1 2.1 1.6 2.1 5.4 0.0 2.1 Adenomatous polyps Cancer/ polyps 6/60 6/37 10/29 9/85 5/23 4/10 40/244 Focal cancer (%) 10.0 16.2 34.5 10.6 21.7 40.0 16.4

Fig. 1. The evolution of endo-organ surgery.

deep margin assessment difficult. In addition, sufficient hemostatic technology is lacking for active upper gastrointestinal bleeding, especially for posterior wall duodenal ulcers. In 1981 Ponsky and Gauderer [17] first described PEG for stomach decompression and feeding in the pediatric age group. Since then, it has become a common procedure in adults as well as children. It soon was realized that modified PEGs also could provide minimally invasive surgical access to the stomach. Lesions of the stomach traditionally have required gastrectomy, involving significant morbidity related to discomfort and associated hospitalization. In this era of minimally invasive surgery, newer approaches to the treatment of stomach diseases have developed. Ohashi [42] presented a study of early gastric cancer using a technique described as laparoscopic intragastric surgery, a method of access into the stomach for laparoscopic instruments using conventional laparoscopic guidance. Since then, others have reported intragastric surgery for bleeding ulcers [30, 48], gastric leiomyoma excision [10, 53], and bezoars [14]. At Creighton University, Filipi [34] developed a safe method of access for intraluminal surgical instruments using modified large-bore PEGs. To distinguish these techniques from laparoscopic intraluminal procedures, we propose the use of the term endo-organ surgery. These advancements have the potential to decrease significantly operative time, intraoperative blood loss, postoperative discomfort, length of stay, and cost of treatment. The evolution of gastric surgery is depicted in Fig. 1. Modified PEGs and better laparoscopic instruments have increased the indications for intraluminal gastric surgery for both laparoscopic intragastric and endoorgan surgery.

foys lesion as a separate entity for consideration because it occasionally may require excision for hemostatic control. Many of the mentioned lesions overlap with gastric carcinoma. However, these lesions are considered separately. With each, we have included indications for endo-organ surgery over endoscopy and laparoscopy, with respect to size and location of the lesion.

Gastric polyps Gastric polyps are mucosal growths that protrude into the gastric lumen. With the widespread use of endoscopy, excision now is being performed for symptoms directly attributable to polyps, and for lesions larger than 1 cm incidentally found during routine examination. The overall incidence of gastric polyps now is estimated to be at 0.4% to 0.7%(40). Gastric polyps are of two types: hyperplastic and adenomatous polyps. Contrary to their name, epithelial polyps involve a definite risk of having malignant foci. Hyperplastic polyps have a significantly lower incidence of focal carcinoma than adenomas [12, 27, 31, 35, 43, 55]. Orlowska [43] showed a dramatic increase in the incidence of carcinoma with the size of polyps. These results are briefly summarized in Tables 1 and 2. The possibility of malignancy requires that each polyp be properly evaluated. It also is important to distinguish polyps from other polypoid lesions that may represent gastric wall tumors. Hupper and Prustly [26] as well as others [9, 13, 20, 36] have shown a survival advantage for patients if these polyps are removed. Hyperplastic polyps constitute approximately 75% of all gastric epithelial polyps. These may be solitary or multiple, sessile or pedunculated. Hyperplastic polyps usually are seen in association with atrophic gastritis [55] caused by a hypergastrinemic state. Increased incidence also has been reported over the past 10 to 15 years, resulting from widespread use of acid suppression therapy [18, 49]. These, for the most part, are hyperplastic polyps or entrochromaffin cell hyperplasia. It has been hypothesized that these may harbor malignancy, but no evidence of this effect has been reported. They occur as a result of regenerative glandular proliferation. Histologically, elongated, tortuous, and dilated gastric glands characterize hyperplastic polys. The risk of malignancy has been reported as 0% to 5.4% (mean, 2.1%; Table 1).

Current status of endo-organ excision Lesions for excision need to be divided on the basis of histology, size, and location. For ease of description, we have categorized the indications according to histology. It is evident that polypoid lesions of the stomach are better classified as epithelial polyps and gastric wall tumors. Polyps with their varied clinical appearances and potential for malignancy are considered as a separate group. Gastric wall tumors are further subclassified on the basis of their histology (i.e., lymphoma, leiomyoma/sarcoma, carcinoid, lipoma, and pancreatic rest). We also have included Dieula-

320 Table 2. Incidence of focal cancer in gastric polyps based on size Hyperplastic polyps Size Smaller than 5 mm 610 mm Larger than 11 Orlowska [42] Cancer/ polyp 1/277 3/148 6/48 Focal cancer (%) 0.4 2.0 12.5 Adenomatous polyps Cancer/ polyp 0/33 1/15 5/6 Focal cancer (%) 0.0 6.7 83.3

Adenomatous polyps constitute most of the remaining 25% of gastric epithelial polyps. More likely to be solitary, these usually are located near the antrum. Histologically, they have atypical glands and pseudostratified epithelium with varying degrees of nuclear atypia. Adenomatous polyps can be tubular or villous. The risk of malignant focus (Table 2) in adenomatous polyps is significantly higher (10 40%; mean, 16.4%) than in hyperplastic polyps. Diagnosis. Gastric polyps are diagnosed by endoscopic biopsy. However, 50% to 75% of the biopsy diagnoses are modified after histologic examination of the specimen [41, 51]. This creates confusion as to what constitutes an appropriate workup for benign epithelial gastric polyps. Treatment. Currently, treatment includes endoscopic biopsy and polypectomy, which is followed by open surgical resection, if indicated. Endoscopic polypectomy is a safe procedure. However, because of technological limitations, it is not possible to excise lesions in the fundus, cardia, or posterior body. It also is difficult to excise lesions larger than 2 cm because the risk of bleeding or gastric wall perforation is increased. In addition, there is concern as to the validity of clear margins for polypectomy specimens because they are excised using either electrocautery or laser, both destructive energy sources. King and Jonathan [29], of the Mayo Clinic, recommend that lesions larger than 2 cm as well as symptomatic polyps, causing bleeding, pain, or obstruction, and those containing a focal malignancy must be resected widely. Endo-organ surgery allows excision of polyps without cauterization of tissue. It also provides easy access to lesions in the fundus, prepylorus, and posterior body. Wide full-thickness resection of up to 6 cm of stomach can be performed for large polyps, polyps with focal carcinoma, or confluent polyps, thus avoiding the need for an open procedure. We present an algorithm for gastric polyp treatment in Figs. 2 and 3. Gastric wall tumors The exact incidence of gastric wall tumors is unknown because most are asymptomatic. However, Yamagiwa et al. [58] found a 3.1% (173/5,451) incidence of submucosal tumors in resected stomach specimens. Widespread use of endoscopy has led to increased detection of many asymptomatic and symptomatic gastric wall lesions. Bandoh and Isoyama [4], in a prospective study, showed that nearly 45% of these lesions are malignant (leiomyosarcoma, 23%; nonHodgkins lymphoma, 22%), and that surgery alone for

Fig. 2. Therapeutic approach to hyperplastic gastric polyps.

Fig. 3. Therapeutic approach to adenomatous gastric polyps.

321

low-grade lesions and surgery in conjunction with chemotherapy for high-grade lesions provided a significant survival advantage. With this in mind, they concluded that it is not possible to follow gastric wall tumors without surgical intervention, which until now meant laparotomy for either enucleation or wide-margin excision. For some lesions, especially those located in the anterior gastric wall, this is possible by using laparoscopy. Endoscopy can only identify the presence of the gastric wall lesion and take a superficial biopsy. A deeper biopsy usually is required for histologic diagnosis. Laparoscopic excision is best suited for anterior and posterior (more technically demanding) body lesions. Open excisional procedures for neoplasms near the fundus and pylorus are extensive. Endo-organ surgery is possible for prepyloric, posterior wall, and fundal lesions. Intraluminal surgery allows enucleation, mucosal resection, and full-thickness gastric wall excision of lesions up to 6 cm in diameter. Gastric carcinoids were first reported by Askenazy in 1923. Until recently, it was common surgical thinking that gastric carcinoids comprised 1% of all gastric cancers and 2% to 4% of all gastrointestinal carcinoids. However, it is now evident that gastric carcinoids are far more prevalent in the stomach, up to 10% to 30% [49]. This finding is the result of increased endoscopy and better immunohistochemical staining of biopsy specimens. It is widely accepted that enterochromaffin cell hyperplasia is found in hypergastrinemic conditions. Proton pump inhibitors have been shown to increase the incidence of gastric carcinoids in laboratory animals [7, 19]. Concerns, although unproved, remain regarding a similar increase in humans. Gastric carcinoids compose a heterogeneous group of neoplasms, each with a different natural history [1, 18, 49]. They are classified on the basis of their association with hypergastrinemia, multiple endocrine neoplasia (MEN syndrome), or neuroendocrine tumors, or as sporadic occurrences. Characteristically, on endoscopic examination, carcinoids appear smooth, round, and submucosal. They may have an erythematous depression or ulceration [40]. Biopsy is required for diagnosis. The clinician also must assess the remainder of the stomach for additional lesions and evaluate for stigmata of hypergastrinemic state such as atrophic gastritis or mucosal hyperplasia. Previously, treatment of a gastric carcinoid meant wide surgical resection with lymph node dissection. It is now evident that the tumors associated with hypergastrinemia behave in a less malignant manner, and that a more conservative approach is acceptable. Hypergastrinemia is the stimulus for most carcinoids, so antrectomy plays an important role in the management of multiple or recurrent carcinoids. Gastric carcinoids associated with hypergastrinemic states tend to have a benign nature, and lesions as large as 2 cm can be excised safely by endo-organ surgery. Larger lesions need wider excision if found to be malignant. Lesions occurring either sporadically or with multiple endocrine neoplasia have a higher likelihood of being malignant, so endo-organ surgery is limited to palliative resection of symptomatic unresectable lesions [1]. Gastric lymphoma comprises 5% of all gastric malignancies. Their presentation generally is heralded by vague symptoms of upper gastrointestinal discomfort, melena, and

dyspepsia. However, they may be completely asymptomatic or present with massive hemorrhage. Primary gastric lymphoma needs to be differentiated from lymphoma of other organs with the stomach being secondarily involved. Diagnosis generally is established by endoscopic biopsy. Wide surgical resection is the treatment of choice for stages I and II gastric lymphoma. Currently, it is proposed that chemotherapy, radiotherapy, or both also achieve comparable results, but bleeding and perforation leading to acute surgical emergency can occur if chemoradiation therapy is given without resection [19, 32]. Excision of gastric lymphomas before chemotherapy and radiotherapy decreases the incidence of gastric perforation during treatment [16]. Excision of small lesions by endo-organ surgery may avert these complications. Moreover, symptomatic advanced lesions with bleeding or pain can possibly be palliated by local excision or vessel ligation using the endo-organ approach. It is estimated that 10% of biopsy specimens diagnosed as lymphoma actually are pseudolymphoma that may not be distinguishable by the specimen alone. Pseudolymphoma has only mucosal spread without any lymph node disease. Wide local excision with endo-organ surgery not only provides complete tissue for diagnosis, but also prevents transformation to lymphoma. Gastric leiomyogenic tumors include most gastric wall neoplasms. Comprising only 2% of excised gastric wall tumors, they usually are smaller than cm and asymptomatic [6]. Benign forms such as leiomyoma must be differentiated from their malignant counterparts, leiomyosarcoma and leiomyoblastoma. Bandoh [36] reported that 47% (20/43) of leiomyogenic tumors were found to be malignant after excision. A similar incidence of malignancy (30%) was reported by Yamagiva [58]. Endoscopic ultrasound has been used preoperatively with great reliability to determine depth, size, and histopathology of the lesion [5, 59]. This technique is highly user dependent and has not gained wide acceptance. Distinction between malignant and benign forms requires demonstration of invasion. Degree of differentiation, rate of mitosis, and grade are good but not absolute indicators of malignant potential. Most leiomyogenic tumors are asymptomatic. They usually are diagnosed during endoscopy/laparotomy/ laparoscopy for other indications. When symptomatic, they commonly present with upper gastrointestinal hemorrhage or nonspecific upper abdominal pain. Occasionally, they may grow large enough to be manually palpable or cause pyloric obstruction. Gastric wall tumors require histologic evaluation to exclude malignancy. Endoscopic biopsy fails to get submucosal tissue and therefore is of limited value. Laproscopic biopsy/excision is most applicable for gastric body lesions. Endo-organ excision plays a pivotal role in the diagnosis of these tumors, allowing complete excision of small (2- to 3-cm) tumors. Symptomatic small benign or tumors with low-grade malignancy are treated adequately by a 1-cm lateral-margin endo-organ excision. An open procedure is thereby avoided. In medically unfit patients, symptomatic large or malignant tumors may be debulked for palliation. There still exists controversy regarding enucleation versus a 2-cm margin in the management of benign leiomyomas. Many believe that

322

the inherent difficulty in distinguishing malignant lesions warrants a 1 to 2-cm margin excision for all benign lesions. Lymph node dissection usually is not indicated because these tumors spread hematogenously. Choi [10] recently reported a series of 12 gastric submucosal tumor excisions using laparoscopic intragastric surgery [10]. Lesions were 2.3 to 5.5 cm in size. All patients had a rapid postoperative recovery.

Table 3. Indications for endo-organ surgery for gastric cancer Current indications for endo-organ excision for gastric cancer TI (mucosal) well-differentiated with no vascular or lymphatic invasion Parts of stomach inaccessible to endoscopic mucosal resection (i.e., fundus/posterior body/cardia) Elevated less than 2.5 cm Depressed less than 1.5 cm (not ulcerated) Any part of the stomach Elevated 2.54.5 cm Depressed 1.52.5 cm (not ulcerated) TI (ulcerated or larger than 5 cm) with T2 lesions Medically unfit patients Advanced lesions with complications

Dieulafoys lesion Dieulafoys lesion, also known as exulceration simplex with submucosal arterial malformation, is the cause of nearly 1% of nonvariceal upper gastrointestinal bleeding [44]. Because the bleeding is intermittent, Dieulafoys lesion frequently is not diagnosed at initial endoscopy. Many authors have reported that multiple endoscopies are required for diagnosis, and that this delay increases transfusion requirements. No etiologic factors have been identified, and treatment requires prompt intervention. Dieulfoys lesion usually is located in the upper third of the stomach [2, 45]. Endoscopic sclerotherapy has been used with varied success to control massive bleeding. Orumo-Cortes et al. [44] reported a 78% success rate with one or two endoscopic sclerotherapy treatments, but there was one death. Mitru [36] published reports of 10 cases, all requiring surgical intervention. Attempts at endoscopic electrocoagulation and angiographic embolization generally have been unsatisfactory [47]. Endoscopic hemoclip application is another alternative hemostatic technique. Binmoller [8], in a series of 88 patients, reported a high success rate in controlling active upper gastrointestinal bleeding by the use of endoscopically applied hemoclips. Surgical intervention includes open gastrostomy with ligation of the bleeding vessel, wedge resection, or gastrectomy (if the lesion cannot be localized). Another option for hemostatic control is suture ligation or wedge resection after location by double transillumination using simultaneous endoscopy and laparoscopy [37]. With endo-organ surgery, the bleeding vessel can be sutured under direct vision. If this fails to control bleeding, partial-wall-thickness excision can be performed.

terior body lesions, wide tumor margins are ensured. Endoorgan surgery also is an alternative for T2 and T3 lesions in medically unfit patients. Despite a decreasing incidence, gastric cancer remains a serious problem in the United States, causing 13,700 deaths annually, primarily because most gastric cancers are discovered at an advanced stage [6, 57]. Approximately 36% and 31% are discovered as stages III and IV lesions, respectively. Until cost-effective screening of the high-risk population allows early detection of gastric cancer, survival benefit will be limited to a few patients whose diagnosis is established incidentally. In more advanced cancers, endoorgan surgery may be useful for performing limited palliative resection (Table 3).

Surgical technique
For endo-organ procedures, two or three gastric ports are placed under endoscopic guidance. The exact position is determined by the location of the lesion that requires excision. Zones of difficulty, shown in Fig. 4, are based on the attainable distance from the port to the lesion. We use a special operative port developed in our laboratory (Cook Surgical Bloomington, IN, USA). The port is a modified PEG, which allows safe and rapid access to the gastric lumen for 5- and 10-mm instruments. The port is designed with a foam rubber stent inside a deflatable balloon that prevents dislodgement.

Port placement
A suitable location is identified on the anterior gastric wall. The operating port is assembled, and care is taken to be sure the bowel is not between the peritoneum and the stomach. A 2.5-cm incision is made, and T fasteners can be placed if needed. A needle sheath assembly is placed through the incision and into the gastric lumen. A looped guidewire is introduced through the sheath and engaged by an endoscopic snare. The guidewire then is pulled out of the mouth. It is in turn connected to the port assembly, which is pulled by the operating surgeon through the mouth, esophagus, and abdominal wall. A fascial incision is made incrementally using a #15 blade guided by the longitudinal slot in the dilator. The port is pulled carefully above the skin level, and the dilator is carefully detached by a to-and-fro rocking motion. The cuff is inflated and the skin bolster applied. This compresses the balloon and foam rubber stent against the gastric wall, preventing leakage of insufflated carbon dioxide and gastric contents. The entire procedure is performed under endoscopic guidance. A valved cannula 10.8 mm lumen in diameter is placed inside the port. An optional 5-mm reducer is available. Other ports are similarly placed as desired. Laparoscopic instruments are used for intraluminal procedures. Preoperative antibiotics, an antiseptic mouth wash, and coating of the PEG with an antibiotic ointment before introduction decreases the possibility of wound infection.

Early gastric cancer The prognosis of gastric cancer has improved significantly in Japan, primarily because an aggressive screening process has become commonplace. A high percentage (40%) of gastric malignancies in Japan are mucosal lesions. For purposes of treatment and prognosis, early gastric cancer is regarded as a separate entity. Endoscopic mucosal resection (EMR) has proved to be adequate treatment for T1 and raised T2a lesions [23, 24, 38], but is not technically feasible for lesions larger than 2 cm in diameter, or for those located in the fundus, posterior body, and, in some instances, the lesser curvature side of the antrum. Ohashi [42] has shown that mucosal resection can be achieved through the intraluminal approach [42]. This allows excision of lesions larger than 2 cm and some depressed lesions. Because this approach provides excellent visualization and access for fundic and pos-

323

Fig. 4. Zones of difficulty for endo-organ gastric excision: A Zones for port placement. B Most accessible zones. C Difficult access zones. D Inaccessible zones.

Partial-thickness excision
For excision procedures involving mucosal lesions, the stomach is insufflated with carbon dioxide gas at a pressure of 6 mmHg. The lower pressure provides adequate distention and does not stop venous bleeding that needs to be recognized at the time of excision. A 1:10,000 adrenaline normal saline solution is injected into the submucosal space to assist with hemostasis. The surgeon should wait approximately 2 min before cutting tissue. The Harmonic Scalpel as well as unipolar or bipolar scissors can be used, but we prefer simply to cut the tissue with a scissors, avoiding coagulation and thus providing an accurate lateral margin. The bleeding that results can be controlled with cautery after the lesion is removed. It is essential to mark the borders of excision before dividing the mucosa, and color-coded clips can be placed at the margins to assist the pathologist in orienting the specimen. Lesions near the gastroesophageal junction and pylorus are particularly suited for the endo-organ approach because the ports can be placed well away from the lesion, and because the internal view ensures that the lumen will not be compromised by suturing. Some feel that closing the mucosal defect is not necessary if hemostasis is attained easily. An endoscopic view with the stomach partly collapsed is appropriate to ensure adequate hemostatis.

treated with caution, and a laparoscopic skeletonization of the gastroesophageal junction may be in order before excision if the lesion is immediately adjacent to the lesser curvature. This helps to spare the vagal nerves and makes external suturing possible if necessary. Our preferred method of closure, however, is internal suturing using monofilament polypropylene sutures tied extracorporeally. These sutures need to placed and tied as the excision progresses, thus minimizing the full-thickness hole and allowing good internal visualization to persist. Extracorporeal tension allows the suture to pull the gastric opening together despite the insufflation. Intracorporeal knot-tying does not work because the gastric margins pull even a surgeons knot apart before the next suture throw can be applied. After the lesion is separated from the gastric wall, it is removed with an endopouch, and frozen sections may be in order. The integrity of the gastric wall closure is tested by insufflating the stomach. If there is any doubt, viewing the closure laparoscopically with irrigation fluid covering the closed gastrotomy is appropriate.

Port removal
Port removal is accomplished by deflating the balloon and removing the skin bolster. In thin patients, Sen retractors and Babcock forceps are used to deliver the anterior gastric wall above the skin level, and gastrostomy is closed directly with sutures. In obese patients, ports are pulled out of the stomach lumen, but the cuff remains inside the abdominal cavity. Laparoscopic instruments can be placed through the port to suture the gastrotomy. The procedure may be aided by the use of T fasteners to retract and identify the gastrotomy. Alternatively, the gastrostomy can be downsized by a smaller gastrostomy tube, which is removed later. The T fasteners allow earlier gastrostomy tube removal. A novel acute gastrotomy closure technique described by Hepworth et al. [21] calls for two additional sutures to be placed through the abdominal

Full-thickness excision
Lesions requiring full-thickness excision can be treated safely with the endo-organ approach. In our laboratory we were able to develop a technique particularly appropriate for fundic and antral lesions. High-flow high-pressure gastric insufflation is used. It is not necessary to be concerned about small bowel distention because the gas is absorbed rapidly. An adrenaline solution is used, and the Harmonic Scalpel is particularly helpful in attaining hemostasis. Lesions with less curvature should be

324 wall fascia and stomach wall adjacent to the port. Their intragastric ends are pulled out the port and tied to each other. Then the knot is retracted back into the gastric lumen. The port is removed, and the external strands are tied, thus closing the gastrotomy and fascia while also securely apposing the stomach to the peritoneum. Leaving the skin open in obese patients after a prolonged procedure will help to prevent wound infection.

vances in minimally invasive surgery, endo-organ surgery now is an alternative for the effective management of gastric diseases. Direct access to the lumen of the stomach affords a magnified high-resolution image for precise excision using laparoscopic instruments. These advances have decreased operative blood loss, length of hospital stay, and postoperative patient discomfort. References
1. Ahlman H, Kolby L, Lundell A, Olbe L, Wangberg B, Granerus G, Grimelius L, Nilsson O (1994) Clinical management of gastric carcinoid tumors. Digestion 55(Suppl 3): 7785 2. Arora A, Mehrotra R, Patnaik PK, Pande G, Ahlawat S, Bhargava DK (1991) Dieulafoys lesion: a rare cause of massive upper gastrointestinal hemorrhage. Trop Gastroenterol 12: 2530 3. Baettig B, Haecki W, Lammer F, Jost R (1993) Dieulafoys disease: endoscopic treatment and follow-up. Gut 34: 14181421 4. Bandoh T, Isoyama T (1992) Submucosal tumors of the stomach: a study of 100 operative cases. Surgery 113: 498506 5. Berenstein E, Ghigliani M, Caro L, Uehara U (1998) Endoscopic ultrasonography in the diagnosis of submucosal tumors of the upper digestive tract. Acta Gastroenterol Latinoam 28: 58 6. Bernan MF, Karpeh Jr MS (1996) Surgery for gastric cancer: the American view. Semin Oncol 23: 352359 7. Bilchik AJ, Nilsson O, Modlin IM, Sussman J, Zucker KA, Adrian TE (1989) H2-receptor blockade induces peptide YY and enteroglucagonsecreting gastric carcinoids in mastomys. Surgery 106: 11191126 8. Binmoeller KF, Thonke F, Soehendra N (1993) Endoscopic hemoclip treatment of gastrointestinal bleeding Endoscopy 25: 167170 9. Bone GE, McClelland RN (1976) Management of gastric polyps. Surg Gynecol Obstet 142: 933938 10. Choi YB (1998) Laparoscopic management of submucosal tumor of the stomach. Presented at the annual meeting of Society of Laparoscopic Surgeons at San Diego, 1998 11. Therapeutic endoscopy and bleeding ulcers. JAMA 262: 13691372 12. Daibo M (1987) Malignant transformation of gastric hyperplastic polyps Am J Gastroenterol 82: 10161025 13. Dutta SK, Costa BS (1979) Umbilicated gastric polyposis: an indicator of metastatic gastric tumor. Am J Gastroentrol 71: 598600 14. Filipi CJ, Perdikis G, Hinder RA, DeMeester TR, Fitzgibbons Jr RJ, Peters J (1995) An intraluminal surgical approach to the management of gastric bezoars. Surg Endosc 9: 831833 15. Filipi CJ, Wetscher G, DeMeester TR, Peters JH, Hinder RA, Fitzgibbons RJ Jr (1994) Development of endo-organ surgery and potential applications. In: Peters JH, DeMeester TR (eds) Minimally invasive surgery of the foregut. Quality Medical Publishing, St Louis, pp 288 308 16. Fleming ID (1982) The role of surgery in the management of gastric lymphoma. Cancer 49: 1135 17. Gauderer MW, Ponsky JL, Izant RJ Jr (1980) Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 15: 872875 18. Gilligan C, Lawton GP (1995) Gastric carcinoid tumors: the biology and therapy of an enigmatic and controversial lesion. Am J Gatroentrol 90: 338352 19. Hakanson R, Tielemans Y, Chen D, Andersson K, Mattsson H, Sundler F (1993) Time-dependent changes in enterochromaffin-like cell kinetics in stomach of hypergastrinemic rats. Gastroenterology 105: 1521 20. Hay LJ (1956) Surgical management of gastric polyps and adenomas. Surgery 39: 114 21. Hepworth CC, Gong F, Kadirkamanathan SS, Swain CP, Roger J (1998) Operating gastrostomy tubes: insertion and removal for minimally invasive transgastric ulcer surgery. Minim Ivasive Ther Allied Technol 7: 371377 22. Hiki V, Sakakibara Y, Mienoti H, Shimao H, Kobayashi N, Katada N (1991) Endoscopic treatment of gastric cancer. Surg Endosc 5: 1113 23. Hiki Y (1996) Endoscopic mucosal resection (EMR) for early gastric cancer. Nippon-Geka-Gakkai-Zasshi 97: 273278 24. Hiki Y, Shimao H, Mieno H, Sakakibara Y, Kobayashi N, Saigenji K (1995) Modified treatment of early gastric cancer: evaluation of en-

Contraindications and limitations Any condition that precludes safe PEG placement is a contraindication for endo-organ surgery. Massive ascites, morbid obesity, peritonitis, local infection, coagulopathy, peritoneal dialysis, and undilatable esophageal stricture all are contraindications for large-diameter PEG placement. Endoorgan surgery has limited application to anterior body lesions. With the current limitation in experience, endo-organ surgery probably is not suitable for unstable patients. The laparoscopic T-bar lesion-lifting technique popularized in Japan by Oghami (personnel communication, 1999) allows full-thickness excision for lesions in the midportion of the stomach. It is guided by endoscopic visualization and has proved to be safe and effective. Vagal nerve dissection is necessary if the lesion is on the side of lesser curvature. If lymph node dissection is required for staging or treatment, a laparoscopic or open procedure is preferable.

