Case reports

Surg Endosc (1997) 11: 277279

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic resection of posterior gastric leiomyoma

I. M. Ibrahim,1 F. Silvestri,1 B. Zingler2
1 2

Section of Laparoscopy, Department of Surgery, Englewood Hospital and Medical Center, 375 Engle Street, Englewood, NJ 07631, USA Section of Gastroenterology, Department of Medicine, Englewood Hospital and Medical Center, 375 Engle Street, Englewood, NJ 07631, USA

Received: 17 February 1995/Accepted: 7 September 1995

Abstract. Laparoscopic gastric surgery is gaining momentum, especially in the treatment of benign disease. Simultaneous endoscopy and laparoscopy allow precise localization of lesions. Because of the stomachs size, mobility, and distensibility, relatively large lesions can be safely excised. Wedge resection for anterior lesions and a transgastric or intragastric approach for posterior lesions are feasible laparoscopically. Two cases of posterior gastric leiomyomas successfully resected laparoscopically are presented. The use of stapling devices greatly facilitates this procedure. Key words: Gastric leiomyoma Laparoscopy Resection

upper posteromedial aspect of the fundus of the stomach. Pathology revealed a well-encapsulated 4.5 3 cm leiomyoma with superficial mucosal and tumor ulceration, multifocal hemorrhage and necrosis. There were zero to one mitoses per ten HPF (Fig. 2). Liquid intake was instituted on the 3rd postoperative day and the patient was discharged on the 5th. A superficial infection at the specimen retrieval site was drained and healed uneventfully.

Case II A 61-year-old male presented with syncope, profound anemia, and gastrointestinal bleeding. Endoscopy revealed a leiomyoma of the posterior wall of the stomach. Following transfusion, the patient underwent semielective laparoscopic transgastric resection of the lesion. The tumor was located on the midposterior gastric wall. Pathology revealed a 5 3.8 cm leiomyoma with a central 5-mm umbilical ulcer (Fig. 3). No mitoses were seen. Liquid intake was started on the 3rd postoperative day and he was discharged on the 5th. Superficial infection of two port sites developed but healed following drainage.

Benign gastric leiomyoma is an uncommon lesion that usually presents with upper gastrointestinal bleeding [1, 6]. Preoperative endoscopic identification of the lesion allows proper planning for laparoscopic resection. The location of the lesion in the stomach should not generally preclude such an approach. We present two cases of posterior gastric leiomyomas managed in this fashion.

Case studies

Method
The patient was placed in the supine position with the surgeon standing between the legs. Four 10-mm port sites (umbilical, right upper quadrant, left upper quadrant, and right epigastric) were used. Simultaneous endoscopy and laparoscopy (Fig. 4) allowed precise localization of the lesion with respect to the anterior wall of the stomach. Anterior gastrotomy was made with electrocautery and enlarged with Endo GIA30 (USSC, Norwalk, CT). (Fig. 6A). The lesion protruding from the posterior wall was identified. Endobabcocks (USSC, Norwalk, CT) were applied proximally and distally to the lesion, thus elevating it through the anterior gastrotomy. In case II, brisk arterial bleeding was encountered from the tumor (Fig. 5) as the Endobabcocks were being applied. The bleeding was directed away from the scope and resection was resumed. The lesion was resected by sequential application of an Endo-GIA30, which

Case I A 54-year-old female presented with melena. Upper endoscopy revealed an upper posterior gastric leiomyoma. CT scan (Fig. 1) confirmed the site of the lesion. Following 1 day of antibiotic bowel preparation, laparoscopic transgastric resection was performed. The tumor was located on the
Correspondence to: I. M. Ibrahim

278

Fig. 4. Simultaneous endoscopic and laparoscopic appearance (inset) of mass. Fig. 5. Tumor is elevated through anterior gastrotomy. Note bleeding from the dome of tumor.

by suture. It was then stapled with the Endo-GIA60 (Fig. 6C). Frozen sections were performed to rule out malignancy. The stomach was then distended with Methylene Blue and saline to ensure the integrity of all staple lines. The abdomen was then irrigated and the port sites were closed.

Discussion Benign tumors of the stomach account for less than 2% of all surgically excised gastric neoplasms [1]. They comprise over 80% of symptomatic smooth muscle tumors of the GI tract [6]. Leiomyomas arise from smooth muscle cells and grow exophytically, or, more commonly, in an endogastric direction. Small leiomyomas are clinically insignificant, but as they enlarge they can cause outlet obstruction if close to the pylorus, or, more commonly, ulceration of mucosa and bleeding that at times may be massive. Clinically they cannot be differentiated from leiomyosarcoma except that the larger masses tend to be malignant. Confirmation of the benign nature of the lesion is mandatory either prior to or at the time of surgery. Endoscopically the lesion is seen pro-

Fig. 1. CT of abdomen showing endogastric mass. Fig. 2. Microscopic (400 ) appearance of leiomyoma. Note mitotic figure (arrow). Fig. 3. Microscopic (20 ) appearance of leiomyoma. Note ulcer (arrow).

simultaneously staples and cuts the posterior gastric wall (Fig. 6B). Any bleeding sites through the staple line were easily electrocauterized. The lesion was then placed in an Endocatch (USSC) and delivered through the right upper quadrant port, which was slightly enlarged to accommodate the size of the tumor. The edges of the anterior gastrotomy were approximated

279

truding into the lumen of the stomach with overlying ulcerated mucosa. CT scanning can further localize the tumor and detect metastatic lesions when present. Malignant lesions require more extensive surgical extirpation that is more demanding with our present laparoscopic techniques. Wedge resection with a margin of normal tissue is sufficient for the benign ones. Different approaches to resection of benign gastric tumors have been utilized depending on the location of the lesion. For lesions on the anterior surface or either curvature wedge resection is optimum [5, 8]. For posterior mid or distal gastric lesions, dissection of the greater curvature and resecting through the lesser sac is feasible [7]. Intragastric resection [3, 4] is attractive especially for small lesions, but may be difficult for large lesions or if bleeding should ensue during the surgery, as occurred in one of our patients. The transgastric approach is simple [2]. The anterior gastrotomy allows wide exposure of the inner stomach. Instruments can be introduced and exchanged without difficulty. The lesion can be grasped and elevated adequately to apply the staplers. These devices simplify the procedure while affording excellent hemostasis. We find it simpler to use staplers for the initial gastrotomy. The use of the GIA30 allows for more flexibility in resecting around the tumor, thus excising a smaller margin of normal tissue.

References
1. Akwari OE (1991) Benign tumors of the stomach. In: Sabiston DC (ed) Textbook of surgery. W.B. Saunders, Philadelphia, P pp 790 791 2. Basso N, Silecchia G, Materia A, Fantini A, Genco A, Surgo D, Pizzuto G (1994) Laparoscopic transgastric excision of a leiomyoma of the posterior wall of the gastric fundus. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 599 3. Hashimoto S, Munakata Y, Hayashi K, Sarvano S, Kawasaki S, Makuuchi M (1994) Laparoscopic intraluminal resection for the submucosal tumor in the cardia. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 547 4. Kanehira E, Mori A, Watanabe T, Ishikawa T, Ishikawa N, Yoshino Y, Omura K. (1994) A technique of laparoscopic intragastric surgery in the treatment of gastric carcinoma in situ. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 547 5. Llorente J (1994) Laparoscopic gastric resection for gastric leiomyoma. Surg Endosc 8: 887889 6. Morgan BK, Compton C, Talbert M, Gallagher WJ, Wood WC (1990) Benign smooth muscle tumors of the gastrointestinal tract. A 24 year experience. Ann Surg 211: 6366 7. Ming Q, Yanning S, Zhangwei K, Junyi S. (1994) Laparoscopic resection of gastric leiomymas. Abstracts of 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 6008. 8. Otani Y, Ohgami M, Hoshiya Y, Kutoba T, Kumai K, Kitajima M (1994) Laparoscopic wedge resection of the stomach for carcinoid tumor using a lesion lifting method. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 546

Fig. 6. Schematic pictures: A anterior gastrotomy, B application of EndoGIA, C closure of gastrotomy.

Guidelines
Surg Endosc (1997) 11: 308314

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Guidelines for conducting economic outcomes studies for endoscopic procedures

developed by Dr. Alfred Cuschieri, President European Association for Endoscopic Surgery (EAES) Dr. Eugenio Ferreira, President Federacion Latinoamericana de Cirugia (FELAC) Dr. Peter Goh, President, Endoscopic and Laparoscopic Surgeons of Asia (ELSA) Dr. Yasuo Idezuki, President, Japan Society for Endoscopic Surgery (JSES) Dr. Guy Maddern, President, Endosurgery Group, Royal Australasian College of Surgeons (RACS) Dr. Gerald Marks, President, International Federation of Societies of Endoscopic Surgeons (IFSES) Dr. Greg Stiegmann, President, Society of American Gastrointestinal Endoscopic Surgeons (SAGES) Dr. Bryce Taylor, President, Canadian Association of General Surgeons (CAGS)

Preamble The revolution in laparoscopic and thoracoscopic minimally invasive surgical technology and techniques has introduced a constellation of new issues, not the least of which is economic outcomes and cost benefits. The IFSES Executive Committee and Corporate Council, recognizing that cost will be addressed by interested surgeons, insurance carriers, HMOs, managed care directors, and governmental legislative bodies, decided to develop a proper approach to the problem. The Corporate Council allocated the funds to enlist the services of two separate, authoritative consulting groups: McKinsey & Company, Inc., and the Deloitte & Touche Consulting Group. Working with representative surgeons from each of the component societies of IFSES, the consultants developed the following document that should be helpful as a guide to parties interested in evaluating economic outcomes. The Executive Committee of IFSES, the members of the IFSES Corporate Council, and the executives of the component societies and associations take pride in presenting this document, the results of a twoyear study.

introduction, the lap chole became the standard procedure offered to patients. Since the acceptance of lap chole, a host of other endoscopic procedures have been developed with the hope of capturing the many benefits offered by the minimally invasive technique. While many of these procedures are currently available, they have not been widely accepted by the medical community due to the absence of a complete clinical and economic fact base to support their adoption. While studies on the clinical aspects of the newer procedures have been widely published in the international literature, attempts to publish economic outcomes studies on these procedures have created a high level of confusion about the procedures costs and benefitsthe inconsistent study designs and approaches to defining economic variables have led to conflicting and inconsistent results in the published literature. In addition, relatively few outcome studies have been conducted and published. To reverse this trend, this Guidelines Document was developed to provide guidance to researchers interested in conducting clinical and economic outcomes studies for endoscopic procedures. The remainder of this document will cover three areas: 1. General approach for assessing clinical and economic impact of endoscopic procedures 2. Special considerations for study design 3. Case example: applying guidelines to conducting an outcomes study for laparoscopic inguinal herniorrhaphy

Laparoscopic cholecystectomy (lap chole) was a technological advance that gained widespread adoption because it offered clinical efficacy and safety combined with lower cost for the hospital and payor. Within a few years of its

309

1. General approach for assessing clinical and economic impact of endoscopic procedures To encourage acceptance of clinically efficacious procedures that are also financially viable, economic outcomes studies should be conducted using uniform guidelines that facilitate the development of a clear and credible fact base. To help build this fact base and to promote greater comparability across published reports on clinical and economic outcomes, this document will outline the two key elements of conducting and documenting the economic outcomes: What to measurethe economic criteria that should be used to assess the merits of a procedure How to collect/develop data basethe process for producing statistically credible and appropriately normalized study results

What to measure The total economic impact of a surgical procedure includes the direct medical expenses charged to the payor (i.e., health service), the patients lost wages, and the employers lost revenues and/or incremental costs incurred due to the patients absence. To capture all these elements, the section below groups the cost items into two categories: (1) direct medical costs, which include the initial medical care and treatment and any complications or recurrences caused by the treatment; and (2) indirect costs, which are related to a patients surgery-driven absence from work/normal activity. The framework outlined below is intended to be comprehensive so that all procedures can be assessed using the same basic approach. Accordingly, not all cost items are applicable for each procedure and should not receive the same level of focus or detailed examination. In fact, each outcomes study should focus on measuring the five to 15 areas that have the greatest impact on the economic outcomes of the procedure in a detailed manner. 1. Direct medical costs should include all short-term expenses incurred during the delivery of the medical procedure (e.g., when the patient is in the hospital) and the follow-on costs incurred after the patient leaves the care facility. a. Short-term direct costs. Short-term direct costs are incurred from the procedure and any complications that arise during the time between admission to and discharge from the hospital. These costs can be divided into professional labor, supplies, diagnostic tests, and capital requirements. Studies should consider both the quantity of items used and the cost associated with each item in the categories listed below. This level of detail is critical to facilitating comparisons among institutions whose cost structures and accounting approaches may differ significantly. Professional labor. Measurement of professional labor requirementsthe number of hours required and cost per hourneeds to differentiate the skill types of personnel associated with the surgery and subsequent care: surgeon, anesthesiologist, surgeons assistant, nurse, etc.

Supplies. Measurement of supply inputs should focus on the quantity used and cost of each supply type. Supplies consist of medical instruments, which are either reusable or disposable, and various disposable supplies (e.g., pharmaceuticals, IV solutions, etc.) Medical instruments. Because hospitals typically mark up these critical supplies at widely varying rates, and because instruments are billed in a number of different ways, it will be important to conduct two measurements of equipment cost cost to the hospital and charge to the patient or payor. 1. The quantity measurement of medical instruments should identify each type of productnot a specific manufacturer or whether the product is reusable or disposable (i.e., number of 10-mm trocars). 2. Cost to the hospital for a disposable instrument consists of the price paid for the instrument. For reusable instruments, the cost measurement should include all direct labor costs for cleaning and preparing the instruments plus a fair allocation of the purchase price and overhead. Labor costs consist of some allocated portion of staff time and materials; allocation of the purchase price and required back-up instruments should be based on the expected use lifetime of the reusable instruments (i.e., number of uses before the instrument is broken, a part becomes missing, etc.). 3. Charge to the payor for a disposable instrument consists of the amount billed for the instrument. This figure is equal to the purchase price of the instrument multiplied by some markup factor, which is determined by the hospital for each instrument. For reusable instruments, which are generally aggregated with other charges, the study should use the instrument cost to the hospital multiplied by the same markup factor as the corresponding disposable. The method discussed in Exhibit 1 could be helpful in making this calculation. Other supplies. All other supplies should be measured as patient charges obtained from the bill and the hospitals information systems. Diagnostic tests. Tests should be measured in the same way as suppliesquantities and charges as posted to the patient bill. These sometimes include separate charges for administration of the test and interpretation of the results. Capital. For most surgeries, the most significant capital items are charges for the operating theater, ICU bed, and hospital bed (i.e., length of stay). These charges are typically posted to the patient bill, but may also be calculated based on the patients length of stay and the charge for each hour/day in the facility. These charges should include a fair allocation of overhead (e.g., nursing, administration, utilities, janitorial). All other capital items should be measured on the same basis as suppliesquantities and charges as posted to the patient bill. b. Follow-on direct costs. Follow-on direct costs consist of costs incurred after the initial procedure. Examples

310

Exhibit 1. Calculation of cost and charge for reusable instruments.

of follow-on cost items include pain relievers, followup visits for complications management, etc. 2. Indirect medical costs should take into account the surgery- and convalescence-related costs of absence from work/normal activity. While the outcomes studies may choose not to measure these indirect costs, they need to present the one critical piece of information required for interested parties to calculate their cost impactthe recovery time required before the patient can resume normal activity. a. For employers, these indirect costs may include replacement workers, lost revenues (opportunity cost), and disability costs. Not all cost items are applicable to all employers (e.g., professional service firms such as law firms are unlikely to incur replacement worker costs but will instead lose revenues from having a professional out of work). b. For payors of workers compensation and disability, indirect costs may include payments made to a recovering worker. c. For those patients who do not have workers compensation/disability coverage, indirect cost may be lost wages. How to collect/develop database To develop a comprehensive fact base, clinicians are likely to need a new process to collect the required dataa process that combines information currently tracked in the hospitals information systems with new tools to track other information not currently captured. This section discusses the likely sources for required information for a detailed economic outcomes study: (1) any existing hospital information systems; (2) additional information to be collected by clinicians; and (3) additional information to be collected by patients.

1. Existing hospital information systems. While some institutions have integrated information systems that link together all information on a patient, most hospitals today are likely to have at least three separate information systems: a clinical/medical patient record system, an operating room system, and a financial system. Each of the three systems provides a portion of the valuable information needed to conduct economic outcomes studies. a. Clinical/medical patient record systems typically track the course of the patients stay, except for the details of surgery tracked by the OR system. Most systems track the following: Case number, admission date, medical record number, physician IDall needed to match with information from other systems Patient demographics (age, gender, etc.) Patient type (inpatient, outpatient, observation, etc.) Diagnosis codes (ICD-9 codes) Information on the procedures performed and resources consumed Total charges and, at times, costs Some hospitals with more sophisticated systems may track the following information as well: Length of stay (med/surg) Length of stay (ICU) Length of stay (other) Discharge status (home, nursing home, deceased, etc.) Readmission code (including return to OR) b. Operating room information systems typically track only information about a surgery case. Nearly all OR systems track the following data, which provides basic information on labor and supplies used during the surgery. Case number, admission date, medical record

311

number, and physician ID (needed to match with information from other systems). Room enter time and room exit timeto calculate OR usage time. Procedure begin time and end time, which could be used to calculate labor requirements. Anesthesia type, begin time, and end time. Some OR systems track additional information that could bring greater clarity to OR charges. Case type (elective, emergency, etc.) Case designation code (teaching, etc.) Scheduled and actual procedure code/name could be used to determine rate of conversion from endoscopic to open procedures. OR supply cost and revenuescould be used to itemize required supplies and related costs/ charges (may provide duplicate information to that captured in the hospitals financial systems). Severity/acuity rating of patients presenting conditioncould be used to normalize for variation among cases. Intraoperative and postoperative complicationscould be used to identify rate of complications associated with a procedure c. The financial information system tracks the detailed inputs required for the patients care during the stay capturing information at a more detailed level than that presented on the patient bill. This system typically captures the quantity and charge for all inputs such as chest x-rays, 1-hour block of OR time, 10-mm trocars, individual doses of pharmaceuticals, etc. The key challenge in working with information from existing information systems will be manipulating the potentially large volume of information for analysis and grouping the costs into the categories discussed above. 2. Additional information to be captured by clinicians. To supplement the data provided by the hospitals information systems, clinicians will likely need to keep additional records to help with the subsequent data analysis. This is likely to involve creation of a worksheet to be used at the time of surgery and may involve some of the following information: Case number, admission date, medical record number, physician IDneeded to match worksheet with information from hospitals information systems Basic information about the patient (e.g., age, sex, profession, etc.) Type of procedure (e.g., traditional vs endoscopic, severity of medical condition, etc.)this will help to normalize the data during the analysis stage For each case, the surgeons experience with performing the procedure (e.g., fifth time with procedure) For all individuals that participate in the surgery, list the name, type of individual (surgeon, internist, nurse, etc.), time entering OR, and time exiting OR. This is needed to determine the actual OR labor content. 3. Information to be collected by patients. Although existing outcomes studies pay little attention to events after discharge, developing a complete understanding of a procedures economic impact will require greater knowl-

edge of what happens to the patient after he/she leaves the care facility. To develop this fact base, a procedurespecific questionnaire could be developed for patients to complete once home. Items contained on this questionnaire could include: Questions about bodily function that the patient would fill out each day after the surgery (for a specific time period appropriate for each procedure). This is needed to determine the time required to the patient to recover and resume normal activity after a procedurefrom a medical-necessity standpoint. Quantities of medications and other supplies consumed each day after discharge. Amount of time spent each day with home-care professionals or with the surgeon or other physician (if applicable). Qualitative questions about the patients health to develop a more subjective quality of life assessment. The obvious challenge of conducting these economic outcomes studies is the sheer amount of data that will need to be analyzed and interpreted. While various sophisticated statistical approaches could be used, the basic analysis should determine the average and standard deviation for the critical measures and whether the results obtained for traditional vs laparoscopic procedures are statistically significant. 2. Special considerations for study design Adoption of the guidelines outlined above will significantly improve understanding of the economic costs and benefits of endoscopic surgery. To further enhance these studies and maximize their value, a number of other factors should be considered as part of the study design. 1. Learning curve effect. To avoid the confusion created in the past by comparing physicians early results from endoscopic procedures to results from well-practiced open procedures, two actions should be taken: a. A surgeons first 2050 times performing a new endoscopic procedure (this number will vary depending on the procedures difficulty) should be considered to constitute the learning curve. Results from those procedures performed while the surgeon is still learning the procedure should be reported in groups of five (e.g., first five surgeries, second five, etc.), but only to demonstrate the improvement that results from practice with the endoscopic approach. These results should not be compared with those from the wellpracticed open approach. b. Results from procedures beyond a surgeons first 20 50 surgeries should be presented as outcomes for actual endoscopic vs open comparisons. 2. Statistical credibility. To demonstrate accurately the expected range of results for a given procedure, several steps should be taken: a. A researcher should consider the fact base to be adequate only if results are available from 2550 patients receiving the open procedure and 2550 patients receiving the endoscopic procedure. Once this many cases are available, the researcher should ana-

312

lyze the data and prepare the publication. When at least five comparably designed publications are available for a procedure, a review article should be prepared to summarize all findings. This article should serve as the definitive documentation for comparing the endoscopic vs open procedures. b. To account for differences in utilization controls and other factors that may vary by institutional settings, surgeons from a variety of care settings should conduct economic outcomes studies. The ideal participation level would involve two publications from each core care setting (e.g., community hospital, academic medical centers, outpatient surgery centers, etc.). Geographical distribution of institutions selected should also be considered. c. Similarly, for procedures that can be performed using two or more techniques that differ substantially in surgical approach, instruments used, or support required, an adequate number of cases should be included to obtain statistically viable results for each technique. 3. Presentation of economic data. To provide greater clarity and consistency in presenting the results of economic outcomes studies, publications should take two steps in the future: a. When presenting their results, authors should aggregate their detailed findings to three major cost groups: Direct medical costs. Professional labor. The research should summarize the time, cost, and associated charge from the surgeon(s), other physicians, nurses, and other technical support. Supplies. Major categories for the supplies summary should include quantity (when appropriate), cost, and charge for surgical instruments, other OR supplies (if appropriate), pharmaceuticals, and other medical supplies as needed. Diagnostic tests. Quantity, cost, and charge for appropriate tests required for the procedure should be presented. Capital use. The research should summarize the usage time for the capital equipment used for the procedureincluding hospital length of stay, time in OR, recovery room, and ICU, and any capital equipment use (e.g., videoscopes). Indirect patient costs. Time to resume normal activity. The researcher should report the medically required time to resume normal activity for the traditional and endoscopic approaches. Average cost to patient for each lost day of work. The researcher should ask the patients and present findings on whether and how much the patient loses each day he/she is prevented from working. Indirect employer costs. The researcher should present findings on costs to the employers for each day a patient is prevented from returning to work (e.g., lost revenues, replacement worker costs, etc.) This information, coupled with the time to resume normal activity (above), should provide a complete understanding of the costs to employers.

b. As support for their aggregated findings, papers should also provide full detail of the individual items measured (quantities, cost, and charge if measures are available). Perhaps an effective method is to create an average patient bill (Exhibit 2). This average patient bill would provide all appropriate line items to enable the readers to understand the details behind the key differences between endoscopic vs open approaches or between early and late experience with a procedure. The author should explain the methods used to measure them and any assumptions made or estimates used.

3. Application of guidelines to inguinal herniorrhaphy The use of endoscopic techniques in hernia repairs has been controversialthis has been the principal result of the absence of an accurate fact base on the procedures true economic impact. To build the definitive fact base and assess the merits of the traditional vs laparoscopic herniorrhaphy, the guidelines discussed above could be applied to design the appropriate economic outcomes study. To illustrate how the guidelines could be used, this section will outline the critical aspects that must be measured by a well-designed economic outcomes study for inguinal herniorrhaphy. Classification of hernia types. Since there are different forms of inguinal hernias and approaches for treating them, identifying the subgroups and determining the accurate measures for each subgroup are critical. Accordingly, hernia cases should be classified along the following three dimensions: Unilateral, bilateral, or recurrent hernia Size of herniasmall (<XX mm) vs large (>XX mm) Procedure typelaparoscopy, shouldice, other Critical criteria to be measured. In using the principles set forth in the guidelines, the herniorrhaphy study would need to assess the direct medical costs, the indirect costs to patients, and the indirect costs to employers. Direct medical costs. The study should estimate the total hospital bill for the laparoscopic vs open procedures by following the approach described above. Particular attention should be paid to specific data findings for the following variables, which should be presented: OR timetime required from each person (e.g., surgeon, assistants, nurses, etc.) to conduct the procedure Instrumentation costsnumber and cost for each major type of medical instrument used (e.g., trocars, graspers, etc.) Anesthesia costtype of anesthesia (e.g., local vs general) and cost Hospitalization timenumber of inpatient hospitalization days Indirect patient costs Number of days required to return to normal daily activitybased on clinical measures (e.g., ability to climb stairs, pain level, etc.) Number of days before patient actually returns to work

313

Exhibit 2. Suggested presentation of economic data.

314

Average cost to patient per day absent from work Indirect cost to employer Estimated average cost to employer of having an employee absent from work Data collection. To collect all required data, the researcher will need to have access to the hospitals information systems (e.g., patient bills, supply costs, etc.) as well as use simple surveys to collection information from the surgical team and the patient. Survey instruments should be used to capture certain information from the surgeon at the time of surgery and from the patient after the surgery. For hernia, these instruments should include the type of hernia, type of surgery used, and patients ability to resume usual daily activities. Data from the hospitals system and from survey responses should be combined into a database that could

be analyzed by an appropriate statistical analysis system (e.g., SAS, SPSS, JMP, etc.). The researcher should focus on developing the total economic impact of the laparoscopic vs open procedures, present the average standard deviation for the key variables presented above, and determine whether differences between the two approaches are statistically significant. Since the whole field of economic outcomes research is relatively new, we hope that the preparation of this Guidelines Document provides a useful starting point for researchers interested in economic outcomes research. We recognize that this document is likely to be incomplete. However, we did want to provide this early document to provide guidance for interested researchers until a more inclusive revision can be published.

Editorial
Surg Endosc (1997) 11: 225

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Combining minimal access procedures expands the potential of laparoscopic surgery

Performing flexible endoscopy as part of a general surgical procedure is not a new concept. The value of intraoperative gastroscopy and colonoscopy has long been recognized in selected open abdominal operations for localization of lesions in the small bowel or colon and for identification of bleeding points. In this issue of Surgical Endoscopy, new and innovative applications of combining intra-operative gastroscopy with laparoscopic surgery are described [15]. Five teams of surgeons have used gastroscopy and laparoscopy to treat large leiomyomas and benign polyps of the stomach and esophagus which otherwise would have required a formal laparotomy or thoracotomy. Two techniques for these combined endoscopic and laparoscopic procedures are described. The most commonly performed technique is laparoscopic wedge resection of the stomach. Gastroscopy is first performed to localize the lesion. Then the region of the stomach with the tumor is grasped and elevated with Tfasteners or with Babcock forceps and resected with an endoscopic stapler (30 mm or 60 mm) using one or several applications as appropriate. This approach is suitable for most tumors of the anterior wall and the accessible portions of the lesser and greater curvature of the stomach. This stapled resection technique can also be used for some tumors of the posterior wall [1]. In this instance, an anterior gastrostomy is made with cautery and/or an endoscopic stapler for exposure and the lesion with adjacent stomach is grasped, elevated, and resected. The second technique is laparoscopic intragastric surgery with gastroscopy to guide trocar placement [2]. The stomach is insufflated via the endoscope and special 5 mm and 12 mm trocar-cannulas with balloons on the shaft are introduced through the abdominal wall directly into the distended stomach. These balloon cannulas hold the stomach against the abdominal wall and provide access to the lumen for standard laparoscopic instruments. An epinephrine solution is injected into the submucosal layer first for hemostasis. The tumor with the overlying mucosa is directly dissected from the gastric wall and removed either with a gastroscope or through a cannula. This technique of mucosal resection has also been described for early gastric cancers with good results [6]. For the esophageal leiomyomas, a thorascopic approach through the right thorax was used. Endoscopic trans-illuminations of the esophagus identified the tumor and guided the dissection. Choice of approach and the location of tumor The location of the gastric lesion determines which laparoscopic approachwedge resection versus intragastric dissectionis favored. Laparoscopic wedge resection is most suitable for mobile lesions in the anterior wall and for lesser and greater curvature lesions of the stomach which generally can be elevated and excised with sequential applications of an endoscopic stapler. Wedge resection is also possible for some lesions of the posterior wall using a transgastric approach. The direct intragastric laparoscopic approach with gastroscopy guidance is preferred by some for large benign tumors, for posterior wall tumors and for tumors near the cardia and the pylorus. Precautions The authors emphasize that these laparoscopic procedures are most appropriate for benign tumors of the stomach and esophagus. The histology of the tumor should be known prior to operation, if possible, and confirmed by frozen section at the time of resection. Large tumors, 45 cm in diameter, can be resected by these techniques, but may be malignant when full thickness biopsies are studied. Therefore preparations should be made for a more extended gastrectomy. For laparoscopic intragastric mucosal resection of early gastric cancer, specific characterization of the size, the histology, and depth of penetration of the tumor, (usually by biopsy and ultrasound), is mandatory. Finally, laparoscopic wedge resection with an endoscopic stapler should be used with caution near the cardia and pylorus, where the lumen can be compromised, and in the posterior wall of the stomach where the pancreas could be injured. References
1. Ibrahim IM, Silvestri F, Zingler B (1997) Laparoscopic resection of posterior gastric leiomyoma. Surg Endosc 11: 275277 2. Gurbuz AT, Peetz ME (1997) Brief clinical report: Resection of a gastric leiomyoma using combined laparoscopic and gastroscopic approach. Surg Endosc 11: 285286 3. Taniguchi F, Kamiike W, Iwase K, Nishida T, Akashi A, Ohashi S, Matsuda H. (1997) Thoracoscopic enucleation of a large leiomyoma located on the left side of the esophageal wall. Surg Endosc 11: 280282 4. Siu WT, Leong HT, Li MK (1997) Laparoscopic resection of bleeding gastric polyps. Surg Endosc 11: 283284 5. Taniguchi E, Kamiike W, Yamanishi H, Ito T (1997) Laparoscopic intragastric surgery for gastric leiomyoma. Surg Endosc 11: 285287 6. Ohgami M, Otani YA, Kumai K, Kubota T, Kitajima M (1996) Laparoscopic curative surgery for early gastric cancer. Surg Endosc 10: 251

G. M. Larson
Department of Surgery School of Medicine University of Louisville Louisville, Kentucky 40292, USA

Surg Endosc (1997) 11: 285286

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Brief clinical report

Resection of a gastric leiomyoma using combined laparoscopic and gastroscopic approach
A. T. Gurbuz, M. E. Peetz
North Colorado Medical Center, 2020 16th St, Greeley, CO 80631, USA Received: 15 December 1995/Accepted: 22 April 1996

Abstract. Laparoscopy has added a new perspective to the diagnosis and treatment of abdominopelvic disease. A wide variety of gastric procedures have been completed with laparoscopy in the past several years. The authors here present successful resection of a submucosal gastric leiomyoma laparoscopically with the combined use of intraoperative gastroscopy for localization. A 2.5 2.0 cm submucosal gastric nodule is resected with ample margins laparoscopically. Intraoperative endoscopy is used for accurate localization because the lesion was not visible to the laparoscope on the serosal surface of the organ. Laparoscopic surgery can be applied to the traditional surgical principles with equal efficacy in selected patients. Key words: Laparoscopy Gastroscopy Endoscopy Smooth muscle tumors Stomach

Laparoscopic resection is a safe alternative to laparotomy for wedge resection of accessable gastric leiomyomas. Case report
An 88-year-old female presented with symptoms of postprandial epigastric fullness and dyspepsia of 2-month duration. She was given a 6-week course of antacidH2-blocker treatment which did not improve her symptoms. A barium upper gastrointestinal series showed a luminal filling defecy without ulceration. She subsequently underwent an esophagogastroduodenoscopy which showed a 2.5 2.0 cm submucosal nodule in the corpus of the stomach on the anterior surface midway between greater and lesser curvature (Fig. 1). There was no ulceration or bleeding on the lesion. Several biopsies of the overlying mucosa showed chronic inflammation without definitive diagnosis. She was referred to surgery for resection. Because of the anterior location of the lesion, a laparoscopic resection is planned. The patient is taken to the operating room; general inhalational anesthesia is induced. CO2 pneumoperitoneum is achieved using the Veress needle. A laparoscope is inserted. Initial exploration did not show a deformity or mass on the serosal surface of the stomach. The liver appeared normal. No other intraabdominal pathology is identified. At this time a gastroscope is inserted into the stomach and the lesion is localized. The light of the endoscope was easily visible through the serosa and the nodule was identified as a dark shadow easily differentiated from the surrounding normal gastric tissue. The lesion is grasped with an endo-Babcock grasper and the position again was confirmed with the gastroscope. A 16-mm port is then placed in the right upper quadrant and an Endo-GIA 60 is inserted through the sheath. The stapler is fired on each side, approximately 2 cm beyond the nodule. The mass is resected with 2 cm of wedge of normal stomach tissue. A specimen is taken out from the right upper quadrant port without difficulty. Complete removal of the lesion is again confirmed with the gastroscope (Fig. 2). There was no air leak at the suture line following intragastric air insufflation. The patient tolerated the procedure well. Nasogastric drainage is discontinued on the 1st postoperative day. She was given liquids the same day, which was rapidly advanced to soft diet. She was discharged home on 2nd postoperative day. Pathological examination of the specimen revealed well-differentiated smooth muscle cells without malignant features suggesting a benign leiomyoma.

Leiomyomas are the most common benign tumors of the stomach found at autopsy although they constitute only 2.5% of gastric neoplasms in the surgical series [10, 12]. Most are benign and remain asymptomatic. They can on occasion enlarge and undergo central necrosis, which may result in massive upper gastrointestinal bleeding. Differentiation of benign leiomyomas from the malignant leiomyosarcomas is also important. Leiomyomas are usually diagnosed as an incidental filling defect at the time of gastroscopy or barium upper gastrointestinal series. The presenting symptoms, if any, are postprandial fullness and gastrointestinal bleeding. Recommended treatment is wedge resection for tumors smaller than 4 cm in diameter. Resection with wide margins is required for tumors larger than 4 cm [15].

Correspondence to: A. T. Gurbuz, Saint Joseph Hospital Medical Center, 1835 Franklin Street, Denver, CO 80218, USA Presented at the 47th annual meeting of the Southwestern Surgical Congress, April 1995, San Antonio, TX

Discussion The spectrum of laparoscopy is enlarging every day. Laparoscopy is being used increasingly in the diagnosis and

286

Fig. 1. Endoscopic view of submucosal gastric nodule.

Fig. 2. Endoscopic view of gastric wound after removal of nodule. 3. Congreve DP (1992) Laparoscopic paraesophagial hernia repair. J Laparendosc Surg 2: 4548 4. Hashimoto S, Munakata Y, Hayashi K, Sarvano S, Kawasaki S, Makuuchi M (1994) Laparoscopic intraluminal resection for the submucosal tumor in the cardia. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 547 5. Kanehira E, Mori A, Watanabe T, Ishikawa N, Yoshino Y, Omura K (1994) A technique of laparoscopic intragastric surgery in the treatment of gastric carcinoma in situ. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 547 6. Lacy AM, Tabet J, Grande L, Garcia-Valdecasas JC, Fuster F, Delgado S, Visa J (1995) Laparoscopic assisted resection of a gastric lipoma. Surg Endosc 9: 995997 7. Llorente J (1994) Laparoscopic gastric resection for gastric leiomyoma. Surg Endosc 8: 887889 8. Lointier P, Leroux S, Ferrier C, Dapoigny M (1992) A technique of laparoscopic gastrectomy and Billroth II gastro-jejunostomy. J Laparendosc Surg 2: 331334 9. Lukaszczyk JJ, Preletz RJ (1992) Laparoscopic resection of benign stromal tumor of stomach. J Laparendosc Surg 2: 331334 10. Marshall SF (1975) Symposium of surgery of digestive tract. Gastric tumors other than carcinoma. Surg Clin North Am 35: 693697 11. Ming Q, Yanning S, Zhangwei K, Junyi S (1994) Laparoscopic resection of gastric leiomyomas. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 600 12. Morgan BK, Compton C, Talbert M, Gallagher WJ, Wood WC (1990) Benign smooth muscle tumors of the gastrointestinal tract. A 24 year experience. Ann Surg 211: 6366. 13. Otani Y, Ohgami M, Hoshiya Y, Kutoba T, Kumai K, Kitajima M (1994) Laparoscopic wedge resection of the stomach for carcinoid tumor using a lesion lifting method. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 546 14. Parikh SS, Kapadia KB, Mandke JV, Prabhu SR, Pai MV, Desai HG (1992) Leiomyoblastoma of the stomach. J Assoc Physicians India 40: 133134 15. Sabiston DC (ed) (1991) Textbook of surgery. 14th ed. WB Saunders, Philadelphia

treatment of a variety of gastroesophagial pathology [2, 3, 8]. Laparoscopic resection of a gastric lipoma [6] and a gastric carcinoid [13] has been described. Resection with the endoscopic staplers can be accomplished easily for small and well-circumscribed gastric stromal tumors. Lesions located on the posterior wall of the organ can be excised either through the gastrocolic omentum or via transgastric approach [1, 4]. Laparoscopic intragastric resection of a gastric carcinoma in situ has also been described [5]. Smooth muscle tumors of the stomach can be both diagnosed [14] and resected [9, 11] via laparoscopy. Simultaneous intraoperative gastroscopy can be used as an adjunct for transillumination of the lesions that are not visible on the serosal surface of the organ [7]. This will not only aid in localization of the lesion but also will confirm complete resection. This is the second case of combined laparoscopic and gastroscopic resection of a gastric smooth muscle tumor reported in the literature. Laparoscopy with or without simultaneous endoscopy is an acceptable alternative for resection of gastric stromal tumors. References
1. Basso N, Silecchia G, Materia A, Fantini A, Genco A, Surgo D, Pizzuto G (1994) Laparoscopic transgastric excision of leiomyoma of the posterior wall of the gastric fundus. Abstracts of the 4th World Congress of Endoscopic Surgery. Surg Endosc 8: 599 2. Colin-Jones DG, Rosch T, Dittler HJ (1993) Staging of gastric cancer by endoscopy. Endoscopy 25: 3438

EndoScope: world literature reviews

Surg Endosc (1997) 11: 290294

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Original articles from a wide range of international surgical journals are selected by our editors and presented here as a structured summary and critical review. EndoScope serves as a quick and comprehensive survey of the expansive endoscopic literature from all the corners of the globe.

Section Editor: J. M. Sackier

Laparoscopic techniques for fecal diversion

lecystectomy was never scrutinized using the standards of prospective randomized trials.

Laparoscopic rectopexy using mesh fixation with a spiked chromium staple

Solomon MJ, Eyers AA Dis Colon Rectum (1996) 39:279284 Objective: To determine the feasibility of laparoscopic abdominal rectopexy using a solitary spiked chromium staple to fix the mesh to the sacrum and to compare results to historical controls. Methods: Duration of operation, day of toleration of solid diet, day of discharge, and patient morphine requirements in the first 48 h were documented prospectively and compared to results from chart review of historical abdominal rectopexy cases. Results: The results from a total of 21 laparoscopic cases were compared to 24 historic open abdominal rectopexies. The laparoscopic rectopexy group had lower morphine requirements when administered as PCA (38.2 vs 100.6 mg), an earlier tolerance of solid diet (2.7 vs 5.8 days), and an earlier discharge from the hospital (6.3 vs 11 days). Operating time was longer for the laparoscopic group (mean 198 vs 130 min). Complications included one patient with pseudomembranous colitis, three port-site hematomas, with one patient requiring transfusion, and one port-site hernia in the laparoscopic group. In the historical control group, there were seven reported complications (one death, one DVT, one dehiscence, one pneumonia, and three prolonged ileus). Neither group has had a recurrence to date. Conclusions: Laparoscopic rectopexy is feasible. Its benefits may include decreased postoperative pain, decreased length of hospital stay, and earlier tolerance of normal diet. Comment: Results of this study are clearly limited by the use of historical controls. However, the authors are to be commended for using these results as a stepping stone for future prospective randomized studies, which they reportedly are undertaking at this time. Clearly, the timing for such a prospective randomized study is ideal since laparoscopic management of rectal prolapse is far from being the standard of care, and there are enough reports similar to the one summarized here to support and justify such a trial.

Ludwig KA, Milsom JW, Garcia-Ruiz A, Fazio VW Dis Colon Rectum (1996) 39:285288

Objective: To assess the safety and efficacy of laparoscopic stoma procedures. Methods: Using a two-cannula technique, 24 stoma procedures (16 loop ileostomies, six end sigmoid colostomies, and one transverse and one sigmoid loop colostomy) were attempted. Indications included rectovaginal fistula, perianal sepsis, incontinence, advanced rectal or colon carcinoma, and complicated pelvic infection. Each operation began by placing the camera port through an eventual ostomy site created using the standard techniques. One or two additional trocars were used to complete the procedure. Results: Median operative time was 60 (20120) min. Median blood loss was 50 ml (0150). No intraoperative complications were encountered. There was one conversion because of dense adhesions. Median interval for passage of flatus and stool following ileostomies was 1 (13) day. Following colostomies, the median intervals for flatus and stool following surgery were 2 (24) and 3 (26) days, respectively. Median length of hospitalization was 6 days, usually delayed by the primary disease or colostomy-care training. The only major postoperative complication was one death following pulmonary embolus in a patient with metastatic colon carcinoma. All stomas functioned well with no need for revisions. Conclusions: Laparoscopic fecal diversion procedures can be performed safely, simply, and effectively. Possible advantages are avoidance of laparotomy and rapid return of bowel function. Comment: This is clearly an appealing application of laparoscopy to colorectal surgery. In the era of cost containment, information regarding comparative cost of hospitalization would be most valuable. The problem with such expectation is that those surgeons who are utilizing this laparoscopic technique at this time (such as the authors of this report) may find it unjustified to participate in a randomized study that would deny half of their future patients laparoscopic stoma procedures. After all, laparoscopic cho-

291

Transanal electroresection of small rectal cancer: a sole treatment?

Faivre J, Chaume J-C, Pigot F, Trojani M, Bonichon F Dis Colon Rectum (1996) 39: 270278 Objective: Evaluation of full-thickness transanal electroresection of rectal tumors as a technique for local treatment of distal rectal carcinoma. Methods: Between 1983 and 1990, 227 patients qualified for local resection of their rectal carcinoma. Transanal resection was performed if the tumor was less than a quarter of the rectal circumference, mobile, and if no perirectal nodules were noted on digital examination. Criteria for cure were absence of extramural invasion and negative margins, excluding 61 of the 227 patients. Six patients who underwent postoperative radiation treatment were also excluded. Postoperative morbidity, mortality, and local and distant metastases were studied. Results: Median tumor size was 35 mm. Distance from anal verge was less than 6 cm for 68% and 610 cm for 30% of patients. Ninety-four percent of the patients had uneventful postoperative course, including one with peritoneal violation requiring transanal suture repair. There was one postoperative death from myocardial infarction. Median followup was 88 months. Cancer-specific survivals of 85 and 81% at 5 and 10 years are reported. Long-term complications included five cases of incontinence and two stenoses. Twenty-four patients required further treatment for recurrence. Cancer-specific survivals following recurrence were 72 and 50% for 1 and 2 years, respectively. The only independent factors adversely affecting survival were mucinous contingent and intratumoral vascular invasion. Conclusions: Local treatment of rectal carcinoma in selected patients is favorable. Comment: This report, which contains an excellent discussion section, sheds significant light on patient selection, technique, and outcome of local resection for superficial rectal carcinoma.

nonsedated colonoscopies were performed by the author. Following the procedure, the patients were asked to rate their maximum pain on a scale of 0 to 3 (no pain, mild pain, moderate pain, and severe pain). Those patients who had previous colonscopy under sedation were asked to compare their experiences. Following the procedure, all patients were asked for their preference of sedation or no sedation for a future colonoscopy. Results: Average procedure time was 13.7 (0.585) min. Cecal intubation was achieved in 97% of the patients. Sixtyone percent of the patients rated their pain as 0 or 1. Of the patients who had prior colonoscopy with sedation, 82% preferred colonoscopy with no sedation. Eighty-four percent of the patients would again choose colonoscopy with no sedation. Those who would choose sedation for future colonoscopies were more likely to be female (25% of females vs 11% of males). Three patients required sedation for the completion of the procedure. In 30% of the patients, polypectomy or biopsy was performed. The single reported complication was hematochezia 5 days following hot biopsy eradication of a diminutive polyp, requiring hospitalization, transfusion, and selective intraarterial vasopressin infusion. Conclusions: Nonsedated colonoscopy is safe, effective, and well accepted by the patients. Comment: Every colonoscopist who can consistently achieve similar satisfactory results with nonsedated colonoscopy should use the results of this report and seize the opportunity to switch to nonsedated colonoscopy in those patients who tolerate it well.

Colon surveillance after colorectal cancer surgery

Khoury DA, Opelka FG, Beck DE, Hicks TC, Timmcke AE, Gathright JB Jr. Dis Colon Rectum (1996) 39:252255 Objective: To determine cost-effective guidelines for postresection surveillance colonoscopy in patients with colorectal cancer. Methods: A retrospective review of 389 patients who underwent curative resection of colorectal cancer, followed by colonoscopy, between 1984 and 1994 is presented. Requirements for the study included complete clearance of the colon from all lesions using perioperative colonoscopy. In addition, a first surveillance colonoscopy was performed no later than 24 months following the original resection. Findings of recurrent or metachronous neoplastic lesions, including polyps, constituted as a positive colonoscopy. Results: Results of first through fourth colonoscopies were compared. A consistent drop in the percentage of patients who underwent additional colonoscopic examinations was noted (66.6%, 42.4%, and 21.3% of patients had second, third, and fourth colonoscopies). First postoperative colonoscopies were performed a median of 13 months from the initial surgery. The subsequent colonoscopies were a median of 15, 14, and 14 months apart, respectively. The posi-

Colonoscopy without sedation

Cataldo PA Dis Colon Rectum (1996) 39:257261 Objective: A prospective evaluation of 258 consecutive nonsedated colonoscopies is presented. Special attention is paid to patients assessment of pain and to their overall satisfaction. Methods: Nonsedated colonoscopy was performed in 258 patients. Sedation was withheld for the following reasons: patient preference, medical problems, prior resection, surgeon preference, lack of intravenous access, and age. All

292

tive examination rates for the 4 years were 18.3, 18.5, 16.4, and 14.5 percent, respectively. In the patients studied, after a normal colonoscopy, the risk for a positive next annual colonoscopy was 10.4%. After a positive colonoscopy, this risk increased to 43%. The risk for a positive study after finding multiple neoplastic polyps was 70.4%. Conclusions: Annual surveillance colonoscopy for the first 2 years following resection of colorectal cancer is beneficial for detecting neoplastic lesions. The interval for the subsequent studies may be increased based on the findings of the most recent colonoscopy. Comment: This paper clearly provides interesting epidemiologic information. However, in the absence of statistical analysis and patient outcome, very few conclusions can be drawn. In addition, the study failed to demonstrate any information regarding cost, and therefore did not truly meet the objective set by the authors.

Randomized, prospective, single-blind comparison of laparoscopic versus small-incision cholecystectomy

Majeed AW, Troy G, Nicholl JP, Smythe A, Reed MWR, Stoddard CJ, Peacock J, Johnson AG Lancet (1996) 347:989994 (Commentary p 985) Objective: To prospectively compare laparoscopic cholecystectomy to small-incision cholecystectomy. Methods: Two hundred patients were randomized to undergo either laparoscopic or small-incision cholecystectomy. Cholangiography was attempted in all cases. To minimize patient-related bias, patients were blinded to the technique of operation. Caretakers were also blinded by using identical dressings following both type of procedures. Results: Laparoscopic cholecystectomy took longer than small-incision cholecystectomy (65 vs 40 min, median). Median length of hospital stay was 3.0 nights for both groups (117 for laparoscopic, 114 for small incision). Time back to work for employed patients and time to full activity were not statistically different between the two groups. Complications occurred with similar incidence in the two groups. (The nature of the complications was individually presented for both groups and appeared to be of similar magnitude.) The rate of conversion to open cholecystectomy was 20%. Conclusions: Laparoscopic cholecystectomy takes longer than small-incision cholecystectomy and offer no advantages with regard to length of hospital stay or postoperative recovery. Comment: As claimed, this is one of the better-designed randomized studies addressing the advantages (or lack thereof) of laparoscopic over small-incision cholecystectomy. If these results are reproduced, the factors examined by this study clearly refute all previously claimed advantages of laparoscopic cholecystectomy. Two of the issues not addressed by this report are the surgeons opinion of the operative techniques (i.e., operative advantages of laparoscopic cholecystectomy vs its disadvantages as compared to open or small-incision cholecystectomy), and patient satisfaction with the surgical scar! I like to believe that for most of the surgeons around the world, it is the former and not the latter that encourages them to continue to embrace the laparoscopic approach to cholecystectomy. The editorial comment on this paper is based on the cost of health care in countries less financially fortunate than those in the Western Hemisphere. This editorial clearly raises a serious concern regarding the direction the healthcare industry has taken following the popularization of laparoscopic general surgery. It is, however, reasonable to acknowledge the differences in availability of resources in different regions of the world and to allow for sensible variations in provision of health care (in this case laparoscopic vs open cholecystectomy). Clearly, even in 1996, for laparoscopic cholecystectomy to gain worldwide acceptance, the operative cost for this operation has to approach its open counterpart.