Complications The safety of percutaneous gastrostomy ports for endoorgan surgery was reviewed by Tomonaga [56] in a recent presentation. A total of 53 ports were placed in 28 patients for various intraluminal procedures. Potential problems related to assembly and introduction of the PEG along with the intra- and postoperative complications were recorded. No inadvertent dilator assembly detachments nor any premature port extractions occurred. The most common complication was wound infection (5/53) ranging from mild cellulitis to subfascial abscess. A gastrocutaneous fistula did develop, which was attributed to prolonged use of a gastrostomy rather than the procedure itself. It spontaneously closed in 4 weeks. The intraperitoneal spillage of gastric contents can be prevented by using ports with inflatable balloon and the foam stent, which snugly appose the anterior gastric wall to the abdominal wall. Preoperative antibiotics along with antiseptic mouthwash may help decrease the wound infection rate, as also may the use of antibiotic ointment on the port. Leaving the skin open, especially in obese patients, also may reduce the rate of wound infection.

Conclusions Extensive surgical procedures are required for effective treatment of a wide range of gastric diseases including polyps, submucosal tumors, lymphoma, and cancer. With the widespread use of endoscopy, many lesions can be diagnosed and managed. The advent of laproscopic surgery has allowed a wide variety of abdominal procedures to be performed with improved surgical outcomes. Because of ad-

325 doscopic treatment of early gastric cancers with respect to treatment indication groups. World J Surg 19: 517522 Hughes Jr RW (1992) Diagnosis and treatment of gastric polyps. Gastro Endo Clinic North Am 2: 457467 Huppler EG, Prustly JT (1960) Diagnosis and results of treatment in gastric polyps. Surg Gynecol Obstet 110: 309 Kamiya T (1981) Histoclinical long-standing follow-up study of hyperplastic polyps of the stomach. Am J Gastroenterol 75: 275281 Kandel G (1990) Management of nonvariceal upper gastrointestinal hemorrhage. Hosp Pract 25: 167184 King RM, Jonathan AVH (1982) The management of gastric polyps. Surg Gynecol Obstet 155: 846848 Kitano S, Kawanaka H, Tomikawa M, Hirabayashi H, Hashizume M, Sugimachi K (1995) Bleeding from gastric ulcer halted by laparoscopic suture ligation. Surg Endosc 8: 405407 Laxen F (1981) Gastric polyps and gastric cancer. Ann Clin Res 13: 154155 Liang R, Todd D, Chan TK, Chiu E, Lie A, Kwong YL, Choy D, Ho FC (1987) Gastrointestinal lymphoma in Chinese: a retrospective analysis. Hemat Oncol 5: 115 Marshall SF (1952) Gastric polyps. Surg Clin North Am 32: 857 Martinez-Serna T, Filipi CJ (19xx) Gastric endo-organ access: technique and complications.Gastric endo-organ access: technique and complications. Surgical Technology International VII Ming SC (1977) The classification and significance of gastric polyps Monogr Pathol 18: 149175 Mitru N, Csendes A, Burdiles P, Smok G, Braghetto I, Chiong H, Gallo G (1990) Dieulafoys vascular malformation as a cause of massive and recurrent gastric hemorrhage. Rev Med Chil 118: 988992 Mixter III CG, Sullivan CA (1992) Control of proximal gastric bleeding: combined laparoscopic and endoscopic approach. J Laparoendosc Surg 2: 105109 Mizumoto S, Misumi A, Harada K, Arima K, Hirata T, Yoshinaka I, Ogawa M (1995) Evaluation of endoscopic mucosal resection (EMR) as a curative therapy against early gastric cancer. Nippon-GekaGakkai-Zasshi 93: 10711074 Modlin IM, Gilligan CJ, Lawton GP, Tang LH, West AB, Darr U (1995) Gastric carcinoids: the Yale experience. Arch Surg 130: 250 256 Namakura S, Lida M, Yao T, Fujishima M (1991) Endoscopic features of gastric carcinoids. Gastroenterol Endosc 37: 535538 Neimark S, Rogers A (1982) Gastric polyps: a review. Am J Gastroenterol 77: 585587 Ohashi S (1995) Laparoscopic intraluminal (intragastric) surgery for early gastric cancer. Surg Endosc 9: 169171 43. Orlowska J, Jarosz D, Pachlewski J, Butruk E (1995) Malignant transformation of benign epithelial polyps. Am J Gastroentrol 90: 2152 2159 44. Ortuno-Cortes JA, Quintana-Tomas L, Garcia-Garcia A (1996) Endoscopic sclerotherapy is useful in Dieulafoys disease. Gastroenterol Hepatol 19: 4751 45. Perez-Machado L, Arcos-de-Gonzalez B (1990) Dieulfoys lesion as a cause of massive gastrointestinal hemorrhage. G E N 44: 5962 46. Pertini JL (1998) Endoscopic therapy for gastrointestinal bleeding. Postgrad Med 84: 239245 47. Polit SA (1990) The Dieulfoy gastric lesion: an infrequently recognized cause of upper gastrointestinal hemorrhage in the elderly. J Am Geriatr Soc 38: 5355 48. Potvin M, Gagner M, Pomp A (1995) Laparoscopic transgastric suturing for bleeding peptic ulcers. Surg Endosc 10: 400402 49. Rappel S, Altendorf-Hofmann A, Stolte M (1995) Prognosis of gastric carcinoid tumors. Digestion 56: 455462 50. Sedillot C (1849) Operation de gastrostomie, Pertiquee pour la premiere fois le. Gaz Med Strassbourg 9: 566 51. Seifert E, Elster K (1975) Gastric polypectomy. Am J Gastroentrol 63: 451456 52. Sugawa C (1989) Endoscopic diagnosis and treatment of upper gastrointestinal bleeding. Surg Clin North Am 69: 11671183 53. Taniguchi E, Kamiike W, Yamanishi H, Ito T, Nezu R, Nishida T, Momiyama T, Ohashi S, Okada T, Matsuda H (1997) Laparoscopic intragastric surgery for gastric leiomyoma. Surg Endosc 11: 287289 54. Thomas RM, Baybick JH, Elsayed AM, Sobin LH (1998) Management of patients with gastric carcinoid tumors. Gastroentrology 108: Selected summaries 55. Tomasulo J (1971) Gastric polyps: histologic types and their relationship to gastric carcinoma. Cancer 27: 13461355 56. Tomonaga T (1998). A new form of access for endo-organ surgery: a review of initial experience. Presented at the 6th World Congress of Endoscopic Surgery, Rome, Italy, June 1998 57. Weese JL, Nassbaum ML (1992) Gastric cancer: surgical approach. Hematol Oncol 10: 3135 58. Yamagiwa H, Matsuzaki O, Ishihara A, Yoshimura H (1978) Clinicopathological study of gastric leiomyogenic tumors. Gastroenterol Jpn 13: 272280 59. Yasuda K, Nakajima M, Yoshida S, Kiyota K, Kawai K (1989) The diagnosis of submucosal tumors of the stomach by endoscopic ultrasonography. Gastrointest Endosc 35: 1015

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

39. 40. 41. 42.

Surg Endosc (2000) 14: 336339 DOI: 10.1007/s004640020059

Springer-Verlag New York Inc. 2000

Microlaparoscopic cholecystectomy
Less invasive gallbladder surgery
S. W. Unger, J. C. Paramo, M. Perez
Department of Surgery, Mount Sinai Medical Center of Greater Miami, 4302 Alton Road #820, Miami Beach, FL 33140, USA Received: 16 February 1999/Accepted: 8 October 1999

Abstract Background: We set out to compare a prospective evaluation of microlaparoscopic cholesystectomy (MLC) using 5-mm ports for the scope and operating ports and two 2-mm ports for retracting to the historic results of standard laparoscopic cholecystectomy (SLC). Methods: Fifty-six consecutive patients were operated electively for symptomatic gallstones between June 1997 and July 1998. Demographics, history of prior abdominal surgery, operative time, resident level, need to convert, length of stay, and postoperative analgesia were recorded for each case. In all, 43 women and 13 men aged 21 to 89 (average, 51 years) underwent MLC. Average weight was 78 kg (range, 48119) and average height was 163 cm. Results: Operative time for MLC was 72 25 min (range, 35140), somewhat less than the referenced standard of 79 27 min (p 0.1). The skin-to-trocar time (6 2 vs 13 77 min) and intraoperative cholangiogram time (9 8 vs 11 6 min) were significantly shorter (p < 0.01 and p < 0.05, respectively) for MLC. Other partial times were not significantly different. PGY2 residents averaged 74 21 min (range, 44118) compared to 75 27 min (range, 35140) for PGY3 and 53 5 (range, 4359) for PGY5. Patient weight influenced time. Patients <65 kg averaged 56 12 min; 6680 kg, 72 24 min; 8195 kg, 78 26 min; and >95 kg, 85 22 min. Previous abdominal surgery did not affect operative time. Nine patients (16%) required conversion from 2- to 5-mm ports because of adhesions, wall thickening, or need for better retraction. Time in these patients was 95 26 min vs 68 21 min in other patients (p < 0.01). No patient was converted to an open procedure. Three patients (5%) had a positive cholangiogram and common bile duct exploration that required placement of an extra 5-mm trocar. Five patients (9%) required insertion of an additional 2-mm port. All patients received patientcontrolled analgesia (PCA). Morphine use was 0.21 0.19 mg/kg (range, 00.8). Hospital stay was 1.31 days (range,

0.54). Subjective satisfaction was excellent because of smaller incisions. No additional morbidity was seen with MLC. Conclusion: MLC is a feasible and safe approach that provides similar times to SLC with better cosmesis, a less painful recovery, and possibly an earlier return to normal activity. Key words: Gallbladder Laparoscopy Microlaparoscopic cholecystectomy Needlescopic surgery

At present, the worldwide standard of care for patients with symptomatic gallbladder disease is laparoscopic cholecystectomy. In the United States, this procedure was popularized by Olsen and Reddick, who performed the operation using two 10-mm ports and two 5-mm ports [5]. However, as medical technology has continued to evolve, smaller instruments as well as better light sources and optics for 5-mm laparoscopes and 5-mm clip appliers have become available. It is now possible to perform microlaparoscopic needlescopic surgery with these fine-caliber instruments [1, 7]. Therefore, we started applying these technologic advances to standard laparoscopic cholecystectomy (SLC). We call this less invasive procedure microlaparoscopic cholecystectomy (MLC) [9]. This report summarizes our experience with MLC.

Materials and methods

From June 1997 to July 1998, we prospectively evaluated 56 consecutive patients operated electively for symptomatic gallbladder disease. Patients with evidence of acute cholecystitis, peritonitis, or sepsis, and those requiring emergent surgical intervention were excluded from the study. Selected patients underwent MLC with intraoperative cholangiogram (IOC) at the Mount Sinai Medical Center of Greater Miami. All cases were evaluated for history of previous abdominal surgery and patient demographics were recorded. All operations were accomplished with the same surgical attending as

Correspondence to: S.W. Unger

337 Table 1. Average operative times for each segment of the cholecystectomy Skin incision to trocar insert MCL (n 56) SLC (n 226) p value 62 13 7 <0.01 Trocar insert to start IOC 20 15 17 11 0.1 (NS) Start IOC to end IOC 98 11 6 <0.05 End IOC to amputation 19 9 19 11 >0.5 (NS) Amputation to extraction 53 66 0.5 (NS) Extraction to skin closure 13 6 13 6 >0.5 (NS) Total (min) 72 25 79 27 0.1 (NS)

IOC, intraoperative cholangiogram; MLC, microlaparoscopic cholecystectomy; SLC, standard laparoscopic cholecystectomy; NS, not significant

the first assistant and with different surgical residents as the operative surgeon. The level of training of the surgical resident was noted in all cases. Cholecystectomy was performed using two 5-mm ports and two 2-mm ports, as follows: One 5-mm port was placed above the umbilicus in the midline and used for the insertion of a 5-mm laparoscope and for removal of the gallbladder. Another 5-mm port was inserted in the epigastrium in the subxyphoid midline position. This was the working port used for dissecting and clipping with 5-mm instruments, as well as for laparoscopic visualization of the gallbladder removal through the supraumbilical port. One 2-mm port was placed in the right midclavicular line subcostally. Another 2-mm port was inserted in the anterior axillary line subcostally. These ports were used to insert 2-mm graspers that were employed to retract and dissect the gallbladder. The cholecystectomy was performed in the usual laparoscopic fashion, following the same basic principles of SLC. Patients requiring conversion to larger ports or to an open procedure were noted. A transcystic IOC was performed in all cases using an Olsen cholangiocath (Cook Companies, Bloomington, IN, USA). Fluoroscopic visualization of the cystic duct, common bile duct, common hepatic duct, right and left hepatic ducts, and adequate emptying of the contrast into the duodenum were required for a cholangiogram to be considered normal. A positive cholangiogram was defined as inadequate visualization of any of the above structures, presence of permanent filling defects within the biliary tree, or failure of emptying of the contrast into the duodenum. Patients with positive cholangiograms underwent laparoscopic common bile duct exploration. Gallbladder removal was initiated by moving the laparoscope to the epigastric port. A grasper was placed through the umbilical port, and the gallbladder was grasped as close to the cystic duct as possible. This umbilical port was then removed, and the opening in the fascia was dilated first with a fine clamp and then with a Kelly clamp. The gallbladder was pulled through the incision, grasped with a Kelly clamp, and opened at its neck. The bile and small stones were suctioned, enabling the gallbladder removal. If this was not possible, a sponge forceps was used to break and remove bigger stones, and the gallbladder was extracted. On rare occasions, the umbilical incision was enlarged sharply at the fascial level in order to remove the gallbladder. This was almost always <10 mm. The skin incision was usually sufficient without enlargement. For the purpose of timing and in conformity with prior studies [8], the operation was divided into six parts, as follows: from the skin incision to the final insertion of all trocars, from the trocar insertion to the initiation of the IOC, from start to end of the IOC, from the end of the IOC to the gallbladder amputation, from the gallbladder amputation to extraction, and from the gallbladder extraction to skin closure. All of these individual parts were timed by an independent nurse who was present in the operating room at all times. The total operative time was then calculated. All of these variables were compared with the national average times for SLC, as described in a study in which we participated [8]. All patients were given morphine as patient-controlled analgesia (PCA), and the total dose was recorded. The total length of hospital stay was documented in every case. For all continuous data, the mean and standard deviation were calculated. Comparison between groups was done using two-sample and Bonferroni t-tests. Statistical significance was defined as a p value of <0.05.

Results Of the 56 patients enrolled in this study, 43 were female and 13 were male. Age distribution was 2189 years (average

51). Average weight was 78 kg (range, 48119) and average height was 163 cm. Average operative times are shown in Table 1. Total operative time for MLC was 72 25 min (range, 35140), as compared to the national average for SLC of 79 27 min (p 0.1). Among the different parts of the procedure, the skin incision to trocar insertion and the IOC had a statistically significant difference in favor of MLC. Twenty-nine patients (52%) had had previous abdominal surgery. These patients had a longer operative time73 27 min vs 71 21 min in patients who did not have previous surgery. This difference was not significant (p >0.5). Nine patients (16%) required conversion from 2- to 5-mm ports because of adhesions, severe chronic cholecystitis, and subacute inflammation with gallbladder wall thickening, all of which mandated better retraction and thus larger (5-mm) graspers. Average operative time in these patients was 95 26 min vs 68 21 min in the other 47 patients (p < 0.01). No patient required conversion to an open procedure. Three patients (5%) had a positive IOC and underwent laparoscopic common bile duct exploration, which required placement of a 5-mm trocar. Five patients required insertion of an extra 2-mm port for additional retraction. These patients either were obese or had stiff livers that would not retract cephalad and therefore required downward retraction of the duodenum. Weight influenced operative time, as heavier patients required longer operative times (Fig. 1). Average operative times for patients with weights <80 kg was 66 22 min vs 84 27 min for patients heavier than 81 kg (p 0.01). Total pain medication varied widely among patients, averaging 0.21 0.19 mg/kg of morphine, with a range between 0 and 0.8 mg/kg. Subjective patient satisfaction was excellent due to the smaller incisions. With regard to the level of training of the resident performing the operation, PGY2 residents averaged 74 21 min (range, 44118) vs 75 27 min (range, 35140) for PGY3 and 53 5 min (range, 4359) for PGY5 (Fig. 2). The difference between PGY2 and PGY3 residents was not significant (p 0.2). However, as compared with PGY2 and PGY3, PGY5 residents performed the operation significantly faster (p 0.01 and p 0.05, respectively). Hospital stay averaged 1.31 days (range, 0.54). Fourteen patients (25%) stayed > 24 h, yet only three patients stayed >48 h. The prolonged stays were due to urinary retention in one patient, migraine headache in one patient, and preexisting medical conditions in one patient with coronary artery disease and a history of pulmonary tuberculosis.

338

Fig. 1. Operative times by weight of patient. Heavier patients required more time. Fig. 2. Operative times by resident level. More experienced surgeons required less time.

Discussion Laparoscopic cholecystectomy has proven advantages over open cholecystectomy [4]. It is a less invasive and less painful procedure with similar morbidity and shorter patient recovery. However, new technology has made this type of surgery even less invasive; this is the basis for MLC [7, 9, 10]. Although some surgeons have argued that smaller is not necessarily better [1], our study proves that MLC is a feasible approach that can be completed with operative times similar to those reported for SLC. Nor does the MLC technique result in any additional morbidity. Initially, we were concerned that the smaller instruments would make certain parts of the operation more difficult. We considered the following questions: (1) Can we retract the gallbladder well enough to adequately expose the triangle of Calot with these small instruments? (2) Can we perform the IOC accurately through the epigastric trocar, given that this is the only port large enough to admit the cholangiogram clamp? (3) Will it be more difficult to remove the gallbladder through a small incision? To address these concerns, we decided to use a timed study and compare our findings to those of Traverso et al. [8]. Our results showed that the times compare well and the IOC could be performed adequately through the epigastric port. Furthermore, we developed techniques to assist in the gallbladder extraction. The 2-mm graspers are adequate to retract the gallbladder and provide exposure of the triangle of Calot in most cases. In difficult cases, the 5-mm dissector was used to grasp Hartmans pouch, and dissection was done with the 2-mm instruments. We believe that our results validate the use of MLC. MLC requires a more delicate dissection and advanced technical skills. Like any new procedure, MLC has a learning curve. Thus, more experienced surgeons, such as PGY5 residents, take less time to perform the operation. Factors that increase operative time include adhesions, thickening of the gallbladder wall, chronic inflammation, anomalous anatomy, the presence of common bile duct stones, obesity, and the surgeons expertise. In this study, the surgical attending was the first assistant during MLC and performed the manipulation of 2-mm instruments for retraction and exposure. This task requires advanced laparoscopic skills. Likewise, the operative surgeon had to perform a more delicate dissection and was faced with the need to interact con-

tinuously with the first assistant. Both of these tasks have a learning curve and pose a higher degree of difficulty for both attending and resident. Heavier patients had longer operative times because instrument manipulation through the abdominal wall is more difficult in these patients, as is the retraction of the liver upward and the abdominal viscera downward. However, since the same problem is encountered with SLC, we think that patient weight is not a limiting factor for the performance of MLC. Obese patients often need a fifth trocar to retract downward, whether MLC or SLC is being performed. Only nine of our patients (16%) required conversion to 5-mm ports for better retraction. Other authors have commented on the superior cosmetic outcome seen with MLC [3, 6]. In our experience, patient satisfaction was excellent because of the even smaller incisions and the potentially less painful recovery. Actually, four patients required no pain medication at all postoperatively. Unfortunately, because this is not a prospective comparison between MLC and SLC with respect to analgesic use, we cannot state definitively that there is less pain with MLC. However, a reduction in the use of postoperative narcotics has been documented by other investigators [10]. Hospital stay averaged 1 day, but this is not significantly different from SLC. Some surgeons have used MLC in pediatric patients with good results [2]. MLC would seem to be a promising procedure for this population because of its minimally invasive nature. In conclusion, MLC is a feasible procedure that provides operative times similar to those reported for SLC. MLC has a better cosmesis, a potentially less painful recovery, and possibly an earlier return to normal activities. The expanding availability of this technology and the continuous search for less invasive procedures should lead to further studies confirming the benefits of MLC. References
1. Berci G (1998) Laparoscopic cholecystectomy using fine-caliber instruments: smaller is not necessarily better. Surg Endosc 12: 197 197 2. Kasirajan K, Obermeyer RJ, Kehris J, Lopez J, Lopez R (1998) Microinvasive laparoscopic cholecystectomy in pediatric patients. J Laparoendosc Adv Surg Tech 8: 131135 3. Kimura T (1998) Laparoscopic cholecystectomy using fine-caliber instruments. Surg Endosc 12: 283286