Laparoscopic surgery for duodenal ulcer: first results of a multicentre study applying a personal procedure

Gomez-Ferrer F, Balique JG, Azagra S, Bicha-Castelo H, Castro-Sousa F, Espalieu P, Rodero D, Estour E Br J Surg (1996) 83:547550 Objective: To retrospectively evaluate the outcome of laparoscopic posterior truncal vagotomy (PTV) and anterior linear gastrectomy (ALG) in treatment of chronic duodenal ulcer. Methods: A total of 136 patients from 14 centers were evaluated. Mean age was 48.1 years. The mean duration of the disease was 9.5 years. Associated gastroesophageal reflux and biliary lithiasis were observed in 17 and 13 patients, respectively. For details of the laparoscopic approach, the readers are referred to the detailed description in the original article. Results: The mean operative time (not including the time required for concurrent Nissen fundoplication or cholecystectomy) was 65 (range 25180) min. There were no conversions to laparotomy. The mean hospital stay was 3.1 (range 213) days. There were no perioperative mortalities. The morbidity rate was 2.9% (one case each of atelectasia, thrombophlebitis, omphalitis, and parietal hematoma). Of the 131 patients who were evaluated 6 to 33 months postoperatively, 96.2% were graded as Visick I or II, 3% were Visick III, and 0.8% (one patient) was considered Visick IV. Conclusions: Laparoscopic PTV with ALG is a simple, safe, and efficient method with which to treat chronic duodenal ulcer. Comment: The discussion section in this paper provides a good comparison between the different surgical techniques currently used for treatment of chronic duodenal ulcer. Clearly, prospective studies of a large number of patients are necessary to identify one procedure as the superior operative approach to chronic duodenal ulcer disease.

293

Laparoscopic mobilization of the stomach for oesophageal replacement

Jagot P, Sauvanet A, Berthoux L, Belghiti J Br J Surg (1996) 83:540542 Objective: To improve the postoperative pulmonary course, the role of laparoscopic gastric mobilization and abdominal lymphadenectomy for treatment of esophageal cancer was evaluated. Methods: Nine patients of mean age 61 with moderate to severe airway obstruction underwent laparoscopic gastric mobilization and abdominal lymphadenectomy for treatment of esophageal cancer. In six patients, laparoscopy was combined with right thoracotomy. The other three had a laparoscopic transhiatal approach with a left cervicotomy. Results: The mean operative time and time for the laparoscopic step were 511 and 210 min, respectively. No conversions to laparotomy were required. The mean number of resected abdominal lymph nodes was 8.1. Two patients were extubated at the end of the procedure, and the other seven on postoperative day 1. Mean duration of hospitalization was 10.3 (range 818) days. There were no operative mortalities. Conclusions: The laparoscopic approach for gastric mobilization and abdominal lymphadenectomy is safe and can be used in patients with impaired pulmonary function. Comment: This is an encouraging first step toward a laparoscopic surgical approach for the treatment of esophageal cancer. The timing may be ideal for a well-designed prospective randomized trial to compare different approaches to esophageal replacement surgery, eliminating the biases inherent to all nonrandomized studies.

75 and 95 min, respectively. There were no statistically significant differences between the two groups with regard to the incidence of or positive predictive value for bile duct stones, rate of retained stones, intraoperative or postoperative morbidity, or incidence of bile duct anomalies. In no patient was the cholangiographic finding of a biliary anomaly crucial for the safe execution of cholecystectomy. Conclusions: Preoperative infusion cholangiography and IOC are comparable, but routine use of either method does not change the safety of elective cholecystectomy and is thus not warranted. Preoperative infusion cholangiography is more expensive and requires support by IOC in 20% of patients. However, PIC reduces the length of the operative procedure as compared to IOC. Comment: Although this study retrospective appears to predate the laparoscopic era, it does provide readers with an interesting comparison between the two techniques of biliary ductal evaluation. Extrapolation of these results to the laparoscopic technique of cholecystectomy may not be possible. Considering the ever-increasing facility of laparoscopic management of common bile duct stones, the use of real-time IOC (using C-arm fluoroscopy) provides the surgeon with all the necessary information for successful management of common bile duct stones. As indicated by this paper, PIC alone may not provide the surgeon with enough information necessary for not just detection, but successful management of biliary ductal stones.

Stapled laparoscopic splenectomy: initial experience

Saldinger PF, Matthews JB, Mowschenson PM, Hodin RA J Am Coll Surg (1996) 182:459461 Objective: An initial experience with stapled transection of the splenic hilum during laparoscopic splenectomy is reported. Methods: Results from seven patients who underwent laparoscopic splenectomy between August 1993 and May 1995 is presented. Indications included idiopathic thrombocytopenic purpura (ITP) (n 6) and hereditary spherocytosis (n 1). Ages ranged from 19 to 74 years. Preoperative platelet count ranged from 16,000 to 225,000/mm3. With the patient in the right lateral decubitus position, a total of four 11.5-mm trocarsone through the left rectus sheet lateral to the umbilicus and the other three positioned subcostally are used. After exploring the splenic hiluma and the omentum for accessory spleen, the lienocolic and lienophrenic ligaments are divided. The spleen is lifted up using atraumatic retractors. The splenic hilum is sequentially ligated and divided using several firings of an Endo-GIA device. Following the take-down of the short gastric vessels, the spleen is placed in a protective bag, morcelated, and aspirated out of the abdominal cavity. Results: There were no reported complications. Two patients had accessory spleens in the hilum which were removed laparoscopically. Operative time ranged from 80

Routine preoperative infusion cholangiography versus intraoperative cholangiography at elective cholecystectomy: a prospective study in 995 patients
Hammarstrom LE, Holmin T, Stridbeck H, Ihse I J Am Coll Surg (1996) 182:408416 Objective: To compare routine preoperative infusion cholangiography (PIC) to routine intraoperative cholangiography (IOC) at elective cholecystectomy. Methods: A retrospective review of 1,042 patients who underwent elective cholecystectomy between 1985 and 1991 is presented. Of these patients 694 had PIC, and the remaining were randomly allocated to IOC. The age of patients, history of biliopancreatic complications, and laboratory findings were similar for the two groups. Results: The cost of PIC in Sweden is nearly five times greater than the cost of IOC. Preoperative infusion cholangiography required support by IOC in 19.5% of patients. The median operative times for PIC and IOC groups were

294

min to 5 h (mean 157 min). Two patients required transfusion with one unit of packed red blood cells each. The length of hospital stay ranged from 1 to 10 (median 3) days. Comment: Considering the nonrandomized nature of this study, it is hard to determine whether the authors achieved their goal: i.e., reducing the complexity of the operation and hence the length of the operative procedure or the amount of blood loss. In addition, although enbloc ligation and division of the splenic hilar vessels is an acceptable practice in

emergency splenectomy (i.e., abdominal trauma), the standard open technique of elective splenectomy involves separate ligation and division of the hilar vessels. Therefore, for the purists, this deviation from the routine open technique of splenectomy may not be readily acceptable.

Reviewers for this issue: J. Sackier, V. Shayaui

Surg Endosc (1997) 11: 283284

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic resection of bleeding gastric polyps

W. T. Siu, H. T. Leong, M. K. W. Li
Department of Surgery, Pamela Youde Nethersole Eastern Hospital, 3 Lok Man Road, Chai Wan, Hong Kong

Abstract. Gastric polyps account for a small proportion of gastric neoplasms. Many of them presented with acute or chronic bleeding. Endoscopic polypectomy is the preferred treatment modality. However, endoscopic polypectomy is often incomplete and impossible for large polyps. Large and intramural polyps require laparotomy and excision. From June to October 1995, four patients with bleeding gastric polyps received laparoscopic resection in our unit. All were women, aged 4078, with polyps 26 cm in diameter. One patient underwent emergency surgery for active bleeding. The others received elective operations. Operating time ranged from 90 to 120 min. There was no perioperative morbidity or mortality. Full diet was resumed by the 4th postoperative day. The average postoperative hospital stay was 5.3 days (47). Laparoscopic resection of gastric polyp is a preferable treatment option when endoscopic treatment fails. Key words: Gastric polyps Gastric neoplasms Bleeding

was 8.6 g/dl. Upper endoscopy showed a 5-cm pedunculated polyp in the gastric fundus with an overlying ulcer covered with adherent blood clot; 10 ml of 1:10,000 adrenaline was injected around the ulcer. However, during the same night, the patient went into shock with fresh hematemesis. Emergency laparoscopic operation was performed using the gasless laparoscopy technique (Laparolift). The bleeding gastric polyp was excised laparoscopically by Endo GIA via gastrototomy. Seven units of blood were transfused. Full diet was started on the 3rd postoperative day and she was discharged 1 week after operation.

Case 2
A 49-year-old lady presented with a 1-day history of tarry stool after NSAID was prescribed for her gout. He hemoglobin on admission was 7.6 g/dl. Upper GI endoscopy revealed a 2-cm leiomyoma with surface ulceration sited on the lesser curve. Attempted endoscopic polypectomy confirmed the intramural nature of the lesion. Laparoscopic wedge excision was performed with preservation of the anterior nerve of the Lartarget. Three units of blood was transfused. Diet was resumed on the 3rd postoperative day and she was discharged 4 days after operation.

Case 3 Gastric polyps usually present with epigastric pain, acute bleeding, or anemia. Diagnosis is usually made by endoscopy or barium meal. Small lesions may be treated with endoscopic polypectomy using a diathermy snare. However, this often results in incomplete excision and perforation is a definite risk. Large polyps and intramural lesions are currently treated by open gastrotomy and complete excision of the lesion. Laparoscopic approach minimizes the access trauma of laparotomy and is theoretically superior for upper gastrointestinal benign lesions [1]. Laparoscopic resection of benign stromal tumor of the stomach has been reported. This article reports different approaches to laparoscopic resection of benign gastric polyps. Case studies Case 1
A 78-year-old lady with chronic obstructive airway disease was admitted to our unit with 2 days history of tarry stool. Her hemoglobin on admission A 65-year-old lady with chronic rheumatic heart disease (aortic, mitral, and tricuspid regurgitation) presented with 3 days history of tarry stool. Hemoglobin on admission was 4.7 g/dl. Upper GI endoscopy discovered a 5-cm ulcerative polyp over the fundus. She received 12 units of blood preoperatively. Laparoscopic gastrotomy followed by resection was performed. She made an unremarkable recovery and discharge on the 6th postoperative day.

Case 4
A 40-year-old lady presented with iron-deficiency anemia. Upper GI endoscopy detected a 3.5-cm fundal leiomyoma. Diathermy snare was unable to engage the whole polyp. Laparoscopic resection was considered an appropriate option. The preoperative hemoglobin was 8 g/dl. She underwent laparoscopic resection of the gastric polyp via a gastrotomy. She made a quick recovery and was discharged on the 4th postoperative day.

Surgical technique and results Preoperative localization was obtained with gastroscopy. The patients were placed supine in Lloyd-Davis position with the surgeon standing in between the patients legs. The

Correspondence to: M. K. W. Li

284

min). The average number of intramuscular pethidine injection was 234 doses, (range 04 doses). The pain was mainly caused by the wound extension for specimen retrieval. Full diets were resumed by the 3rd postoperative day and the average postoperative stay was 5 days (47 days).

Discussion Most patients with large gastric polyps presented with bleeding, either acute or chronic. The advent of therapeutic endoscopy has revolutionized the management of the majority of small pedunculated polyps. Nevertheless, diathermy snare has its size limitation and is not possible for sessile polyps. Many of the endoscopic polypectomies are incomplete, especially when the lesion is intramural. Postpolypectomy bleeding and perforation can be life threatening. Open gastrotomy and complete excision are the timehonored approach when the endoscopic method fails. A laparoscopic approach minimizes the access trauma of laparotomy wound and hastens postoperative recovery. The application of laparoscopic surgery in the management of gastric polyp has not yet been popularized. Isolated case reports have been published [2, 3]. Emergency laparoscopic excision of bleeding gastric polyps was successful. The availability of laparoscopic mechanical stapling devices allows rapid access, resection, tissue approximation, and simultaneous hemostasis. Gasless laparoscopy is beneficial for patients with poor cardiorespiratory reserve. It also allows continuous aspiration to be applied during dissection. The choice between gastrotomy and closed technique mainly depends on the location of the polyp. Centrally located anterior lesions can be excised by wedge excision of stomach with minimal dissection of the greater or lesser omentum. Gastrotomy and direct grasping control of the activity are more suitable for posteriorly sited lesions. In conclusion, laparoscopic resection of bleeding gastric polyps is definitely feasible, has a comparable operating time, and is less painful to patients. It should be the preferred procedure for excision of benign large, sessile, or intramural polyps.

Fig. 1. Theater setup and trocar placement.

theater setup and the trocar positions are shown in Fig. 1. A 30 laparoscope was introduced via the umbilical port. Because of the proximity of spleen and short gastric vessels to the fundus, wedge excision of posteriorly located fundal gastric polyp was difficult. Gastrotomy approach was preferred. The polyp was palpated with blunt forceps to determine the optimal site for gastrotomy. Air was introduced through nasogastric tube to distend the stomach. A gastrotomy was then made by needle knife and enlarged with Endo GIA through the 12-mm cannula. Endo GIA provided quick access and simultaneous hemostasis. This minimized bleeding from the gastric wall that would otherwise obscure the operative field. The polyps were grasped with forceps and delivered through the gastrotomy. The normal stomach adjacent to the lesion was exposed by inverting the posterior wall of the stomach. Endo Catch was employed to engage the polyp and prevent contamination. The polyp was excised by sequential application of Endo GIA. Multiple applications were required. Complete excision was confirmed and hemostasis ensured before closure of the gastrotomy. The gastrotomy was closed with Endo TA. The polyp was then retrieved using the Endo Catch. For the patient with an intramural lessercurvature polyp, a closed technique was adopted. The part of the stomach containing the polyp was resected without a gastrotomy. After identification of the vagal trunks the anterior leaf of the lesser omentum was separated from the lesser curve, carefully preserving the anterior nerve of Latarget. The polyp was then excised by sequential application of Endo GIA. The average operating time for our patients with laparoscopic resection of gastric polyps was 118 min (90140

References
1. Cuschieri A (1995) Whither minimal access surgery: tribulations and expectations. Am J Surg 169: 919 2. Llorente J (1994) Laparoscopic gastric resection for gastric leiomyoma. Surg Endosc 8(8): 887889 3. Motoson RW, Fisher PW, Dawson JW (1995) Laparoscopic resection of a benign intragastric stromal tumour. Br J Surg 82: 1670

Surg Endosc (1997) 11: 235238

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Does laparoscopy increase bacteremia and endotoxemia in a peritonitis model?

C. A. Jacobi,1 J. Ordemann,1 B. Bo hm,1 H. U. Zieren,1 H. D. Volk,2 W. Lorenz,3 E. Halle,4 J. M. Mu ller1
1 2

Department of Surgery, University of Berlin, Charite , Germany Institute of Medical Immunology, University of Berlin, Charite , Germany 3 Institute of Theoretical Surgery, University of Marburg, Germany 4 Institute of Microbiology, University of Berlin, Charite , Germany Received: 28 May 1996/Accepted: 25 July 1996

Abstract Background: Laparoscopy is increasingly used in patients with intraabdominal bacterial infection although pneumoperitoneum may increase bacteremia by elevated intraabdominal pressure. Methods: The influence of laparotomy and laparoscopy on bacteremia, endotoxemia, and postoperative abscess formation was investigated in a rat model. Rats received intraperitoneally a standardized fecal inoculum and underwent laparotomy (n 20), or laparoscopy (n 20), or no further manipulation in the control group (n 20). Results: Bacteremia and endotoxemia were higher after laparotomy and laparoscopy compared to the control group (p 0.01) 1 h after intervention. One hour after intervention, aerobic and anaerobic bacterial species were detected in the laparotomy group while only anaerobic bacteria were found in the other two groups. Although bacteremia and endotoxemia did not differ among the three groups after 1 week, the mean number of intraperitoneal abscesses was significantly higher (p < 0.05) after laparotomy (n 10) compared with laparoscopy (n 6) and control group (n 5). Conclusion: Laparoscopy does not increase bacteremia and intraperitoneal abscess formation compared to laparotomy in an animal model of peritonitis. Key words: Laparoscopy Peritonitis Bacteremia Endotoxemia

tality [6]. But there is still some theoretical concern that pneumoperitoneum may cause enhanced bacteremia and endotoxemia due to increased intraperitoneal pressure. Only few and even controversial data exist from experimental studies which have investigated the effects of pneumoperitoneum on bacteremia and physiological changes during peritonitis and sepsis [2, 4, 8]. Furthermore, peritonitis was not caused by different bacterial species but by intraperitoneal inoculation of Escherichia coli alone in all studies. Thus, it remains questionable whether laparoscopic surgery may be harmful in patients with diffuse peritonitis. Since there are to our knowledge no studies which have compared the influence of surgical intervention on early and late outcome in peritonitis, the influence of laparotomy and laparoscopy on bacteremia, endotoxemia and postoperative intra- and extraperitoneal abscess formation was investigated in a peritonitis model. Materials and methods
A standardized fecal inoculum of human stool was chosen for this study because it produces a nonfatal bacteremia after intraperitoneal instillation in rats [10]. It further has been demonstrated that inoculation of heatinactivated stool suspension did not cause intraabdominal peritonitis in this model. Therefore peritonitis was caused by bacterial infection and not by toxic noninflammatory effects of the stool suspension. After stool injection, rats were randomized into three groups (group I: laparotomy; group II: laparoscopy; group III: control group). The hypothesis of the experiment was that laparoscopy with carbon dioxide leads to enhanced bacteremia, endotoxemia, and development of intraperitoneal abscess formation. The endpoints of the study were perioperative changes in bacteremia and endotoxemia and the incidence and number of intraperitoneal abscess formation 1 week after intervention.

Laparoscopic surgery is currently performed for benign and malignant intraabdominal diseases. Some authors reported successful treatment of inflammatory processes like appendicitis, Crohns disease, perforated peptic ulcer, or diverticular disease [1, 3, 9, 11, 13]. Furthermore, laparoscopic surgery appears to be feasible in patients with abdominal sepsis without increasing postoperative morbidity and morCorrespondence to: C. A. Jacobi

Animals
Sixty male inbred 2-month-old Wistar rats (Charles River, Sulzfeld, Germany) were acclimated to a climate- and light-cycle-controlled environment for at least 7 days prior to investigations. The animals were allowed standard laboratory food and water ad libitum. All studies were performed under protocols approved by the local committees of Animal Use and Care.

236

Experimental course and operative procedures

All animals were anesthetized by intraperitoneal injection of pentobarbital 60 mg/kg under sterile conditions. It was verified by a chromogenic Limulus amoebocyte lysate assay (Whittaker Bioproducts, Wakersville, Maryland, USA) that pentobarbital did not contain endotoxins. Stool suspension (1 0.5 ml/kg) was then intraperitoneally applied transcutaneously under sterile conditions in all animals. The rats were randomized into three different groups. A 10-cm midline laparotomy was accomplished in the first group (n 20) and the abdomen was closed after 30 min. In the second group, pneumoperitoneum was performed through a Veress needle with insufflation of carbon dioxide (n 20) at a pressure of 10 mmHg over 30 min. The control group (n 20) underwent no further manipulation after stool injection. Blood samples were taken from the femoral vein of the rats and placed into sterile heparinized vials (pyrogen-free) before, 1 h, 2 and 7 days after surgery to determine endotoxin levels of the blood by using a chromogenic Limulus amoebocyte lysate assay. Microbiologic phenotypical identification of bacteria in the blood cultures was performed before, 1 h, and 1 week after stool application. The animal underwent laparotomy on postoperative day 7 to determine the number and location of intraperitoneal abscess formations. Abscesses were excised, analyzed microbiologically, and confirmed histologically using HE stains. Additionally, microbiologic analysis of peritoneal swabs was performed to evaluate the differences between intraperitoneal and intravenous bacterial species on day 7.

Fig. 1. Incidence of positive blood cultures after laparotomy (n 20), laparoscopy (n 20), and control group (n 20).

Microbiological analysis
The microorganisms were grown on chocolate agar (Tryptic Soy agar supplement with 10% defibrinated sheep blood, heated for 10 min to 80C), blood agar (Columbia agar supplemented with 5% defibrinated sheep blood), Endo agar, and Sabouraud agar in an aerobic and anaerobic atmosphere. The phenotypical identification of all strains was verified by testing the carbohydrate fermentation reactions or by using commercially available enzyme activity and fermentation tests (API, Bio Me rieux, Nu rtingen, Germany).

Measurement of endotoxin blood levels

Plasma was separated from blood samples by centrifugation at 3,000g at 4C for 10 min. Plasma samples were diluted tenfold and heated to 70C for 5 min to eliminate endotoxin inhibitors. All samples were stored in 2-ml pyrogen-free polypropylene screw cap tubes (Sarstedt, NumbrechtRommelsdorf, Germany) at 85C until final analysis. A chromogenic Limulus amoebocyte lysate (LAL) assay (Whittaker Bioproducts, Walkersville, Maryland, USA) was used to determine endotoxin serum levels; 50 l of each plasma sample and 50 l of LAL were pipeted into wells of a sterile microtiter plate (Falcon No. 3072, Becton Dickinson, Lincoln Park, New Jersey, USA). After incubation at 37C for 10 min, 100 l of chromogenic substrate (Acetyl-Ile-Glu-Gly-Arg-pnitroanilide, 1.6 mmol/l) was added. The reaction was stopped by 100 l of 25 % (v/v) acetic acid after 6 min and absorbance of each well was measured at 405 nm with a microplate reader. Endotoxin standards at concentrations of 0.1, 0.25, 0.5, 0.75, and 1.0 U/ml were prepared from E. coli 0111:B4 (Whittaker Bioproducts) and two blanks were performed in parallel with each test. In this assay, 12 endotoxin units correspond to 1 ng of E. coli 0111:B4 lipopolysaccharide (LPS).

Statistics
Data are given as mean and standard deviation. Data between groups were compared using a Kruskal-Wallis test for continuous data and Fishers exact test for categorical data, if appropriate. P values less than 0.05 were considered significant.

Results Bacteremia was not found in any animal at the beginning of the experiment. One hour after stool application, incidence

of bacteremia increased after laparotomy and laparoscopy compared to the control group (p < 0.001) (Fig. 1). There was no statistical difference in bacteremia between all groups 1 week after fecal inoculum and surgical intervention. Positive peritoneal cultures did not significantly differ between laparotomy (20/20), laparoscopy (15/20), and control group (18/20) at the end of the experiment. Qualitative microbiologic analyses have detected 42 different bacteria in assays of the human stool inoculum. Analysis of the stool taken from the intestinum of the rats have found 35 different bacterial organisms. Sixteen different species out of human fecal inoculum and rat stool species were detected in blood culture and peritoneal swabs with different distribution between the three groups (Table 1). Five of these bacteria were not found in fecal inoculum but in stool of the rats. While rats in the laparotomy group showed both anaerobic and aerobic bacteria in blood cultures, aerobic bacteria were not detected in blood cultures after laparoscopy and in the control group 1 h after intervention. Anaerobic Bifidobacterium spp. and Bifidobacterium adolescentis were most often found in blood culture 1 h after intervention in all groups. Although the number of positive blood cultures decreased on postoperative day 7, various anearobic and aerobic bacteria were detected in all groups. Microbiologic analysis of peritoneal fluid detected 12 different organisms after laparotomy, nine species after laparoscopy, and only eight species in the control group 1 week after stool inoculation. E. coli and Enterococci were the most frequent species found intraperitoneally in all groups. These species showed only in part coincidence with organisms detected in blood cultures on postoperative day 7. No animal had endotoxemia at the beginning of the experiment. Plasma endotoxin levels increased within 1 h after stool inoculum in all groups (p < 0.01) (Fig. 2). Levels were higher (p < 0.01) in animals after laparotomy compared with laparoscopy and control group 1 h after intervention but were not different during the following postoperative course. Endotoxin levels already had reached their maximum peak 1 h after intervention in all groups followed by a slight decrease toward the end of the observation on day 7. No animal died before scheduled autopsy 1 week after stool inoculation. Iatrogenic injury from insertion of the Veress needle or from abdominal incisions did not occur in any animal. Intraperitoneal abscess formation was con-

237

Fig. 2. Endotoxin level in blood after laparotomy (n 20), laparoscopy (n 20), and control group (n 20) (mean and standard deviation, asterisk indicates p < 0.01: laparotomy vs laparoscopy and control group).

Fig. 3. Incidence and organ distribution of intraperitoneal abscess formations after laparotomy (n 20), laparoscopy (n 20), and control group (n 20) (asterisk indicates p < 0.05: laparotomy vs laparoscopy vs control group).