339 4. Neugebauer E, Troidl H, Kum CK (1995) The E.A.E.S. consensus development conferences on laparoscopic cholecystectomy, appendectomy and hernia repair. Surg Endosc 9: 550563 5. Reddick EJ, Olsen DO (1989) Laparoscopic laser cholecystectomy. Surg Endosc 3: 131133 6. Regier H (1998) Needlescopic cholecystectomy leads to better cosmesis, less use of postoperative narcotics. Gen Surg News 19: 4 7. Tanaka J, Andoh H, Koyama K (1998) Minimally invasive needlescopic cholecystectomy. Surg Today 28: 111113 8. Traverso LW, Koo KP, Hargrave K, Unger SW (1998) Standardizing laparoscopic procedure time and determining the effect of patient age/ gender and presence or absence of surgical residents during operation. Surg Endosc 11: 226229 9. Unger SW, Paramo JC, Perez-Izquierdo M (1998) Smaller is better. Surg Endosc 12: 1450 10. Watanabe Y, Sato M, Ueda S (1998) Microlaparoscopic cholecystectomythe first 20 cases: is it an alternative to conventional LC? Eur J Surg 164: 623625

Surg Endosc (2000) 14: 407 DOI: 10.1007/s004640000032

Springer-Verlag New York Inc. 2000

A simple technique for delivery of bulky gallbladder during laparoscopic cholecystectomy

A cautionary note
Saber and La Raja are to be commended for their efforts in developing a simple solution to a common problem that can be quite challenging [1]. To ease the extraction of a bulky gallbladder through the port wound, they use a Kelly clamp holding a scalpel between its jaws. As the clamp is passed along the gallbladder, the blade cuts through the constricting fascial layer. I have tested this technique on a number of patients and recently encountered an incidence of bowel injury, therefore, I feel compelled to strike a note of caution. First of all, plunging a knife deep into the peritoneal cavity is always a hazardous procedure, particularly when it is done blindly, as is the case with the method described by the authors. I would therefore strongly advise that this procedure be done under visual guidance, with the telescope repositioned in the subxiphoid port to allow a peritoneal view of the umbilical wound. In practice, the grip of the clamp on the blade (made similarly of metal) is not always secure enough, particularly when cutting the thick fascia. In my experience, the blade often slipped, so that the sharp tip of the blade projected out of the greater curvature of the curved Kelly clamp. In this situation, the gallbladder or, worse still, an underlying loop of bowel may be punctured. It would be safer to close the jaws of the clamp and slide the scalpel blade across the groove between the closed jaws, using the locked clamp to protect the gallbladder. But even in spite of these efforts, the blade sliding across the shallow gutter can easily go astray. A fistula director with a deep groove is more effective in protecting the gallbladder [2], and a broader grooved director serves this purpose even better [3]. Unfortunately, the right instrument is not always readily available when it is needed. A simpler, and much safer, solution is to insinuate the clamp between the wound edge and the gallbladder. With the jaws open and lift upward, the fascia is cut open using a pair of scissors inserted between the jaws. The cutting of the desired fascial layer is aided by the tactile sensation. Saber and La Raja claim that their technique has the advantage of facilitating the delivery of the bulky gallbladder without enlarging the skin incision (as depicted in the accompanying figure). Obviously, the skin wound, of necessity, must be big enough to allow for the eventual passage of the bulky gallbladder. Following the extension of the skin incision, it becomes easier to insert the scissors, even in obese patients. References
1. Saber AA, La Raja RD (1998) A simple technique for delivery of bulky gallbladders during laparoscopic cholecystectomy. Surg Endosc 12: 1003 2. Viswanath YKS, Wynne KS (1999) Use of fistula director to enlarge the port site opening to retrieve a stone packed bulky gall bladder during laparoscopic cholecystectomy: a simple and safe technique. J R Coll Surg Edinb 44: 179180 3. Williams IM, Rees BI, Ivey V (1992) Easy delivery of the gallbladder in laparoscopic cholecystectomy: a grooved director. Br J Surg 79: 344

W. T. Ng
Department of Surgery Yan Chai Hospital 7-11, Yan Chai Street Tsuen Wan Hong Kong SAR

Surg Endosc (2000) 14: 395399 DOI: 10.1007/s004640020066

Springer-Verlag New York Inc. 2000

A new remote-controlled endoscope positioning system for endoscopic solo surgery

The FIPS Endoarm
G. F. Buess,1 A. Arezzo,1 M. O. Schurr,1 F. Ulmer,1 H. Fisher,2 L. Gumb,2 T. Testa,1 C. Nobman1
1

Section for Minimally Invasive Surgery, Department of General Surgery, Eberhard Karls University, Waldho rnlestrasse 22, D-72072 Tu bingen, Germany Department of Technical Engineering, Karlsruhe Research Center, Karlsruhe, Germany Received: 9 July 1999/Accepted: 21 October 1999

Abstract. In the field of endoscopic solo surgery, the assistance received by the surgeon from ergonomical positioning devices is extremely important. They aid in both the retracting of instruments and the positioning of the endoscope. However, passive systems derived from open surgery have not proved satisfactory. Therefore, we set out to develop a remote-controlled arm capable of moving a rigid endoscope with about four degrees of freedom, while maintaining an invariant point of constraint motion coincident with the trocar puncture site through the abdominal wall. The system is driven by means of speaker-independent voice control or a finger-ring joystick clipped onto the instrument shaft close to the handle. When the joystick is used, the motion of the endoscope is controlled by the fingertip of the operating surgeon, which is inserted into the small ring of the controller in such a way as to make the motion of the fingertip correspond directly to the motion of the tip of the endoscope. A study was performed to compare the two different interfaces available for the system. With both interfaces, the guiding system allows for transparent and intuitive operation. Its set-up is easy; it is safe and reliable to use during the intervention; and it is faster than human assistance. With its improved ergonomy, this new generation of remote-controlled endoscope positioning system represents a further step toward the diffusion of solo surgery techniques in minimally invasive therapy. In our opinion, this prototype creates a valid compromise between human and robotic control of rigid endoscopes. Key words: Endoscopic surgery Instrumentation Retractors Robotics Solo surgery

In recent years, advances in technology have improved the quality and efficiency of many operative procedures in minimally invasive surgery. Since the advent of endoscopic surgery, the vision of the operating surgeon has depended on the help of an assistant surgeon whose responsibility it was to position the endoscope. To perform this task, the assistant has to keep the surgical point of interest in the center of the video image, thus providing sufficient target magnification and maintaining a stable horizontal image. Experience has shown that the assistant can seldom achieve what the operating surgeon would consider an optimal position. The application of robotic technology to solo surgery has the potential to increase both the precision of action and cost-effectiveness of endoscopic surgery. Although several prototypes have been introduced, there are currently only two remote-controlled serial systems available on the market: the AESOP system (Computer Motion, Goleta, CA, USA) and the Endoassist (Armstrong Healthcare, UK). In collaboration with the Karlsruhe Research Center (Germany), we have developed a prototype of a new robotic system (Fig. 1), with the aim of improving the system architecture and the human-machine interface.

Materials and methods Technology

The geometry of an endoscope guiding system for laparoscopic surgery should take account of the principle of the invariant point of motion [3, 5] where the trocar enters the abdominal cavity. The kinematic principle chosen for the FIPS project was similar to that of the passive TISKA Endoarm [6]. It establishes a remote center of motion, assuring that no lateral force is exerted around the trocar puncture site. The principle of maintaining the invariant point of motion through mechanical constraints was defined by Mueglitz et al. [3]. They described a robotic motion principle along virtual axes, intersecting at the point of trocar insertion. This

Correspondence to: G. F. Buess

396 phantom models with integrated animal organs; these conditions represent reliable and reproducible experimental conditions. The Tu bingen LapTrainer (Coburger Lehrmittelanstalt, Coburg, Germany) was equipped with porcine liver segments including the gallbladder to simulate laparoscopic cholecystectomy. Conventional laparoscopic OR equipment (video unit, HF, etc.) and a regular set of instruments were used. The prototype was used to guide the optic in combination with a TISKA Endoarm used as a retracting instrument during the procedure. All devices were positioned on the right side of the operating table, opposite the surgeon. The experiments involved three surgeons, all skilled in endoscopy. The voice-controlled FIPS Endoarm used in the speaker-independent mode was compared with the joystick controller. A control group working with human assistance was also included. Each surgeon performed 15 experiments randomized among the three groups, for a total of 45 experiments. At the end of each experiment, intuitiveness of handling, practicality of use under OR conditions, and mechanical stability of the FIPS system were judged subjectively on a scale from 1 to 10. The procedure time was broken down into segments related to different tasks, including: set-up time, for draping the phantom and setting up external devices; positioning time, for positioning trocars under vision, connecting devices to their instruments, and positioning the optic; intervention time, for performing the complete dissection of the gallbladder; extraction time, for extracting the gallbladder; breakdown time, for removing trocars under vision and placing external devices aside. Statistical analysis was conducted with the JMP system (SAS Institute, Cary, NC, USA) using ANOVA, the Wilcoxon test, and the Tukey-Kramer test.

Results
Fig. 1. The FIPS Endoarm, current prototype. set-up is of particular importance for the control scheme of electrically driven robotic devices. The FIPS camera guiding system (Fig. 2) is composed of a power supply unit, an operating table attachment, and an endoscope guiding device with two different human-machine interfaces: a finger-ring joystick (Fig. 3) and a voice control facility. The finger-ring joystick can be used as hand-held controller, or it can be clipped onto the handle of the operating instrument used. The tip of the second finger of the operating hand is inserted into the ring; its intuitive movements along the three spatial axes correspond to identical movements of the endoscope. The voice control device (DASA, Germany) can either be used as a speaker-independent system, or it can be programmed by the surgeon to recognize his or her voice. The kinematic design of the system (Fig. 2) comprises a first axis right through the point of incision. A second axis intersects the first perpendicularly at the point of incision. A small C-arch mechanism connects the two axes. This geometry permits only movements around the arch, whose center coincides with the point of incision. Thus, a remote center of motion is established at the point of intersection of the two axes. When the trocar tube and the inserted endoscope are moved, they are guided precisely through the invariant point of trocar insertion, without any force being exerted on the abdominal wall. Translation and rotation around the longitudinal axis of the instrument are assisted by electric motors, completing the four degrees of freedom necessary to guide the endoscope in the abdominal cavity. After the device has been attached to the operating table, it can be covered with a sterile plastic sac. A special technique allows the precise definition of the sterile area. The carrier system of the FIPS device is attached to the standard rail by means of a larger rail, which is screwed under the sterile drape covering the operating table. This larger rail is shaped with round edges to prevent tears to the draping tissue and to allow an easy translation of the base along the operating table. The whole arm is covered by a transparent plastic tube, which is fixed to the base of the carrier system. Thus, the boundaries of the sterile field are clearly defined. The entire arm is made of stainless steel and may be gas-sterilized if required.

The evaluation of the prototype was carried out on phantom models and confirmed by animal experiments. We found that the overall handling of the system was simple and did not require specific training. The trocar tube can be mounted easily on the guiding device. The results are shown in Figs. 47. All statistical results yielded by tests for interval variables were confirmed by rank tests. Both interfaces were judged to be intuitive, with an advantage for the joystick solution (7.3 vs 6.5, p < 0.05). A slight advantage in intervention time was found for the voice-control solution (11:57 vs 13:40, NS). Voice control was limited by the small movement of the endoscope that could be achieved; repetition of a sequence of identical commands was required to cover long distances. This system turns out to be quite cumbersome during the positioning and extraction processes, when long distances need to be covered by the endoscope to follow the insertion of the trocars. In this phase, voice control proved to be slower than the finger-ring joystick (4:28 vs 6:14, p < 0.0001; 3:29 vs 4:11, p < 0.0001). On the other hand, the interface, although speaker-independent, proved as safe and reliable as other speaker-dependent interfaces previously tested. The global time registered by the control group was significantly shorter than with the combinations involving the FIPS and TISKA Endoarms (p < 0.0001). This was the result of shorter set-up, positioning, extraction, and breakdown times (each p < 0.0001). As regards intervention time, the three groups scored comparable times, with no statistically significant difference.

Discussion Laparoscopic assistance often demands tiring standing positions and monotonous tasks. The use of mechanical positioning systems can be unsatisfactory. The movement of the retractors is often cumbersome and unsafe, since in most

Experimental evaluation
Evaluation of the prototypes was performed simultaneously with the technical development of the device. The experiments were carried out on

397

Fig. 2. Basic design of the FIPS Endoarm. Fig. 3. Finger-ring joystick interface of the FIPS Endoarm.

cases both hands of the surgeon are required for changes in position. The introduction of robotic technologies seems to be the major step toward the solution of this problem. The use of positioning and holding devices in laparoscopic surgery returns direct control of the whole procedure to the operating surgeons. Solo surgery is meant to allow for an increased precision of action. Several laparoscopic camera-driving systems have been devised in recent years. Currently, the Endoassist (Armstrong Healthcare) [1] and the AESOP system (Computer Motion) [8] are on the market. The rapid introduction of several different architecture and interface solutions reflects the growing interest in these developments. Since 1994, prototype work has been performed in cooperation with the Research Center in Karlsruhe; it led first to the design of the passive system TISKA Endoarm [6] and later to the remotecontrolled FIPS Endoarm presented here. The FIPS Endoarm allows motion in four degrees of freedom, consisting of two spatial axes and translation and rotation about the endoscope longitudinal axis, all of which are electrically driven. The FIPS Endoarm prototype has several basic ad-

vantages, which were confirmed during our phantom and animal experiments:

Negligible force exerted around the trocar puncture site: The system keeps an invariant point of constraint motion. Reduction of dangerous interferences with other positioning systems and the surgeons movements: The system architecture has the shape of an arch, which comes on the operative field from above; this design allows free space around the optic. Reduction of space requirement: The arm is attached directly to the trocar, close to the insertion of the instrument into the body; the movement of the optic does not affect the position of the arm. Remote control of the rotation of the optic: This is mandatory when angulated optics are used. Without it, the surgeon is forced to rotate the angulated optic by hand; a sudden drop of the light cable would cause involuntary rotation of the optic and consequent dangerous loss of the image of the point of surgical interest. Different intuitive interfaces, consisting either of a speaker-independent voice control device or a finger-ring

398

Fig. 4. Set-up and break-down time requirements with human assistance, FIPS joystick control and FIPS voice control, respectively. Fig. 5. Positioning and extraction time requirements with human assistance, FIPS joystick control and FIPS voice control, respectively. Fig. 6. Intervention time requirements with human assistance, FIPS joystick control and FIPS voice control, respectively. Fig. 7. Global time requirements with human assistance, FIPS joystick control and FIPS voice control, respectively.

joystick clipped close to the handle of the working instrument. The voice interpreter system developed by DASA proved as reliable as the speaker-dependent systems available today. At the same time, it obviates the need to program the system. A single command generates a motion along the axis of the optic of 10 mm or a motion of 5. Thus, the surgeon is obliged to repeat the command in sequence to obtain long motions. The joystick controller, on the other hand, is capable of long and combined movements. Although it is still in a prototype version, its ergonomy proved to be sufficient. A safety button placed inside the ring obviates involuntary movements. The speed is set to a reasonably slow rate to avoid abrupt movements and to adapt well to the operator. Other human-machine interfaces have been designed previously. Hand controls consisting of different buttons for different movements were ergonomically deficient. Foot pedals are more intuitive because they free the hands from the camera guidance task. Because the feet are inherently more clumsy than hands for precise tasks, they have found limited acceptance among surgeons, although there are a number of types of foot-operated switches in use [1, 8]. Voice control systems consisting of a recognition system for

synthesized speech have proved to be an extremely useful means of providing information and short instructions to surgeons. On the input side, speech recognition systems are now reliable and fast enough to be useful, but recognition of accuracy and response time are critical to their acceptance, since confusion could result, especially in stressful situations when control is critical [1, 8]. A natural method of indicating the way to an object is simply to point to it. In the act of pointing to the object, the movement is straight rather than being split into different vectors of motion, as is the case with robots, including those guiding the endoscope. This method is not only time-consuming, but often also results in less precision when pointing to the target. Track visual markers have been suggested as a possible improvement, but according to some authors [7], these devices do not always provide an optimal view in depth. This drawback is due to the difficulty in calculating the depth of the track, which can result in suboptimal magnification of the visual field. Moreover, it is sometimes inconvenient to have to keep the marker always in sight, as it could accidentally be covered by grasped or overlying tissue. Finally, visual track recognition systems require the use of expensive three-dimensional optics. A different solution proposed for a human-machine in-

399

terface is the head controller, which consists of a helmet with a light pointer held by the surgeon and a visual detector placed just over the monitor in front of him or her. Thus, the movements of the surgeons head are picked up by the visual detector and transferred to the robot guiding the optics. Tests with head controllers have shown that compound motion was sometimes confusing to the surgeon. Therefore, the controller was limited to detecting the dominant head gesture and powering only this axis of the manipulator [6]. We believe that the finger-ring joystick solution supplying the FIPS Endoarm is currently the best interface because it does not require any training due to its intuitive construction, and it is supported by a sufficiently safe technology. As with all other systems, the FIPS Endoarm entails the inconvenience of requiring a considerable amount of time for its set-up and breakdown. More effort needs to be directed toward streamlining this aspect of positioning systems. On the other hand, the longer time required for tasks other than intervention was particularly influenced by the relatively short time required for dissection in phantoms. In real-life clinical situations, the intervention time will account for more than half the procedure; consequently, the longer time needed for other phases will be less influential in the calculation of the overall time required. In phantom and animal models, the system proved to be safe, compact, user-friendly, and compatible with existing surgical equipment. It does not interfere with the ergonomic work space of the surgeon but, on the contrary, returns to him or her the freedom to determine a personal view, which was lost when laparoscopy replaced the open method. By replacing the camera assistant, the FIPS system en-

ables solo surgery for standard laparoscopic procedures. The ability to perform solo surgery in community hospitals and private institutions can be expected to alleviate some of the pressure due to limited resources, as well as reducing the need for extra personnel [3]. At the same time, residents in university and teaching hospitals would no longer be asked to perform tiring and boring assistant duties and therefore should be able to pay more attention to the maneuvers performed by the operating surgeon during the operation. References
1. Finlay PA, Ornstein MH (1995) Controlling the movement of a surgical laparoscope. IEEE Eng Med Biol 289291 2. Morgan ME (1993) Stationary and automated laparoscopically assisted technologies. J Laparoendosc Surg 3: 221227 3. Mueglitz J, Kunad, Dautzenberg P, Neisius B, Trapp R (1993) Kinematic problems of manipulators in minimally invasive surgery. Endosc Surg Allied Technol 1: 160164 4. Sackier JM, Wang Y (1994) Robotically assisted laparoscopic surgery: from concept to development. Surg Endosc 8: 6366 5. Schurr MO, Buess G, Rininsland HH, Holler E, Neisius B, Voges U (1996) ARTEMISManipulatorsystem fu r die endoskopische Chirurgie. Endoskopie Heute 9: 245251 6. Schurr MO, Arezzo A, Neisius B, Rininsland H, Hilzinger HU, Dorn J, Roth K, Buess GF (1999) Trocar and instrument positioning system. Surg Endosc (in press) 7. Taylor RH, Funda J, LaRose D, Treat M (1992) An experimental system for computer assisted endoscopic surgery. Proceedings of the IEEE Satellite Symposium on Neurosciences, Lyons, France, November 1992 8. Taylor RH, Funda J, Eldridge B, Gomory S, Gruben K, LaRose D, Talamini M, Kavoussi L, Anderson J (1995) A telerobotic assistant for laparoscopic surgery. IEEE Eng Med Biol 279288

Surg Endosc (2000) 14: 388394 DOI: 10.1007/s004640020064

Springer-Verlag New York Inc. 2000

The clinical suitability of laparoscopic instrumentation

A prospective clinical study of function and hygiene
T. W. Fengler, H. Pahlke, S. Bisson, E. Kraas
Department of Surgery, Krankenhaus Moabit, Lehrkrankenhaus der Humboldt Universita t zu Berlin, Turmstrasse 21, D-10559 Berlin, Germany Received: 9 July 1999/Accepted: 2 August 1999

Abstract. On the basis of experience gained from 6,000 laparoscopies (73% cholecystectomies) at the Moabit Hospital in Berlin, we carried out a cohort study to analyze the failure rate and decontamination of labeled tracer instruments processed in three test trays that were each subjected to 100 cycles. The majority of repairs focused on the functional parts of separable scissors and damaged or lost components. At 4%, the repair index after laparascopic use was less than that of a previously documented investigation period covering 1990 to 1996. A comparison of the costs of disposable and reusable instruments showed that reusable instruments were more cost-effective by a factor of 10, indicating that the price gap reported in our previous calculation for 1992 and 1994 has closed only slightly. After 100 cycles, we found traces of proteinaceous material in the eluate on every fourth instrument inspected (eight of 32); half of them (four) gave a positive reading when tested with a hemoglobin pseudoperoxidase test stick. It must be said, however, that similar residual contamination has been found on instruments used in conventional open surgery, with no indication of clinical relevance. This study was designed to examine the clinical suitability of laparoscopic instruments in terms of function and hygiene. Improvements in instrument design and cleanability must focus in particular on the reproducibility of cleaning results, because cleaning is the most important step in processing sterile supplies. As the number of minimally invasive operations has risen considerably, a mere visual check no longer meets the requirements prescribed by modern quality assurance. A multicenter study of residual proteins found on tracer instruments in all surgical fields is now in progress. Key words: Cleaning Complications Instruments Laparoscopy Sterilization

Correspondence to: T. W. Fengler

The objective of sterilization is to kill or irreversibly deactivate all microorganisms and virusesin particular, bacterial sporesin or on the surfaces of an object [3]. To achieve this level of hygiene, thousands of surgical instruments and parts are regularly transported, disassembled, checked, andin so far as they are able to withstand heat sterilized with the aid of saturated steam. Legal considerations have led to a strict interpretation of the term sterilization, even in the absence of a standardized approach dictating the means by which this objective might be met through disinfection and sterilization [1, 4, 7, 10, 11, 12, 16, 23]. Adequate cleaning of all surfaces, including lumens, is a prerequisite to ensure that any residual contamination will be permeable to steam or disinfectants, since there is a critical thickness of biofilms [25, 28]. The rapid spread of laparoscopic methods of surgery since the late 1980s has necessitated a reappraisal of the process of preparing sterile supplies in terms of both decontamination and instrument wear. By now, minimally invasive techniques are widely applied in most advanced societies. In 1996, two-thirds of all cholecystectomies and onethird of all appendectomies performed in Germany were done using minimally invasive methods [9]. Of the 1.8 million cholecystectomies performed annually in the United States, 8090% are now done laparoscopically. In addition, minimally invasive techniques are used in 3040% of the 200,000 thoracic procedures performed annually, 10 15% of the 1.5 million appendectomies, 1015% of the 2.0 million hernia repairs, and 510% of the 1.8 million hysterectomies. According to Malchesky et al., the cleaning process itself may involve many processes, but the most important outcome is the removal of protein material [15]. The intensive use of tubular instruments in this type of surgery requires a more systematic appraisal of their surgical and hygienic properties. Videoscopic surgery is highly dependent on technical reliability, and a switch to conventional methods of surgery due to technical failure can never be ruled out. So far, only a limited number of clinical studies have dealt with technical failure or examined the hygienic reuse of sterile instruments [2, 7, 14]. These reports

389

Fig. 1. Sterile supply processing (see [8]): manual and automated steps from dismantling to cleaning to sterilization.

have generally focused on surgical complications, such as the accidental puncture of blood vessels when introducing the trocar and intestinal bleeding [5, 6, 11, 20, 21, 24]. In the past, hospitals did not routinely examine instruments for wear and tear and residual contamination according to quality assurance criteria. When instruments are steam-sterilized, technical parameters can be checked using biological, chemical, and physical indicators (test spores, Bowie-Dick test, temperature, pressure); however, there are no indicators available to monitor cleaning, which represents the single most important decontamination stage. Therefore, hospitals had to rely solely on visual checks. The physical and chemical documentation and evaluation of cleaning parameters in the processing of sterile supplies is an area that has been ignored largely; at best, it has been a spinoff of test systems created by the food industries [18, 22, 23, 27]. Furthermore, it is important for any evaluation of surgical functionality and hygienic suitability to be quantifiable. A test protocol that yields clinical data on medical products should be devised to establish manufacturer liability, as suggested by the Medical Devices Directive (MDD 1998), which differentiates only between the liability of manufacturers and operators [17]. In recent years selected laparascopic instruments were examined for quality control reasons at the Moabit Hospital to determine their functional and hygienic suitability for the entire cycle, from the operating table to the Central Sterile Supplies Department (CSSD) [8]. Experience gained from 6,000 surgical laparoscopies performed during the period 199096 enabled the CSSD and the first surgical ward to carry out practical experiments in conjunction with the Free University of Berlin and a number of companies involved in the manufacture of medical products (Surgical Instruments Working Group). The objective of this work was to collate the results of a clinical cohort study involving three test trays subjected to 100 cycles each and evaluate the results in terms of instrument suitability (handling, functionality, economy, reusability).

Materials and methods

After use in an operation, the instruments requiring sterilization were sorted into suitable dry containers that were then sealed. No manual precleaning by the theater staff was allowed either during or after the operation. In the case of tubular laparoscopic instruments, although early rinsing of lumens would be beneficial to the cleaning process, disinfection, in particular with aldehydes, was avoided so as not to complicate intramural cleaning. Aldehydes promote the fixation of proteinaceous material, the removal of which is the main part of the cleaning process. The water-tight containers were then transported on trolley to the CSSD several times a day using elevators and road vehicles. In the CSSD, the instruments were manually sorted, disassembled wherever possible, and processed in an ultrasonic bath (except for laparascopes and other fragile components) before subsequently being arranged in or connected to the inserts of a washer-disinfector using the necessary adapters. During cleaning, the instruments, which are generally made from chromium-nickel steel, are in contact with water, electrolytes, and chemical detergents. For the subsequent sterilization process, temperatures of 134C and pressures of up to 3.1 bar are reached (Fig. 1). In compiling the criteria for instrument suitability, the instrument circuit should be documented in its entirety, with all its properties and on a continuing basis, including any irregularities that may occur. First, the instrument stock and any instances of failure of the laparoscopic instruments were recorded. This log contained information regarding the type of instruments, time of use, and parameters of the cleaning process (tray contents, machine identification, cleaning program, trolley code, type of detergents). The selection of instruments was restricted to steam-sterilizable mechanical and electromechanical components that are typical for the surgical techniques required for laparoscopic cholecystectomy and regularly used at the sterile surgeon/patient interface. The criteria of interest were as follows: Instrument design (functionality and handling at the operating table and in the CSSD, disassembly, cleaning, maintenance, mounting) Frequency of failure (fragility, separable instruments) Reusability (susceptibility to soiling, cleanability, ability to withstand wear and tear) Economic aspects (as compared with those of disposable instruments with identical function) Three test trays with new laparoscopic tracer instruments (Karl Storz, Tuttlingen, Germany) were made up for 100 cycles; the instruments included straight scissors, straight traumatic (insulated) and atraumatic grasping forceps, monopolar hooks, and grasping forceps. Each instrument was indelibly marked, and no transfer of instruments between trays was allowed. These precautions allowed us to document any individual signs of wear and tear on each instrument or section (mileage). A repair index

390

Fig. 2. Flow chart for the individual instruments in the three test trays during 100 clinical cycles from laparoscopic operating theater to CSSD: prospective instrument study 199596.

was then assigned to the assembled instrument of which the repaired components were a part (Fig. 2). The degree of postoperative visible soiling and the visible cleaning results achieved were entered for each cycle on an instrument tracer slip. Three degrees of postoperative soiling (heavy, medium, light) and soiling after cleaning (heavy, light, clean) were distinguished. The information included all operative parameters (e.g., ultrasound mobilization, detergent, washer-disinfector). The test to determine the extent of residual contamination was performed after 100 cycles by means of elution of the surfaces of the instrument sections (handle, working tip, lumens) for 10 min in 5 ml of the common detergent sodium dodecyl sulphate solution (SDS). The solution (eluate) underwent photometric analysis to detect protein content in the form of primary amino groups using the highly sensitive modified OPA method (reaction of tho-thaldialdehyde in the presence of N, Ndimethyl-2,3 mercapto ethyl ammonium chloride, a thiol component with and terminal amino groups of proteins). In a parallel test, the eluate was also checked with the standard erythrocyte quick-test sticks used in the diagnosis of microhematuria to detect erythrocytes (pseudoperoxidase reaction) [18, 19]. A comparison of the costs of disposable and reusable instruments used in similar procedures (scissors, three pairs of grasping forceps, four trocars, clip applicator) was performed for 1996. Sterile processing was disregarded as waste disposal, because laparoscopic operations accounted for <30% of all operations and disposal methods differ considerably from site to site.

Results Most of the repairs were related to the sharpness of scissor blades andto a lesser extentto fractures on the blades or tongue sections that might lead to a risk of sections breaking off and falling into the peritoneal cavity. The occurrence rate was <4%. Wear was most clearly evident on the insulation after several cycles of use. Charred or open-circuited electromechanical components represented a second source of defects that indicated improper use of the instrument. The loss of components and damage to them has become more significant since the introduction of separable instruments. A similar phenomenon was previously observed with trocars. The repair index of the tracer instruments processed in the test inserts was much lower than that recorded for the retrospective survey [9]. Due to the statistically low incidence rate, we found no particular clusters of defects needing repair in the three test trays after 300 cycles; these defects can be ascribed to the frequency and nature of usage (Table 1). Interestingly, only

185 pairs of scissors (62%) were used in 300 laparascopic operations: 61 (tray 1), 59 (tray 2), and 65 (tray 3). Traumatic grasping forceps were used in a total of 235 operations (78%): 80 (tray 1), 78 (tray 2), and 77 (tray 3). Straight atraumatic grasping forceps were used 228 times (76%): 77 (tray 1), 68 (tray 2), and 83 (tray 3). The curved atraumatic grasping forceps were used relatively infrequently; in all, they were used 81 times (27%): 28 (tray 1), 27 (tray 2), and 26 (tray 3). Bipolar grasping forceps were used 80 times (tray 1), 77 times (tray 2), and 81 times (tray 3)hence, in all, 238 times (79%). Due to an instrument mixup, figures on the actual frequency of use were not reliable and were therefore not evaluated. Alternatively, three ceramic monopolar hooks were used. These remained free from damage throughout the 100 test cycles and are still in circulation (62 operations during the study; 21%). The evaluation of postoperative contamination after 100 cycles and residue visible to the human eye after cleaning in a washer-disinfector provided interesting findings for the individual components, such as working tips, tubular sections, and handle pieces (Fig. 3). There were significant differences that could be ascribed to instrument design, function (thermal coagulation), influence of ultrasound (US), and water flow of the pump (automated device). Dismantable instruments were easier to clean; thermal coagulation made it difficult to judge the surface and to clean it safely. Ultrasound mobilizes soils and enables their removal. Pump pressure influences cleaning, as compared with similar automated cleaning systems. The differences between insulated and noninsulated tubes were not significant, possibly due to differences in the type of surface (metallic vs black). Residual contamination detected via SDS elution and the chemical-photometric OPA method of analysis was similar to the residue contained in fingerprints. It was recorded on eight of the 36 eluted instrument sections used in clinical operations. In five cases, the contaminated areas consisted of the inner surfaces of various instrument tubes. In four of these cases, a quick-test using sticks yielded evidence of the presence of hemoglobin. On balance, the results of a comparative study of the

391 Table 1. Wear and tear and contamination of three test trays (handle, tube, working tip) of tracer instruments each subjected to 100 cycles Instrument tray no. Tray 1 Scissors, straight Forceps, sharp isolated Forceps, blunt Forceps, bent and blunt Tray 2 Scissors, straight Forceps, sharp isolated Forceps, blunt Forceps, bent and blunt Tray 3 Scissors, straight Forceps, sharp isolated Forceps, blunt Forceps, bent and blunt OT cycles 61 80 77 28 59 78 68 27 65 77 83 26 Repair 4 3 1 2 1 5 3 Working tip 4 1 1 1 3 1 Tube 1 1 1 1 Handle 1 1 1 Contaminationa Handle after 36 cycles Tube Tube Tube Working tip after 19 cycles Tube

a Elution was executed in 37 of 56 parts of clinically used tracer instruments (66%). Six of 37 mechanical instrument parts were contaminated (16%). For methodical reasons, two contaminated parts of a bipolar forceps are not included (tube, handle)

commercial costs of disposable and reusable instruments (excluding manufacturing, processing, and disposal costs) were overwhelmingly in favor of reusable instruments. Considering the current prices and actual failure rates during 1996 for the relevant items (scissors, three grasping forceps, four trocars, Veress cannula), reusable instruments were cheaper by a factor of >10. Discussion During their useful life in the operating theaterCSSD operating theater cycle, surgical instruments spend only a brief time in the hands of surgeons. Their fitness for use must therefore be measured not only in terms of operational safety and precision of handling but also safe transportation, simple disassembly, cleanability, and quality of materials. Wear and tear is unacceptable if it impairs surgical functionality. The replacement of (sterile) instruments (or parts thereof) cannot be recommended in the dim light of a laparoscopic operation. The practical requirements of the surgeon must be weighed against the need for hygienic processing. Hence, although instrument separability facilitates cleaning, it has in part resulted in impaired transmission of muscular force and reassembly errors. Functionality and precision are paramount to the surgeon, whereas separability and durability (availability) are important in hygienic, logistical, and economic terms. Suitable instrument design and the use of composite materials still pose a challenge to instrument manufacturers, although great progress has already been made in these areas. The results of this prospective study confirm the existing repair statistics that we collected for all instrument cycles between 1990 and 1996 (although the total population of instruments available for the former survey was unknown). A shift toward an increased loss of parts and incorrect reassembly was noticed. The influence of frequency of use on individual instruments is negligible, as has already been demonstrated [8].

Together with the pH value of detergents and disinfectants, the quality and texture of the surface (smooth, homogeneous) has a major effect on corrodibility. Experience has shown that tubular sleeves require additional cleaning with a brush (with no metal wires) because some (visible) traces of detritus remain on the walls of the shaft and on fulcrum joints and blades despite the high rate of flow of the cleaning solution (15 L/min). It seems to be important to have a turbulent, nonlaminar flow at sites that are difficult to access. HF instruments used to carry an electric current should generally be both pre- and postcleaned. Many instances of damage testify to the improper use of delicate and fragile instruments as a result of general rough handling and as a consequence of separability. The degree of cleaning that can be achieved depends on the nature of the instrument in questione.g., a pair of scissors is easier to clean than serrated forceps. A comparison of cleaning results for the functional section of instruments (scissors/ bipolar forceps) after 100 cycles clearly shows the problems associated with composite materials (insulation sleeves), as well as the thermomechanical load to which these instruments are subjected and its effects on residual contamination. The interrelationship between soiling and thermal adhesion to insulating material makes cleaning unpredictable. There is no justification for reverting to the use of only disposable instruments, given the vast number of successful minimally invasive operations performed over the last decades. There is no precedent for adopting such a policy for economic reasons, as we and others have already proven [8, 13]. Furthermore, the quality of disposable instruments is generally poorer than that of reusable instruments because cheap mass-production methods demand compromises in terms of materials and dimensional accuracy. We believe that a number of high-quality disposable items might well be reclassified as reusable instruments. However, this is a matter for the manufacturer, since the provision of CSSD-treated instruments by a hospital is legally tantamount to placing a new medical product on the

392

Fig. 3. Profile of cleaning efficacy (after operation/after automated cleaning) of working tips. Cleanliness was assessed visually as one of three grades. After cleaning: , Heavy soiling; , light soiling; , clean. A Influence of instrument design on cleanability (percentage): scissors compared with blunt (atraumatic) forceps. B Influence of instrument design on cleanability: blunt (atraumatic) forceps compared with monopolar hooks and bipolar forceps. C Ultrasound mobilization and cleanability: blunt (atraumatic) forceps.

market [4, 17]. Disposable instruments can be useful for specific tasks, such as clip stapling or as a reserve pair of sharp scissors. This point is particularly valid in view of the fact that disposable scissors were possibly used in every third operation in this prospective study.

For reusable laparoscopic instruments, the cost per operation is dictated by the absolute number of inserts and available trays (acquisition costs). Faulty but repairable instruments are typically discarded more readily when a repurchase has been announced. Careful monitoring of trays

393

and inserts helps to reduce the repair incidence rate, as indeed the low repair index in the test inserts (4% vs 6%) shows [8]. Reuse, however, only makes sense if the instruments in question can be cleaned reliably, since comprehensive sterilization of all instrument surfaces (including lumens) cannot be guaranteed. Plastic coating used as electric insulation material also poses a problem in terms of the cracks and crevices that generally occur after a number of sterilization cycles. If they are properly positioned and adapted to the nozzles, even trocars can be processed in a washer-disinfector. Because cleaning is the single most important step in reducing microbial count, it must be quantifiable (separable instruments or ones with an irrigation connector, decontamination indicators). Proneness to soiling and cleanability are closely relateda fact which should be heeded by instrument development engineers. The clinical relevance of the detection of protein residue needs to be studied in further experiments. Congealed organic surface detritus is very nearly completely dissolved by the solvent SDS (in this case, >90% recovery rate). The modified OPA method is reproducible, sensitive, and specific to amino groups typically present in proteins [18, 19]. The necessary wet chemical stages, including photometric extinction measurements, involve potential sources of error (preparation of the solution, dilution, transfer, batch errors), as does any other method in this field [10]. Although the exact quantity of contamination is generally unknown and its actual constituents vary, elution and detection using the OPA method is easy to perform. With this process, the instrument to be examined remains in the operating theaterCSSDoperating theater cycle. However, because of the thousands of instruments that pass through CSSD daily, the OPA method could only be performed on random samples of selected surgical instruments. Proteinaceous material, the most important residual surface adhesion, is not confined to laparoscopic instruments [8]. In all the cases observed in this survey, only small quantities of contamination were found. The clinical relevance in terms of clusters of postoperative complications such as impaired wound closure or fever has not yet been proven. Evidence of a causal link to a contaminated instrument requires that suspicion be aroused and tests carried out. Given the praxis of antibiotic prophylaxis and the generally low infection rate associated with laparoscopic operations, we cannot conclude that the degree of instrument sterility is currently insufficient [6, 8, 14, 27]. One of the very few reports of possible cross-infections describes an outbreak of wound infection caused by contaminated bone drills used in podiatric surgery [26]. The instruments now used in minimally invasive endoscopic surgery are both modular and reusable for almost all functions and for a wide range of operations; they have an external diameter of 3 mm. There are only a few specialized functions that require the use of disposable instrument sections, which in turn requires that the instruments be separable. In these cases, greater precision in the transmission of force must also be achieved. Discussions of the economic and environmental aspects of disposable, semidisposable, and reusable instruments are likely to be increasingly concerned with the use of individual components (for example, the blade of a pair of

scissors or a handle section); therefore, documented tests need to be done in the real-life working environment of the hospital. Test soiling and contamination that reflect field conditions must be studied under scientific laboratory conditions. In addition to laparoscopic instruments, we are seeing an increasing use in therapeutic fields of sensitive and thermally sensitive intelligent instruments ranging from gastroscopes and coloscopes to dental systems and the microinstruments used in neurosurgery or in ear, nose, and throat treatment diagnostics. These instruments could also be the source of nosocomial infections [26]. A quick-test device to provide an assessment of the cleaning stage as part of the overall decontamination effort would be highly welcome but it is more difficult to develop than the chemobiological bioindicators (test spores) used in steam sterilization, since the main problem with the cleaning process is the localization of detritus and layer thickness. It is therefore necessary to run comprehensive correlation tests between clinical and laboratory contamination, such as those initiated by the Surgical Instruments Working Group, Berlin. A multicenter study is currently in progress to collect data on the amount of residual contamination on instruments after processing.

References
1. Arbeitskreis Instrumenten-Aufbereitung (1997) InstrumentenAufbereitung richtig gemacht. Selbstverlag 6., Auflage, Tuttlingen 2. Bittner R (1995) Die laparoskopische Hernioplastik (TAPP): Komplikationen und Rezidive bei 900 Operationen. 3. Bundesgesundheitsblatt (1992) Anforderungen der Hygiene an die Aufbereitung von Medizinprodukten. Anlage zu Ziffer 7 der Richtlinie fu r Krankenhaushygiene und Infektionspra vention, Bundesgesundheitsblatt 35: 642 4. Canadian Healthcare Association: Holubitsky DJ, et al (1996) The reuse of single-use medical devices: guidelines for healthcare facilities. CHA Ottawa, American Association for the Advancement of Medical Instrumentation (AAMI), Arlington VA 5. Champault G, Cazacu F (1995) Laparoskopische Chirurgie: schwere Verletzungen durch Einsatz des Trokars. Analyse von 103 852 Eingriffen einer franzo sischen Umfrage 1994. Akt Chir 30: 233236 6. Collet D, Edye M, Pe rissat J (1993) Conversions and complications of laparoscopic cholecystectomy: results of a survey conducted by the French Society of Endoscopic Surgery and Interventional Radiology. Surg Endosc 7: 334338 7. Des Coteaux J-G, Poulin EC, Lortie M, Murray G, Gingras S (1995) Reuse of disposable laparoscopic instruments: a study of related surgical complications. Can J Surg 38: 497500 8. Fengler TW, Pahlke H, Kraas E (1998) Sterile and economic instrumentation in laparoscopic surgery. Surg Endosc 12: 12751279 9. Gashinger (1997) 10. Kelsey JC (1972) The myth of surgical sterility. Lancet 2: 13011303 11. Kennedy RJ, Clements WDB, Diamond T (1995) Cystic duct laceration by metallic clips: a cautionary note. Br J Surg 82: 15401543 12. Kirk B (1996) Entwicklungen bei biologischen und chemischen Sterilisationsindikatoren. Zentr Steril 4: 297307 13. Lefering R, Troidl H, Ure BM (1994) Entscheiden die Kosten? Einweg- oder wiederverwendbare Instrumente bei der laparoskopischen Cholecystektomie? Chirurg 65: 317325 14. Manger T, Fahlke J, Settmacher U, Zanow J, Lippert H (1994) Komplikationen bei 1000 laparoskopischen Eingriffen: eine prospektive Studie zur Fehleranalyse. Min Invas Chir 2: 6480 15. Malchesky PS, Chamberlain VC, Scott-Conner C, Salis B, Wallis C (1995) Reprocessing of reusable medical devices. ASIO J 41: 146151 16. Marshburn PB, Rutala WA, Wannamaker NS, Hulka JF (1991) Gas and steam sterilization of assembled and disassembled laparoscopic equipment: microbiological studies. J Reprod Med 36: 483487

394 17. Medizinprodukte-Gesetz (1996) Handbuch fu r Medizintechnik. Ecomed-Verlag, Landsberg berpru 18. Michels W, Frister H, Pahlke H, Fery R (1996) U fung der Reinigung minimalinvasiver Instrumente nach maschineller Dekontamination. Hyg Med 21: 324330 berpru 19. Michels W (1997) Bewertung eines Schnelltests zur U fung des Reinigungserfolgs aufbereiteter chirurgischer, minimal-invasiver Instrumente. Hyg Med 22: 173184 20. Montz FJ (1995) Complications of laparoscopic ports. Min Invas Ther 4: 310 21. Morgenstern L, McGrath M, Carroll B, Paz-Partlow M, Berci G (1995) Continuing hazards of the learning curve in laparoscopic cholecystectomy. Am Surg 161: 914918 22. Nystro m B (1981) Disinfection of surgical instruments. J Hosp Infect 2: 363368 23. Ojaja rvi J (1993) Grundlagen der Dekontamination. Zentr Steril 1: 277282 24. Oza KN, ODonnell N, Fisher JB (1992) Aortic laceration: a rare complication of laparoscopy. J Laparoendosc Surg 2: 235237 25. Rioufol C, Devys C, Cachefo A, Meunier G, Forestier AV, Perraud M, Goullet D (1996) Bakterielle Biofilme und Endotoxine. Zentr Steril 4: 143150 26. Rutula WA, Weber DJ, Thomann CA (1987) Outbreak of wound infections following outpatient podiatric surgery due to contaminated bone drills. Foot Ankle 7: 350354 27. Rutula WA, Gergen MF, Jones JF, Weber DJ (1998) Levels of microbial contamination on surgical instruments. Am J Infect Control 26: 143145 28. Schmidt R (1995) Sterilita t: ein materialwissenschaftliches Problem? Zentr Steril 3: 7585

Surg Endosc (2000) 14: 382387 DOI: 10.1007/s004640020010

Springer-Verlag New York Inc. 2000

The Tu bingen balloon

A new method for adjusting the tension of the fundic wrap during laparoscopic Nissen fundoplication
D. Kalanovic, G. F. Buess, J. Kayser, B. Mentges, K. Roth, H. Raestrup, L. Tijerina, H. Kaczorowski
Section for Minimally Invasive Surgery, University of Tu bingen, Tu bingen, Germany Received: 9 July 1999/Accepted: 2 August 1999

Abstract. An adequate fundic wrap is fundamental to the success of conventional and laparoscopic Nissen fundoplications. Nevertheless, up to now there has been no standardized method for the surgeon to determine intraoperatively the width and tension of the fundic wrap according to objective criteria. With the support of Ru sch (Kernen, Germany), we developed a measurement balloon for use in laparoscopic Nissen fundoplication. The balloon allows the surgeon to define the width of the wrap and predetermine its length, as well as to measure its tension. Depending on the measured balloon pressure, the surgeon can perform fundic sutures more or less tightly. On the basis of 41 fundoplication model tests, we found that a fundic wrap typically described as loose and floppy produced a balloon pressure of 5060 mmHg. In 10 laparoscopic Nissen fundoplications on domestic pigs, we were able to adjust the fundic wrap intraoperatively to a balloon pressure in this range (mean; 53.5; SD; 2.25). After the optimal intraoperative balloon pressure in humans has been investigated in a prospective study that is in progress, the Tu bingen balloon is expected to serve as an instrument for quality assurance in reflux surgery. Key words: Antireflux surgery Gastroesophageal reflux Intraoperative adjustment Laparoscopic fundoplication Nissen fundoplication Pressure recording Quality assurance

passing two fingers between the esophagus and the wrap [11]. The wrap must be loose and floppy. In the hands of an experienced surgeon a 91% success rate over a 10-year period can be achieved with this method [3]. But what is loose and floppy? And how can this (subjective) description be applied to laparoscopic fundoplications? The surgeon encounters two specific difficulties in laparoscopic Nissen fundoplication. The first is loss of palpation, which makes it difficult to assess the width and tension of the fundic wrap. The second problem concerns the objective measurement of the dimensions of the wrap via the endoscopic image. Long-term studies have shown that the length of the wrap should not exceed 23 cm [5, 10]. Up to now there has been no standardized method that gives the surgeon objective information intraoperatively about the width, tension, and length of the fundic wrap. We have developed, with the support of Ru sch (Willy Ru sch AG, D-71385 Kernen, Germany), a measurement balloon (the Tu bingen balloon) that may help to resolve the problem.

Materials and methods The Tu bingen balloon

We have evaluated and tested different materials and balloon designs. The final version of the Tu bingen balloon set is shown in Fig. 1. It consists of the measurement balloon, an introduction device for a 10-mm trocar, a Y piece, a special bolt-syringe, and a long, thin measurement tube. Figure 2 shows details of the balloon design. The balloon has a saddle shape in side view; in cross section, it is sickle-shaped. The body of the balloon is made of a single layer of yellow latex. The two edges of the saddle have an additional layer of blue latex. The length of the balloon is 4 cm. The length of the yellow section, which constitutes the measurement zone, is 2.5 cm. The tubes of the measurement system have an overall length of 140 cm and a diameter of 1.4 mm. Figure 3 shows a schematic drawing of the complete measurement system. A syringe is included in the set and comprises an integrated bolt to guarantee the exact application of 17 ml of air. The components F, G and J are not part of the Tu bingen balloon set.