Table 1. Bacterial species in blood cultures and peritoneal fluidsa Blood culture 1 hour Bacteria Aerob gram-negative Escherichia coli Proteus mirabilis Aerob gram-positive Enterococcus faecalis Bacillus spp. Staphylococcus aureus Staphylococcus coag. neg. Streptococcus viridans Anaerob gram-negative Bacteroides spp. Bacteroides fragilis Bacteroides uniformis Bacteroides ovatus Anaerob gram-positive Prevotella spp. Clostridium perfringens Propionibacterium spp. Bifidobacterium spp. Bifidobacterium adolescentis
a

Blood culture 7 days Laparotomy + + + + + Laparoscopy Control +

Intraperitoneal Laparotomy +++ ++ +++ + + Laparoscopy +++ ++ +++ +++ + + + + + Control +++ + +++ + + ++ ++ ++

Human feca inoculum

Rat intestinum

Laparotomy + ++ + +

Laparoscopy

Control

+ +

+ ++ ++ +

+ + + + + + + +++ +++ +++ ++ ++ ++ + +

+ +

+ +

+ + +

Key to symbols: : detected in human fecal inoculum and rat intestinum; : only detected in fecal inoculum; : only detected in rat intestinum; +: 13 positive cultures; ++: 46 positive cultures; +++: >6 positive cultures.

firmed in all rats after laparotomy, in 15 of 20 rats after laparoscopy, and in 18 of 20 rats in the control group 1 week after inoculation. The mean number of abscesses in each animal was 10 6.2 after laparotomy, 6 5.1 after laparoscopy, and 5 4.8 in the control group and differed (p < 0.05) between the first and the two other groups. The localization of intraabdominal abscesses also differed (p < 0.01) between the groups (Fig. 3). In the laparotomy group, abscesses were most often localized at the peritoneal surface of the liver, on the bowel surface and the abdominal wall. Interenteric abscesses were only found in five rats after laparotomy and did not occur in the two other groups. In laparoscopic group, abscesses were increasingly found at the peritoneal surface of the bowel and the abdominal wall while in the control group abscesses were most often found in the omentum majus.

Discussion Laparoscopic techniques are sometimes utilized in patients with diffuse or localized peritonitis. It has been demonstrated in prospective randomized trials that laparoscopic appendectomy is superior or at least does not differ compared to open appendectomy in terms of postoperative complications, hospital stay, and recovery [1, 11]. Urbano et al. further demonstrated that laparoscopic surgery of perforated peptic ulcers is simple, rapid, and followed by a quick recovery [13]. In contrast to these results, Eypasch et al. reported higher morbidity after laparoscopic treatment of perforated peptic ulcer than after a conventional approach [5]. Although the reported results may also be influenced by selection of patients and experience in laparoscopic techniques, it is hypothetical that continuous elevated intraab-

238

dominal pressure promotes bacterial translocation and thus increases postoperative septic complications. This is supported by experimental studies which showed an increase in the extent and severity of peritonitis as well as bacteremia after perforated peptic ulcer and laparoscopy compared to the control group in a rat model [2]. However, the data of experimental studies regarding laparoscopic surgery during peritonitis remain controversial. Eleftheriadis et al. [4] have demonstrated that elevated intraabdominal pressure (15 mmHg) leads to intestinal ischemia, to oxygen free-radical production, and to increased bacterial translocation in rats, while Gurtner et al. [8] did not find an increase of bacteremia or endotoxemia after pneumoperitoneum of 12 mmHg in rabbits. Unfortunately, laparotomy and laparoscopy were not compared in these studies. In the present study, a nonfatal peritonitis was induced, stimulating the clinical stiuation with intraabdominal pus, abscess formations, and positive blood cultures of different bacterial species. The spectrum of bacterial species (Escherichia coli, Enterococcus faecalis, Staphylococcus aureus) found in blood cultures and peritoneal fluids of the rats is similar to the spectrum commonly detected in patients with abdominal sepsis [15, 16]. Overall, 16 different bacteria were found in blood cultures and peritoneal fluid of all animals. Five of the detected bacteria were not found in fecal inoculum, which may indicate that these germs are probably translocated from the rat intestinum during the development of peritonitis. The incidence of positive blood cultures was significantly higher after laparoscopy compared to control group 1 h after intervention but showed no difference after 7 days. It is well known that intraabdominal fluids, cells, and particles are removed from the peritoneal cavity via large terminal lymphatics connected to the thoracic duct and venous system. The openings of these lymphatics are located on the peritoneal surface of the diaphragm between the lateral borders of peritoneal mesothelial cells [14]. It has been demonstrated that elevated intraabdominal pressure leads to an increase of the patency of these lymphatic openings in laparoscopy [12], which may explain the difference in bacteremia between the laparoscopic and control group. However, laparotomy promoted bacterial translocation and bacteremia even more. Blood cultures of rats detected more different bacterial species after laparotomy than after laparoscopy. Furthermore, blood cultures of the rats showed both anaerobic and aerobic bacteria after laparotomy while aerobic bacteria were only detected once in blood cultures after laparoscopy. This is confirmed by higher plasma endotoxin levels after laparotomy compared to laparoscopy and the control group 1 h after intervention. Since carbon dioxide is bacteriostatic on aerobic bacteria [7] this may explain why aerobic bacteria were only detected in blood cultures of one animal after laparoscopy. Interestingly, there was no statistical difference between all groups in bacteremia and endotoxemia after 1 week. The early translocation may be promoted by the different operations while late translocation is caused by the development of peritonitis. The difference in early endotoxemia between rats underling laparotomy and those in either the laparoscopy or control group may be caused by a transient bacteremia of specific aerobic organisms. It also may be that the incision of the abdominal wall with injury of blood vessels caused higher translocation of

bacteria in the laparotomy group. The enhanced perioperative strain during laparotomy is confirmed by the significantly higher number of intraperitoneal abscess formations in this group. Nevertheless, perioperative treatment with antibiotic may reduce the differences between the groups. Therefore, further experimental and clinical trials are needed to analyse different adjuvant therapeutic strategies in this model. In summary, laparoscopy does initially increase the incidence of positive blood cultures compared to control group in an animal model of peritonitis. But endotoxemia and the development of intraabdominal abscesses are not increased by laparoscopy compared with the control group. In contrast to this, laparotomy significantly promotes transient translocation of aerobic and anaerobic species, endotoxemia, and development of intraperitoneal abscesses compared to laparoscopy and the control group. References
1. Attwood SE, Hill AD, Murphy PG, Thorton J, Stephens RB (1992) A prospective randomized trial of laparoscopic versus open appendectomy. Surgery 112: 497501 2. Blo chle C, Emmermann A, Achilles E, Treu H, Zornig C, Broelsch CE (1995) Einflu des Penumoperitoneums auf Ausdehnung und Schwere einer durch peptische Ulkusperforation induzierten Peritonitis der Ratte. Langenbecks Arch Suppl: 15 3. Bo hm B, Schwenk W, Mu ller JM (1996) Laparoskopische kolorektale Chirurgie. Dtsch Med Wochenschr (submitted) 4. Eleftheriadis E, Kotzampassi K, Papanotas K, Heliadis N, Sarris K (1996) Gut ischemia, oxidative stress, and bacterial translocation in elevated abdominal pressure in rats. World J Surg 20: 1116 5. Eypasch E, Menningen R, Paul A, Troidl H (1993) Die Bedeutung der Laparoskopie bei der Diagnostik und Therapie des akuten Abdomens. Zentralbl Chir 118: 726732 6. Geis WP, Kim HC (1995) Use of laparoscopy in the diagnosis and treatment of patients with surgical abdominal sepsis. Surg Endosc 9: 178182 7. Gill CO, Delacy KM (1991) Growth of Escherishia coli and Salmonella typhimurium on high > pH beef packed under vacuum or carbon dioxide. Int J Food Micribiol 13: 2130 8. Gurtner GC, Robertson CS, Chung SCS, Ling TKW, IP SM, Li AKC (1995) Effect of carbon dioxide pneumoperitoneum on bacteraemia and endotoxaemia in an animal model of peritonitis. Br J Surg 82: 844848 9. Hansen JB, Smithers BM, Schache D, Wall DR, Miller BJ, Menzies BL (1996) Laparoscopic versus open appendectomy: A prospective randomized trial. World J Surg 20: 1721 10. Lorenz W, Reimund KP, Weitzel F, Celik I, Kurnatowski M, Schneider C, Mannheim W, Heiske A, Neumann K, Sitter H, Rothmund M (1994) Granulocyte colony-stimulating factor prophylaxis before operation protects against lethal consequences of postoperative peritonitis. Surgery 116:925934 11. Tate J, Dawson JW, Chung SC, Lau WY, LI AK (1993) Laparoscopic versus open appendectomy: prospective randomised trial. Lancet 342: 633637 12. Tsilibary EC, Wissig SL (1983) Lymphatic absorption from the peritoneal cavity: regulation of patency of mesothelial stomata. Microvasc Res 25: 2239 13. Urbano D, Rossi M, De Simone P, Berloco P, Alfani D, Cortesini R (1994) Alternative laparoscopic management of perforated peptic ulcers. Surg Endosc 8: 1208 14. Walker AP, Condon RE (1989) Peritonitis and intraabdominal abscesses. Principles of surgery. 5th ed. Mc Graw-Hill, New York, pp 14591489 15. Walker AP, Krepel CJ, Gohr CM, Edmiston CE (1994) Microflora of abdominal sepsis by locus of infection. J Clin Microbiol 32: 557558 16. Zubkov MN, Menshikov DD, Gugutsidze EN, Chegin VM, Vasina TA (1995) Microbiologic diagnosis of mixed anaerobic and aerobic surgical infections. Antibiot Khimioter 40: 4650

Original articles
Surg Endosc (1997) 11: 226229

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Standardizing laparoscopic procedure time and determining the effect of patient age/gender and presence or absence of surgical residents during operation
A prospective multicenter trial
L. W. Traverso,1 K. P. Koo,1 K. Hargrave,1 S. W. Unger,2 T. S. Roush,3 L. L. Swanstrom,4 M. S. Woods,5 J. H. Donohue,6 D. J. Deziel,7 I. B. Simon,8 E. Froines,9 J. Hunter,10 N. J. Soper11
1 2

Department of Surgery, Virginia Mason Medical Center, 1100 Ninth Avenue, Seattle, WA 98011, USA Department of Surgery, Mt. Sinai Hospital, 4302 Alton Road, Miami, FL 33140, USA 3 Department of Surgery, US Navy Hospital, Bremerton, WA 98312-1898, USA 4 Department of Surgery, Emanuel Hospital, Portland, OR 97227, USA 5 Department of Surgery, Wichita Clinic, Wichita, KS 67208, USA 6 Department of Surgery, Mayo Clinic, Rochester, MN 55905, USA 7 Department of Surgery, Rush-Presbyterian Hospital, Chicago, IL 60612-3833, USA 8 Department of Surgery, Sunrise Hospital, Las Vegas, NV 89109, USA 9 Department of Surgery, Group Health Cooperative, Seattle, WA 98112, USA 10 Department of Surgery, Emory University, Atlanta, GA 30322, USA 11 Department of Surgery, Washington University, St. Louis, MO 63110, USA Received: 22 April 1996/Accepted: 8 July 1996

Abstract Background: Most of the expense of laparoscopic cholecystectomy (LC) is incurred while the patient is in the operating room (OR). Half of this operating room cost is equipment and the other half is personnel. What is an acceptable LC procedure time and how much variation is there? What are the effects of age, gender, and expertise on the mean LC procedure time? Methods: A prospective, multicenter gathering of LC procedure times and task component times was performed through the cooperative effort of members of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) at 11 hospitals. The effect of LC time of age, gender, and surgical resident was recorded. Results: The mean LC time for 359 cases was 73 28 min. The percent of this LC time for the following component tasks included: to place and remove trocars, 34%; total dissection time, 40%; intraoperative cholangiogram, 15%; and removing the gallbladder, 7%. Age and gender did not change LC time, but the presence of a surgical resident prolonged LC time from 53 to 79 min due to an increase in all LC component task times.

Conclusions: LC time was globally calibrated in 11 North American hospitals and was found to be affected by expertise but not by gender or age. The mean and standard deviation of LC time can be used for purposes of selfassessing quality performance. Key words: Gallbladder Cholelithiasis Laparoscopy Surgical residents Costs Quality assessment

Correspondence To: L. W. Traverso

Any new procedure will become standardized over time. With the intention of interrupting spiraling health-care costs, surgeons and hospitals have become more involved in cost-effectiveness measures. Most hospital laparoscopic cholecystectomy (LC) costs are incurred while the patient is in the operating room [3]. Half of these operating room costs are due to equipment and the other half are due to personnel. Most hospitals estimate costs and then charge for an operation based on the amount of time the patient is in the operating room. This operating room time consists of in-the-room nonoperating time (30%) and time actually spent performing the operation (70%) [3]. The average procedure time for an LC therefore is an important statistic, useful for hospitals and surgeons in at-

227

tempting to determine whether they are operating within an acceptable range of variation. This type of information must be obtained through analysis of a large number of procedures. The variation around an acceptable procedure time can be quantitated and best analyzed by surgeons because they understand the significant variables. Examples of such factors are severity of disease, presence of anomalous ducts, age, weight, sex, and surgical expertise. Therefore, a prospective multicenter gathering of information for LC procedure times was accomplished through a cooperative effort of members of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES). The hope was to globally calibrate the average procedure time, including minimal and maximal range, of LC in North America. We wished to establish this gold standard so that an acceptable range of performance could be defined. The methods of those with efficient procedure times might then be assessed and made available to others. Our primary goal was to standardize LC procedure time. We were also able to assess the influence of age, sex, and expertise (comparison of the presence of surgical residents vs no residents).

Fig. 1. The mean LC time of 73 28 min is broken down into four major components: dissection or total time for dissection; trocars in/out or time to incise the skin and place trocars plus removing the trocars and closing the skin; IOC or placing and removing the IOC catheter plus performing the IOC; and to remove GB or removing the gallbladder from the abdomen. These figures do not include the setup time between removing the IOC catheter and beginning the gallbladder dissection.

Methods
Each LC case had to meet the following criteria for entry into the study. The LC had to be completed without conversion to an open procedure. An intraoperative cholangiogram (IOC) had to be completed in all cases, and these cholangiograms had to be negative. Once these criteria were met, a procedure-time questionnaire was completed during the operation by the circulating personnel. We also recorded the age and sex of the patient plus the presence and position of a resident surgeon. The components of the skin-to-skin procedure times were as follows: trocar placement from the skin incision to the beginning of the cystic duct dissection; cystic duct dissection until the cystic ductotomy was completed; the intraoperative cholangiogram time from the cystic ductotomy until the cholangiogram catheter was removed; the setup for gallbladder resection from the time the IOC catheter was removed until the gallbladder dissection began; the gallbladder dissection until it was removed from the gallbladder bed; removal of the gallbladder from the time it was removed from the gallbladder bed until it was removed from the abdomen; and then, the time to remove the trocars and close the skin incision. All of these prospectively completed questionnaires were sent for compilation and data analysis to a central site (Virginia Mason Medical CenterVMMC). The following were calculated for the above components of procedure times as well as the total procedure timethe average, standard deviation, minimum, and maximum. In addition, we compared the various components of procedure times using the following variablesage <65 vs 65 years, male vs female, no surgical resident present vs surgical resident on the patients left side (most common position). Significance was tested with an unpaired, two-tailed Students t-test assuming unequal variances. A significant value was assigned, a priori, as p < 0.01. Finally, a correlation coefficient was calculated in an attempt to determine which component task time changed the most as the total procedure time changed. The more the task time varied in parallel with the total procedure the closer the coefficient would be to 1.0.

Table 1. Participating SAGES members and hospitals SAGES member Deziel, D Donohue, J Froines, E Hunter, J Roush, T Simon, I Soper, N Swanstrom, L Traverso, W Unger, S Woods, M Hospital Rush-Presbyterian, Chicago, IL Mayo Clinic, Rochester, MN Group Health Hosp, Seattle, WA Emory Univ, Atlanta, GA US Navy Hosp, Bremerton, WA Sunrise Hosp, Las Vegas, NV Washington Univ, St. Louis, MO Emmanuel Hosp, Portland, OR Virginia Mason MC, Seattle, WA Mt. Sinai Hosp, Miami, FL Wichita Clinic, Wichita, KS n 359 cases 19 25 15 11 35 18 8 31 94 72 31

Table 2. LC procedure time and component task times (n 359)a Task Total LC time Trocars In Cystic duct IOC Setup GB dissection GB removal Trocars out
a

Mean 73 13 15 11 3 14 5 12

Standard Deviation 28 7 11 7 3 8 5 6

Minimum 25 3 2 2 0 1 1 1

Maximum 176 49 64 65 31 53 50 55

Data in minutes

Results The number of cases contributed by participating hospitals is listed in Table 1. The total procedure and component task times are listed in Table 2. The component task times were combined into four LC task groups: trocars in and out; dissection time; IOC time; and time for removal of gall-

bladder. The task times for these four groups are shown in Fig. 1. Did a change in total procedure time correlate more (coefficient closer to 1.0) with one of the component task times or one of the combined LC task groups than with another? The closest correlation was observed with the combined LC task dissection time, which showed a correlation coefficient of 0.89. Further correlation was observed

228 Table 3. Specific component task times: resident versus no residenta Task Total LC time Trocars in Cystic duct dissection IOC Setup Gallbladder dissection Gallbladder removal Trocars out
a

No resident (n 67) 53 20 10 4 96 83 32 11 7 32 95

Resident (left side, n 226) 79 27* 13 7* 17 11* 11 6* 43 15 8* 6 6* 13 6*

Data in minutes: mean standard deviation * p < 0.01, resident vs no resident

Fig. 2. The four major components of LC time as (described in Fig. 1) are compared when a resident was not present and when a resident participated on the left side of the patient (the most common position in this study, n 226). All components were significantly prolonged when a resident was present (p < 0.01).

in decreasing order with cystic duct dissection (0.74) and gallbladder bed dissection (0.64). Trocars in and out (0.52) and IOC times (0.45) had a lower correlation coefficient. The four LC task groups of Fig. 1 were also compared when a resident was present on the left side of the patient (the most common resident participation, n 226) vs no resident (n 67). Significant differences were found in all LC task groups as well as total procedure time (Fig. 2). Specific component task times for the resident vs no resident comparison are listed in Table 3. Significant differences were noted in all times except setup. In 341 cases the sex was recorded and 20% (70/341) were men. In 212 cases the age was recorded and 18% (39/212) were 65 years or older. We found no significant differences when comparing the mean total procedure times for men (78 30 min) vs women (72 27 min) or age <65 years (78 28 min) vs 65 years (78 27 min).

Discussion This prospective, multicenter SAGES cooperative study indicated that in these North American hospitals the average procedure time for laparoscopic cholecystectomy was 73 28 min. Once the mean LC time was quantitated it was a surprise to observe that the mean time to place and remove trocars was 24 min or approximately 34% of the total procedure time. These findings have directed us to place new emphasis on evaluating trocar placement techniques. In regard to dissection, it was a prestudy assumption that dissecting the gallbladder free from the gallbladder bed required a large amount of time as compared to the other tasks during LC. Another surprise was the finding that the total dissection tasks required only 40% of the LC procedure time, which was evenly split between dissecting out the cystic duct and dissecting the gallbladder from its bed. It was assumed that cystic duct dissection proceeded slowly to prevent a bile duct injury due to anomalous bile ducts or a

dangerous dissection through acute and chronic inflammation. It must be remembered that a change in LC procedure time was most correlated with the task of dissection (correlation coefficient 0.89). Therefore the total time spent dissecting was most likely to be the cause of prolonging LC time, even though only 40% of LC time was due to dissection. Finally, since the participants in this trial usually had practiced routine intraoperative cholangiography, it was valuable to learn that the intraoperative cholangiogram required an average time of 10 3 min or 14% of the procedure time. These surgeons felt the information gained from IOC was worth the 10 min, e.g., presence of anomalous ducts, malignant or benign stricutres, and bile duct stones. Factors that might increase the LC procedure time include: the presence of inflammation disease severity resulting in a difficult dissection; presence of anomalous anatomy that slows the dissection; inadequate equipment; and finally, limited expertise of the operating team. It is hoped that the number of cases in this study are sufficient to contain a representative number of patients with a difficult dissection due to disease severity. Although we did not measure disease severity in this study it is felt that the mean procedure time reflects all patients presenting to a general surgeons practice. Some bias for exclusion of difficult cases did exist in our protocol. (We excluded those patients requiring conversion to an open procedure.) The effect of the expertise of the operating team on procedure time became apparent when the presence of a resident slowed the procedure significantly. The LC procedure time with a resident on the left side of the patient was increased by 49% from 53 to 79 min (Table 3). The increased time with resident teaching appeared to be distributed during all components of the LC (Fig. 2). One efficient surgeon without surgical residents submitted a case load with a mean LC time of 37 6 minutes. This data at least quantitates for the first time the increased time spent by surgical attendings during surgical teaching. Another surprise was the lack of effect of patient age and gender on LC procedure time. The likelihood of conversion to an open procedure (suggesting difficult dissection by long-standing cholecystitis) had been shown to be increased by age [1] or male gender [1, 2]. Perhaps this discrepancy can be explained by the studies analyzing a larger number of patients (n 1,676 [1] and 587 [2]). This global North American average procedure time may allow comparison of LC procedure times between in-

229

stitutions. If there is too much variance from the mean time, technical details of the performance of others with shorter procedure times may be worth study. Innovations learned from these observations may help to further narrow and decrease the standard LC time. It must be emphasized that variation is inherent in any biologic process, and this variation, once defined, can be improved. In the case of LC, improvement in procedure times will come with surgical innovation and sharing of methodology.

References
1. Fried GM, Barkun JS, Sigman HH, Joseph L, Clas D, Garzon J, Hinchey EJ, Meakins JL. (1994) Factors determining conversion to laparotomy in patients undergoing laparoscopic cholecystectomy. Am J Surg 167: 3541 2. Hutchinson CW, Traverso LW, Lee FT (1994) Laparoscopic cholecystectomy. Do preoperative factors predict the need to convert to open? Surg Endosc 8: 875878 3. Traverso LW, Hargrave K (1995) A prospective cost analysis of laparoscopic cholecystectomy. Am J Surg 169: 503507

Surg Endosc (1997) 11: 245248

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The laparoscopic management of post-transplant lymphocele

A critical review
W. S. Melvin,1 G. L. Bumgardner,2 E. A. Davies,2 E. A. Elkhammas,2 M. L. Henry,2 R. M. Ferguson2
1 2

Division of General Surgery, Department of Surgery, Ohio State University, 410 W. Tenth Avenue, Columbus, OH 43210, USA Division of Transplantation, Department of Surgery, Ohio State University, 410 W. Tenth Avenue, Columbus, OH 43210, USA

Received: 15 March 1996/Accepted: 3 July 1996

Abstract Background: The management of lymphocele in patients following kidney (KT) and kidney pancreas (KPT) transplants is evolving. Open surgery has been the traditional treatment, but some authors have advocated laparoscopic drainage in selected patients. Methods: We retrospectively reviewed our results in lymphocele treatment since developing a laparoscopic program at our institution. Results: Between May 1994 and June 1995, 186 KTs and 48 KPTs were performed, and 1,354 patients are currently being followed. Eight patients developed symptomatic lymphoceles an average of 26 months (range 459) following 6 KTs and 2 KPTs. All patients diagnosed were successfully drained laparoscopically, with no conversions to open surgery. Laparoscopic ultrasound was used to help with localization of the fluid collection. Operative time averaged 59 min, median hospital stay was 1 day (range 14), and there were no perioperative complications. Follow-up imaging was obtained on six patients, 316 months following their procedures, and no recurrences were noted. A review of the literature demonstrates a 5.3% rate of major complications and a 7% incidence of lymphocele recurrence. Conclusions: Intraoperative laparoscopic ultrasound can help localize fluid collections and prevent organ injuries. Laparoscopic drainage of lymphocele following transplantation results in minimal disability and an acceptable complication rate, although it is higher than with open drainage. Therefore, laparoscopic drainage should be considered as primary treatment for all patients with symptomatic posttransplant lymphocele. Key words: Lymphocele Laparoscopy Renal transplantation Ureteral injury

Renal transplantation is complicated by lymphocele formation in 118% of transplantations [3, 12, 20, 23, 28]. Surgical technique, retransplantation, and rejection episodes have been suggested as risk factors [10]. Therapeutic intervention is necessary for symptomatic lymphocele causing allograft obstruction, lower extremity edema, or symptoms related to the mass. Suggested treatment modalities have been percutaneous aspiration with or without placement of drains, injection with sclerosing agents or fibrin sealant, and surgical transperitoneal fenestration. Advances in laparoscopic instrumentation and technical skills now allow a minimally invasive approach to transperitoneal drainage. Many reports now advocate this technique, espousing a low morbidity, short hospital stay, and effectiveness of this technique. Herein, we report our institutions experience in the treatment of post-transplant lymphocele and critically review the available data on the management of this problem.