Whether for conventional or laparoscopic Nissen fundoplication, a properly performed wrap is a decisive factor in the success of the operation [7]. In open surgery, a 2-cm wrap is formed over a large stomach tube. At the end of the operation, the surgeon estimates the width of the wrap by
Correspondence to: D. Kalanovic

383 The proximal part of the stomach and the abdominal esophagus were exposed. The abdominal esophagus was dissected, and the anterior and posterior vagus nerves were identified. Behind the esophagus, the right and the left crura of the diaphragm were exposed. With further dissection, a window was established behind the esophagus for passage of the wrap. The upper part of the great curvature was mobilized, and the short gastric vessels were dissected. A curved laparoscopic grasper was positioned behind the esophagus, and the fundus at the back wall was grasped. This part of the fundus was pulled through the retroesophageal window. A second grasper approximated a fundic part from the stomach front wall. The two parts of the fundus were sutured to create the wrap. After every fundic suture, a measurement was made with the Tu bingen balloon.

Measurement with the Tu bingen balloon

At the start of the operation, the long measurement tube (Fig. 3; E) was connected to the pressure module of the anesthesia monitor. After the first fundic suture, the introducer sheath containing the balloon was placed through the lower left trocar. The balloon was brought into the space between the esophagus and the fundic wrap. Figure 6 shows a schematic drawing of the measurement position of the Tu bingen balloon. The concave side of the balloon is lying on the esophagus; the convex side is directly below the fundic suture line. The esophagus contained a 35-Fr stomach tube during the measurement. The deflated balloon was placed in the described measurement position (Fig. 7A) and then filled with 17 ml of air (the complete volume of the delivered bolt-syringe). Due to the saddle shape, the fundic wrap was held between the two blue borders of the balloon (Fig. 7B). After the correct position of the balloon was checked, a measurement was performed. Depending on the size, width, and tension of the fundic wrap, the pressure inside the balloon increased a given amount. This balloon pressure could be read on the pressure curve of the connected manometer (anesthesia monitor). The pressure after every suture was noted or printed out. The balloon can be left in place during suturing (Fig. 7C). The fundoplication model tests revealed that a wrap performed using established techniques [2, 11] and considered loose and floppy produced a balloon pressure between 50 and 60 mmHg. Therefore, we tried during the animal study to adjust the wrap to a balloon pressure between these values. Any suture that produced a balloon pressure >60 mmHg was removed and replaced with a less tight suture. With the third suture, the balloon pressure had to be in the range of 5060 mmHg. After the laparoscopic fundoplication was finished, we opened the pig to assess the wrap that had been achieved.

Fig. 1. Tu bingen balloon set (Ru sch, Kernen, Germany).

Before it can be introduced into the abdomen, the balloon is placed inside an introducer sheath (Fig. 3, B, C). The introducer sheath has a compressible net (Fig. 3, B1) at its distal end. After the balloon is pulled inside the net, the movable cover sleeve (Fig. 3, C) is pushed over the net. By that means, the balloon is compressed to an outer diameter of 10 mm during the introduction procedure (Fig. 4).

Evaluation in fundoplication models

During the development and testing of the prototype, two models for testing the fundoplication were used. The first was a rigid wrap model simulated by pipe clamps of different diameters. This model was used to evaluate the suitability of the balloon material for pressure measurement, the influence of the size of the intraesophageal tube on the measurement, and the influence of the pneumoperitoneum on the measurement. The second fundoplication model consisted of an ex vivo pigs stomach with esophagus from a slaughterhouse integrated in the Body Form laparoscopic trainer (Limbs & Things, Bristol, England). We used the models to evaluate the design of the balloon and the range of pressure changes inside the balloon during a fundoplication. Table 1 shows the requirements and specifications for the measurement balloon that were defined during the development phase. In the early stages of development, a blood pressure hand manometer was used for pressure measurements. In the final version, the pressure module of the anesthesia monitor was used for this purpose (HP M 1006 Hewlett-Packard). We performed 31 open and 10 laparoscopic Nissen fundoplications using established techniques [2, 11] in this model. After each fundic suture, a measurement was made with the balloon. From the results of this model testing, we have defined a target pressure for the intraoperative adjustment of the wrap during the animal operation.

Results Fundoplication models During the prototype testing using the rigid wrap model and pig stomachs, it became clear that the balloon has to meet a number of requirements (Table 1). These findings led to the final version of the balloon. The only modification after OR testing was the addition of a strap (Fig. 3, A1) to the balloon to facilitate intraoperative handling. It was found that the pressure changes inside the balloon were most sensitive when a balloon volume of 17 ml and a 35-Fr stomach tube were used. The fundoplication model tests revealed that a fundic wrap performed using established techniques [2, 11] and considered loose and floppy produced a balloon pressure between 50 and 60 mmHg (with the CO2 pneumoperitoneum at 15 mmHg). A balloon pressure >70 mmHg was associated with a rather tight wrap. A pressure <40 mmHg indicated a very loose fundoplication. Physical evaluation The physical characteristics of the measurement set-up were evaluated in cooperation with a physicist in our section. We found that:

Animal operations
The goal of the animal operations was to investigate the following parameters: The practicability of the measurement under operative conditions The range of pressure changes inside the balloon between each fundic suture The suitability of the measurement balloon for adjusting the fundic wrap to a defined pressure All operations were carried out on domestic pigs (4555 kg, female) under general anesthesia with volatile narcotics (isoflurane, nitrous oxide) in endotracheal intubation (animal study approval C2/93, C3/95 by the Regierungspraesidium Tu bingen). Preoperatively, a 35-Fr stomach tube was placed in the esophagus. The CO2pneumoperitoneum was at 15 mmHg. Figure 5 shows the positioning of the trocars. The lower left trocar contained a 10-mm atraumatic grasper to retract the stomach. This instrument was removed after the first fundic suture and replaced by the Tu bingen balloon introducer sheath.

384

Fig. 2. Schematic drawings of the Tu bingen balloon in side view (A) and in cross section (B). The length of the yellow part is indicated by an asterisk. Fig. 3. Schematic drawing of the complete measurement system. Components: Balloon (A) with strap (A1); introducer sheath (B) with compressible net (B1) and movable cover sleeve (C); tube for pressure measurements (D) with Y piece (D1); long tube end for pressure measurements (E); sterile connection (F), which fits onto the pressure transducer (G) of the anesthesia monitor (J); 17-ml syringe with bolt (H). Fig. 4. The introducer sheath with the balloon inside, ready for passage through a 10-mm trocar.

Table 1. Specifications for the measurement balloon defined during the development phase The material must be sensitive for pressure measurements The balloon must fit exactly into the space between the esophagus and the sutured fundus (sickle-shaped) The balloon must be saddle-shaped to prevent slippage The balloon must define an exact measurement zone by its shape and color The material must be able to withstand the necessary mechanical forces The balloon and its introducer device must fit in a 10-mm trocar

constantly low because the tube inside the esophagus was misplaced. A measurement was performed only when both blue borders of the balloon were visible. Thus, none of the wraps exceeded 2.5 cm in length. We did not observe a significant influence on the balloon pressure as a result of anesthesia or surgical manipulation in the region of measurement. The total number of cases evaluated was 10. Results of the measurements. The balloon pressure after the first suture was in the range 1482 mmHg (SD, 21.6); after the second suture, it was 4175 mmHg (SD, 10.9). At the end of the fundoplication (third suture) the pressure was in the range 5056 mmHg (mean, 53.5; SD, 2.25). Therefore, an adjustment to the defined target balloon pressure (5060 mmHg) was feasible. Two characteristic types of pressure curves have been noted: a stepwise rise in pressure after every fundic suture (Fig. 9A) and a relatively high balloon pressure after the first suture due to too tight a wrap (Fig. 9B). In the latter case, the first suture (suture 1) was replaced by a second less tight suture (suture 1a). The wrap assessment in the opened pig confirmed that a fundic wrap that was adjusted intraoperatively to a balloon pressure between 50 and 60 mmHg was loose and floppy according to the traditional criteria [2, 3]. Discussion In the 40 years of research into the efficacy, technique and side effects of Nissen fundoplication, it has always been emphasized that attention must be paid to technical details. Small modifications can have a significant effect on the outcome of the operation. The shortening of the wrap length from 6 to 3 cm dramatically reduced persistent postopera-

The filled balloon (17 ml air) produces a basic pressure of 14 mmHg inside a pneumoperitoneum of 15 mmHg. The influence of the difference between room and body temperature is negligible on the volume and the pressure inside the balloon (Table 2). The compression of air inside the balloon during the measurement is tolerable in the area of practical measurement (Fig. 8). The use of water in the measurement system was abandoned because of air bubble formation in the measurement tubes, which prevented successful measurement, and because of inconvenience in the handling of the system.

Animal operations Practicability of the measurement system. No additional trocars were needed for the measurement procedure in any of the operations. With some experience, the additional time for the adjustment with the Tu bingen balloon was 1015 min4 min for introduction of the balloon and 2 min for each measurement. In two cases, the balloon was punctured by the suturing needle. In one case, the balloon pressure was

385

Fig. 5. Trocar positions: 1, laparoscope; 2, curved grasper/needle holder; 3, multifunction instrument by Buess/Melzer (Wolf, Knittlingen, Germany); 4, liver retraction; 5, stomach retraction/introducer sheath of the Tu bingen balloon. Fig. 6. Schematic drawing of the Tu bingen balloon in the measurement position. OE, esophagus with 35-Fr tube inserted. Fig. 7. A Tu bingen balloon inside the fundic wrap (deflated). B Tu bingen balloon during measurement. The yellow body of the balloon indicates the correct length of the wrap. The saddle design prevents slippage of the balloon. C The balloon can be left inside the wrap during suturing.

tive dysphagia and gas bloat symptoms [10]. Switching to a larger intraesophageal bougie (60 Fr rather than 36 Fr) reduced temporary postoperative dysphagia by half [4]. In the latter study, the difference in tube sizes is only 0.8 cm. DeMeester et al. [3] have shown that their technique of Nissen fundoplication can effectively prevent reflux in 91% of cases over a 10-year period. In the same paper, they describe the moment of wrap assessment as follows: if [the surgeon] feels tight bands over his finger, the wrap is too small . . . if there is excessive space, the wrap is too loose.

Table 2. Calculated relationship between volume, pressure, and temperature inside the balloon Balloon with air filling Room temperature (20C) Body temperature (37C) Difference Combined change of volume and pressure (situation in vivo) 17 ml/14 mmHg 17.34 ml/14.28 mmHg 2%

386

Fig. 8. Relationship between volume and pressure inside the balloon during the measurement process (air filling) Fig. 10. Laparoscopic view of a fundic wrap.

Fig. 9. A, B Typical balloon pressure curves (air filling) during fundoplication in an animal.

In open Nissen fundoplication, the surgeon can judge the correctness of the tactile feedback with his fingers. But in the laparoscopic technique, there is no such tactile feedback, so the surgeon must rely on the visual impression of the finished wrap (Fig. 10). With the support of Ru sch, we have developed a measurement balloon for laparoscopic Nissen fundoplication. It enables the surgeon to determine the length and width of the fundic wrap intraoperatively and to measure the tension of that wrap. We were able to demonstrate that the Tu bingen balloon allows a sensitive intraoperative adjustment of the fundic wrap. Depending on the measured balloon pressure, the surgeon can make the sutures more or less tight. Though the method of measurement is quite simple, there are many physical aspects that can affect the exactness and reliability of measurement and have to be taken into

account. It is crucial that the user of the Tu bingen balloon adhere to the defined measurement set-up. In particular, the use of a 35 Fr bougie and the pressure of the CO2 pneumoperitoneum at 15 mmHg are mandatory. Attention has also to be paid to the gas tightness of the system. The measurement can be performed with a conventional blood pressure hand manometer. But for exact and easy documentation, we propose that a Tu bingen balloon be connected to an anesthesia monitor with a pressure transducer. This method requires close cooperation with the anesthetist or a person in the nonsterile area during the intraoperative measurement. The influence of intraoperative manipulation (dissection, mobilization) and anesthesia on the measurement was a problem for other groups who tried to calibrate the fundoplication with intraoperative esophageal manometry [6, 9]. With our method, this problem does not occur, probably because the Tu bingen balloon measures the pressure directly below the suture line and the muscular tone of the esophagus has no influence on the balloon pressure. With some training, the additional time for adjustment of the wrap is between 10 and 15 min. Ru sch has not yet set a price for the Tu bingen balloon, but because of the inexpensive materials used we expect it to be in the low-price range for catheters. To date, we have only the results from the fundoplication model evaluation and the animal study. The optimal balloon pressure for an operation in humans is currently being investigated in a prospective multicenter trial. In this continuing trial, experienced surgeons perform laparoscopic fundoplications according to established techniques [5]. At the end of the procedures, a measurement is performed with the Tu bingen balloon using the defined measurement setup. The pressure values obtained will be compared with the clinical course of the patients after 6 and 12 months. Only after this will it be possible to define the optimal intraoperative balloon pressure in humans. With the results of this study, the surgeon will then be able to adjust the sutures of the fundic wrap intraoperatively depending on the current balloon pressure. The beginner in the field of fundoplication will have a tool that guides him or her during a crucial part of the operation. The oftenmentioned learning curve [1, 8, 12] of laparoscopic Nissen fundoplication was significantly reduced. This shortened

387

time is even more important now that laparoscopic fundoplication is being performed in less experienced centers. The successful introduction of our measurement system will promote the standardization of the operational technique. It may therefore improve quality assurance in antireflux surgery. References
1. Dallemagne B, Weerts JM, Jehaes C, Markiewicz S (1996) Causes of failures of laparoscopic antireflux operations. Surg Endosc 10: 305 310 2. Dallemagne B, Weerts JM, Jehaes C, Markiewicz S, Lombard R (1991) Laparoscopic Nissen fundoplication: preliminary report. Surg Laparosc Endosc 1: 138143 3. DeMeester TR, Bonavina L, Albertucci M (1986) Nissen fundoplication for gastroesophageal reflux disease. Ann Surg 204: 920 4. DeMeester TR, Stein HJ (1992) Minimizing the side effects of antireflux surgery. World J Surg 16: 335336 5. Fuchs KH, Feussner H, Bonavina L, Collard M, Coosemans W (1997)

6.

7. 8.

9.

10.

11.

12.

Current status and trends in laparoscopic antireflux surgery: results of a consensus meeting. Endoscopy 29: 298308 Johnsson F, Ireland AC, Jamieson GG, Dent J (1994) Effect of intraoperative manipulation and anaesthesia on lower oesophageal sphincter function during fundoplication. Br J Surg 81: 866868 Nissen R (1956) Eine einfache Methode zur Behandlung der Refluxkrankheit. Schweiz Med Wochenschr 86: 590592 Peters JH, Heimbucher J, Kauer WK, Incarbone R, Bremner CG, DeMeester TR (1995) Clinical and physiological comparison of laparoscopic and open Nissen fundoplication. J Am Coll Surg 180: 385 393 Puglionisi A, Asole F, Clemente G, Villani (1984) Effectiveness of intraoperative LES pressure measurement during Nissen-Rossetti fundoplication. It J Surg Sci 14: 9196 Shirazi SS, Schulze K, Soper RT (1987) Long-term follow-up for treatment of complicated chronic reflux esophagitis. Arch Surg 122: 548551 Siewert JR, Feussner H, Walker SJ (1992) Fundoplication: how to do it. Peri-esophageal wrapping as a therapeutic principle in gastroesophageal reflux prevention. World J Surg 16: 326334 Stein HJ, Feussner H, Siewert JR (1996) Failure of antireflux surgery: causes and management strategies. Am J Surg 171: 3639

Letters to the editor

Surg Endosc (2000) 14: 406 DOI: 10.1007/s004640020011 Springer-Verlag New York Inc. 2000

Teamwork in digestive intervention

I enjoyed Professor Cuschieris provocative and pertinent editorial ( To each his own will fall short of the mark in the next millennium) [1]. His thesis that traditional professional structures are not relevant to the new world of minimally invasive therapy has been the theme of my writing and activities for several years [2, 3]. I challenged the gastroenterology community 5 years ago to embrace restructuring in digestive diseases [2], and moved to the Medical University of South Carolina in 1994 to develop a multidisciplinary Digestive Disease Center. Its stated mission is to provide courteous and cost-effective care for patients with digestive disorders, and to provide a springboard for the clinical research and education necessary to enhance it. Our focus is the patient, not the provider. The tools of the center include mutual respect and goals; shared clinical, administrative, and research facilities; and support structures. We have not separated faculty from their traditional academic roots, but instead developed a matrix/umbrella organization to facilitate the necessary collaboration. One key goal (not currently completed) is the creation of a comprehensive multidisciplinary database, to allow all constituents to understand their contributions to the whole. Cuschieri raises the issue of multiprofessional training. Will we be able to persuade conservative professional organizations and examination boards that things must be done differently? My only quibble with Cuschieri (perhaps not surprisingly, since I am a medical interventionist) concerns his statement that the surgeon interventionist remains the most crucial member of such a team, and for this reason alone, should lead it. This assertion is based on his view of surgery as a fallback position, that is open surgical intervention when the procedure fails or serious life threatening complications develop. I cannot see why such a need (or skill) should determine leadership; the skills of leadership are not technical. It could be argued that this role (fallback open surgery) is so important that surgeons with these skills and burdens should be protected from the day-to-day, timeconsuming tasks of leadership and management. At any rate, Cuschieris statement indicates that turf issues still exist even among the enlightened. The way forward is to focus on what works best for our current patients and to provide an unbiased platform for the development of future management through technical innovation, quality improvement, and evaluation research. Different structures will evolve in different countries and environments, often driven by persuasive individuals. It would be helpful to have a forum in which to learn from the triumphs and disappointments of the pioneers. References
1. Cuschieri A (1999) To each his own will fall short of the mark in the next millennium [Editorial]. Surg Endosc 13: 443444 2. Cotton PB (1994) Interventional gastroenterology (endoscopy) at the crossroads: a plea for restructuring in digestive diseases. Gastroenterology 107: 294299 3. Cotton PB (1997) Endoscopy in crisis: the challenge of new technology. Gastrointest Endosc 46: 189191

P. B. Cotton
Digestive Disease Center Suite 210 Clinical Science Building 96 Jonathan Lucas Street Post Office Box 250327 Charleston, SC 29425 USA

Surg Endosc (2000) 14: 408 DOI: 10.1007/s004649900058

Springer-Verlag New York Inc. 2000

What is the appropriate mesh for laparoscopic intraperitoneal repair of abdominal wall hernia?
The manuscript of A. Bickel and A. Eitan [1] about a new technique for laparoscopic repair of ventral hernias with mesh deserves a special comment: First, positioning a 25 25-cm polypropylene mesh into an intraperitoneal position so that it has full contact with the small bowel does not seem appropiate in our opinion. There is enough data in the medical literature (clinical and experimental) pointing out the necessity of a peritoneal covering for the polypropylene mesh to avoid the direct exposure of this material to the intraperitoneal content because of its profound adhesive and fixing properties [3]. This is also the main concept of the transabdominal preperitoneal (TAPP) technique for laparoscopic repair of the inguinal hernia. Also, despite a complete peritoneal covering of the polypropylene mesh, these have been found in the urinary bladder, small bowel, and other anatomic locations. In our experience with more than 160 cases of laparoscopic repair of abdominal wall hernias, we have had the opportunity to discover polypropylene meshes positioned previously by open surgery in a prefascial position that had penetrated through the abdominal wall. Second, incisional hernias frequently are multicavitated and multiperforated. Often, two or more wall defects are found (mean, 3.5 per patient in our series) that not in the same anatomic plane. Over which abdominal wall defect should the center of the polypropylene mesh be fixed according to the technique described in the referred article? Third, the abdominal wall defects in abdominal hernias are formed by skin and the peritoneal sac alone. The positioning of a tack staple in the center of the defect through the polypropylene mesh probably will perforate the skin. We never use this approach even though the mesh that we use (PTFE Dual Mesh Plus) is thicker than polypropylene mesh. Fourth, there are more feasible technique options for a correct positioning of the mesh. In our early cases, we used external knotting for this purpose. Actually, we made a total intraperitoneal fixing in the beginning by placing the first tackers in the different axes of the mesh previously marked. A double crown of tackers around the hernial ring provided a complete fixing of the mesh and a definitive solution for the abdominal wall defect [2]. Fifth, differents technical options have been proposed for placing meshes intraperitoneally: polypropylene mesh with mayor epiplon covering, polypropylene mesh covering by polyglycolic absorbable mesh [4]. Nowadays, we do not think it necessary to use techniques experimentally proved inappropriate. Today a double-coated mesh is available, which causes a minimum of tissue reaction in the peritoneal face, which can be brought in to direct contact with intraperitoneal content without secondary complications (inclusion, perforation). Moreover, this double-coated is completely reperitonized in the first postoperative days. In our experience, with a follow-up period of 6 years, the PTFE Dual Mesh Plus with holes have all the aforementioned characteristics. References
1. Bickel A, Evitan A (1999) A simplified laparoscopic technique for mesh placement in ventral hernia repair. Surg Endosc 13: 532534 2. Carbajo MA, Mart n JC, Blanco I, Cuesta C, Toledano M, Mart n F, Vaquero C, Inglada L (1999) Laparoscopic treatment versus open surgery in the solution of major incisional and abdominal wall hernias with mesh. Surg Endosc 13: 250252 3. Le Blanc KA, Both WV, Whitaker JM, Baker D (1998) In vivo study of meshes implanted over the inguinal ring and external iliac vessels in uncastrated pigs. Surg Endosc 12: 247251 4. Holtzman MD, Purut CM, Reintgen K, Eubanks S, Pappas TN (1997) Laparoscopic ventral and incisional hernioplasty. Surg Endoscop 11: 3235

M. A. Carbajo J. C. Mart n del Olmo J. Blanco

Department of General and Digestive Surgery Medina del Campo Hospital Cta. De Pen aranda de Bracamonte, km 2 47400 Medina del Campo Valladolid, Spain

Surg Endosc (2000) 14: 345348 DOI: 10.1007/s004640020063

Springer-Verlag New York Inc. 2000

Prospective randomized blinded trial of pulmonary function, pain, and cosmetic results after laparoscopic vs microlaparoscopic cholecystectomy
W. Schwenk, J. Neudecker, J. Mall, B. Bo hm, J. M. Mu ller
Department of General, Visceral, Vascular and Thoracic Surgery, Charite Medical School, Campus Mitte, Humboldt University of Berlin, Schumannstr. 20/21 D-10117 Berlin, Germany Received: 30 April 1999/Accepted: 13 August 1999

Abstract Background: The size of laparoscopic instruments has been reduced for use in abdominal video endoscopic surgery. However, it has yet to be proven that microlaparoscopic surgery will actually result in clinically relevant benefits for patients. Methods: Fifty patients were randomized in a blinded fashion to receive either elective laparoscopic (MINI), (n 25) or microlaparoscopic (MICRO) (n 25) cholecystectomy. Pulmonary function (FVC, FEV 1), analgesic consumption during patient-controlled analgesia (PCA), pain perception by visual analogue score (VAS), and the cosmetic result (by the patients self-assessment) were evaluated postoperatively as clinically relevant end points. Results: Age, sex, body mass index (BMI), preoperative pulmonary function, pain perception, and operative time were similar for the two groups. At 8:00 PM on the day of surgery, FVC (MINI: 1.96 L [range, 1.482.48]; MICRO: 2.13 L) [(range, 1.61.2.50)] and FEV 1 (MINI: 1.17 L/sec) [range, 0.871.48]; MICRO: 1.34 L/sec [range, 1.05.2.14] were also similar (each p 0.5). From surgery to the 3rd postoperative day, cumulative PCA morphine doses were comparable (MINI: 0.15 mg/kg bw [range, 0.090.23]; MICRO: 0.21 mg/kg bw [range, 0.100.42]; p 0.4), but overall VAS scores for pain while coughing were higher in the laparoscopic group (406 [range, 358514]) than in the microlaparoscopic group (365 [range, 215427]; p 0.02). The cosmetic result was judged to be slightly superior by the microlaparoscopic patients (10 [range, 910]), as compared to those in the laparoscopic (9 [range, 810]) group (p 0.04). Conclusion: Because microlaparoscopic cholecystectomy has some minor advantages over laparoscopic surgery, it should be considered for use in selected patients.

Key words: Cholecystectomy Gallbladder Laparoscopic cholecystectomy Microlaparoscopy Miniinstruments

In the last decade, laparoscopic cholecystectomy has become the standard therapy for symptomatic cholecystolithiasis in most hospitals. Not only is there is better preservation of pulmonary function with laparoscopic cholecystectomy, but patients also experience less pain and fatigue than is associated with conventional cholecystectomy [5, 13, 14, 16]. The main reason for these advantages seems to be the smaller access to the abdominal cavity afforded by the laparoscopic approach. Recently, the size of incisions required for the laparoscopic procedure has been further minimized by the introduction of 5-mm laparoscopes and 2-mm instruments [6, 8, 9, 15, 19]. Whether this microlaparoscopic approach actually results in clinically relevant benefits for the patient is still questionable [3, 4]. Therefore, we performed a randomized controlled study to evaluate pulmonary function, analgesic requirements, pain perception, and cosmetic results after laparoscopic and microlaparoscopic cholecystectomy.

Patients and methods Hypothesis, end points, and calculation of sample size
The null hypothesis (H 0) of the study was that postoperative pulmonary function is no different following laparoscopic cholecystectomy than after microlaparoscopic cholecystectomy. The alternative hypothesis (H A) was that pulmonary function is different following these two types of laparoscopic intervention. Forced vital capacity (FVC) was the major end point of the study. Minor end points were postoperative forced expiratory volume in 1 sec (FEV 1); postoperative morphine consumption during patient-controlled analgesia (PCA); pain perception as assessed by a visual analogue scale

Correspondence to: W. Schwenk

346 Table 1. Sex, age, sex, BMI, operative time, and preoperative pulmonary function in patients undergoing laparoscopic (MINI) or microlaparoscopic (MICRO) cholecystectomy (data given as median [95% confidence interval]) MINI (n 25) Sex (F/M) Age (yr) BMI (kg/m2) Operative time (min) Preoperative pulmonary function FVC (L) FEV1 (L/sec) 17/8 52 (3855) 22.9 (20.223.8) 70 (6087) 3.3 (3.03.6) 2.6 (2.02.8) MICRO (n 25) 18/7 44 (3257) 21.7 (19.723.5) 70 (6080) 3.7 (3.24.0) 3.1 (2.33.2) p value 1.0a 0.6b 0.6b 0.9b 0.2b 0.2b

BMI, body mass index; FVC, forced vitality capacity; FEV1, forced expiratory volume in 1 sec a Groups compared by Fishers exact test b Groups compared by Mann-Whitney U test

(VAS) during rest, while coughing, and when sitting up; and cosmetic results. The sample size was calculated before the beginning of the study [1]. It was assumed that FVC at 8:00 PM on the day of laparoscopic cholecystectomy would be suppressed by 40 15% of the preoperative value [10]. A difference of 15% in the postoperative FVC between microlaparoscopic and laparoscopic cholecystectomy was considered clinically relevant. This difference can be detected by a two-tailed test with 0.05, 0.1 (power 90%), and 25 patients in each group.

Study population
Patients with symptomatic cholecystolithiasis who were scheduled for elective cholecystectomy were included in the study. Those who required emergency cholecystectomy for acute cholecystitis were excluded. Further exclusion criteria were severe obesity (body mass index [BMI] >32 kg/m2), operative risk >ASA class III, incorrigible coagulopathy or thrombocytopenia, and known alcohol or drug addiction.

day of surgery, and three times per day from the 1st postoperative day until discharge. The patient-controlled analgesia (PCA) device was explained to all patients on the day before surgery. Immediately after surgery, PCA with morphine sulfate was initiated for every patient using an electronically controlled system (PCA-PACOM Injectomat; Fresenius, Germany). The PCA bolus was set to 0.02 mg/kg morphine with a lockout interval of 10 min. A continuous basal rate was not administered. Pain was assessed by the patient on a visual analogue scale within a range of 0 (no pain) to 100 (unbearable pain) during rest, while coughing, and when sitting up in bed. The morphine dose, the frequency of requests for analgesia, and the amount of analgesic boluses administered were recorded three times a day. Patients were not discharged before the 3rd postoperative day so that all measurements could be completed for this postoperative interval. On the 7th postoperative day, all patients were seen again at the outpatient service, spirometry was performed, and pain was assessed during rest and while coughing. All intra- and postoperative complications and mortalities were recorded. A self-assessment of the cosmetic result of surgery was made by all patients at the end of the study using a scale ranging from 1 (poor) to 10 (excellent).

Study design, randomization, and surgical technique

The study was approved by the local ethics committee. Informed consent was obtained from all patients. Preoperatively, the patients were randomized to either laparoscopic cholecystectomy or microlaparoscopic cholecystectomy. All procedures were performed under general anesthesia using the American technique [10]. For laparoscopic cholecystectomy, a 10-mm infraumbilical port (10mm 30 laparoscope), a 10-mm epigastric port (endoscopic hook, dissector, scissors, 10-mm Endoclip [Autosuture Germany, Toenisvorst, Germany]), and two 5-mm right subcostal ports (endoscopic graspers) were utilized. Microlaparoscopic cholecystectomy was performed using a 5-mm infraumbilical port (5-mm 30 laparoscope), a 5-mm epigastric port (endoscopic hook, dissector, scissors, 5-mm Endoclip [Autosuture Germany]), and two 2-mm right subcostal ports (Microsite; Autosuture Germany). In all cases, the gallbladder was extracted via the infraumbilical incision in an endoscopic bag (Endocatch I; Autosuture Germany). The infraumbilical skin incision was enlarged whenever necessary. In all patients, a 12 Charrie ` re Robinson drain was inserted via the lateral subcostal port incision and placed in the subhepatic area until the 1st postoperative day. All patients were operated on by experienced laparoscopic surgeons.

Statistical analysis
Statistical analysis of all data was performed using SAS 6.10 for Windows. Continuous data are given as median (95% Cl) and were compared between groups using the Mann-Whitney U test. Categorical data were compared using Fishers exact test. A p value of 0.05 was considered significant.

Results A total of 50 patients were randomized to laparoscopic (MINI: n 25) or microlaparoscopic (MICRO: n 25) cholecystectomy. One patient from the MINI group had to be converted to conventional cholecystectomy through a subcostal incision because of unexpectedly severe acute cholecystitis. One patient from the MICRO group was converted to laparoscopic cholecystectomy because handling of the acutely inflamed gallbladder was not possible with the 2-mm instruments. The data for both patients were analyzed according to their randomization as intention to treat. Age, sex, BMI, operative time, and preoperative pulmonary function were not different for the two groups (Table 1). There were no intraoperative complications. A superficial wound infection was observed in one patient (4%) in the laparoscopic group. There were no complications after microlaparoscopic choleystectomy. Postoperative suppression of pulmonary function was

Pulmonary function, analgesic consumption, pain, and cosmetic result

Bedside spirometry (Renaissance Spirometer; Firma Puritan Bennett Hoyer, Gra feling, Germany) was carried out with the patient lying in bed and the upper body elevated by 45. Each test was repeated three times and the best of the three results for FVC and FEV 1 was chosen for further analysis [2]. Spirometry was performed preoperatively, at 8:00 PM on the

347

Fig. 1. Forced vital capacity (FVC) after laparoscopic and microlaparoscopic cholecystectomy. Fig. 2. Forced expiratory volume within 1 sec (FEV1) after laparoscopic and microlaparoscopic cholecystectomy.

comparable after MINI and MICRO cholecystectomy. At 8:00 PM on the day of surgery, FVC decreased to 1.96 L (range, 1.482.48) in the MINI group and 2.13 L (range, 1.612.50) in the MICRO group (p 0.5). At the same time, FEV 1 fell to 1.17 L/sec (range, 0.871.48) in the MINI group and 1.34 L/sec (range, 1.052.14) in the MICRO group (p 0.5). During the postoperative period, pulmonary function was similar for the two groups (Figs. 1 and 2). From surgery to the 3rd postoperative day, there was no significant difference in the cumulative morphine dose between the MINI group and the MICRO group (Table 2). Requested, rejected, and given analgesic boluses were likewise not different for the two groups. At 8:00 PM on the day of surgery, the VAS score for pain during rest and when sitting did not differ between the MINI and MICRO groups, but the VAS pain score while coughing was lower in the MICRO group. From surgery to the 3rd postoperative day, there were no differences in the cumulative VAS pain scores during rest and when sitting up in bed, but again the cumulative VAS pain score while coughing was lower in the MICRO group. When contacted at home, patients of both groups judged the cosmetic results as very good; but on a scale from 1 to 10, MICRO patients rated the cosmetic results slightly better (10 [range, 910]) than MINI patients (9 [range, 810]) (p 0.04). When the analysis was repeated without the intention -to- treat approach (i.e., excluding patients who were converted from MICRO to MINI and from MINI to conventional cholecystectomy), there were no relevant changes in the results (data not shown).

Discussion Although some of the advantages of laparoscopic over conventional cholecystectomy have been challenged by a prospective randomized double-blinded study [11], better pulmonary function and less pain are genuine benefits of laparoscopic cholecystectomy that have not been disproved by this trial [18]. Laparoscopic cholecystectomy remains the standard approach for elective cholecystectomy in most hospitals because several prospective randomized studies

have shown relevant benefits for postoperative pulmonary function and pain [12, 17]. Recently, several authors have reported their initial experience with mini-laparoscopic [8, 9], needleoscopic [6], microlaparoscopic [19], and gasless laparoscopic cholecystectomy with mini-instruments [15]. All of them concluded that micro-invasive cholecystectomy was a safe and feasible procedure. Although none of these studies was performed as a randomized controlled trial and some of them even lacked a control group, one of them stated that the potential for decreased recovery and increased cosmesis without untoward outcome makes mini-laparocopic cholecystectomy an attractive option [8]; another claimed that postoperative complaints were reduced after microlaparoscopy [19] and recommended a four-cannula microlaparoscopic approach to make surgery more comfortable and safer [15]. Because the scientific evidence to support these hypotheses is rather scarce, Berci [3] commented in an editorial note that it is not clear that smaller instruments will produce better or safer results, and Bo hm et al. [4] stated that this [microlaparoscopic] regimen must be tested in randomised controlled studies. Golder and Rhodes [7] investigated the influence of the size of the epigastric port (5-mm or 10-mm) on postoperative outcome after 53 laparoscopic cholecystectomies, but they did not use any micro-instruments. Length of operation, pain scores, hospital stay, and time to return to normal activity were similar for the two groups. The authors concluded that unless objective benefits to patients in terms of pain and overall recovery could be shown, a change to smaller ports could not be recommended. Because the results of one prospective randomized double-blind trial of laparoscopic and conventional cholecystectomy [11] suggested that the patients knowledge of the operative technique might influence outcome, we analyzed postoperative pulmonary function, analgesic requirements, pain perception, and cosmetic results in a blinded patient population. When planning the study, we considered the possibility of double-blinding patients, physicians, and nurses to the kind of surgery, but this approach was not feasible within the clinical routine of our department. After some microlaparoscopic cholecystectomies had been performed in a pilot study, it became evident that the learning

348 Table 2. Results of PCA therapy and VAS scores for pain during rest, while coughing, and while sitting up after laparoscopic (MINI) and microlaparoscopic (MICRO) cholecystectomy (data given as median [95% confidence interval]) MINI (n 25) PCA morphine boluses requested 9 (514) rejected 0 (01) given 7 (514) Cumulative PCA MSO4 dose until 3rd POD (mg/kg bw) 0.15 (0.090.23) VAS pain scores at 8:00 PM on the day of surgery during rest 29 (2246) while coughing 73 (6484) while sitting up 68 (5578) Cumulative VAS pain scores until 3rd POD during rest 158 (96228) while coughing 406 (358514) while sitting up 381 (289503) MICRO (n 25) 13 (627) 1 (12) 11 (519) 0.21 (0.100.42) 34 (2745) 57 (5068) 57 (4275) 100 (82192) 352 (219390) 365 (215427) p value 0.4a 0.4a 0.1a 0.4a 0.5a 0.06a 0.2a 0.5a 0.02a 0.1a

PCA, patient-controlled analgesia; VAS, visual analogue scale; POD, postoperative day a Groups compared by Mann-Whitney U test

curve for microlaparoscopic cholecystectomy was minimal. However, it was the impression of all surgeons that handling of the gallbladder was more difficult with the microinstruments, especially in obese patients and those with acute inflammation. During the postoperative course, only minor differences were observed between the groups. Because the postoperative suppression of pulmonary function was similar for both groups, we were not able to reject our null hypothesis. Smaller differences between the groups in suppression of pulmonary function might be detected with a larger sample size, but we do not believe that a difference of <15% has clinical relevance. In summary, elective microlaparoscopic cholecystectomy is a safe and feasible procedure, and conversion to laparoscopic or conventional cholecystectomy is rare. Although the handling of the anatomical structures may be more difficult with micro-instruments, the operative time is not increased when microlaparoscopy is performed by an experienced laparoscopic surgeon. However, postoperative pulmonary function, as well as degree of postoperative discomfort during rest and mobilization, are similar, whereas pain while coughing and cosmetic results are slightly improved by the microlaparoscopic approach. Because of these results, we conclude that microlaparoscopic cholecystectomy performed by experienced laparoscopic surgeons should be considered for selected patients.
Acknowledgments. This work incorporates data derived from the thesis of H. Scheerat, M.S., and R. Motz, M.S.

References
1. Altman DG (1991) Practical statistics for medical research. Chapman & Hall, London 2. Standardization of spirometry1987 update. Am Rev Respir Dis 136: 12851298 3. Berci G (1998) Laparoscopic cholecystectomy using fine-caliber instruments. Surg Endosc 12: 197 4. Svanvik J (1998) Microlaparoscopic cholecystectomythe first 20 cases: is it an alternative to conventional laparoscopic cholecystectomy? (editorial commentary) Eur J Surg 164: 625

5. Frazee RC, Roberts JW, Okeson GC, Symmonds RE, Snyder SK, Hendricks JC, Smith RW (1991) Open versus laparoscopic cholecystectomy: a comparison of postoperative pulmonary function. Ann Surg 213: 651654 6. Gagner M, Henniford BT, Mayes JT, Garcia-Ruiz A, Crook S, Malm J (1998) Needleoscopic cholecystectomy versus laparoscopic cholecystectomy: a comparative study. Surg Endosc 12: 502 7. Golder M, Rhodes M (1998) Prospective randomized trial of 5- and 10-mm epigastric ports in laparoscopic cholecystectomy. Br J Surg 85: 10661067 8. Ikramuddin S, Gourash W, Nguyen N, Luketich J, Schauer P (1998) Mini-laparoscopic cholecystectomy is a safe alternative to the standard laparoscopic approach. Surg Endosc 12: 504 9. Kimura T, Sakuramachi S, Yoshida M, Kobayashi T, Takeuchi Y (1998) Laparoscopic cholecystectomy using fine-caliber instruments. Surg Endosc 12: 283286 10. Kum CK, Eypasch E, Aljaziri A, Troidl H (1996) Randomized comparison of pulmonary function after the French and Amercan techniques of laparoscopic cholecystectomy. Br J Surg 83: 938941 11. Majeed AW, Troy G, Nicholl JP, Smythe A, Reed MWR, Stoddard CJ, Peacock J, Johnson AG (1996) Randomised, prospective, singleblinded comparison of laparoscopic versus small-incision cholecystectomy. Lancet 347: 989994 12. McMahon AJ, Russell IT, Ramsay G, Sunderland G, Baxter JN, Anderson JR, Galloway D, ODwyer PJ (1994) Laparoscopic and minilaparotomy cholecystectomy: a randomized trial comparing postoperative pain and pulmonary function. Surgery 115: 533539 13. Peters JH, Ortega A, Lehnerd SL, Campbell AJ, Schwartz DC, Ellison C, Innes JT (1993) The physiology of laparoscopic surgery: pulmonary function after laparoscopic cholecystectomy. Surg Laparosc Endosc 3: 370374 14. Poulin EC, Mamazza J, Breton G, Fortin CL, Wabha R, Ergina P (1992) Evaluation of pulmonary function in laparoscopic cholecystectomy. Surg Laparosc Endosc 2: 292296 15. Queiroz-Medeiros MJ, Menezes-Silva A, Raposo dAlmeida JR, Vasconcoles Dias H, Nunes JL, Balhuna C (1998) Gasless laparoscopic cholecystectomy with mini-instruments (2mm) video. Br J Surg 85(Suppl 2): 204 16. Schauer PR, Luna J, Ghiatas AA, Glen ME, Warren JM, Sirinek KR (1993) Pulmonary function after laparoscopic cholecystectomy. Surgery 114: 389399 17. Schulze S, Thorup J (1993) Pulmonary function, pain, and fatigue after laparoscopic cholecystectomy. Eur J Surg 159: 361364 18. Squirrel DM, Majeed AW, Troy G, Peacock JE, Nicholl JP, Johnson AG (1998) A randomized, prospective, blinded comparison of postoperative pain, metabolic response, and perceived health after laparoscopic and small incision cholecystectomy. Surgery 123: 485495 19. Watanabe Y, Sato M, Ueda S, Abe Y, Horiuchi A, Doi T, Kawachi K (1997) Microlaparoscopic cholecystectomythe first 20 cases: is it an alternative to conventional LC? Eur J Surg 164: 623625

Surg Endosc (2000) 14: 362366 DOI: 10.1007/s004640000155

Springer-Verlag New York Inc. 2000

Effects of pneumoperitoneum on cardiac autonomic nervous activity evaluated by heart rate variability analysis during sevoflurane, isoflurane, or propofol anesthesia
N. Sato,1 M. Kawamoto,2 O. Yuge,2 H. Suyama,1 M. Sanuki,1 C. Matsumoto,1 K. Inoue1
1 2

Department of Anesthesia, Hiroshima General Hospital, Hatsukaichi-Jigozen 1-3-3, Hiroshima 738, Japan Department of Anesthesiology and Critical Care Medicine, Hiroshima University, Kasumi 1-2-3, Minami-ku, Hiroshima 734, Japan

Received: 22 January 1999/Accepted: 22 March 1999

Abstract Background: The effects of pneumoperitoneum on the activity of the cardiac autonomic nervous system have not been completely understood. Methods: In this study, 45 unpremedicated adult patients who underwent laparoscopic cholecystectomy were anesthetized with either 3.5% sevoflurane, 2% isoflurane, or 8 mg/kg/h propofol (15 patients in each group). The status of cardiac autonomic nervous activity was evaluated by heart rate variability analysis three times: once when the patient was awake, once after induction of general anesthesia, and once after insufflation for pneumoperitoneum. Intraabdominal pressure was maintained automatically at 10 mmHg by a carbon dioxide (CO2) insufflator. For each measurement, electrocardiogram was recorded for 256 s and played back offline to detect R-R intervals. Power spectral analysis of heart rate variability was applied, and the lowfrequency (LF, 0.040.15 Hz) and high-frequency (HF, 0.150.40 Hz) bands of the spectral density of the heart rate variability were obtained from a power spectra of R-R intervals using the fast-Fourier transform algorithm. The HF/ LF ratio also was analyzed. Results: Measurements of heart rate variability in the three groups showed similar change. Although the power of HF, which represents parasympathetic nervous activity, did not change, the power of LF, which represents both sympathetic and parasympathetic nervous activity, decreased during the anesthetized stage and increased during the insufflated stage. The HF/LF ratio, which represents the balance of parasympathetic and sympathetic activity, increased after induction of general anesthesia, and decreased after insufflation. Conclusions: Our results suggest that pneumoperitoneum increases sympathetic cardiac activity. The choice of general anesthetic did not seem to have a major influence on the

change in the cardiac autonomic nervous system after induction of pneumoperitoneum for laparoscopic cholecystectomy. Key words: Cardiac autonomic nervous system Heart rate variability Isoflurane Pneumoperitoneum Power spectral analysis Propofol Sevoflurane

Pneumoperitoneum has significant implications for cardiovascular and respiratory status [3, 23]. However, the effects of pneumoperitoneum on the cardiac autonomic nervous system (ANS) have not been reported. Moreover, it is possible that the effects of pneumoperitoneum on the cardiac ANS differ when patients are anesthetized with different anesthetics. The purpose of this prospective study was to evaluate the effects of pneumoperitoneum on the activity of the cardiac ANS in patients under anesthesia with sevoflurane, isoflurane, or propofol, which are commonly used anesthetics in clinical settings. The cardiac ANS regulates heart rate (HR) variability. Power spectral analysis of the R-R interval in the electrocardiogram is a noninvasive tool that quantifies the contributions of the parasympathetic and sympathetic systems in the ANS [14]. The spectral components can be divided into several frequency ranges. The low frequency (LF) band is mediated by both the parasympathetic and sympathetic systems, and the high frequency (HF) component is mediated mainly by the parasympathetic system. To assess the effects of pneumoperitoneum on the ANS, power spectral analysis was used to evaluate sympathetic and parasympathetic control of HR variability. Methods
This study was approved by our institutional ethical committee for clinical investigation, and informed consent was obtained from 45 adult patients