Patients and methods

The records of all patients who had undergone kidney or combined kidney/pancreas transplants at our institution from 1985 to 1995 were evaluated. All patients diagnosed with postoperative symptomatic pelvic lymphocele were evaluated. Operative records, discharge summaries, radiographic reports, and outpatient records were reviewed. Currently, all patients who present with an elevated creatinine or symptoms of mass effect near the transplanted kidney undergo diagnostic ultrasound. If ultrasound demonstrates a fluid collection, percutaneous aspiration confirms the diagnosis of lymph vs urine leakage and results in definitive therapy for some. If the fluid collection recurs, then the patient undergoes definitive internal surgical drainage. Prior to March 1994, all indicated patients underwent laparotomy with open fenestration of the lymphocele into the peritoneal cavity. After March 1994, patients had transperitoneal laparoscopic fenestration of their lymph collections.

Presented as a Platform Presentation at the Fifth World Congress of Endoscopic Surgery of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA, March 1996 Correspondence to: W. S. Melvin

Results Our institution currently performs approximately 200 renal transplants and 40 combined renal/pancreas transplants per

246

year. One thousand three hundred fifty-four patients with functional renal allografts were included in the review. Between April 1992 and March 1994, eight patients underwent open drainage of post-transplant lymphoceles. There were six males and two females with an average age of 46.9 years. Lymphocele was diagnosed an average of 3.3 months following transplantation with a range of 1.5 to 6 months. The average time for the surgical procedure was 55 min ranging from 34 to 107 min. Patients were discharged an average of 5.2 days postoperatively with a range of 3 to 11 days. The only complication was a postoperative pneumonia. There were no complications related to the surgical procedure. One patient was treated for acute rejection during the same hospitalization. Between March 1994 and July 1995, eight patients underwent laparoscopic drainage of their lymph collections. There were six males and two females with an average age of 52, who had six cadaveric renal transplants and two combined kidney/pancreas transplants. Diagnosis of lymphocele was made from 4 to 59 months following operation with an average of 26 months. The preoperative size of the fluid collection, determined by ultrasound, was fairly consistent. The smallest collection was measured at 10 6.4 cm; the two largest collections were estimated to be greater than 3,000 cc. Presenting symptoms were pain and/or pressure in five patients; five patients also had hydronephrosis with urinary symptoms and an elevated creatinine. The diagnosis was made by ultrasound in all eight patients and further delineated by computerized tomography in three. Operative time averaged 59 min with a range of 33 to 88 minutes. Intraoperative ultrasound was used in the last two patients. There were no conversions to open and no complications from the surgery. Combined laparoscopic cholecystectomy was performed in one patient, and one patient had a saphenous vein ligation while still anesthetized. Five patients stayed in the hospital overnight; one patient each stayed 2, 3, and 4 days, yielding an average hospital stay of 1.75 days. The minimum follow-up was 3 months; average follow-up was 9.1 months and the longest was 16 months. No patients have recurrent symptoms and six patients have had repeat diagnostic studies that have demonstrated no collection of fluid.

Table 1. Lymphocele drainage: review of the literature Open Reports No. of patients Recurrence Complications Total problems 7 52 2 (3.8%) 2 (3.8%) 4 (7.6%) Laparoscopic 20 57 4 (7.0%) 3 (5.3%) 7 (12.3%)

carefully incising the thickened wall in layers. A generous portion of lymphocele wall was removed to allow adequate drainage. Careful inspection of the interior of the lymphocele with the laparoscope allowed any loculations to be broken up. Irrigation of the cavity was performed with saline. Repeat ultrasound then confirmed the fact that the entire lymphocele had been drained. In some cases Foley catheterization and distention of the bladder with methylene blue helped identify the bladder, especially in cases of a low lymphocele and in the two patients who had combined kidney/pancreas transplants. Fascial incisions were always closed with nonabsorbable sutures and infiltrated with local anesthetic. Patients were monitored overnight and given oral liquids and analgesics on the operative day.

Discussion During the logarithmic growth phase of laparoscopic surgery, some authors enthusiastically embraced laparoscopic surgery as a tool to eliminate morbidity in surgical procedures. Applying laparoscopic techniques to transplant lymphocele, many surgeons were quick to report their accomplishments and, based on their experiences, to make recommendations as to the most appropriate treatment of the post-transplant lymphocele. After 5 years of experience reported in the literature, it is now reasonable to critically review the data and discuss the appropriateness of this treatment technique. The treatment of post-transplant lymphocele has evolved over time. Initially, repeated aspirations were done, and if drainage was required, external marsupialization and drainage with Penrose drains were performed [28]. With experience, it was realized that this technique led to a high incidence of wound-care problems and infected fluid collections. Open transperitoneal drainage then became the standard technique, including the practice of excising portions of the lymphocele wall and, sometimes, omentopexy [8, 12, 20]. Our experience paralleled these results; eight patients underwent open drainage during our study period with no complications or recurrences and an average postoperative stay of 5.25 days. Combining six previous reports of open drainage from the literature, a total of 52 cases were included with only two lymphocele recurrences and two operative complications (3.8%). These included a late wound hernia and an intraoperative bladder injury [8, 10, 12, 17, 20, 23] (see Table 1). In comparing our patients undergoing open drainage to our laparoscopic group, the diagnosis of lymphocele was made considerably longer after surgery in recent years. The reason for this discrepancy is not completely clear; however, it may because of our changing patient population, or because changes in the immunosup-

Operative technique Open drainage was accomplished through a midline laparotomy in all patients. This was performed under either spinal or epidural or general anesthesia. After identification, a portion of the lymphocele wall was excised and drainage was allowed into the peritoneum. The laparoscopic technique was performed under general anesthetic. Insufflation of the abdomen was usually accomplished using an open technique, and a 10-mm telescope was used. One or two additional 5-mm working ports were placed away from the site of the lymphocele. Intraoperative ultrasound was used to delineate the anatomy and location of the lymphocele. Location of the lymphocele was confirmed by aspiration of fluid. Following this, electrocautery and sharp dissection were used to incise the lymphocele wall. After opening the lymphocele, the ureter was avoided by making a linear incision in craniocaudad direction and

247

pression regimen have decreased the incidence of early rejection, a recognized risk factor for lymphocele formation. Laparoscopic treatment of post-transplant lymphocele was first reported in 1991, and the worldwide experience is still somewhat limited. The original report was quickly followed by many other institutions reporting one or two cases. At the end of 1995, only 19 series were found in the English literature and, of these, only five reports had more than two patients [1, 2, 47, 11, 1316, 18, 19, 21, 22, 2428]. Two reports had the maximum experience of nine cases each [5, 11]. Including our current series of eight patients, a total of 62 attempts at laparoscopic drainage have been reported with 57 successful laparoscopic drainages. Complications were not rare. Division of the transplanted ureter occurred in four cases, or 7% of all cases reported. This significant complication resulted in open laparotomy and repair of the transplanted ureter in each case. In one case the diagnosis was not made intraoperatively and a second operation was necessary. Ureteral division is by far the most common complication reported; no other technical complications were identified. As the risk of ureteral injury was recognized, a variety of techniques have been described to avoid it. Preoperative stent placement has been used, and placement of a fiberopticlighted stent has even been proposed [18]. However, this is a quite invasive procedure, and retrograde cannulation of the transplanted ureter may be a technically difficult procedure. Percutaneous filling of the lymphocele with methylene blue allows better identification of the lymphocele itself; however, identification of the ureter may still be difficult, especially in the thick-walled lymph cavity [22]. Transillumination of a lymphocele via a second flexible scope passed through an accessory port has been used in a single case in a patient with a kidney/pancreas transplant to help avoid injury while draining a multiloculated lymphocele [21]. Extracorporeal intraoperative ultrasound has been used extensively to diagnose and therapeutically aspirate lymphocele and can be used to help facilitate the dissection in open drainage [20]. However, it cannot visualize beyond the interface created by the pneumoperitoneum of laparoscopy. We have used the intraoperative laparoscopic ultrasound probe to identify the lymphocele and adjacent structures but acknowledge the difficulty in identifying the transplanted ureter itself. The method of choice to avoid the common and serious complication of ureteral division is careful dissection and precise identification of all tissues prior to sharp dissection, with appropriate use of some of these ancillary techniques as needed, combined with sound surgical judgment. Recurrence following intraabdominal drainage of lymphocele is recognized as an uncommon occurrence. Accordingly, the recurrence rate following laparoscopic drainage should be low, if this procedure is to be advocated. Fortynine of the laparoscopic cases reported had follow-up identified and most were confirmed with imaging studies. Three recurrences were diagnosed 3, 12, and 20 months following laparoscopic drainage. This incidence of recurrence of 6% (3/49) seems comparable to our review of open drainage procedures, in which two recurrences were reported in a total of 52 cases (3.8%). Adequate resection of the lymphocele wall and complete drainage of a multiloculated collection are important in preventing recurrence. More informa-

tion on recurrence rates following laparoscopic drainage will become available as follow-up times lengthen, and further reports are published. The overall rate of conversion of laparoscopic procedures to open is difficult to estimate at this time. Only one previous report has identified an institutions series on the treatment of all lymphoceles done during a set time period. At the University of Minnesota, all 14 patients diagnosed with lymphoceles had attempted laparoscopic drainage; five patients required conversion to open [5]. We have been fortunate to be successful in performing laparoscopic drainage in all patients diagnosed at our institution during the last 18 months. The risk factors for conversion to open include previous abdominal surgery, multiple fluid collections, and lymphocele location. Gruesner et al. identifies a lymphocele located in the lateroposterior or lateroinferior position as a risk factor for conversion to open surgery [5]. Continued experience and perhaps the use of the intraoperative ultrasound probe may continue to improve these results and decrease the incidence of conversion to open surgery. Overall, the treatment of post-transplant lymphocele continues to evolve. Laparoscopy is a safe technique which can be used to perform intraperitoneal fenestration of the lymphocele. Surgeons must maintain an acute awareness of the risk of injury to the transplanted ureter and take active steps to avoid this unfortunate event. Despite the increased risk of complications, laparoscopic techniques remain the choice of treatment for lymphocele because of the overriding decrease in patient discomfort, disability, and hospital stay. References
1. Ancona E, Rigotti P, Zaninotto G, Comandella MG, Morpurgo E, Costantini M (1991) Treatment of lymphocele following renal transplantation by laparoscopic surgery. Int Surg 76: 261263 2. Bardot SF, Montie JE, Jackson CL, Seiler JC (1992) Laparoscopic surgical technique for internal drainage of pelvic lymphocele. J Urol 147: 908909 3. Braun WE, Banowsky LH, Straffon RA, et al. (1974) Lymphoceles associated with renal transplantation: report of 15 cases and review of the literature. Am J Med 57: 714 4. Fahlenkamp D, Raatz D, Schonberger B, Loening SA (1993) Laparoscopic lymphocele drainage after renal transplantation. J Urol 150: 316318 5. Gruessner RWG, Fasola C, Benedetti E, Foshager MC, Gruessner AC, Matas AJ, Najarian JS, et al. (1995) Laparoscopic drainage of lymphoceles after kidney transplantation: indications and limitations. Surgery 117: 288295 6. Iselin CE, Rochat CH, Morel P, Merlini M (1994) Laparoscopic drainage of postoperative pelvic lymphocele. Br J Surg 81: 274275 7. Ishitani MB, DeAngelis GA, Sistrom CL, Rodgers BM, Pruett TL (1994) Laparoscopic ultrasound-guided drainage of lymphoceles following renal transplantation. J Laparoendosc Surg 4: 1 8. Kay R, Fuchs E, Barry JM (1980) Management of postoperative pelvic lymphoceles. Urology 15: 345347 9. Khauli RB, Mosenthal AC, Caushaj PF (1992) Treatment of lymphocele and lymphatic fistula following renal transplantation by laparoscopic peritoneal window. J Urol 147: 13531355 10. Khauli RB, Stoff JS, Lovewell T, Ghavamian R, Baker S (1993) Post-transplant lymphoceles: a critical look into the risk factors, pathophysiology and management. J Urol 150: 2226 11. Lange V, Schardey HM, Meyer G, Illner WD, Petersen P, Land W (1994) Laparoscopic deroofing of post-transplant lymphoceles. Transpl Int 7: 140143 12. Langle F, Schurawitzki H, Muhlbacher F, et al. (1990) Treatment of lymphoceles following renal transplantation. Transplant Proc 22: 14201422 13. Manncke KH, Stoblen F, Hopt UT, Buess G (1993) Laparoscopic

248 fenestration of a lymphocele following combined pancreas-kidney transplantation. Surg Laparosc Endosc 3: 10911 McCullough CS, Soper NJ, Clayman RV, So SSK, Jendrisak MD, Hanto DW (1991) Laparoscopic drainage of a posttransplant lymphocele. Transplantation 51: 725741 Mourad M, Bertin D, Gigot JF, Squifflet JP (1994) Laparoscopic treatment of lymphoceles following kidney transplantation by intraperitoneal fenestration and omentoplasty. Surg Endosc 8: 14271430 Mulgaonkar S, Jacobs MG, Viscuso R, Lyman N, Klein P, Bravo B, Clavello A, et al. (1992) Laparoscopic internal drainage of lymphocele in renal transplant. Am J Kidney Dis 19: 490492 Olsson CA, Willscher MK, Filoso AM, Cho SI (1976) Treatment of posttransplant lymphoceles: internal versus external drainage. Transplant Proc 8: 501504 Paolucci V, Meyer W, Schaeff B, Monico R (1994) Laparoscopic drainage of a lymphocele after renal transplantation. Surg Endosc 8: 111113 Persson NH, Almquist P, Ekberg H, Kallen R, Loren I, Montgomery A (1994) Laparoscopic drainage of renal transplant lymphocelewith and without complications. Transplant Proc 26: 1765 Pollak R, Veremis SA, Maddux MS, Moxes MF (1988) The natural history of and therapy for perirenal fluid collections following renal transplantation. J Urol 140: 716720 21. Ratner LE, Bender JS (1994) A novel approach to the drainage of loculated perirenal allograft lymphoceles. Transplantation 58: 961 964 22. Schilling M, Abendroth D, Kunz R (1995) Treatment of lymphocele in renal transplant recipients by laparoscopic fenestration after transcutaneous staining. Br J Surg 82: 246248 23. Schweizer RT, Cho S, Kountz SL, Belzer FO (1972) Lymphoceles following renal transplantation. Arch Surg 104: 4245 24. Seiler Ch, Horber F, Czerniak A (1994) Laparoscopic intraperitoneal drainage of lymphocele after renal transplantation. Nephrol Dial Transplant 9: 185186 25. Shokeir AA, Eraky I, El-Kappany H, Ghoneim MA (1994) Accidental division of the transplanted ureter during laparoscopic drainage of lymphocele. J Urol 151; 16231625 26. Slavis SA, Gardner LD, Swift C, Gross ML (1992) Laparoscopic drainage of lymphocele after renal transplantation. J Urol 148: 9697 27. Voeller G, Butts A, Vera S (1992) Kidney transplant lymphocele: treatment with laparoscopic drainage and omental packing. J Laparoendosc Surg 2: 1 28. Zincke H, Woods JE, Aguilo JJ, et al. (1975) Experience with lymphoceles after renal transplantation. Surgery 77: 444450

14. 15. 16. 17. 18. 19. 20.

Surg Endosc (1997) 11: 261263

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Nissen and Toupet fundoplications effectively inhibit gastroesophageal reflux irrespective of natural anatomy and function
W. S. Richardson,1 T. L. Trus,1 S. Thompson,2 J. G. Hunter1
1 2

Department of Surgery, Emory University School of Medicine, 1364 Clifton Road, N.E., Atlanta, GA 30322, USA Ethicon Endosurgery, Cincinnati, OH, USA

Abstract Background: The physiology of Nissen fundoplication (NF) and Toupet fundoplication (TF) is controversial. The aim of this study was to determine the contribution of elevated intragastric pressure to the antireflux mechanism after surgically created fundoplication in explanted porcine stomachs. Methods: The stomachs and 68 cm of distal esophagus were removed from 15 pigs and placed in anatomic position. Five NF, 2 cm in length with three interrupted sutures, were performed, taking full-thickness bites of stomach and partial-thickness bites of esophagus around a 60 French dilator. Five 270 TF 2 cm in length with six interrupted sutures were performed taking full-thickness bites of stomach and partial-thickness bites of esophagus around a 60 French dilator. Each stomach served as its own control. The pylorus was tied off and the stomach was inflated with Ringers lactate while the pressure was monitored. Results: Before NF, reflux could be easily induced with a mean intragastric pressure of 5.5 3.7 mmHg. After NF reflux could not be induced but the sutures pulled out of the stomach at a mean pressure of 36.8 11.7 mmHg (p < 0.01 vs control). Before TF, reflux could easily be induced with a mean intragastric pressure of 3.0 3.0 mmHg. After TF, reflux could not be induced and the sutures pulled out of the esophagus or stomach with a mean pressure of 30.8 9.0 mmHg (p < 0.01 vs control). Porcine stomachs in vivo are resistant to reflux, but when explanted they reflux easily. NF and TF are so effective at interrupting reflux that the sutures tear out instead of allowing reflux. Conclusions: While not yet statistically significant, it appears that sutures tear out of the esophagus (TF) more readily than they tear out of the stomach (NF). TF and NF prevent reflux in the absence of anatomic or functional components of the lower esophageal sphincter.
This study was performed in cooperation with and with support from Ethicon EndoSurgery, Cincinnati, OH Presented at the 5th World Congress of Endoscopic Surgery of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA Correspondence to: W. S. Richardson

Key words: Nissen fundoplication Toupet fundoplication Gastroesophageal reflux

The Nissen fundoplication is generally considered the most competent surgically created antireflux valve with the greatest augmentation of resting lower esophageal sphincter pressure (LESP) [2], yet the physiology is incompletely understood. Aspects of the fundoplication thought to be important to its function as an antireflux valve are the creation of a cardioesophageal angle, the length of the intraabdominal esophagus, and distension of the wrap compressing the distal esophagus. The aim of this study is to determine the contribution of elevated intragastric pressure to the maintenance or failure of the antireflux mechanism after surgically created fundoplication in explanted porcine stomachs.

Methods
The stomachs along with 68 cm of distal esophagus from ten 3040-kg pigs were removed. The stomachs were placed in anatomic position. Each stomach served as its own control. The pylorus was tied off around a piece of IV tubing which was attached to a reservoir of Ringers lactate and was used to inflate the stomach. The pressure was transduced through IV tubing and a DTX pressure transducer (Visso-Spectramed, Oxnard, CA) which was brought into the stomach via a 14-gauge needle. A monitor (Spacelabs Medical, Inc., Redmond, Washington) was used to determine the pressure. Ringers lactate was infused into untreated stomachs until reflux of fluid occurred out the open end of the esophagus, and at that point the intragastric pressure was recorded. In the first part of the experiment, five Nissen fundoplications were performed around a 60 French dilator. Three 2-0 silk sutures were placed 1 cm apart to create a 2-cm-long fundoplication taking full-thickness bites of stomach and partial-thickness bites of esophagus (Fig. 1). Five 270 fundoplications (Toupet) were fashioned around a 60 French dilator using two rows of 2-0 silk sutures on either side of the anterior vagus nerve, taking full-thickness bites of stomach and partialthickness bites of esophagus (Fig. 2). Three sutures were placed in each row, 1 cm apart, to create a 2-cm-long fundoplication. The stomachs were again inflated until either reflux of fluid or breakdown of the suture line occurred. The intragastric pressure at the time of failure was recorded. Each stomach served as its own control. In the second part of the experiment esophageal manometry was per-

262

Fig. 1. A schematic of the experiment. The stomach is inflated with Ringers lactate through the pylorus while the intragastric pressure is constantly monitored through the side of the stomach. The end of a manometry catheter is seen exiting the esophagus for the second part of the experiment.

Fig. 2. The Nissen fundoplication partially inflated. Fig. 3. The Toupet fundoplication partially inflated.

formed. Seven stomachs and distal esophagi were harvested from 4050-kg pigs. A Zinectics EMC radial esophageal manometric catheter EMC8-R (Zinectics Medical, Salt Lake City, UT) was used to measure maximal distal esophageal pressure at the level of the LES before operation and after Nissen fundoplication with an empty stomach. Each measurement was taken three times and averaged (Synectics 8 channel PC polygraph HR upper GI motility system and Gastrosoft version 5.0 6C2 software, Irvine, TX). The stomachs were then inflated with Ringers lactate as described above, and intragastric pressure was held constant at 0, 2.5, 5, 10, 20, and 30 mmHg while maximum distal esophageal pressure at the level of the LES was measured.

Results Before Nissen fundoplication, reflux occurred at 5.5 3.7 mmHg of intragastric pressure. After Nissen fundoplication, reflux was not induced but suture lines pulled out of the stomach at a pressure of 36.8 11.7 mmHg (p < 0.01 vs control). This extreme valve competence occurred because the fundus filled with fluid, expanded, and blocked the esophagus. Before Toupet fundoplication reflux could eas-

ily be induced with a mean pressure of 3.0 3.0 mmHg. After Toupet fundoplication (Fig. 3) reflux could not be induced but sutures pulled out of either the stomach or the esophagus with a mean pressure of 30.8 9.0 mmHg (Table 1). With an empty stomach average resting pressure in the distal esophagus before and after Nissen fundoplication was 14.8 mmHg and 14.9 mmHg, respectively (Table 2). As intragastric pressure was increased, maximum pressure in the distal esophagus linearly correlated with intragastric pressure, and a best-fit line gives a slope of 4.6 and a yintercept of 14 (Fig. 4). At an intragastric pressure of 30 mmHg the average maximal pressure of the LES was 152 mmHg.

Discussion This study demonstrates that standard antireflux valves (Nissen or Toupet) are 100% competent (when properly

263 Table 1. Mean pressure of reflux or disruption of fundoplication+ Nissen Control Fundoplication 5.5 3.7* 36.8 11.7+ p < 0.0005 Toupet 3.0 3.0* 30.8 9.0+ p < 0.0005 p ns ns

Table 2. Resting pressure in region of LES with an empty stomach before and after Nissen fundoplication Intragastric pressure Control Nissen fundoplication 0 0 Maximal pressure at LES 14.8 3.6 14.9 2.8 ns Fig. 4. Maximum pressure at the level of the LES vs intragastric pressure after Nissen fundoplication. Intragastric pressure is held constant at ascending pressures during manometry.

constructed) even in the absence of external forces such as the crura of the diaphragm, vagal stimulation, intraabdominal positioning, or sphincter muscle viability. This explanted porcine preparation helps us understand how these valves work. As the stomachs were slowly inflated the fundoplication inflated like inflating an inner tube around a straw, while the intragastric pressures stayed low. As the stomach expanded, the wall tension transmitted to the enwrapped lower esophageal sphincter increased rapidly. As the intragastric pressure increased, the distal esophageal pressure increased linearly at nearly five times the rate. The increasing pressure gradient between stomach and esophagus renders the fundoplication more competent as the stomach fills. In this study, the freshly explanted unaltered porcine LES had a resting pressure of 15 mmHg. Elastic elements in the LES may be responsible for the muscle tone. Despite a normal sphincter pressure, reflux occurred when the intragastric pressure was 35 mmHg. As the stomach was inflated the esophageal orifice dilated, until the LES pressure dropped to the level of intragastric pressure, allowing reflux to occur. After fundoplication the esophageal orifice did not dilate (perhaps the mechanism of Angelchik prosthesis effectiveness), and instead, extrinsic pressure was generated, increasing pressure in the LES. Predictably, failure occurred when wall tension was so great that the gastric muscle fibers started to separate and the sutures pulled out of the esophagus. Similar results have been shown in a cadaver study [1].

The clinical implications of this are clear in the postfundoplication stomach. The more dilated the stomach (and it is difficult to generate more than 10 mmHg pressure in the normal stomach with an open pylorus) the less likely one is to reflux. As well, the more bloated the individual, the harder it is to belch. Further maneuvers, such as decreasing the length of the valve or decreasing the circumference of the fundoplication, need to be performed in this model in order to decrease the competency of the wrap and determine thresholds for reflux. In conclusion, fundoplication function results from augmentation of LESP. Poor gastric motility and increased intragastric pressure may induce a very elevated resting LESP that could result in dysphagia, poor relaxation of the LES, bloating, and inability to belch.
Acknowledgment. We would like to acknowledge Mary Mootoo and Julianne Meister for their help in obtaining the manometry data.