Correspondence to: N. Sato

363 scheduled to undergo laparoscopic cholecystectomy. The subjects had no significant systemic or cardiopulmonary disease such as diabetes mellitus, arrhythmia, coronary artery disease, or chronic obstructive lung disease. None of the patients was receiving beta-adrenergic-blocking drugs, calcium-channel blockers, digitalis, antihypertensives, or vasodilators preoperatively. By a closed envelope method, patients were allocated to one of three groups: isoflurane, sevoflurane, or propofol group (15 patients in each). Patients received no preanesthetic medication. On arrival in the operating room, patients were placed in the supine position, and routine cardiovascular monitoring was applied, including electrocardiogram (ECG, lead II) and pulse oximetry, and indirect arterial pressure (M1166A, Hewlett-Packard, Palo Alto, CA, USA). After the peripheral intravenous line was secured, lactated Ringers solution was administered at a rate of 5 ml/kg/h. A continuous 256-s ECG was recorded as the awake stage. In the isoflurane and sevoflurane groups, anesthesia was induced with fentanyl (3 g/kg) and thiamylal (3 mg/kg), then maintained with 3.5% sevoflurane (endexpiratory concentration) or 2% isoflurane in oxygen. In the propofol group, anesthesia was induced with fentanyl (3 g/kg) and propofol (2 mg/kg), then maintained with propofol (8 mg/kg/h) according to Giradis et al. [9]. A 1.75 MAC concentration of sevoflurane and isoflurane was used to compare their effects with those of propofol because 1 MAC is comparable to 2% sevoflurane and 1.15% isoflurane [2, 6]. Endotracheal intubation was facilitated with 0.15 g/kg vecuronium. Ventilation was mechanically controlled (NAD IIb, North American Dra ger, Telford, PA, USA) at a frequency of 18 cycles/min, with a tidal volume (VT) sufficient to maintain end-tidal CO2 (ETCO2) between 30 to 35 mmHg. The ETCO2 was measured by a capnometer (M1026A, HewlettPackard, Palo Alto, CA, USA). After arterial pressure and heart rate (HR) became stable for 10 min, a continuous 256-s ECG was recorded as the anesthetized stage. Pneumoperitoneum was introduced by insufflation of CO2 via a Veres needle inserted into a small umbilical incision. Intra-abdominal pressure was maintained automatically at 10 mmHg by a CO2 insufflator. After the completion of pneumoperitoneum, a continuous 256-s ECG was recorded as the insufflated stage. Ventilator settings were not changed, and the patients were kept in a horizontal position during the study. Measurements including arterial pressure, HR, (at all three stages) SpO2, ETCO2, and peak airway pressure (PAP) (anesthetized and insufflated stages) were obtained at the start and end of each ECG recording. The VT (during anesthetized and insufflated stages) was obtained at the middle of ECG recordings. The PAP and VT were observed using a built-in monitor. Mean arterial pressure (MAP) was calculated as: MAP ([ systolic pressure) + 2 (diastolic pressure)] 3 Power spectral analysis of the R-R intervals was performed as previously described [14]. In brief, the ECG channel output was recorded onto a floppy diskette (RD-F1, TEAC, Tokyo, Japan) and digitized at 500 Hz for offline analysis. The computer program processed the digitized data using a 14-bit A-D converter-equipped desktop computer (PC98, NEC, Tokyo, Japan) and analyzed electrocardiographic wave. The program measured the time difference between two R waves to create an R-R interval tachogram. The contamination of artifacts was erased beat by beat manually. Instantaneous 1024-HR data from 256-s R-R interval segments were converted to 1/R-R interval by sampling at 4 Hz, and the 256-s segment of R-R intervals were subjected to offline power spectral analysis by fast-Fourier transform. A rectangular local window periodogram method was used as a low-pass antialiasing digital filter at a point above the Nyquist sampling rate (2 Hz), which allowed spectral estimates between 0 and 1 Hz to be computed reliably. The power spectra at frequencies less than 0.5 Hz were standardized as the square of the mean HR (Hz). The spectra were quantified by examining two areas of the spectrum: the LF (0.040.15 Hz) and HF (0.150.40 Hz) band areas. The peak areas of the power spectral densities were integrated, and the HF/LF ratio was computed. Then the log power of these peak areas and the HF/LF ratio were calculated by taking their common logarithm. Data were expressed as mean standard deviation (SD). Demographic data were analyzed by one-way analysis of variance (ANOVA). Proportions of gender in each group were compared by the chi-square test. For data analysis of cardiovascular and respiratory measures and the measures from power spectral analysis, intergroup and intragroup differences were evaluated by two-way ANOVA for repeated measures. Once the intragroup differences were identified, the effects of the successive steps in the procedure were analyzed by paired t-test. A p of less than 0.05 was considered significant. Table 1. Patient demographic characteristics Sevoflurane group Age (year) Sex (male/female) Height (cm) Weight (kg) 47 10 6/9 160 8 59 10 Isoflurane group 47 13 6/9 161 10 63 14 Propofol group 45 10 5/10 157 9 63 10

Values are mean SD or n

Results The three groups were similar with respect to demographic data including age, gender, height, and body weight (Table 1). Mean arterial pressure and HR were decreased at the anesthetized stage, then restored to the control level at the end (MAP) or start (HR) of the insufflated stage (MAP: p < 0.0001; HR: p < 0.0001) (Table 2). Peak airway pressure and ETCO2 were increased at the start (PAP) or end (EtCO2) of the insufflated stage (PAP: p < 0.0001; EtCO2: p < 0.0001). As a result of insufflation, VT decreased (p < 0.0001). There were no intergroup differences among the three groups in MAP, HR, PAP, ETCO2, and VT. Log HF, which indicates parasympathetic activity, did not change throughout the study (p 0.08; Fig. 1). Log LF, which represents sympathetic activity, decreased at the anesthetized stage, then increased to the control level at the insufflated stage (p < 0.0001; Fig. 2). Log HF/LF, the balance of the sympathetic and parasympathetic activity, increased with induction of anesthesia and decreased to the control level after pneumoperitoneum (p < 0.0001; Fig. 3). These three measurements made by power spectral analysis did not differ among the three groups [log HF: p 0.19; log LF: p 0.16; log HF/LF: p 0.97].

Discussion The results of this study suggested that pneumoperitoneum with the patient under general anesthesia increased cardiac sympathetic activity. The increase in MAP and HR during pneumoperitoneum also may reflect the increase in cardiac sympathetic activity. Three possible mechanisms may explain the increase in cardiac sympathetic activity. First, the increase in intra-abdominal pressure may have caused initial reduction in venous return and cardiac output [10, 11], resulting in reflexed increase of sympathetic activity. However, some authors have reported that pneumoperitoneum alone does not cause significant change in cardiac output [17, 19]. Therefore, further investigation is required to clarify whether this mechanism is possible when pneumoperitoneum is applied. Second hypercarbia itself may have directly stimulated the sympathetic nervous system, resulting in increased cardiac output and MAP [18, 21]. Then hypercarbia also may have stimulated sympathetic nervous system indirectly by increasing plasma catecholamine concentration, including epinephrine and norepinephrine [21]. Indeed, the plasma concentration of dopamine, vasopressin, epinephrine, norepinephrine, rennin, and cortisol increased shortly after in-

364

Fig. 1. Change in log HF, which indicates cardiac parasympathetic activity in each group. Each point represents the mean of 15 patients, and the error bar indicates standard deviation. sevoflurane group, isoflurane group, propofol group. Fig. 2. Change in log LF, which indicates cardiac sympathetic activity in each group. Each point represents the mean of 15 patients, and the error bar indicates SD. sevoflurane group, isoflurane group, propofol group. *Significantly different (p < 0.05) when compared with awake stage (paired t test). Fig. 3. Change in log HF/LF, the balance of the sympathetic and parasympathetic activity in each group. Each point represents the mean of 15 patients, and the error bar indicates standard deviation. sevoflurane group, isoflurane group, propofol group. *Significantly different (p < 0.05) when compared with awake stage (paired t test).

Table 2. Changes in cardiovascular and ventilatory measures Awake Start MAP (mmHg) Sevoflurane Isoflurane Propofol HR (beats/min) Sevoflurane Isoflurane Propofol PAP (mmHg) Sevoflurane Isoflurane Propofol ETCO2 (mmHg) Sevoflurane Isoflurane Propofol VT (ml) Sevoflurane Isoflurane Propofol 96.3 (13.3) 99.5 (13.0) 99.1 (16.6) 75.1 (13.1) 78.8 (19.2) 83.9 (14.8) End 95.8 (13.8) 96.1 (13.9) 99.6 (15.3) 76.4 (15.1) 78.9 (18.4) 85.2 (14.9) Start 65.4 (12.2)a 66.2 (10.5)a 74.2 (11.6)a 68.0 (16.6)a 72.5 (15.0)a 62.3 (9.5)a 9.7 (2.2) 10.0 (1.8) 10.5 (2.4) 32.2 (1.8) 32.1 (1.7) 32.3 (2.1) Anesthetized End 64.1 (14.8)a 60.3 (11.3)a 73.2 (11.8)a 68.4 (17.0)a 68.6 (13.2)a 61.1 (9.2)a 9.7 (2.2) 10.1 (1.8) 10.7 (2.4) 31.8 (1.6) 31.6 (1.8) 31.8 (2.2) 370.8 (105.0) 379.5 (66.1) 334.7 (40.5) Start 82.8 (11.2)a,b 81.5 (10.8)a,b 92.6 (17.7)a,b 76.7 (16.9)b 76.3 (14.1)b 70.0 (10.1)a,b 13.7 (2.4)b 13.8 (3.1)b 14.2 (2.3)b 32.6 (2.5) 32.0 (2.1) 32.5 (2.4) Insufflated End 91.9 (14.3)b 104.2 (16.0)b 105.8 (17.5)b 79.2 (19.5)b 83.1 (13.6)b 71.2 (9.2)a,b 14.1 (2.8)b 13.9 (3.0)b 14.5 (2.3)b 35.0 (3.1)b 34.5 (3.4)b 35.3 (2.9)b 355.4 (101.2)c 369.0 (65.3)c 316.2 (44.2)c

Values are mean SD; n 15 in each group; MAP mean arterial pressure; HR heart rate; PAP peak airway pressure; EtCO2 end-tidal CO2 concentration; Vt tidal volume a Significantly different (p < 0.05) from start of awake stage (paired t test) b Significantly different (p < 0.05) from start of anesthetized stage (paired t test) c Significantly different (p < 0.05) from anesthetized stage (paired t test) Start, end; measurement at the start or end of each EEG recording

365

duction of pneumoperitoneum [1, 4, 12]. However, Oleary et al. [20] demonstrated that only renin-aldosterone increased in parallel with the MAP increase after induction of pneumoperitoneum. Epinephrine, norepinephrine, and cortisol increased only after deflation of the pneumoperitoneum. Therefore, the contribution of neumoendocrine response to the increase in sympathetic activity shortly after induction of pneumoperitoneum still is uncertain. Third, the increase in log LF might have represented direct stimulation with pneumoperitoneum. Distention of the abdominal muscles might have produced pain [15]. Furthermore, some authors consider that insufflated carbon dioxide, diaphragmatic distention, or both may be the cause of irritation at the phrenic nerve distribution area, resulting in postoperation pain after laparoscopic cholecystectomy [5, 16]. Similarly, it is speculated that mechanical stimulation and stimulation by insuffulated carbon dioxide may directly cause sympathetic activation as nociceptive stimulation. In this study, although log LF decreased in all three groups after induction of anesthesia, log HF did not change. Several reports showed the effect of anesthetics used in this study on the cardiac autonomic nervous system, evaluated using spectral analysis of HR or systemic arterial pressure. Kato et al. [13] showed dose-related decreases in three bands of spectral analysis of HR variability (low, 0.040.09 Hz; mid, 0.090.15 Hz; and high, 0.150.4 Hz) during isoflurane anesthesia. Galletly et al. [8] also demonstrated a decrease in three frequency bands of HR variability (low, 0.020.08 Hz; mid, 0.080.15 Hz; and high, 0.150.45 Hz) under 1.5% isoflurane with 66% nitrous oxide anesthesia. Scheffer et al. [22] showed both a low (0.060.12 Hz) and a high (0.120.36 Hz) frequency spectral power decrease after 2.5 mg/kg bolus administration of propofol. Furthermore, Galletly et al. [7] demonstrated that propofol induces a significant reduction in all component frequencies (low, 0.020.08 Hz; mid, 0.080.15 Hz; and high, 0.15 0.45 Hz) of spectral power of HR variability under continuous infusion of propofol (mean infusion rate, 0.19 mg/kg/ min). However, Wang et al. [24] showed a significant decrease in powers of very low frequency (0.000.08 Hz), low frequency (0.080.15 Hz), and very high frequency (0.80 1.60 Hz), but not in powers of high frequency (0.150.25 Hz) under continuous infusion of propofol (5 and 10 mg/ kg/min) using spectral analysis of systemic arterial pressure. Little is known about sevoflurane anesthesia. Therefore, direct comparison between current and previous studies is difficult because study conditions such as the age of the patients, doses of the anesthetics, coadministered drugs, definition of each frequency bandwidth, and mode of ventilation differ from study to study. However, because the log HF is coupled with respiration, the relatively unaltered log HF in the current study may be caused partly by the controlled respiration [24]. Three different anesthetics commonly used in daily anesthesia practice were employed in this study. Propofol anesthesia was prepared by the same recipe that Giradis et al. [4] used. The cardiovascular and ventilatory measurements in the current propofol group almost agreed with their results. For sevoflurane and isoflurane anesthesia, 1.75 MAC of each were used to match the cardiovascular response to pneumoperitoneum. Indeed, no difference was noted among

the three groups in cardiovascular measurements as a response to the pneumoperitoneum. Neither was any difference noted in ventilatory measurements among the three groups. Therefore, cardiovascular and ventilatory response to pneumoperitoneum seemed to be identical among the three groups. Three measurements from HR variability analysis, log HF, log LF, and log HF/LF, showed similar change in response to pneumoperitoneum in all three groups. The different anesthesias used in this study did not have major effect on cardiac ANS as measured by HR variability analysis. These results indicate that changes in HR variability are similar no matter what anesthesia is used when the cardiovascular and ventilatory changes are equivalent. The authors suggest that measurements from HR variability analysis mainly reflect cardiovascular and ventilatory change rather than the characteristics of the anesthetics used. In summary, the findings of this study showed that pneumoperitoneum increased sympathetic cardiac activity as evaluated by HR variability analysis. The choice of general anesthetic did not seem to have a major influence on the changes in cardiac ANS induced by pneumoperitoneum.

References
1. Aoki T, Tanii M, Takahashi D, Tateda T, Miyazawa A (1994) Cardiovascular changes and plasma catecholamine levels during laparoscopic surgery. Anesth Analg 78: S8 2. Campbell C, Nahrwold ML, Miller DD (1995) Clinical comparison of sevoflurane and isoflurane when administered with nitrous oxide for surgical procedures of intermediate duration. Can J Anaesth 42: 884 890 3. Cunningham AJ, Brull SJ (1993) Laparoscopic cholecystectomy: anesthetic implications. Anesth Analg 76: 11201133 4. Felber AR, Blobner M, Goegler S, Senekowitsch T, Jelen-Esselborn S (1993) Plasma vasopressin in laparoscopic cholecystectomy. Anesthesiology 79: A32 5. Fredman B, Jedeikin R, Olsfanger D, Flor P, Gruzman A (1994) Residual pneumoperitoneum: a cause of postoperative pain after laparoscopic cholecystectomy. Anesth Analg 79: 152154 6. Frink EJ Jr, Malan TP, Atlas M, Dominguez LM, KiNardo JA, Brown BR (1992) Clinical comparison of sevoflurane and isoflurane in healthy patients. Anesth Analg 74: 241245 7. Galletly DC, Buckley DHF, Robinson BJ, Corfiatis T (1994) Heart rate variability during propofol anesthesia. Br J Anaesth 72: 219220 8. Galletly DC, Westenberg AM, Robinson BJ, Corfiatis T (1994) Effect of halothane, isoflurane and fentanyl on spectral components of heart rate variability. Br J Anaesth 72: 177180 9. Giradis M, Broi UD, Antonutto G, Pasetto A (1996) The effect of laparoscopic cholecystectomy on cardiovascular function and pulmonary gas exchange. Anesth Analg 83: 134140 10. Johannsen G, Andersen M, Juhl B (1989) The effect of general anesthesia on the haemodynamic events during laparoscopy with CO2 insufflation. Acta Anaesthesiol Scand 33: 132136 11. Joris JL, Noirot DP, Legrand MJ, Jacquet NJ, Lamy ML (1993) Hemodynamic changes during laparoscopic cholecystectomy. Anesth Analg 76: 10671071 12. Joris J, Lamy M (1993) Neuroendocrine changes during pneumoperitoneum for laparoscopic cholecystectomy. Br J Anaesth 70: A33 13. Kato M, Komatsu T, Kimura T, Sugiyama F, Nakashima K, Shimada Y (1992) Spectral analysis of heart rate variability during isoflurane anesthesia. Anesthesiology 77: 669674 14. Kawamoto M, Tanaka N, Takasaki M (1993) Power spectral analysis of heart rate variability after spinal anesthesia. Br J Anaesth 71: 523 527 15. Koivusalo AM, Kellokumpu I, Lindgren L (1996) Gasless laparoscop-

366 ic cholecystectomy: comparison of postoperative recovery with conventional technique. Br J Anaesth 77: 576580 Lindgren L, Koivusalo AM, Kellokumpu I (1995) Conventional pneumoperitoneum compared with abdominal wall lift for laparoscopic cholecystectomy. Br J Anaesth 75: 567572 Liu SY, Leighton T, Davis I, Klein S, Lippman M, Bongard F (1991) Prospective analysis of cardiopulmonary responses to laparoscopic cholecystectomy. J Laparoendosc Surg 1: 241246 Marshall BE, Cohen PJ, Klingenmaier CH, Neigh JL, Pender JW (1971) Some pulmonary and cardiovascular effects of enflurane (Ethrane) anesthesia with varying PaCO2 in man. Br J Anaesth 43: 9961002 Marshall RL, Jebson PJR, Davie IT, Scott DB (1972) Circulatory effects of carbon dioxide insufflation of the peritoneal cavity for laparoscopy. Br J Anaesth 44: 680684 20. OLeary E, Hubbard K, Tormey W, Cunningham AJ (1996) Laparoscopic cholecystectomy: haemodynamic and neuroendocrine responses after pneumoperitoneum and changes in position. Br J Anaesth 76: 640644 21. Rasmussen JP, Dauchot PJ, DePalma RG, Sorensen B, Regula G, Anton AH, Gravenstein JS (1978) Cardiac function and hypercarbia. Arch Surg 113: 11961200 22. Scheffer GJ, Ten Voorde BJ, Karemaker JM, Ros HH, de Lange JJ (1993) Effects of thiopentone, etomidate, and propofol on beat-to-beat cardiovascular signals in man. Anaesthesia 48: 849855 23. Wahba RWM, Beique F, Kleiman SJ (1995) Cardiopulmonary function and laparoscopic cholecystectomy. Can J Anaesth 42: 5163 24. Wang H, Kuo TBJ, Chan SHH, Tsai TH, Lee TY, Lui PW (1996) Spectral analysis of arterial pressure variability during induction of propofol anesthesia. Anesth Analg 82: 914919

16. 17. 18.

19.

Editorial
Surg Endosc (2000) 14: 317 DOI: 10.1007/s004640000153 Springer-Verlag New York Inc. 2000

Are radially dilating trocars better than the alternatives?

A new type of trocar with a radially dilating tip has been introduced in recent years. The concept is a good one; namely, abdominal access is gained with a small 2 or 3 mm instrument, and that hole is dilated up to 5 to 10 mm to accommodate conventional trocars. Theoretic advantages of such a trocar over conventional cutting trocars or open trocar access is less tissue trauma and a tighter fascial seal [5]. Issues of cost, not only for the disposable trocar, but also savings in terms of operative time for closing the fascia of the conventional port site must be considered. The traditional approach for trocar placement has been a closed one with the establishment of a pneumoperitoneum by the Verres needle. The tip configuration for reusable trocars is usually conical or triangular. The trauma caused by these tips has been studied on a scientific basis with the conical tips requiring higher insertion pressure but leaving less of a defect [3]. Disposable alternatives are also in use, most of which have a safety shield that comes out over the sharp tip as soon as the peritoneum is entered. This supposedly diminishes the risk of intraabdominal injury, although the latter claim has not been proved [2]. A more recent device utilizes direct observation through the trocar as it is transversing the fascia, preperitoneal space, and peritoneum. Again, this latter device is a disposable instrument. Many surgeons use the open trocar access popularized by Dr. Hasson more than two decades ago. Many surgeons believe that this is the safest approach to minimize the risk of injury to intraperitoneal organs. Although it has been stated that this approach totally avoids vascular injury, there have been two case reports of such injury using this open approach [4]. Closure of the fascial defect is usually easier with this approach because the sutures have been placed in the fascia prior to insertion of the trocar. The cost is not inconsiderable. Disposable trocars significantly increase the cost of a laparoscopic procedure [1]. However, they are generally easier to use than reusable trocars. The cost of operating room time is also an issue. If radially expanded trocars can save 10 minutes of operative time because their fascial defects do not need to be closed, this would be a cost consideration in their favor. The issue of postoperative pain from trocar sites is important. Studies using fine caliber instruments have shown a lower need for narcotic analgesia postoperatively [6]. The question remains: Does a statistical difference correspond to a clinically significant difference? Both groups of patients had low analgesic requirements. I do not think there is a clinical difference. The security of access may be significant for these trocars. In prolonged laparoscopic cases, conventional cutting trocars can become dislodged because the fascial defect enlarges around the trocar sheath, especially with manipulation of that sheath. This is one of the criticisms of open trocar sheath placement. The radially expanded trocar would have an advantage here because it is held snugly by the fascia because of the way it is placed. However, there are no clinical or even anecdotal reports of this theoretic superiority. In summary, the radially expanded trocar technique offers both practical and theoretical advantages over conventional trocar technique for those surgeons who prefer closed peritoneal access. However, the cost of these disposable trocars outweighs their advantages. Development of reusable or reposable radially expanded trocars may necessitate a future reanalysis. References
1. Apelgren KN, Blank ML, Slomski CA (1994) Reusable instruments are more cost-effective than disposable instruments for laparoscopic cholecystectomy. Surg Endosc 8: 3234 2. Apelgren KN, Scheeres DC (1994) Aortic injury: a catastrophic complication of laparoscopic cholecystectomy. Surg Endosc 8: 589691 3. Bo hm B, Knigge M, Kraft M (1998) Influence of different trocar tips on abdominal wall penetration during laparoscopy. Surg Endosc 12: 1434 1438 4. Hanney RM, Cormolt HL, Merrett N (1999) Vascular injuries during laparoscopy associated with the Hasson technique. J Am Coll Surg 188: 337 5. Schulam PG, Hedican SP, Docino SG (1999) Radially dilating trocar system for open laparoscopic access. Urology 54: 727729 6. Turner DJ (1996) A new, radially expanding access system for laparoscopic procedures versus conventional cannulas. J Am Assoc Gynecol Laparosc 3: 609615

K. N. Apelgren
Michigan State University Department of Surgery 1200 East Michigan Avenue Suite 655 Lansing, MI, 48912, USA

Surg Endosc (2000) 14: 409410 DOI: 10.1007/s004649900059

Springer-Verlag New York Inc. 2000

The author replies

We thank Dr. Carbajo et al. for their response to our article. We also thank the editorial board of Surgical Endoscopy for enabling us to respond, clarify, and stress several critical points concerning our article. The aim of our article was to suggest a novel technique for mesh placement (without relating to its material), and to present the concept of central orientation and centrifugal fixation (COCF). It was not our purpose to present our experience of laparoscopic hernia repair. Therefore, we first answer the comments concerning the technique. Second comment The concept of COCF is valid for many types of abdominal wall defects. The technique is very easy to perform in dealing with a solitary abdominal wall defect. With a compound multicavitated defect, the solution is to find its common center, to which the center of the mesh should be attached first. We performed this successfully several times without any technical difficulties. If defects are not on the same anatomic plane or remote from each other, we use a supplementary prosthetic mesh, applying the COCF concept to each defect separately (by defining more than one center point). Third comment In our article, we present three different techniques for attaching the center of the mesh to the center of the defect. The criticism relates only to the second technique, which we used rarely in the past. In any event, we had never attached the tacker to the skin above the defect [2]. The only technique we use currently, having gained a lot of experience, is to mark the center of the mesh with an absorbable knot and loop to be pulled outside through the center of the defect by a suture needle device, (Endoclose, USSC, CT, USA) that penetrates through the skin (1-mm cut). This attaches the mesh over the defect precisely in the proper position, which is followed by centrifugal fixation of the mesh to the abdominal wall by tackers, and finally by cutting of the protruding loop. In this way, no penetration of the skin by tackers should occur. Fourth comment We are confident that Dr. Carbajo et al. realize that we are well acquainted with the various techniques of mesh placement. Our small series reflects only our experience in using the current technique between January 1997 and September 1998. Having used different laparoscopic techniques in the past, we have expanded our experience using the current technique over the past year. Our aim was to suggest an additional technique that, in our opinion, presents a very accurate and logical method requiring only a few minutes to perform. First and fifth comment Although we think that the discussion about the appropriate mesh material for laparoscopic repair of incisional ventral hernias is not very relevant to our article (although it is an important issue), it does deserve a few clarifying comments. We are fully aware of the clinical and experimental data concerning the consequences of using nonabsorbable material for intra-abdominal mesh reinforcement, particularly the relatively high adhesion formation associated with polypropylene (PP) use. However, this material incorporates very well into the abdominal wall, and the rate of intestinal obstruction (which reflects the clinical significance of adhesion formation) is extremely low. Besides, the PP mesh was not directly implicated as the cause of the obstruction in those reported cases [6, 7]. We, along with others, have extensive and long-term experience using intraperitoneal PP mesh for open repair of incisional hernias, with no complications related to adhesions (e.g., intestinal obstruction). The only problem we faced, only rarely, was related to patients that for other reasons required laparotomy, which might have been technically tedious in some cases because of extensive intestinal adhesions to the abdominal wall and the mesh. (We have the impression that placing the omentum between the mesh and the bowel may lessen this phenomenon.) We already have operated laparoscopically on approximately 50 patients for incisional ventral hernias without facing any significant complication related to the mesh. Concerning the transabdominal preperitoneal (TAPP) technique mentioned by Dr. Carbajo, it should be stressed that according to animal studies, attempts to approximate the peritoneum over preperitoneal mesh may enhance rather than reduce adhesion formation [7]. In addition, it should be mentioned that the use of PP mesh for laparoscopic repair of ventral hernias is very well known and has been presented in recent medical literature [1, 4, 5, 8, 9]. On the basis mainly of experimental data, the

410

use of dual mesh (several known types) has increased recently, and being aware of this trend, we also use this prosthesis currently with the same technique for incisional ventral hernia repair. However, the use of these prostheses is not without technical faults (e. g., need for larger fascial opening) and need more years of follow-up evaluation. Finally, we want to add that we enjoyed reading the recently published comprehensive article written by M. A. Carbajo et al. [3]. To our surprise, we found that in 22 of 30 patients, PP mesh was used for open repair of incisional abdominal wall hernias. References
1. Arca MJ, Heniford BT, Pokorny R, Wilson MA, Mayes J, Gagner M (1998) Laparoscopic repair of lumbar hernias. J Am Coll Surg 1987: 147152 2. Bickel A, Eitan E (1999) A simplified laparoscopic technique for mesh placement in ventral hernia repair. Surg Endosc 13: 532534 3. Carbajo MA, Martin del Olmo JC, Blanco JI, Cuesta C, Toledano M, Martin F, Vaguero C, Imglada L (1999) Laparoscopic treatment vs open

4. 5. 6. 7. 8. 9.

surgery in the solution of major incisional and abdominal wall hernias with mesh. Surg Endosc 13: 250252 Helfrich RB, Gianturco C (1995) Abdominal wall hernia repair: use of the Gianturco-Helfrich-Ederbach hernia mesh. J Laparoendosc Surg 5: 9195 Holzman MD, Purut CM, Reintgen K, Eubanks S, Pappas TN (1997) Laparoscopic ventral and incisional hernioplasty. Surg Endosc 11: 32 35 McDonald S, Gagic N (1984) Intraperitoneal prolene mesh in hernia repair: a comparison of two techniques. Can J Surg 27: 157158 Morris-Stiff GJ, Hughes LE (1998) The outcome of nonabsorbable mesh placed within the abdominal cavity: literature review and clinical experience. J Am Coll Surg 186: 352367 Park A, Gagner M, Pomp A (1996) Laparoscopic repair of large incisional hernias. Surg Laparosc Endosc 6: 123128 Saiz A, Willis H (1994) Laparoscopic ventral hernia repair. J Laparoendosc Surg 4: 365367

A. Bickel A. Eitan
Department of Surgery Western Galilee Hospital Nahariya, Israel

New technology
Surg Endosc (2000) 14: 375381 DOI: 10.1007/s004640020067 Springer-Verlag New York Inc. 2000

Robotics and telemanipulation technologies for endoscopic surgery

A review of the ARTEMIS project
M. O. Schurr,1 G. Buess,1 B. Neisius,2 U. Voges3
1 2

Section for Minimally Invasive Surgery, Eberhard Karls University, Waldhoernlestrasse 22, D-72072 Tu bingen, Germany Department of Engineering, Karlsruhe Research Center, Karlsruhe, Germany 3 Institute of Applied Informatics, Karlsruhe Research Center, Karlsruhe, Germany Received: 25 February 1998/Accepted: 20 April 1999

Abstract. In endoscopic surgery, the ability to guide the instrument is significantly decreased compared with open surgery. Rigid laparoscopic instruments offer only four of the six degrees of freedom required for the free handling of objects in space. Robotics technology can be used to restore full mobility of the endoscopic instrument. Therefore, we designed a master-slave manipulator system (ARTEMIS) for laparoscopic surgery as a prototype. The system consists of two robotic arms holding two steerable laparoscopic instruments. These two work units are controlled from a console equipped with two master arms operated by the surgeon. The systems and its components were evaluated experimentally. Laparoscopic manipulations were feasible with the ARTEMIS system. The placement of ligatures and sutures and the handling of catheters were possible in phantom models. The surgical practicability of the system was demonstrated in animal experiments. We conclude that robotic manipulators are feasible for experimental endoscopic surgery. Their clinical application requires further technical development. Key words: Endoscopic surgery Instrumentation Robotics Telemanipulation Telesurgery

Following the advent of endoscopic surgery and its routine clinical use in laparosopic cholecystectomy, a wide series of applications has been developed. From a technical standpoint, most operations in general surgery can now be performed endoscopically. As endoscopic surgery has become increasingly complex a demand has arisen for improved instrumentation. Besides developing improved endoscopic vision systems [2, 10, 29], leading research groups worldwide have focused their effects on increasing instrument
Correspondence to: M. O. Schurr

functionality. Robotics has been recognized as a major driving force in the technological advance of endoscopic surgery. The initial use of robotics for increasing instrument functions was in the field of endoscope guidance, where robotic instrument holders were employed to direct the endoscope during surgery [3, 26, 29]. Robotic endoscope manipulators proved to be safe and efficient in various fields of use and are now accepted as assisting devices among endoscopic surgeons [1, 16]. The use of robotics for enhancing surgical instrumentation did not emerge as quickly due to its greater technical complexity and safety questions linked to the use of robotic devices for tissue manipulation. The first mention of the use of robotics for surgical instrumentation occurred in the early 1970s, when the interest in providing medical care during space missions triggered the idea of using teleoperated robots for remote surgery in space [24]. The first practical attempts to integrate robotic manipulators into surgical procedures were made in the fields of neurosurgery urology and orthopedic surgery [4, 11]. The application of robotic manipulation technologies to the field of laparoscopic surgery, and the respective developments in this area since the early 1990s, is the outgrowth of improvements in the steerability and dexterity of surgical instruments [20, 30]. As more complex laparoscopic procedures involving ligation, suturing, and knotting techniques were performed, the need for instruments with increased degrees of freedom of motion became apparent [8, 20, 30]. The human arm features seven degrees of freedom, which allow the hand (and a hand-held tool) to be guided precisely to any point in the reachable work space. In open surgery, the full mobility of arm and hand can be used to direct the surgical instrument. In endoscopic operations, the mobility of the instrument is decreased significantly due to the invariant point of insertion through the patients abdominal or thoracic wall (Fig. 1) [21, 22]. Only four of the seven natural degrees of freedom (DOF) remain [30]. This reduction in instrument mobility is not an issue in the simple

376

Fig. 1. Work space of a rigid laparoscopic instrument. Only four degrees of freedom remain. A and B, Rotation around the point of insertion; C, Rotation around the shaft axis; X, translation [16]. Fig. 2. Rigid manipulators. A Rigid remote handling tongs for technical manipulation. B Conventional laparoscopic instrument. Fig. 3. Technical master-slave manipulator.

manipulations that dominate laparoscopic cholecystectomy or hernia surgery, but it can lead to significant handling restrictions in interventions that require complex dissection or tissue connection techniques. The problems arising from limited handling capabilities in closed cavities are well known in the world of industrial

manufacturing processes. By the 1960s, remote handling tongs were being used to manipulate toxic or radioactive substances in closed rooms. These devices (Fig. 2A) were pure mechanical instruments with rigid force transmission elements between handle and grasper. They were technically equivalent to todays laparoscopic instruments (Fig.

377

Fig. 4. Evolution of remote handling technology in industry and surgery [20].

2B). The mechanical handling devices were soon enhanced with electrical motors and computers for easier control and, most important, were fitted with additional joints to allow the user to regain full dexterity in space with six DOF [17]. Systems of the 1980s featured sophisticated kinematic joint structures with six DOF and appropriate electric drive and control systems (Fig. 3). These complex robotic systems are called master-slave manipulators. The master-slave mode of operation is a control principle in which all movements done with a master input device are transformed in real time to the slave output device [6, 17]. The entire manipulator system can be guided by the user; there is no preprogrammed robotic motion. Notwithstanding the clear differences between advanced manufacturing processes and endoscopic surgery, the basic principles of electrical master-slave manipulators can be used to enhance endoscopic instrumentation in a way similar to their application in the technical field (Fig. 4) [21, 25]. How this technology was used for the ARTEMIS project and the experimental evaluation that was performed are described in this review paper.

Fig. 5. Graphic simulation of an endoscopic master-slave manipulator scenario. shows the two basic components of the envisioned ARTEMIS (Advanced Robotic TElemanipulator for Minimally Invasive Surgery) manipulator system: the user station (master) and the instrument station (slave). The surgeon sits at a console that integrates endoscopic monitors, communication facilities, and two master devices to control the two slave arms, which are mounted on the operating table.

Materials and methods Technical development

Based on the concept of using manipulator technology to enhance surgical instrumentation in laparoscopic procedures (Fig. 5), our own work in the area of surgical robotics began in 1991, in collaboration with the Karlsruhe Research Center, Karlsruhe, Germany. The scenario illustrated in Fig. 5

The slave subsystem

The first phase of the ARTEMIS project [31] was dedicated to the development of the slave arm and appropriate kinematic units to restore full

378

Fig. 6. Steerable instruments with additional degrees of freedom. A Initial prototypes in a phantom model. B Steerable instrument in an animal. C Optimized joint segment.

instrument mobility with six DOF. Initially, we concentrated on developing new joint structures capable of bending the tip of the instrument inside the human body. Different versions of multi-joints have been developed and were integrated into the tip of mechanically (Fig. 6A) and, later, electrically driven steerable endoscopic instruments [20, 30]. In addition to bending, rotation of the surgical tool at the instrument tip is also possible. The steerable instruments have been evaluated in a series of different laparoscopic procedures in phantoms and animals (Fig. 6B) to optimize design and mechanical properties, such as bending angle and radius or stiffness against manipulation forces [30]. As a result, an optimized joint structure was developed that ensured, on one hand, high mechanical stability and, on the other, mathematically constant behavior in its motion sequence, even under external load (e.g., resulting from tissue manipulation) (Fig. 6C) [22]. The bending angle of this multi-joint system is 90, and the instrument has a diameter of 10 mm. In practice, the steerable instruments proved to be especially helpful in surgical maneuvers that are often difficult with straight rigid instruments, such as placing ligatures around vessels, or encircling the bowel or esophagus during resection procedures. After their experimental evaluation and the redesign of certain technical aspects, the steerable instruments were motorized step by step and integrated into an entire slave arm (Fig. 7A, B) [15, 31]. This slave arm is an external kinematic unit that guides the steerable instrument around the invariant point of insertion in the body of the patient. The arm has two segments and four joints, which are driven by integrated electromotors. The special mechanical structure of the slave arm automatically guides the instrument around the invariant point without exerting force on the ab-

dominal wall. This movement is achieved by mechanical constraints in the joint structure and is fully independent of software or electronic control systems. The steerable instrument is held in an attachment that allows active translation into the operating field. The functional unit of both the steerable instrument and the guiding arm restores full spatial mobility of the instrument tip with six DOF of motion. Two slave units (Fig. 7B) can be attached to the operating table by means of a special carrier system that is compatible with Maquet operating tables (Maquet, Rastatt, Germany).

The master subsystem

Complex kinematic units with several DOF are very difficult to control without electronic control systems. A special control system was used for the ARTEMIS prototype. This system is similar to the MONSUN (Manipulator CONtrol System Utilising Network Technology) universal manipulator control architecture described by Breitwieser and Weber [7]. This control system allows the use of different master devices within ARTEMIS, making it adaptable to the specific ergonomic or functional demands of the surgeon. During the design phase of the ARTEMIS control system, the suitability of the control architecture of technical telemanipulators for surgery was evaluated. In an experimental study, the suitability of control features, such as scaling of the master-slave motion relations and general telemanipulation conditions in endoscopic surgery, was tested. Scaling is a specific control feature of advanced manipulator systems that allows the ratio be-

379

Fig. 7. The ARTEMIS manipulator. A Motion sequence of the slave arm. B System at the OR table.

tween the input and output movements of the system to be changed. A conventional technical telemanipulator, the DISTEL system [6], was modified for surgical use. A laparoscopic instrument attached to the slave arm was teleoperated in phantom models (Fig. 8) with varying settings of the control parameters. The results of this experimental study, which have been published elsewhere [32], indicated that functional similarity between master and slave arms, adjustable scaling ratios in master-slave translation, and the sitting position of the operator are all important requirements for intuitive and efficient system operation. These results were used as basic input for specification of the ARTEMIS master subsystem. The surgeons work place (Fig. 9) consists of an endoscopic three-dimensional monitor for the visualization of the operative field, as well as two additional monitors for the display of a graphic model of the slave arms and various systems data. Using these monitors, additional informationsuch as preoperative findings for the individual patient, laboratory parameters, or radiological imagescould be displayed in the future. The endoscope is controlled by means of a small joystick at the console. The FIPS endoscope guidance system [29] is used to allow remote operation of the laparoscope. Two master arms with six DOF are attached to this console; the surgeon is in a sitting position. Two different master devices are currently applicable for guiding ARTEMIS: the custom-made HT2 master, which was developed for practical surgical manipulations with the ARTEMIS system, and the TeSt master, a research-oriented input device with adaptable system functions. Both master devices are now being tested in a comparative study to establish which is better. The current manipulator prototype does not provide haptic force feedback or tactile sensing functions.

Results of experimental preclinical evaluation The subsystems of the ARTEMIS manipulator have been tested concurrently with the development of the full prototype; the results have been reported elsewhere [30, 32]. The final prototype is currently under experimental evaluation. The evaluation methods include technical phantoms simulating surgical maneuvers such as suturing, ligating, and intraabdominal handling of catheters. A standardized study evaluating of the performance of the prototype is now in progress. After appropriate system function and safety had

been proven in phantom models, two animal experiments were performed in domestic pigs (female, weight 50 kg) under general anesthesia. Transportation of the entire twoarmed prototype out of the engineering facility was technically impossible; therefore only a one-arm version could be used in the animal laboratory. The aim of the animal experiments was to demonstrate the practical feasibility of the system for laparoscopic manipulations. In both cases, the ARTEMIS arm was employed for mobilization of the sigmoid colon and ligation of sigmoid vessels for laparoscopic sigmoidectomy. The manipulator was inserted through a 10-mm port in the left lower abdomen of the pig; the laparoscope and two further instruments were inserted through three additional 10-mm ports in the midline and the right and left abdomen. These instruments were operated directly by an assistant at the operating table. The master arm was positioned at the side of the operating table; the surgeon was in a sitting position (Fig. 10A, B). After dissection of the sigmoid colon and fenestration of the mesentery, the blood vessels were encircled and ligated with the flexible section of the steerable instrument. Several ligatures were placed at different heights of the colon. The maneuver was easy to perform. It was found, however, that geometric changes in the flexible tip section are required to improve the practicality of the device. Further surgical evaluation is planned after modification of the prototype.

Discussion The restoration of the functional freedom available in open surgery is an important step for the advancement of endoscopic surgery. It is particularly important to restore the basic functions needed for human interactions with the en-

380

Fig. 8. Surgical telemanipulation experiment in 1994. Fig. 9. Surgeons console with two master arms to guide the working units. Fig. 10. The manipulator (one-arm version) in an animal experiment. A Master and slave sides. B Slave at the OR table.

vironment, such as spatial vision [2, 10], tactile sense, and instrument mobility in the operative field [8]. Recent developments in the field of endoscopic vision systems [29] have led to better and more natural endoscopic visualization in terms of improved image resolution, illumination, and clear vision maintenance techniques. The most notable advantages of these new vision systems are improved handling accuracy and time savings during endoscopic manipulations [9, 18, 23]. The restoration of full instrument mobility is a further technological challenge in endoscopic surgery. The first discussions of the use of robotic manipulators in endoscopic surgery can be found in the literature of the early 1990s [12, 21, 27], and the first prototypes of steerable endoscopic instruments with two additional DOF were introduced by our group in 1992 [20]. A functional master-slave manipulator for surgery was introduced by Hill et al. [13] (SRI International, Menlo Park, CA, USA). The SRI telemanipulator was not designed for endoscopic use and had only four DOF in its first versions. It was intended to be used mainly for remote surgery through telecommunication links, with particular emphasis on hostile environments, such as military applications [27]. The first applications of the SRI system to open surgery were described by Bowersox et al. [5] in 1996, who used it for vascular surgery in pigs. They described dissection of

the common femoral artery and closure of a 3-cm arteriotomy with a running suture through the master-slave manipulator in nine experimental cases. With this study, Bowersox was able to demonstrate the feasibility of delicate surgical manipulations, such as vascular suturing techniques, via a master-slave manipulator system. However, the SRI system, in the configuration used by Bowersox, does not solve the problem of limited instrument mobility since it provides only four DOF of motion. Notwithstanding these encouraging results, the lack of full six DOF instrument mobility is regarded as a limiting factor for surgical actions even in open techniques [5, 14]. Judging from our own experimental experience with the ARTEMIS system, the primary advantage of having the full six DOF is the extension of the work space of the instrument and the improved alignment of the grasper according to the operative situation [29, 30]. This aspect is of special importance in endoscopic surgery, but it may not be a key motivation for using telemanipulation technology in open microsurgery, where other system functions, such as scaling, seem to be most important [5]. Our own studies on the intuitiveness and ergonomy of surgical telemanipulation also revealed that scaling of the master vs slave motion relation and the selection of the optimal ergonomic working position through reindexing of the master arm are important user demands [31]. Reindexing is a control feature that al-

381

lows repositioning of the master arm to an ergonomic position. In the area of operative ergonomy and master functionality, improvements in the ARTEMIS slave unit are still needed. The restoration of full spatial mobility of the instrument for laparoscopic or thoracoscopic surgery is a complex task. The ARTEMIS system was the first six DOF master-slave manipulator for endoscopic surgery to be reported in the literature [30]. Clinical use of the system, however, will require further development in the area of slave mechanics and the control system. Manipulators for endoscopic microsurgery are currently the focus of intensive research and development efforts by various work groups at scientific institutions and in industry. Several devices have already entered the preclinical testing phase, and it appears highly probable that master-slave manipulators will soon be in practical surgical use. The role of robotics in the further development of surgery might not be as decisive as it was for production technology in the 1970s and 1980s, but it will help to move endoscopic surgery forward and to enhance surgical capabilities in various fields [19]. References
1. Arezzo A (1997) Sistemi per solo-surgery in chirurgia mini-invasiva: risultati di uno studio sperimentale. Dissertation, University of Genoa, Italy 2. Becker H, Melzer A, Schurr MO, Buess G (1993) 3D video techniques in endoscopic surgery. Endosc Surg Allied Technol 1: 3639 3. Begin E, Gagner M, Hurteau R, Santis S, Pomp A (1995) A robotic camera for laparoscopic surgery: conception and experimental results. Surg Laparosc Endosc 5: 611 4. Benabid AL, Cinquin P, Lavallee S, Le Bas JF, Demongeot J, De Rougemont J (1987) Computer driven robot for stereotactic surgery connected to CT scan and magnetic resonance imaging: technological design and preliminary results. Appl Neurophysiol 50: 16 5. Bowersox JC, Shah A, Jensen J, Hill J, Cordts PR, Green PS (1996) Vascular applications of telepresence surgery: initial feasibility studies in swine. J Vasc Surg 23: 281286 6. Breitwieser H (1991) DISTEL: digitales Steuerungssystem fu r Telemanipulatoren. Robotersysteme 1: 160164 7. Breitwieser H, Weber W (1996) MONSUN: a distributed manipulator control system utilizing network technology. Kopacek P, (ed) Humanoriented design of advanced robotics systems. Pergamon Press, Oxford, pp 159167 8. Buess G (1993) Why this journal? [Editorial] Endosc Surg Allied Technol 1: 12 9. Buess G, Bergen P, Kunert W, Schurr MO (1996) 2D und 3D Sichtsysteme in der minimal invasiven Chirurgie. Chirurg 67: 10411046 10. Crosthwaite G, Chung T, Dunkley P, Shimi S, Cuschieri A (1995) Comparison of direct vision and electronic two- and three-dimensional display systems on surgical task efficiency in endoscopic surgery. Br J Surg 82: 846851

11. Davies GL, Hibber RD, Ng WS, Timoney AG, Wickham JEA (1989) The development of a surgeon robot for prostatectomies. J Med Eng Technol 6: 273 12. Dohi T (1993) Robot technology in medicine. Adv Robotics 7: 179 187 13. Green PS, Hill JW, Jensen JF, Shah A (1995) Telepresence surgery. Proc IEEE Eng 14. Gupta SK (1996) Vascular applications of telepresence surgery: initial feasibility studies in swine. J Vasc Surg 23: 287 15. Holler E, Englert M, Neisius B, Trapp R (1994) Telepra senzsysteme fu r die minimal invasive Chirurgie. Kuenstliche Intelligenz 8: 3341 16. Jakobs LK, Shayani V, Sackier JM (1997) Determination of the learning curve of the AESOP robot. Surg Endosc 11: 5455 17. Koehler GW (1981) Typenbuch der Manipulatoren. Verlag Karl Thiemig, Munich 18. Kunert W, Flemming E, Schurr MO, Buess G (1997) Optik mit natu rlich wirkender Zusatzbeleuchtung. Langenbecks Arch Chir X(Suppl 2): 12321234 19. Lirici MM, Papaspyropoulos V, Angelini L (1997) Telerobotics in medicine and surgery. Min Invas Ther Allied Technol 6: 364378 20. Melzer A, Schurr MO, Dautzenberg P, Trapp R, Buess G (1992) Erho hung der Freiheitsgrade bei Instrumenten fu r die laparoskopische Chirurgie. Langenbecks Arch Chir X(Suppl X): 279285 21. Melzer A, Schurr MO, Kunert W, Buess G, Voges U, Meyer JU (1993) Intelligent surgical instrument system ISIS: concept and preliminary experimental application of components and prototypes. Endosc Surg Allied Technol 1: 165170 22. Mueglitz J, Kunad G, Dautzenberg P, Neisius B, Trapp R (1993) Kinematic problems of manipulators for minimal invasive surgery. Endosc Surg Allied Technol 1: 160165 23. Peitgen K, Walz MK, Eigler FW (1996) Praktische und experimentelle Erfahrungen mit dem Einsatz der 3D Technologie in der Laparoskopie. Endoskopie Heute 9 (Suppl): 64 24. Rasor NS, Spickler JW (1973) Endocorporeal surgery using remote manipulators. Herr E, (ed) Remotely manned systems for exploration and operation in space. California Institute of Technology, Pasadena, pp 483492 25. Rininsland H, Trapp R, Becker H (1992) Development of remotehandling technology. Cuschieri A, Buess G, Perissat J, (eds) Operative manual of endoscopic surgery. Springer, Berlin Heidelberg New York, pp 344347 26. Sackier JM, Wang Y (1994) Robotically assisted laparoscopic surgery: from concept to development. Surg Endosc 8: 6366 27. Satava RM (1992) Robotics, telepresence and virtual reality: a critical analysis of surgery of the future. Min Invas Ther 1: 357363 28. Schurr MO, Buess G, Kunert W, Flemming E, Hermeking H, Gumb L (1996) Human sense of vision: a guide to future endoscopic imaging systems. Min Invas Ther Allied Technol 5: 410418 29. Schurr MO, Melzer A, Dautzenberg P, Neisius B, Trapp R, Buess G (1993) Development of steerable instruments for minimal invasive surgery in modular conception. Acta Chir Belg 93: 7377 30. Schurr MO, Buess G, Rininsland H, Holler E, Neisius B, Voges U (1996) ARTEMIS: Manipulatorsystem fu r die endoskopische Chirurgie. Endoskopie Heute 9: 245251 31. Schurr MO, Breitwieser H, Melzer A, Kunert W, Schmitt M, Voges U, Buess G (1996) Experimental telemanipulation in endoscopic surgery. Surg Laparosc Endosc 6: 167175