References
1. Butterfield WC (1971) Current hiatal hernia repairs: similarities, mechanisms and extended indicationsan autopsy study. Surgery 69(6): 910916 2. DeMeester TR, Johnson LF, Kent AH (1974) Evaluation of current operations for the prevention of gastroesophageal reflux. Ann Surg 180: 511523 3. Stein HJ, et al. (1991) Three-dimensional imaging of the lower esophageal sphincter in gastroesophageal reflux disease. Ann Surg 214(4): 374384

Surg Endosc (1997) 11: 268271

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Conversions and complications in 185 laparoscopic adjustable silicone gastric banding cases
E. Chelala,1 G. B. Cadie re,1 F. Favretti,2 J. Himpens,1 M. Vertruyen,1 J. Bruyns,1 L. Maroquin,1 M. Lise2
1 2

Department of G.I. Surgery, University Hospital Center Saint-Pierre, rue Haute, 322 1000 Brussels, Belgium University Hospital of Padoua, Italy

Received: 28 May 1996/Accepted: 25 July 1996

Abstract Background: Kuzmaks gastric silicone banding technique is the least invasive operation for morbid obesity. The purpose of this study was to analyze the complications of this approach. Methods: Between September 1992 and March 1996, 185 patients underwent laparoscopic gastroplasty by the adjustable silicone band technique. A minimally invasive procedure using five trocars was performed. Results: In 11 patients exposure of the hiatus was impeded because of hypertrophy of the left liver lobe which led to conversion in eight patients and abortion of the procedure in three other patients. Anatomical complications: We observed two gastric perforations and one band slippage at the early stage, one infection and three rotations of the access port. Functional complications: There were eight (4%) cases of irreversible total food intolerance resulting in pouch dilation and eight cases (4%) of esophagitis. One fatality on the 45th day in a patient with a Prader-Willi syndrome. Conclusion: The most disturbing complications of gastric banding technique are gastric perforation and pouch dilation. Their incidence may be reduced by improving the technique and by considering pitfalls of the procedure. Key words: Laparoscopic Gastroplasty Conversions Complications

not uncommon [9, 10, 12], and questions concerning pouch volume remain a common matter of debate [5, 12]. In an attempt to minimize the incidence of complications in the future we describe the pitfalls of the operative technique and detail how to avoid pouch dilation by trying to understand its physiopathology.

Patients and methods

In 1992 one of us [1] first showed the feasibility of using the laparoscopic approach for the Kuzmak gastric banding technique in five patients. Few modifications had to be introduced to make the band suitable for laparoscopy. After these changes, between September 1993 and March 1996 a total of 185 patients (150 female and 35 males) selected on basis of the criteria of the American Society of Bariatric Surgery [3, 15] underwent laparoscopic adjustable silicone gastric banding (LASGB). Median age was 38 years (1865). The average body weight was 118 kg (74202), the percentage of excess weight as compared to the ideal body weight was 199 (105268); the mean body mass index was 43 (3467); 37 patients had had previous abdominal surgery. In 136 patients, gastroscopy found no evidence of esophagitis whereas 12 patients had esophagitis stage I, five stage II, and two stage III. Sixteen patients presented a hiatal hernia less than 2 cm. After initiation of the pneumoperitoneum, five trocars are needed. A calibrated balloon-tipped orogastric tube is inserted in the stomach. The balloon is inflated with 25 cc of saline and then pulled back to the esogastric junction. Dissection of the lesser curvature is initiated with the coagulation hook, starting at the equator level of the inflated balloon. Dissection is then resumed 1 cm to the left of the left hiatal pillar. A tunnel is created behind the stomach, joining the two dissected areas. A roticulating grasper is passed through the tunnel and grasps one end of the inflatable prosthesis. The band is looped around the stomach, thus creating a 15-cc proximal pouch. Slipping of the band is prevented by four anterior seroserosal stitches. Finally, a posterior fixation is performed after opening the pars flaccida of the gastrohepatic ligament. At the end of the procedure, an implantable reservoir is fixed on the left anterior rectus sheath, just distal to the costal edge.

Kuzmaks adjustable gastric banding procedure has been well established and has proven efficacy in morbidly obese patients. The procedure by laparoscopic approach has been performed in our institution since 1992 [1] with fewer general complications than the open procedure [7, 11]. However, complications related to the technique and the band are

Results In our study, the mean operation time was 90 min (40240). In three patients laparoscopy was abandoned right from the

Correspondence to: G. G. Cadiere

269 Table 1. Reoperation for late complications (N 8; 4%) Patient No. (N 8) 3 5 10 16 22 24 26 32 Charact BMI 36 Op 12/92 BMI 41 Op 12/93 BMI 35 Op 10/94 BMI 41 Op 11.94 BMI 42 Op 11/94 BMI 36 Op 12/94 BMI 34 Op 12/94 BMI 43 Op 12/94 Reason Defective band Psychological Pouch dilation Pouch dilation Pouch dilation Pouch dilation Pouch dilation Pouch dilation Period postop 36 months 12 months 14 months 5 months 3 months 12 months 4 months 4 months Laparoscopic treatment Replacement Removal Reposition Removal Reposition Reposition Removal Reposition

beginning because of left liver lobe hypertrophy. Eight conversions were necessary because of difficult and risky dissection, short instruments, and incorrect band position. The mean hospital stay was 2 days (210). Gastrograffin swallow was performed on the 1st postoperative day in order to check for leakage. Band-related complications are divided in anatomical and functional. Anatomical complications Two patients had gastric perforation in their early postoperative course. One was caused by traumatic placement of a nasogastric tube on the 1st postoperative day and was treated by closure of the perforation by laparotomy. The second happened on the 5th postoperative day and necessitated laparoscopic reoperation for removal of the band and gastrorraphy. One band that had not been fixed slipped and induced food intolerance early on. One aspiration pneumonia was diagnosed on the 2nd postoperative day (patient 7) and was treated by deflation of the band and antibiotherapy. One patient had an infection of the access port 12 months postoperatively which was treated medically. Three access ports rotated (upside-down), necessitating reposition and refixation under local anesthesia. Functional complications Requiring reoperation. Functional complications requiring reoperation were noted in eight patients (see Table 1). We observed six cases of irreversible total food intolerance resulting in pouch dilation requiring surgery for repositioning in four cases and removal of the band in two cases. In two cases the band had to be replaced. In one of these, the band was removed for psychological reasons after 12 months. The patient was operated on 35 months later using the Mason II procedure. The other patient presented leakage of the band diagnosed radiologically. The band was replaced after 36 months. Requiring conservative treatment. Four percent of the patients had recurrent heartburn with esophagitis (four stage

I, two stage II, one stage III) on control gastroscopy. Successful medical treatment was instituted with omeprazole. There was one mortality: It occurred 45 days postoperatively in a patient with Prader-Willi syndrome. Cause of death was gastric bleeding.

Discussion It is now widely admitted that surgery is the only effective approach for optimizing weight loss in the morbidly obese patient [13, 16]. After a well-established open gastric banding procedure was described by Kuzmak, and after observing his long-term results [11], we started using the laparoscopic approach as of October 1992. The modification of the technique and new devices allowed us to reduce our operating time and to decrease anatomical and functional complications. We cancelled surgery for three patients judged to be contraindicated for the laparoscopic approach because of left liver hypertrophy as documented by subxyphoid ultrasonography measuring the depth of the left liver lobe. Now, however, we think that there is no final correlation between feasibility by laparoscopy and radiological assessment of left liver size. Decision about feasibility should be delayed until after introduction of the laparoscope and retraction of the liver to the right. No further conversions were noted, since we place the liver retractor more to the left of the xyphoid, thereby moving the liver from the left to the right. In eight patients we had to convert. In four cases dissection was judged hazardous because of left liver hypertrophy: In two cases early in our experience in superobese patients the instruments were too short, and in two band position was incorrect. Gastric perforation is a serious early perioperative complication in LASGB. This occurred in two cases in our earlier patients, probably because efforts to create the retrogastric tunnel were too traumatic. In one of the two patients we did remove the band after closure of the perforation by laparoscopy on the 5th day. In the other case successful gastrorraphy was performed without removal of the band because it happened on the 1st postoperative day and

270

Fig. 1. Pathophysiology and management of pouch dilation.

soiling was only minimal. We suggest that removal of the band and gastrostomy should not be done routinely if the perforation is diagnosed early. Choosing reference landmarks (equator of the calibrated tube balloon), dissecting under vision (thanks to the use of 30 optical system), dissecting downward the left crus, reducing the blind area, and resisting the impulse to push the roticulating forceps forcefully may further minimize the risk of perforation. Since too tight a calibration of stoma during operation may contribute to gastric tissue edema, early food intolerance, esophagitis, and aspiration pneumonia, as seen in patient 7, we changed our strategy, as of patient 34, to keep the band completely deflated after surgery. Insufflation is delayed until after barium swallow at 1 month postoperative. A large debate still exists concerning the positioning, infection, and fixation of the access port. Since we observed three instances of rotation of the access port, we stress that any twist, kink, or insufficient length of the port tubing may contribute to port rotation and that fixation should be done by unresorbable suture, with three stitches on the left rectus sheath. Pouch dilation is a common problem associated with gastroplasty procedure, resulting in late functional stenosis [4, 6, 10, 14]. In an effort to understand the mechanism and the physiopathology of food intolerance and pouch dilation, we tried to analyze some factors that possibly contributed to the enlargement of the proximal pouch. An algorithm of pouch dilation summarizes our policy in Fig. 1. One cause of pouch dilation is excessive vomiting. Overeating and ingestion of sparkling drinks may lead to pouch enlargement and may also contribute to excessive vomiting. Therefore, a restrictive liquid diet must be maintained in the early postoperative stage to allow adhesions to form and to prevent slippage of the gastric wall. All patients to be treated by this method need to fully understand what strict alimentary restriction means and what the consequences are of uncontrolled eating behaviors. Therefore, in our opinion the ap-

propriate treatment in Prader-Willi syndrome is biliopancreatic derivation and not LASGB. We and others [5] believe that a pouch of 15 cc reduces the incidence of pouch enlargement problems. (After changing our policy, reducing from 25 cc to 15 cc, we had less pouch enlargement.) Food intolerance can be managed successfully by band deflation (in some cases with gastric aspiration and parenteral nutrition). If the condition is not promptly solved, irreversible total food intolerance will occur and surgical revision by removing or repositioning the band will be unavoidable. We reported six cases of pouch dilation because of posterior slippage of the gastric wall and two replacements in our first 40 cases. We therefore think band dilation is to be recommended to assure posterior fixation of the band. The Endo stitch (USSC, Norwalk, CT) has proven useful in performing this task. We did not encounter any pouch dilation after beginning posterior fixation as of patient 34. The eight patients with slipped band were reoperated. At laparoscopy, posterior slippage below the peritoneal reflection was apparent. There were remarkably few adhesions after the procedure except from left liver to the band. Technically the procedure started with the dissection of the adhesions between the liver and the anterior gastric wall. A tight fibrous capsula was found, mostly on the proximal pouch, and had to be incised horizontally (scissors) on the anterior side of the ring toward the gastrospleninc ligament. Opening the pars flaccida on the lesser curvature gives access to the lesser sac, providing opening of the posterior fibrous capsula. The band was liberated and the slipped gastric wall was reduced. The band was liberated and repositioned on a calibrating pouch of 15 cc and secured by four anterior and one posterior seroserosal stitches. Many (10%) patients develop esophagitis. Large hiatal hernia as well as important esophagitis (stage III Savary) must therefore be considered a contraindication for banding. Esophagitis stage decreases in the long run. In 4% of the cases, however, recurrent heartburn persists, probably because of alimentary stasis. Interestingly, all of the patients responded to medical treatment by omeprazole. Predisposition to reflux esophagitis [8] according to the law of Laplace [2] and incompetence of the lower esophageal sphincter (observed in 30% of patients) should be considered in these patients. In this preliminary report, we conclude that a longer follow-up is needed to establish the role of LASGB. References
1. Cadie ` re GB, Bruyns J, Himpens J, Favretti F (1994) Laparoscopic gastroplasty for morbid obesity. Br J Surg 81: 1524 2. Catona A, Gossenberg M, Mussini G, La Manna L, De Bastiani T, Armeni E (1995) Videolaparoscopic vertical banded gastroplasty. Obes Surg 5: 323329 3. Deitel M, Shahi B (1992) Morbid obesity: selection of patients for surgery. J Am Coll Nutr 11: 457467 4. Deitel M, Jones Ba, Petrov I, Wcodarczyk SR, Basi S. (1986) Vertical banded gastroplasty: results in 233 patients. Can J Surg 29: 322324 5. Desaive C (1995) Influence of the initial volume of the gastric pouch on the rate of complication after adjustable silicone gastric banding. Obes Surg 5(3): 247 (abstract 16) 6. Desaive C (1995) A critical review of a personal series of 1000 gastroplasties. Int J Obes 19(suppl 3): 561565

271 7. Favretti F, Cadie ` re GB (1995) Laparoscopic placement of adjustable silicone gastric banding: early experience. Obes Surg 5: 7173 8. Fisher BL (1991) Erosive esophagitis following horizontal gastroplasty for morbid obesity. Obes Surg 1: 8993 9. Hall JC, Watts J, OBrien PE, Dunstan RE, Walsh JF, Slavotinek AH, Elmalie RG. (1990) Gastric surgery for morbid obesitythe Adelaide study. Ann Surg 419427 10. Ismail T, Kirby, Crowson MC, Baddeley RM (1990) Vertical Silastic ring gastroplasty: a 6 year experience. Br J Surg 77: 8082 11. Kuzmak LI (1991) A review of seven years experience with silicone gastric banding. Obes Surg 1: 4038 12. Lise M, Favretti F, Belluco C, et al. (1994) Stoma adjustable silicone gastric banding: results in 111 consecutive patients. Obes Surg 4: 274278 13. Lichtman SW, Pisarka K, Berman Pestone M, Dowling H, Offenbacher E, Weisel H, Heshka S, Mattews DE, Ileymsfield SB (1992) Discrepancy between self-reported and actual colonic intake and exercise in obese subjects. N Engl J Med 327:18931898 14. Mason EE, Doherty C, Scott D, Faber LA, Maher JW (1989) Vertical banded gastroplasty: an eight year review. American Society for Bariatric Surgery, Sixth Annual Meeting, Nashville 15. National institutes of Health Consensus Development Conference Draft Statement on Gastrointestinal Surgery for severe obesity (1991) Obes Surg 1: 257265 16. Sugarman HJ, Londrey GL, Kellum JM, et al. (1989) Weight loss with vertical banding gastroplasty and Roux-Y gastric bypass for morbid obesity with selective versus random assignment. Am J Surg 157: 93100

Surg Endosc (1997) 11: 299302

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The role of the endoscopic extraperitoneal approach in large inguinal scrotal hernias
G. S. Ferzli, T. Kiel
Department of Laparoendoscopic Surgery, Staten Island University Hospital, 78 Cromwell Avenue, Staten Island, NY 10304, USA Received: 7 May 1996/Accepted: 12 July 1996

Abstract. The role of endoscopic extraperitoneal herniorrhaphy (EEPH) in the management of giant scrotal hernias has not been well defined, and the technical details relating to operations on such hernias have not been described. We present our experience with 17 patients undergoing repair of giant scrotal hernias. Foley catheter bladder decompression was routinely employed. The Retzius space was developed early in the procedure and hernia sac contents were reduced in all cases. The inferior epigastric vessels were likewise divided in all patients. The average operative time was 76 min and all patients were discharged home the same day. There have been no recurrences on follow-up. There was no mortality, and morbidity was limited to seroma formation in two patients. We conclude that with certain technical modifications, EEPH can be safely employed for the treatment of giant scrotal hernias. Key words: Scrotal hernia Endoscopic extraperitoneal herniorrhaphy Retzius space

In comparison with the various laparoscopic methods available today, endoscopic extraperitoneal inguinal hernia (EEPH) repair combines the benefits of conventional preperitoneal repair, as described by Stoppa and Warlaumont [10] and Nyhus et al. [8, 9], with the advantages of minimally invasive surgery. This technique has been previously described [3], and earlier results have been recorded [5, 6]. But its use in cases of inguinal scrotal hernia has not been reported; the authors herein do so. Patients and methods
EEPH is performed with the patient in the supine position, using general or epidural anesthesia. The basic principle in this technique is the creation of a vacuous preperitoneal extraabdominal space through which the myopectineal orifice can be visualized. To create this cavity, a 1-cm incision is

Correspondence to: G. S. Ferzli

made just below the umbilicus. The linea alba is incised transversely, and the preperitoneal fat is visualized. Blunt finger dissection is performed, allowing the placement of two trocars in the midline, a 10-mm one infraumbilically and a 5-mm one halfway between the umbilicus and the pubic symphysis. The preperitoneal space is maintained as a vacuous working environment by the insufflation of CO2 at a pressure of 910 mmHg. Up to this point, the procedure is similar to all extraperitoneal endoscopic dissections. However, with large inguinal scrotal hernias, the first area that must be developed is the Retzius space. A Foley catheter, which is no longer routinely used in small to moderate-sized inguinal hernias, must be placed in these cases. It has a dual role. By deflating the urinary bladder it increases the Retzius space, and by collapsing the bladder it decreases the probability of injury should the bladder be incorporated in the hernia sac. At this point it is important to reduce the hernia sac contents, a process that likewise has a dual purpose. It provides more space in the working cavity, and it allows easier manipulation of the hernia sac. Once the Retzius space has been maximized and Coopers ligament has been identified on each side, and the obturator fossa spaces have also been located, attention can be directed to the femoral and direct spaces. In cases of direct hernia, a rolled edge unfolds nicely, demarcating the separation between the redundant thickened traversalis and the peritoneal sac (Fig. 1). Dissection continues along this fold, which reduces the entire contents of the direct defect. If gentle traction and countertraction are used, and sharp dissection is avoided, this process is usually free of hemorrhage. The edges of the defect, the iliopubic tract, the epigastric vessels, and the iliac vessels will all be easily visible. The margins of a direct hernia defect should be cleaned of preperitoneal, areola, and fatty tissues. Large defects can be imbricated with suture material or stapled, reducing the amount of dead space and minimizing the chance of a postoperative seroma. In the more commonly encountered large indirect inguinal scrotal hernia, the posterior floor is weakened and bulging inward, while the epigastric vessels are displaced. In this case it is preferable to divide these vessels. This allows easier access to the transversalis fascial sling, which must on occasion be divided to allow access to the sac more distally in its course with the cord structures toward the scrotum. This division also allows for the placement of a smooth, nonconvoluted mesh over the myopectineal orifice. This is not possible prior to vessel division, since the bulging of the weakened floor renders the origin of the vessels more proximal. After division of the epigastric vessels, the surgeon should come back cephalad or posterior to these vessels at a level halfway between the umbilicus and the pubic symphysis (Fig. 2). From there, access can be gained to the transversus abdominis muscle laterally at the level of the anterosuperior iliac spine. Once this crural has been identified at that level, the surgeon continues along this lateral aspect toward the indirect ring (Fig. 2). The indirect hernia sac is now situated between two open spaces, the Retzius space medially and Bogross space laterally. It is important always to approach the deep internal ring along its anterolateral aspect. This has several advantages. It avoids injury to the cord structures and allows identification of a preperitoneal cordal lipoma,

300

Fig. 1. Anatomic relations and initial dissection approach in a large direct scrotal hernia.

Fig. 2. Order of dissection in the isolation of a large indirect hernia sac.

Fig. 3. Two methods for separating the cord structures from the hernia sac in a large indirect scrotal hernia.

301 which can be easily suppressed if found. It makes the sac margins readily visible, allowing the surgeon to grasp the indirect sac without injuring its contents or the cord structures. The suppression of a cordal lipoma also increases space within the working cavity. Dissection of the cord structures is done in a direction perpendicular to them, gently sweeping them medially and posteriorly, and, while holding the sac with the opposite grasper, laterally and superiorly (Fig. 3). An alternate and certainly safer technique is to pivot the hernia sac medially and posteriorly and dissect the cord structures in a perpendicular fashion, sweeping in a lateral and posterior direction (Fig. 3). Avoidance of injury to the internal spermatic vessels is an advantage of this approach. The surgeon can then dissect alternately between the medial and lateral approaches, creating a window between the cord structures and the sac. The latter is either totally invaginated and reduced, or transected, which leaves the distal blunt end in situ while the proximal end is sutured after it is ascertained that the contents have not been injured. As stated before, the dissection of the sac can be accomplished more easily in large inguinal scrotal hernias by dividing the transversalis fascial sling and the transversus abdominis. This allows access to the superficial internal ring and occasionally even the external ring at a point at which the sac diameter has usually tapered down, so that the surgeon can encircle it more easily. If the sac must be transected and then sutured closed, the surgeon should not fear leakage of CO2 into the peritoneal cavity, for the working cavity will not collapse in spite of any such leakage. It will be kept open by the weight of the preperitoneal fat, the bladder in the midline, and laterally by the floppiness of the peritoneum and the weight of the redundant hernia sac. Finally, cord structures must be totally parietalized with the posterior wall so that no element closes the preperitoneal prevesical space. Synthetic mesh is placed from the anterosuperior iliac spine to the pubic symphysis and anchored with staples to Coopers ligament, the midline, and the abdominal wall musculature laterally.

Results Of 167 male patients who underwent 464 hernia repairs at our institution, 17 had large inguinal scrotal hernias, two of them direct and 15 indirect. Sliding hernias were identified in two patients. The average operative time was 76 min, and all patients were discharged home the same day. There have been no recurrences on follow-up. There was no mortality, and morbidity was limited to seroma formation in two patients, one of whom required aspiration. The epigastric vessels were divided in all cases, and the sac was opened in three patients with indirect hernias. There were no pantaloon hernias in this group. Three patients were being operated for recurrent hernias.

Some of the variations suggested by experience, such as epidural anesthesia and blunt finger dissection, are still applicable when dealing with large inguinal scrotal hernias. Other modifications routinely used for small to moderatesized groin hernias are also necessary for large inguinal scrotal cases. For one, the placement of a Foley catheter is important, because the manual downward retraction of the bladder employed in the open preperitoneal approach to create a larger working cavity is not feasible endoscopically. The Foley keeps the bladder collapsed. It also avoids potential injury to a bladder that may be contained within the hernia. Another modification suggested by experience is to limit the amount of intravenous fluid given by the anesthesiologist. Even though Nyhus [7] does not routinely divide the epigastric vessels, we feel that this is better done, for it allows access to the deep internal ring without risk of injury to those vessels. Division also allows release of the transversalis fascial sling, which permits dividing the floor toward the external ring. And finally, division of the epigastric vessels allows the mesh to lie relatively smoothly on the posterior aspect of the floor. Without it, a warped placement usually results. In most of our cases the inguinal indirect large sac was totally reduced. But we certainly agree with Nyhus that if extensive dissection is needed to free the sac, it should be divided with the proximal end closed and the scrotal end left in situ, minimizing injury to the cord structures. When the sac is being divided, escape of CO2 into the abdominal cavity is inevitable. The ensuring pneumoperitoneum may cause discomfort to a patient requiring epidural anesthesia. However, this in no way precludes successful completion of the case. It does not cause collapse of the working space, and the surgeon will be able to finish closure of the sac and placement of the mesh. It is always advisable to open the transversalis fascial sling and continue into the transversus abdominis. This allows for distal access to the sac and, since the sac is more tapered distally, easier encircling of it. We concur with Stoppas advice that a large piece of mesh should always be placed. We leave no drains.

Conclusion We feel that the technical modifications described here make the extraperitoneal approach to large inguinal scrotal hernias safer and easier and eliminate its unjustified reputation as being contraindicated in these cases.

Discussion The use of laparoscopy for the repair of primary unilateral inguinal hernia remains controversial. For a large inguinal, large recurrent, or large sliding hernia, several authors [1, 4, 11] recommend the open preperitoneal approach. But this is very much a matter of choice. As we gained experience with the endoscopic procedure, we found that we could always open the properitoneal space with blunt finger dissection instead of balloon devices or operative scopes. We no longer used Foley catheters, epidural anesthesia became routine, and dissection in the proper plane eliminated the need for cautery, hemoclips, or suction irrigation devices [2].

References
1. Condon RE (1972) Posterior repair of groin hernias. Ethicon 2. Ferzli GS, Kiel T (1995) Evolving techniques in endoscopic extraperitoneal herniorrhaphy. Surg Endosc 9: 928930 3. Ferzli GS, Massaad A, Albert P (1992) Extraperitoneal endoscopic hernia repair. J Laparoendosc Surg 2: 281286 4. Greenburg AG (1987) Revisiting the recurrent groin hernia. Am J Surg 154: 3540

302 5. Massaad A, Fiorillo M, Hallak A, Ferzli GS (1996) Endoscopic extraperitoneal herniorrhaphy in 316 patients. J Laparoendosc Surg 6: 1316 6. McKernan JB, Laws HL (1993) Laparoscopic repair of inguinal hernia using a totally extraperitoneal prosthetic approach. Surg Endosc 7: 2628 7. Nyhus LM (1994) The preperitoneal approach and iliopubic tract repair of inguinal hernias. In: Nyhus LM, Condon RE (eds) Hernia. JB Lipincott, Philadelphia, pp 154188 8. Nuhus LM, Condon RE, Harkins NH (1960) Clinical experience with preanesthesia hernia repair for all types of hernia of the groin. Am J Surg 100: 234240 9. Nyhus LM, Pollack R, Bombeck CT, Donohue PE (1988) The preperitoneal approach and prosthetic buttress repair for recurrent hernia. Ann Surg 208: 733737 10. Stoppa RE, Warlaumont CR (1989) The preperitoneal approach and prosthetic repair of groin hernia. In: Nyhus LM, Condon RE (eds) Hernia. JB Lippincott, Philadelphia, pp 199225 11. Wantz EW (1989) Giant prosthetic reinforcement of the visceral sac. Surg Gynecol Obstet 169: 408417

Surg Endosc (1997) 11: 253256

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Thoracoscopic partial pericardiectomy in the diagnosis and management of pericardial effusion

R. Robles,1 A. Pin ero,1 J. A. Luja n,1 J. A. Ferna ndez,1 J. A. Torralba,1 F. Acosta,2 M. Villegas,3 P. Parrilla1
1 2

Servicio de Cirug a general y del A. Digestivo, Virgen de la Arrixaca University Hospital, 30120, El Palmar, Murcia, Spain Servicio de Anestesiolog a y Reanimacio n, Virgen de la Arrixaca University Hospital, 30120, El Palmar, Murcia, Spain 3 Servicio de Cardiolog a, Virgen de la Arrixaca University Hospital, 30120, El Palmar, Murcia, Spain Received: 30 May 1996/Accepted: 27 August 1996

Abstract Background: An effort was made to present our experience with thoracoscopy in the diagnosis and management of pericardial effusions. Methods: Twenty-two partial pericardiectomies were performed with the thoracoscopic approach in patients with pericardial effusions, the etiology of which was uremic (n 7), neoplastic (n 8), idiopathic (n 5), septicemia (n 1), and postpericardiotomy (n 1). All cases had grade III-IV/IV radiological cardiomegaly and ultrasonographic confirmation of the effusion. We found hemodynamic compromise in 17 patients. The operation, requiring the insertion of three trocars, enabled us to remove a large part (approximately 6 10 cm) of the left anterolateral side of the pericardium and aspirate the effusion contents for diagnostic and therapeutic purposes. Results: In five cases we found coexisting pleural effusions. The pericardial effusion had a mean volume of 817 ml, which was serous in 11 cases, hematic in six, serohematic in four, and purulent in one. Cytology of the pericardial effusion was positive for neoplasia in four cases (one pulmonary neoplasia, two breast carcinomas, and one lymphoma). We observed conversion to grade I/IV cardiomegaly in 16 cases and a return to normality in the other six, with the absence of ultrasonographic effusion in all cases. There was no recurrence during the mean follow-up period of 20.5 months (range: 247). Conclusions: The thoracoscopic management of pericardial effusions is a simple and effective technique that allows us to create a large pericardial window that drains the effusion definitively, determines its etiology, and explores and treats coexisting pleural lesions, all without recurrences. Key words: Pericardial effusions Pericardial window Partial pericardiectomy Thoracoscopy Video-assisted thoracoscopic surgery

The creation of a pericardial window has been used to manage pericardial effusions by means of different approaches: subxiphoid [12], sternotomy [15], or left anterior thoracotomy [10, 11, 15]. Endoscopic surgery has witnessed a major development in the last 5 years and has been applied to numerous techniques performed via thoracic [4] or abdominal approach [18]. One such technique creates a pericardial window or thoracoscopic partial pericardiectomy to manage pericardial effusions that resist medical treatment or cause a hemodynamic compromise [13, 20]. The aim of this paper is to present the results obtained in a series of 22 patients managed with this technique, analyzing whether these results are an improvement on those reported in the literature with other approaches (subxiphoid, sternotomy, or left anterior thoracotomy).

Patients and surgical technique Between April 1992 and December 1995 we drained 22 pericardial effusions using partial pericardiectomy via the thoracoscopy approach (Table 1). Mean patient age was 52.4 years (range 1881). Thirteen patients were women and nine were men. The patients presented with pericardial effusions that were resistant to medical treatment (five patients) or had an important haemodynamic compromise (17 patients). Preoperative chest radiology revealed grade IIIIV/IV cardiomegaly in all cases. Preoperative ultrasound demonstrated massive pericardial effusion with a mean ejection fraction of 57.3% (range 2060%) and collapsed right cavities in 17 patients. The pathology responsible for the pericardial effusion was as follows: in seven cases uremic effusion, due to chronic renal insufficiency; in eight cases neoplasia (four due to breast cancer, two to lung cancer, one to epidermoid carcinoma of the esophagus, and one to a nonHodgkins lymphoma); idiopathic in five cases; sepsis in a patient with mediastinitis and pleural empyema;

Correspondence to: R. Robles Campos

254 Table 1. Characteristics of the seriesa No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Sex F M F F M F F M M M F F M M F F F M F F F M Age (years) 18 58 36 29 61 76 39 41 55 67 60 30 35 53 64 59 68 46 62 46 81 69 Etiology Uremic Lung cancer Uremic Uremic Uremic Idiopathic Breast cancer Uremic Postperic Idiopathic Lung cancer Uremic Sepsis Lymphoma Uremic Breast cancer Idiopathic Esoph. cancer Breast cancer Breast cancer Idiopathic Idiopathic Volume (ml) 600 700 700 800 600 500 400 500 1,100 600 800 600 500 1,300 1,500 500 250 1,525 1,400 1,700 1,100 400 Type Serous Serous Serous Serohem. Serous Serous Hemorr. Serous Hemorr. Serous Hemorr. Hemorr. Seropur. Hemorr. Serous Serous Serohem. Serohem. Serous Hemorr. Serohem. Hemorr. E. F. (%) 35 50 50 55 60 60 60 55 50 55 40 60 60 60 60 60 60 20 45 60 60 60 S. T. (min) 50 50 45 50 45 45 40 35 60 45 45 60 45 30 35 30 40 40 25 45 35 30 H. S. (days) 4 9 12 8 3 8 3 4 5 10 8 10 15 10 6 3 2 2 5 2 2 2 Follow-up (months) 25 35 40 24 47 25 26 24 23 21 20 18 18 25 16 14 12 12 12 10 2 2 Exitus No Yes No No No No No No Yes No Yes No Yes Yes No No No No No No No No

a M: male; F: female; E. F.: ejection fraction; Hemorr.: hemorragic; Serohem.: serohematic; Seropur.: seropurulent; S. T.: surgical time; H. S.: hospital stay; Postperic.: postpericardiotomy; Esoph. cancer: Esophagus cancer

and one case of postpericardiotomy effusion due to previous valvular surgery. Surgery was done under general anesthesia and orotracheal intubation. The patient was placed in the decubitus position on his/her right side so that the thoracoscopic approach could be made via the left hemithorax. The surgeon and assistant stood at the patients back with the video monitor opposite on the other side of the patient. Three trocars were inserted: one 10-mm trocar in the fifth left intercostal space (IS) along the posterior axillary line, for the straight telescope, and two 5-mm trocars in the fourth and seventh left IS along the anterior axillary line, for the forceps, hook, scissors, and aspiration. If pulmonary distention hampered vision, this could be avoided by increasing the respiratory frequency and decreasing the inspiratory volume of the ventilator. If this were not sufficient, we could generate a pneumothorax with CO2 at a positive pressure of 6 mmHg through one of the 5-mm trocars. After the identification of the phrenic nerve an orifice was created through which most of the pericardial effusion was drained; after this, a large pericardial window of some 6 10 cm was opened. In all cases we left an aspirative pleural drainage for 48 h. The following determinations were made of the fluid obtained: cell count and cytology, culture and stains (Gram, Ziehl, and Papanicolau), biochemical analysis (Na+, K+, Cl, Glu, BUN), detection of nonorganospecific antibodies (ANOEs), and in the last three patients, Adenosine Deaminase (ADA) analysis. The extirpated pericardium was divided into two fragments: one for culture and the other for anatomopathological study. The results were assessed by chest X-ray and ultrasound, which were performed in the immediate postoperative period and periodically during the follow-up. Mean follow-up was 20.5 months (range 247).

Results Operative findings Exploration of the pleural space enabled us to find loculations and easily removable adhesions in two patients, pleural effusions in five patients, and pleural metastasis in two patients. The mean volume of the pericardial effusion aspirated was 817 ml (range: 2501,700). It was serous in 12 patients, hematic in six, serohematic in three, and seropurulent in the mediastinitis patient. Mean surgery time was 42 min (range: 2560). Diagnostic efficiency The cultures of the pericardial effusion, pleural fluid and hemocultures in the mediastinitis patient were positive for Streptococcus sp. The remaining cultures of pleural and pericardial fluid were negative. Cytology of the pericardial effusion was positive for neoplastic cells in four patients (two breast carcinomas, one lung carcinoma, and one patient with lymphoma), and neoplastic involvement of the pericardium was confirmed in the histological study in two of them. The cytological study in an effusion initially considered idiopathic demonstrated a cellularity compatible with adenocarcinoma (lung cancer). Pleural cytology was negative in the five cases in which associated pleural effusion was found. The histological study of the pericardium, except in the two above-mentioned patients, demonstrated a fibrosis with nonspecific chronic inflammation and mesothelial hyperplasia. Two patients have histological demonstration of pleural metastasis. Morbidity and mortality Only one patient had immediate postoperative complications with the development of a pleural effusion, which

255

required thoracocentesis. There was no intraoperative mortality. Two patients died in the immediate postoperative period: one on day 5 for bilateral pneumonitis due to cytomegalovirus (patient suffering from familial Mediterranean fever) and one on day 15 following sepsis (mediastinitis patient). Hospital stay averaged 6 days (range: 215).

Follow-up In the immediate postoperative period there was radiological conversion to grade I cardiomegaly in 16 patients and a return to normality in the remaining six. In the 20 surviving patients, periodical follow-up with ultrasound assessment showed the absence of recurring effusion. Three patients were lost to the long-term follow-up, all due to the development of an underlying neoplastic disease (at 3 months in a patient with lung cancer, at 10 months in the lymphoma patient, and at 1 year in another lung cancer patient).

Discussion The therapeutic options in the management and treatment of patients with pericardial effusions are very varied [3, 14]: corticoids, external beam irradiation, pericardiocentesis alone or with instillation of chemotherapeutic agents, and surgery. There are three fundamental surgical possibilities: complete pericardiectomy, pericardial window (both performed via the thoracotomic or thoracoscopic approach), and subxiphoid pericardiotomy. Any of these techniques must fulfill the three fundamental objective of any therapy for effusion: resolve it, determine its aetiology, and avoid recurrences, all with the least possible morbidity and mortality [4]. Resolution of the effusion is quick with any of the three techniques. Subxiphoid pericardiectomy requires a mean operating time of some 3045 min [3], similar to that recorded in our thoracoscopic experience. Thoracotomy requires more time than the other two methods. The diagnostic capacity of any of these techniques will depend on the possibility of thoroughly exploring the pericardial cavity and pleural space and on the possibility of obtaining sufficient study material (biopsies and effusion samples). Thoracoscopy has a better diagnostic capacity in these aspects than the other techniques [6], as it enables us to explore the pleural cavity thoroughly and visualize and obtain samples of suspicious lesions and associated effusions (effusions were found in five of our patients and tumoral implants in two). Furthermore, it provides a wide view of the pericardial surface, which gives the most suitable place for biopsies, and the pericardial cavity can be explored easily for septa, localized effusions, or tumoral implants. This diagnostic capacity is important from a therapeutic and prognostic point of view, since many of the patients are neoplastic, but their effusions are not. In our series, of eight neoplastic cases only four had positive cytology for the primary neoplasm, and in one of them it diagnosed the underlying condition (lung cancer). The thoracotomic approach also evaluates the pleural cavity and pericardial surface [10], but less so than thoracoscopy [7]. The subxiphoid approach is limited in its exploratory capacity, because it enables samples of the pericardium and

fluid to be obtained, it is not guided and moreover does not enable the whole of the pericardial surface or pleural space to be visualized. After a follow-up period of 20.5 months we had no recurrence. These data contrast with those reported for the subxiphoid procedures with recurrence rates ranging from 2.5% to 18% (reoperation rates ranging from 0% to 9%) [1, 2, 9, 15, 19, 20]. Thoracotomy presents very low rates of recurrence. According to some authors [4, 15, 19] the recurrence rate is related to the diameter of the window created, which means that the greater the pericardial exposure and resection the lower the rate of recurrences, even if the efficiency mechanism of this procedure is highly debated. Morbidity and mortality with the thoracoscopic approach are minimal; only one of our patients had a pleural effusion requiring thoracentesis. There was no intraoperative or postoperative death related to the technique. Respiratory complications did not exist, and postoperative pain was minimal. The thoracotomic approaches have a greater morbidity rate [9, 14], mainly as a result of respiratory complications. The subxiphoid approach had fewer complications than thoracotomy, although more than thoracoscopy, with morbidity figures of 5 to 20% [3, 9, 14, 17]. One of the main inconveniences attributed to thoracoscopy is that it requires general anesthesia, whereas the subxiphoid approach can be performed with local anesthesia [11], this being one of the main reasons for its more widespread use. However, for greater patient comfort and improvement management in a more stable surgical field, many authors carry out this procedure with general anesthesia and endotracheal intubation [5, 8, 13, 16]. The type of intubationselective or endotrachealis another controversial point. Selective intubation with ventilation of a lung facilitates visualization of the pericardium and reveals possible hemodynamic effects. Even then, since the beginning of our experience and as has been the case for other authors [8], all the procedures have been carried out with general anesthesia and endotracheal intubation. To improve visualization we decrease pulmonary distension by lowering the inspiratory volume and raising the respiratory frequency. When visualization is not correct we perform a low-pressure pneumothorax (6 mmHg) and have had no problem with this technique. Conversely, in the only patient with selective intubation the degree of hypoxemia required a conversion to endotracheal intubation; furthermore, it was noted that contralateral hyperinsufflation and the consequent mediastinal shift made surgical management of the effusion very difficult. In conclusion, we recommend the thoracoscopic approach in the management of pericardial effusions, rather than the subxiphoid or thoracotomic approaches, because of its good diagnostic and therapeutic capacity, the low morbidity and mortality rates, and the absence of long-term recurrences.

References
1. Campbell PT, Van Trigt P, Wall TC, Kenney RT, OConnor CM, Sheikh KH, Kisslo JA, Baker ME, Corey GR (1992) Subxiphoid pericardiotomy in the diagnosis and management of large pericardial effusions associated with malignancies. Chest 101: 938943 2. Gregory JR, McMutrey MJ, Mountain CF (1985) A surgical approach

256 to the treatment of pericardial effusion in cancer patients. Am J Clin Oncol 8: 319 Hawkins JW, Vacek JL (1989) What constitutes definitive therapy of malignant pericardial effusion? Medical vs surgical treatment. Am Heart J 118: 428432 Hazelrigg SR, Mack MJ, Landreneau RJ, Acuff TE, Seifert PE, Auer JE (1993) Thoracoscopic pericardiectomy for effusive pericardial disease. Ann Thorac Surg 56: 792795 Little AG, Kremser PC, Wade JL, Levett JM, DeMeester TR, Skinner DB (1984) Operation for diagnosis and treatment of pericardial effusions. Surgery 96: 738744 Liu HP, Chang CH, Lin PJ, Hsieh HC, Chang JP, Hsieh MJ (1994) Thoracoscopic management of effusive pericardial disease: indications and technique. Ann Thorac Surg 58: 16951697 Mack MJ, Landreneau RJ, Hazelrigg SR, Acuff TE (1993) Video thoracoscopic management of benign and malignant pericardial effusions. Chest 103: 390S393S Moores DWO, Allen KB, Faber LP, Dziuban SW, Gillman DJ, Waren WH, Ilves R, Lininger L (1995) Subxiphoid pericardial drainage for pericardial tamponade. J Thorac Cardiovasc Surg 109: 546552 Naunheim KS, Kessler KA, Fione AC, Turrentine M, Hammel LM, Brown JW, Mohamed Y, Pennington DG (1991) Pericardial drainage: subxiphoid vs transthoracic approach. Eur J Cardithorac Surg 5: 99 104 Olsen PS, Sorensen C, Andersen HO (1991) Surgical treatment of large pericardial effusions. Etiology and long-term survival. Eur J Cardithorac Surg 5: 430432 Okamoto H, Shinkai T, Yamakido M, Saijo N (1993) Cardiac tamponade caused by primary lung cancer and the management of pericardial effusion. Cancer 71: 9398 Ozuner G, Davidson PG, Isenberg JS, McGuinn JT (1992) Creation of a pericardial window using thoracoscopic techniques. Surg Gynecol Obstet 175: 6971 Palatianos GM, Thurer RJ, Pompeo MQ, Kaiser GA (1989) Clinical experience with subxiphoid drainage of pericardial effusion. Ann Thorac Surg 48: 381385 Park SJ, Retschler R, Wilbur D (1990) Surgical management of pericardial effusion in patients with malignancies. Comparison of subxiphoid window vs pericardiectomy. Cancer 67: 7680 Piehler JM, Pluth JR, Schaff HV, Danielson GK, Orszulak TA, Puga FJ (1985) Surgical management of effusive pericardial disease. J Thorac Cardiovasc Surg 90: 506516 Prager RL, Wilson CH, Bender HW (1982) The subxiphoid approach to pericardial disease. Ann Thorac Surg 34: 69 Press OW, Livingston R (1987) Management of malignant pericardial effusion and tamponade. JAMA 257: 10891092 Sastic JW, Stalter KD, Goddard RL (1992) Laparoscopic pericardial window. J Laparoendosc Surg 2: 263266 Sugimoto JT, Little AG, Ferguson MK, Borow KM, Vallera D, Staszak VM, Weinert L (1990) Pericardial window: mechanism of efficacy. Ann Thorac Surg 50: 442445 Van Trigt P, Douglas J, Smith PK, Campbell PT, Wall TC, Kenney RT (1993) A prospective trial of subxiphoid pericardiotomy in the diagnosis and treatment of large pericardial effusion. A follow-up report. Ann Surg 218: 777782

3. 4. 5. 6. 7. 8. 9.

12.

13.

14.

15.

16. 17. 18. 19.

10. 11.

20.

News and notices

Surg Endosc (1997) 11: 315317

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

New Address for the European Association for Endoscopic Surgery (E.A.E.S.)
Effective January 1, 1997, the new correspondence, telephone, and fax numbers of the E.A.E.S. office are: E.A.E.S. Office, c/o Mrs. Ria Palmen Luchthavenweg 81 Unit 1.42 5657 EA Eindhoven The Netherlands or: P.O. Box 335 5500 AH Veldhoven The Netherlands Tel: +31 40 2525288 Fax: +31 40 2523102

For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

Advanced Endoscopic Skills Surgical Skills Unit University of Dundee Scotland, UK

Each month Professor Cuschieri Surgical Skills Unit offers a 412 day course in Advanced Endoscopic Skills. The course is intensely practical with hands on experience on a range of simulated models. The program is designed for experienced endoscopic surgeons and covers advanced dissection techniques, extracorporeal knotting techniques, needle control, suturing, internal tying technique, stapling, and anastomotic technique. Individual workstations and a maximum course number of 10 participants allows for personal tuition. The unit offers an extensive collection of surgical videos and the latest books and publications on endoscopic surgery. In addition, participating surgeons will have the opportunity to see live advanced laparoscopic and/or thoracoscopic procedures conducted by Professor Cuschieri and his team. The course is endorsed by SAGES. Course fee including lunch and course materials is $1850. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

Fellowship in Minimally Invasive Surgery George Washington Medical Center Washington, DC USA
A one-year fellowship is being offered at the George Washington University Medical Center. Interested candidates will be exposed to a broad range of endosurgical Education and Research Center. Active participation in clinical and basic science research projects is also encouraged. For further information, please contact: Carole Smith 202-994-8425 or, send curriculum vitae to: Dr. Jonathan M. Sackier Director of Endosurgical Education and Research George Washington University Medical Center Department of Surgery 2150 Pennsylvania Avenue, N.W. 6B-417 Washington, DC 20037, USA

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

Under the direction of Professor 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 purpose-built 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 Professor Cuschieri, and the latest books on endoscopic surgery are on display in our Resource area. Course fee including lunch and course materials is $860.

The Practical Aspects of Laparoscopic Fundoplication Surgical Skills Unit University of Dundee Scotland, UK
A three-day course, led by Professor Cuschieri, designed for experienced laparoscopists wishing to include fundoplication in their practice. The course covers the technical details of total and partial fundoplication using small group format and personal tuition on detailed simulated models. There will be an opportunity to observe one of these procedures live during the course. Maximum course number is six. Course fee including lunch is $1850. For further details and a brochure please contact: Julie Struthers, Unit Co-ordinator Surgical Skills Unit Ninewells Hospital and Medical School Dundee DD1 9SY Tel: +44 382 645857 Fax: +44 382 646042

316

Courses at George Washington University Endosurgical Educational and Research Center

George Washington University Endosurgical Educational and Research Center is proud to offer a wide range of surgical endoscopy courses. These courses include advanced laparoscopic skills such as Nissen fundoplication, colon resection, common bile duct exploration, suturing, as well as subspecialty courses. Individual surgeons needs can be met with private tuition. The Washington D.C. area is a marvelous destination to visit for recreational pursuits which can be arranged by the facility to suit your personal agenda. For further details please contact: Carole Smith: Department of Surgery 2150 Pennsylvania Avenue NW 6B Washington, DC 20037, USA Tel: (202) 994-8425

nical nuances, and troubleshooting; visual perception problems and solutions; magnified eye-hand coordination; and two-handed (ambidextrous) technique. 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). 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

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 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

IIIrd European Workshop on Digestive Surgery March 1314, 1997 Brussels, Belgium
The IIIrd European Workshop on Digestive Surgery is focused on live operative demonstrations performed and narrated by European experts. Panel discussions, quizzes, and selected communications will take place during the sessions. The aims are to confront alternative procedures and to provide young surgeons with an overview of selected fields. Topics will be: functional anorectal surgery, proctology, colorectal surgery, and updates in laparoscopic surgery. Course direction: J. J. Houben, MD. For further information, please contact: Administrative Secretariat Conference Services s.a. Avenue de lObservatoire 3, bte 17 B-1180 Brussels, Belgium Tel: +32 2 375 16 48 Fax: +32 2 375 32 99

Medicine and the Law for Junior Hospital Doctors April 11, 1997 Middlesbrough, UK
For further information, please contact: Miss Welsh ENTER North Riding Infirmary Newport Road Middlesbrough TS1 5JE UK

Advanced Laparoscopic Suturing and Surgical Skills Courses MOET Institute San Francisco, CA, USA
This intensive hands-on training program is intended to help the surgeon develop proficiency in the essential laparoscopic surgical techniques. A sequence of progressively challenging exercises has been designed to enable the surgeon to improve his or her laparoscopic dexterity, efficiency, and creativity. Exact and meticulous technique is emphasized so that the surgeon can apply these skills with confidence. Personal instruction is provided by Zoltan Szabo, Ph.D., F.I.C.S., Director of the MOET Institute, and surgeons are allowed to progress their own pace. Each participant has sole use of a laparoscopic training station equipped with high-quality clinical laparoscopic equipment and instrumentation. Inanimate, animal tissue, and optional live animal models are utilized. Features of these program include: fluently choreographed instrument movements; economy of movement and flawless technique; needle and suture handling skills (2-0 to 7-0); precision suturing, knotting, ligature, and anastomosis techniques; atraumatic, hemostatic tissue handling and dissection; optimal angles of approach (coaxial alignment of setup and geometry of port positioning); laparoscopic surgical strategy, tech-

First International Baltic Conference of Videosurgery of the Viscerosynthesis Section of the Association of Polish Surgeons April 2427, 1997 Gdansk, Poland
For further information, please contact: Organizing Secretariat Second Department of Surgery Medical University of Gdansk 1 Prof. Z Kieturakisa Street 80-742 Gdansk, Poland Tel/Fax: (0 048 58) 31 87 75

317

4th International Meeting on Laparoscopic Surgery May 17, 1997 Berne, Switzerland
Main topic: Acute appendicitis: Standard treatment and the role of laparoscopic surgery For further information, please contact: Mrs. Caroline Zrcher Klinik fr Viszerale und Transplantationschirurgie Universitt Bern Inselspital CH-3010 Bern, Switzerland Tel: +41 31 632 97 22 Fax: +41 31 632 97 23

SETUR Congress Department Cumhuriyet Cad. No. 107 80230 Elmadag Istanbul, Turkey Tel: (90.212) 23003 36 Fax: (90.212) 240 82 37

9th International Meeting Society for Minimally Invasive Therapy July 1416, 1997 Kyoto, Japan
Scientific program to include: Plenary, Parallel, Poster, and Video sessions. Host Chairman: Professor Osamu Yoshida, Department of Urology, Kyoto University, 54 Shogoin Kawahara-sho, Sakyo, Kyoto 606, Japan. Phone: +81 75 751-3328, Fax: +81 75 751-3740. This meeting coincides with the Gion Festival in Kyoto, one of the greatest festivals in Japan. For further information, please contact: Secretariat of SMIT 9th Annual International Meeting c/o Academic Conference Planning 383 Murakami-cho Fushimika, Kyoto 612 Japan Tel: +81 75 611-2008 Fax: +81 75 603-3816

European Course on Laparoscopic Surgery (French language) May 1316, 1997 (English language) November 1821, 1997 Brussels, Belgium
Course director: G.B. Cadiere For further information, please contact: Administrative Secretariat Conference Services s.a. Avenue de lObservatoire, 3 bte 17 B-1180 Bruxelles Tel: (32 2) 375 16 48 Fax: (32 2) 375 32 99

6th World Congress of Endoscopy Surgery Roma 98 6th International Congress of European Association for Endoscopic Surgery June 36, 1998 Rome, Italy
The program will include: the latest, original high quality research; symposia; plenary lectures; abstract presentations (video, oral, and posters); EAES and SAGES postgraduate courses, OMED postgraduate course on therapeutic endoscopy; working team reports; educational center and learning corner; meeting of the International Society of Nurses and Associates; original and non original scientific reports; and a world expo of new technology in surgery. For further information, please contact: Congress Secretariat: Studio EGA Viale Tiziano, 19 00196 Rome, Italy Tel: +39 6 322-1806 Fax: +39 6 324-0143

Joint Euro Asian Congress of Endoscopic Surgery 5th Annual Congress of the European Association for Endoscopic surgery (EAES) 3rd Asian-Pacific Congress of Endoscopic Surgery June 1721, 1997 Instanbul, Turkey
The Congress will include a joint postgraduate course EAES/SAGES/ ELSA on June 17th. For information and registration:

Surg Endosc (1997) 11: 264267

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic surgery for diverticulitis

M. E. Sher, F. Agachan, M. Bortul, J. J. Nogueras, E. G. Weiss, S. D. Wexner
Department of Colorectal Surgery, Cleveland Clinic Florida, 3000 W. Cypress Creek Road, Fort Lauderdale, FL 33309, USA Received: 25 March 1996/Accepted: 17 July 1996

Abstract Background: Resection of diverticular disease may be quite challenging; the acute inflammatory process, thick sigmoid mesentery, and any associated fistula or abscess can make this procedure technically demanding. The aim of this study was to compare the results between laparoscopic and laparotomy-type resections stratified by disease severity and thereby predict outcome and possibly a subset of patients who may benefit from a laparoscopic approach. Methods: From August 1991 to December 1995, all patients with diverticular disease were classified according to a modified Hinchey classification system. The laparoscopic group included 18 patients who underwent a laparoscopic assisted colectomy, one with a loop ileostomy. The identical procedures were performed in 18 patients by laparotomy. The mean age of the two groups were 62.8 and 67.1 years, respectively (p NS). Results: Seven of 18 patients in whom laparoscopy was attempted (38.9%) had conversion to laparotomy. Six of seven (85.7%) conversions were directly related to the intense inflammatory process. Laparoscopic treated patients with Hinchey IIa or IIb disease had a morbidity rate of 33.3% and a conversion rate of 50% while all patients with Hinchey I disease were successfully completed without morbidity or conversions to laparotomy. However, after the first four cases, the intraoperative morbidity and postoperative morbidity rates were zero and 14.3% and after ten cases they were zero and zero, respectively. Furthermore, the median length of hospitalization for Hinchey I patients after laparoscopy was 5.0 days vs 7 days after laparotomy (p < 0.05). In Hinchey IIa and IIb patients, the median length of hospitalization was almost 50% shorter with a laparoscopic approach (6 days vs 10 days, p < 0.05). Conclusion: In conclusion, laparoscopic resection of diverticulitis can be performed without additional morbidity particularly in Hinchey I patients and with a reduced length of

hospitalization in patients with class I or II disease. Patients with class I disease, and after initial experience even those with class II disease, can benefit from the reduced morbidity and length of hospitalization associated with laparoscopic treatment. Key words: Laparoscopy Diverticular disease Hinchey system

The surgical approach for complicated diverticulitis has undergone significant changes. Traditionally, a three-stage approach was the accepted treatment for acute diverticulitis. However, morbidity and mortality were prohibitively high and sepsis persisted when the diseased segment was left in situ. In the 1980s, it became clear that the perforated segment should be resected whenever possible and the Hartmanns procedure was popularized [1, 10]. After realizing the difficulty of Hartmann reversals, surgeons began to perform resection with primary anastomosis with or without loop ileostomies in select cases. Whether diverticulitis is resected through a laparotomy or a laparoscope, such surgery can be challenging due to the acute inflammatory process and associated fistulae or abscess. We assessed the results of laparoscopic surgery for diverticulitis as stratified by severity of disease. We then sought to compare the results to well-matched patients who had undergone the same procedure for the same indications by laparotomy. Ultimately, we aimed to identify a subset of patients with diverticulitis who may benefit from a laparoscopic approach.

Materials and methods

Presented at the annual meeting of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA 1317 March 1996 Correspondence to: S. D. Wexner We reviewed all colorectal laparoscopic cases performed from August 1991 to December 1995. Patients with diverticulitis were classified according to a modified Hinchey grading system (Table 1) [5]. Medical records from a control group of patients matched for age, procedure, and Hinchey class were reviewed during the same time interval (January 1993

265 Table 1. Modified Hinchey et al. grading system [5] I. IIA. IIB. III. IV. Pericolic abscess Distant abscess amenable to percutaneous drainage Complex abscess associated with fistula Generalized purulent peritonitis Fecal peritonitis

December 1995). All laparoscopic cases were performed by a single surgeon, whereas laparotomies were performed by one of two surgeons, neither of whom routinely performed laparoscopic surgery. All patients with Hinchey I or II disease were offered a laparoscopic approach when cared for by the one surgeon who routinely performed laparoscopy. Age, gender, diagnosis, procedure, operative time, hospital stay, morbidity, and mortality were evaluated for comparison. Comparisons were made among Hinchey classes, between converted vs laparoscopic completed procedures, and between open vs laparoscopic procedures. Patients with Hinchey III or IV disease were not offered a laparoscopic approach and thus were excluded from the study. Laparoscopic procedures were performed in the standard manner as previously described [13]. Mann-Whitney, Kruskal-Wallis nonparametric ANOVA and Fisher exact tests were used for statistical analysis; p < 0.05 was considered significant (Instat Graphpad, San Diego, CA).

Fig. 1. Results: open vs laparoscopy, hospital stay.

Results One hundred eighty-five laparoscopic colorectal procedures were reviewed. Eighteen patients (9.7%) were of a mean age of 62.8 (range 3486) years; eight males and 10 females were operated on for diverticulitis and make up the study group. There were six patients with Hinchey I, seven patients with Hinchey IIa, and five patients with Hinchey IIb disease. There were four patients with fistula in the Hinchey IIb group, one colofallopian, one coloenteric, and two colovesical. Eighteen patients underwent a laparoscopic assisted sigmoid colectomy, one with Hinchey IIb disease who had a concomitant loop ileostomy. Eighteen matched patients with a mean age of 67.1 (range 3377) years, seven males and 11 females, underwent elective laparotomy for diverticular disease. Eighteen patients underwent a sigmoid resection, one of whom had a loop ileostomy. The patients were matched to the laparoscopic group, specifically by age, Hinchey class, procedure, and presence of a fistula. There were four patients with a colovesical fistula in the Hinchey IIb group who were treated in the conventional manner (sigmoid resection, takedown, and repair of the fistula with Foley catheter drainage for 5 days). Converted group vs laparoscopically completed group In seven of the 18 patients (38.9%), a laparotomy was necessary. Six of the seven (85.7%) conversions were directly related to the intense inflammatory process. Two of the seven were due to intraoperative complications (one enterotomy, and one colotomy), both of which occurred during the first four cases. The average operative time for the converted group was 214 min vs 213 min for the laparoscopically completed group (p NS). Furthermore, the median length of hospitalization for the converted group was 8 days vs 5 days for the laparoscopically completed group (p < 0.01) (Fig. 1). The converted group had a postoperative morbidity of 28.6% (2/7 patients) while two of seven patients (28.6%) had intraoperative complications (one ente-

rotomy and one colotomy both requiring conversion) yielding an overall morbidity of 57.1%. Both the anastomotic leak and the enterotomy occurred in the same patient, who was the fourth patient operated on for diverticulitis. The colotomy occurred in the first patient operated upon for diverticulitis; that patient also developed a wound infection. Conversely, there was only one postoperative pneumonia in 11 patients (9.17%) of the laparoscopically completed group (p < 0.05). Laparoscopic group: Hinchey I vs Hinchey IIa/IIb While the patients with Hinchey I disease had no morbidity and no conversions, Hinchey IIa and IIb groups combined had morbidity and conversion rates of 33.3% (p NS) and 50% (p < 0.5), respectively. In addition, the median hospital stay for patients with Hinchey I diverticulitis was 5 days vs 6 days for Hinchey IIa and IIb disease (p NS). Interestingly, there was no difference in the operative time in patients with Hinchey I (215 min) vs Hinchey IIa or IIb pathology (213 min) (p NS). Open vs laparoscopic in Hinchey I patients The laparoscopic operative time for Hinchey I patients was significantly longer (215 min) than it was for patients who underwent a laparotomy (108.3 min) (p < 0.005). Conversely, the median hospital stay was significantly longer in the open group, 7 days vs 5 days (p < 0.05) (Fig. 1), and no mortality or morbidity was noted in either group. Open vs laparoscopy in Hinchey IIa or IIb patients The operative time was much more closely aligned in more advanced disease. Specifically, the operative time was 213 min for the laparoscopic group vs 167 min for laparotomy (p NS). Furthermore, the median length of stay for the laparoscopic group was 6 vs 10 days after laparotomy (p < 0.05) and morbidity was not significantly higher for the laparoscopic group (33.3%) than for the open group (33.3%) being treated for complicated diverticular disease. However, procedure-related morbidity was three for the laparoscopic group and zero for the open group. This consisted of one anastomotic leak, one colotomy, and one enterotomy. The colotomy and enterotomy occurred in the first and

266 Table 2. Laparoscopic colectomy for diverticulitis Total no. patients (all etiologies) 51 65 80 66 185 No. patients with diverticulitis only 13 (25.4%) 10 (15.4%) 26 (32.5%) 19 (28.8%) 18 (9.7%) Overall conversion rate 7.8% 3% 22.5% 41% 23% Conversion rate for cancer 5% 0% 10% 36% 4% Conversion rate for diverticulitis 15.3% 10% 38% 53% 38.9%

Author Phillips et al. [9] Zucker et al. [14] Hoffman et al. [6] Falk et al. [4] Sher and Wexner [12]

fourth cases performed for diverticulitis at this institution. No such injuries have occurred during the last 3 years. Thus, after the learning curve the intraoperative morbidity was zero and the postoperative morbidity 14.3%. There was no postoperative morbidity during the last 10 cases.

Discussion All surgeons know that resection of diverticulitis can be quite challenging. Therefore, it is intuitive that the acute inflammatory process also renders laparoscopy technically demanding, with higher conversion rates (Table 2) [4, 6, 9, 12, 14]. The fact that there were no statistical differences between the operative times for laparotomy and laparoscopy for Hinchey IIa and IIb patients supports the notion that even resection by laparotomy can be difficult. In a multicenter retrospective study, Falk et al. [4] reported a 53% conversion rate for all laparoscopic assisted sigmoid colectomies. Most reports of laparoscopic surgery for diverticular disease combine acute and chronic cases, elective and emergency cases, and cases associated with complications of diverticulitis. In addition, most authors combine the results of both benign and malignant diseases. In the multicenter trial it was impossible to determine outcome data specific for diverticulitis [4]. Phillips et al. [9] reported their experience with 51 colectomies, including 13 for diverticular disease. However, only four patients were treated for complicated diverticulitis. They, too, noted that inflammatory lesions with an indurated mesentery increased the likelihood of conversion and recommended preoperative computerized tomography scans to help select patients. They converted two of 13 patients with acute diverticulitis but only two of 38 for noninflammatory conditions. Hoffman and associates [6] reported 26 colectomies for diverticulitis. There were 10 conversions to open procedures (38.5%)two for recurrent diverticulitis, four with fistula, one with an abscess, and three technical, not related to the inflammatory process. Complications were more prevalent in the converted cases. The overall conversion rate for all 80 patients who had laparoscopic surgery was 22.5%, while 70% of conversions were due to the inflammatory process. Similarly, 38.9% in the current study required conversion to laparotomy. Like other series 85.7% of the conversions were directly related to the inflammatory process including two intraoperative complications. Both enterotomies were related to scarring and inflammation; both were intraoperatively recognized and repaired after converting to an open procedure. Both injuries led to major postoperative

morbidityone anastomotic leak and one wound infection. However, all intraoperative complications and all postoperative septic complications occurred during the first four cases. Importantly, the average total operative time in the converted group was not different from the completed group. Despite rapid conversion if no progress was made, the hospitalization for the converted group was significantly longer than it was for the completed group, primarily due to more advanced disease. The higher morbidity in the converted group parallels previous findings [2, 4, 6, 8, 9, 12, 14]. Hinchey class at operation reflects overall outcome. Auguste et al. [1] reported zero mortality with stage I disease treated by primary resection vs 5% mortality for stage II and 18% for stage III. We categorized patients according to a modified Hinchey grading system. Patients with complex abscesses not amenable to computerized-tomographyguided drainage and patients with a colovesical fistula or other fistulae were included in class IIb. Patients with a pelvic abscess or phlegmon amenable to drainage were classified as Hinchey class IIa and after computerizedtomography-guided drainage were reclassified as class I. As had been previously mentioned, staging by computerized tomography scan may allow selection of patients most likely to benefit from laparoscopy and/or percutaneous abscess drainage [8]. There was a shorter hospitalization in the laparoscopic procedures compared to laparotomies in all Hinchey groups. All of our patients, open and laparoscopic, were fed clear liquids immediately postoperative and progressed to a regular diet as tolerated [3, 10]. None of the patients in the laparoscopic group required a nasogastric tube, while seven of 18 (38.8%) patients treated by laparotomy required a nasogastric tube. This was the single most important factor leading to prolongation of hospital stay. Unlike other series, all of our patients were discharged only after tolerating a regular diet for at least 24 h. Thus the postoperative ileus appears less with a laparoscopic approach. Liberman and Hoffman also noted a decreased time to return to bowel function and discharged patients on 6.2 and 5.2 days, respectively [6, 8]. There was no morbidity in either the laparoscopic or laparotomy groups for uncomplicated Hinchey I patients. Moreover, there was no statistically significant difference between complication rates for laparotomy or laparoscopy for Hinchey IIa or IIb patients. Furthermore, and of crucial importance, is the fact that after the first four cases, there was no intraoperative morbidity and only a 14.3% postoperative morbidity, and that postoperative morbidity has decreased to zero during our last 10 cases.

267

Conclusion Laparoscopic resection of diverticulitis can be performed without additional morbidity in Hinchey I patients and with a reduced length of hospitalization in all patients with diverticulitis. However, if conversion to laparotomy is necessary due to intraoperative complications, such benefits are less likely to be seen. Such problems are more frequently encountered early during ones experience. Morbidity rates of zero to 14.3% can be expected even in these complicated cases, as more procedures are performed. Therefore, patients with class I disease and, after initial experience, even those with class II disease can benefit from the reduced morbidity and length of hospitalization associated with laparoscopic treatment.
Acknowledgment: This study was supported in part by funding from the Eleanor Naylor Dana charitable trust and the David G. Jagelman, M.D. memorial research and education fund.

References
1. Auguste L, Borrero E, Wise L (1985) Surgical management of perforated colonic diverticulitis. Arch Surg 120: 450452 2. Beart RW Jr (1994) Laparoscopic colectomy: status of the art. Dis Colon Rectum 37(suppl): S47S49 3. Binderow SR, Wexner SD. Current surgical therapy for mucosal ulcerative colitis. Dis Colon Rectum. (1994) 37:610624

4. Falk PM, Beart RW Jr, Wexner SD, Thorson AG, Jagelman DG, Lavery IC, Johanson OB, Fitzgibbons RJ Jr (1993) Laparoscopic colectomy: a critical appraisal. Dis Colon Rectum 36: 2834 5. Hinchey E, Schaal PG, Richards GK (1978) Treatment of perforated diverticular disease of the colon. Adv Surg 12: 85109 6. Hoffman GC, Baker JW, Fitchett CW, Vansant JH (1994) Laparoscopic assisted colectomy. Initial experience. Ann Surg 219: 732743 7. Jager R, Wexner SD (eds) (1996) Laparoscopic Colorectal Surgery. Churchill-Livingstone, New York 8. Liberman MA, Phillips EH, Carroll BJ, Fallas M, Rosenthal R (1996) Laparoscopic colectomy versus traditional colectomy for diverticulitis. Surg Endosc 10: 1518 9. Phillips EH, Franklin M, Carroll BT, Fallas M, Ramos R, Rosenthal D (1992) Laparoscopic colectomy. Ann Surg 216: 703707 10. Reissman P, Teoh TA, Cohen SM, Weiss EG, Nogueras JJ, Wexner SD (1995) Is early oral feeding safe after elective colorectal surgery? A prospective randomized trial. Ann Surg 222: 7377 11. Roberts P, Abel M, Rosen L, Cirocco W, Fleshman J, Leff E, Levien D, Pritchard T, Wexner SD, Hicks T, Kennedy H, Oliver G, Reznick R, Robertson H, Robertson W, Ross T, Rothenberger D, Senatore P, Surrell J, Wong D. (1995) The standards task force, American Society of Colon and Rectal Surgeons. Practice parameters for sigmoid diverticulitis: supporting documentation (1995). Dis Colon Rectum 38: 126132 12. Sher ME, Wexner SD (1996) Laparoscopic surgery for benign colorectal disease. In: Cosgrove J, Longo W, Andrus C (eds) Minimally invasive surgery: principles and outcomes. Harwood Academic, New York (in press) 13. Wexner SD, Johansen OB (1992) Laparoscopic bowel resection: advantages and limitations. Ann Med 24: 105110 14. Zucker KA, Pitcher DE, Martin DT, Ford RS (1994) Laparoscopic assisted colon resection. Surg Endosc 8: 1218

Surg Endosc (1997) 11: 272276

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Cost analysis of diagnostic laparoscopy vs laparotomy in the evaluation of penetrating abdominal trauma
J. M. Marks,1 D. F. Youngelman,2 T. Berk1
1

Department of Surgery, The Mount Sinai Medical Center, School of Medicine, Case Western Reserve University, One Mount Sinai Drive, Cleveland, OH 44106, USA 2 Department of Surgery, Harry S. Truman Memorial Veterans Hospital, 800 Hospital Drive, University of Missouri-Columbia, Columbia, MO 65201, USA Received: 11 March 1996/Accepted: 5 July 1996

Abstract Background: Diagnostic laparoscopy for the evaluation of injuries in patients with penetrating abdominal trauma has been shown to decrease the morbidity and mortality associated with mandatory laparotomy. The overall impact on patient care and hospital costs has not been thoroughly investigated. The goal of this study was to determine the economic impact of laparoscopy as a diagnostic tool in the management of patients following penetrating trauma to the abdomen or flank. Methods: Retrospective chart review of all hemodynamically stable patients with penetrating trauma to the abdomen or flank, but without other injuries requiring emergent intervention, admitted to a level I trauma center between January 1, 1992, and September 30, 1994. Those patients who underwent either laparoscopy (DL) or laparotomy (NL) or both (CONV) and who had no intraabdominal organ injuries requiring surgical therapeutic intervention were included in the study. Age, operative time, operative findings, length of hospitalization, Injury Severity Score (ISS), variable costs, and total costs were recorded for each patient. Results: Fourteen patients underwent negative/nontherapeutic laparoscopy (DL), 19 patients underwent negative/nontherapeutic laparotomy (NL), and four patients underwent both laparoscopy and laparotomy, a conversion procedure (CONV). There was no significant difference in age, operative times, or ISS between the DL and NL groups. Mean ISS of CONV patients was significantly greater than that of DL patients, 5.75 1.97 vs 2.43 0.63 (p < 0.05). Mean operative time for CONV patients was also significantly greater than both DL and NL patients, 106.5 17.00 min vs 66.1 6.55 and 47.3 7.50 min, respectively (p < 0.05). The mean length of stay was significantly shorter in

the DL group as compared to the NL or CONV groups, 1.43 0.20 vs 4.26 0.31 and 5.0 0.82 (p < 0.0001). The variable costs for the DL group were significantly lower than those incurred by patients in the NL and CONV groups, $2,917 175 vs $3,384 102 and $3,774 286, (p < 0.05). Variable costs were not significantly different between the NL and CONV groups. Total costs were also significantly lower in the DL group when compared to NL and CONV, $5,427 394 vs $7,026 251 and $7,855 750 (p < 0.005), but again, they were not statistically different between the NL and CONV groups. The overall total costs for laparoscopy, including the costs incurred by conversion patients, was significantly less than the total costs for laparotomy patients, $5,664 394 vs $7,028.47 250 (p < 0.005). This resulted in an overall savings of $1,059.44 per laparoscopy performed. The overall negative/nontherapeutic laparotomy rate during this study was 19.1%, which was significantly lower than the negative or nontherapeutic exploration rate during the time period prior to the use of laparoscopy (p < 0.01, z 2.550). Conclusion: Variable and total costs and length of stay were significantly lower in our population of patients who underwent DL as compared to NL. The rate of negative or nontherapeutic laparotomy was also significantly reduced when compared to the rate identified during the era prior to the use of laparoscopy. Laparoscopy resulted in an overall savings of $1,059 per laparoscopy performed when compared to laparotomy. Key words: Laparoscopy Penetrating abdominal trauma Cost effectiveness

Correspondence to: J. M. Marks Presented at the 5th World Congress of Endoscopic Surgery of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) Scientific Session, Philadelphia, Pennsylvania, USA, 15 March 1996

The use of diagnostic laparoscopy for the evaluation of penetrating abdominal trauma is gaining increasing acceptance. The high rates of negative and nontherapeutic laparotomy and their inherent complications are added incentive

273 Table 1. Patients admitted with penetrating trauma to the abdomen or flank between January 1, 1992, and September 30, 1994 Gunshot/shotgun victims (n 81) Exploratory laparotomy Negative/nontherapeutic Therapeutic Diagnostic laparoscopy Negative/nontherapeutic Therapeutic Conversion (laparoscopy/laparotomy) Negative/nontherapeutic Therapeutic 8 69 3 0 1 0 Stab wound victims (n 50) 11 24 11 1 2 1
a

Table 2. Comparison of the negative/nontherapeutic exploration rates for patients with penetrating trauma to the abdomen or flank Group I 19891991 (laparotomy only) Gunshot/shotgun victims Stab wound victims Total 28.2% (20/71) 42.6% (20/47) 33.9% (40/118) Group II 19921994 (laparotomy only) 11.5% (9/78) 35.1% (13/37) 19.1%a (22/115) Group III 19921994 (laparoscopy or laparotomy) 14.8% (12/81) 48.0% (24/50) 27.5% (36/131)

Group I vs group II (p < 0.01, z 2.550)

for the increasing use of this modality. Laparoscopy has been shown to be useful in excluding or confirming peritoneal penetration and therefore in preventing unnecessary laparotomy [25, 7, 8, 11, 15, 17]. Morbidity and mortality rates have also been shown to be significantly lower in patients undergoing laparoscopy as compared to celiotomy [2, 4, 17]. Length of hospital stay has also been proven to be significantly shorter following laparoscopic evaluation as compared to formal laparotomy [4, 7]. This retrospective study was undertaken to compare the costs incurred and the overall economic impact of laparoscopy as a diagnostic tool in the management of patients with penetrating trauma to the abdomen and flank. Materials and methods
The charts were reviewed of all patients admitted (to the Mt. Sinai Medical Center, Cleveland, Ohio, a level I trauma center) with penetrating trauma to the abdomen or flank between January 1, 1992, and September 30, 1994. These patients were all evaluated by the trauma team, which included a chief surgical resident and the surgical attending on call. Included in this study were all patients who underwent diagnostic nontherapeutic laparoscopy (DL) or negative or nontherapeutic laparotomy (NL). Conversion patients (CONV) were those patients who underwent both laparoscopy and laparotomy. Negative laparoscopy or laparotomy was defined as an exploration without evidence of peritoneal penetration or intraabdominal organ injury. Nontherapeutic laparoscopy or laparotomy was defined as an exploration identifying peritoneal penetration and/or intraabdominal organ injury which did not demand surgical therapeutic intervention. Hemodynamically stable patients (SBP > 90 mmHg, HR < 110) were candidates for diagnostic laparoscopy if they had (1) stab wounds to the abdomen or flank with presumptive peritoneal penetration based on omental evisceration, positive diagnostic peritoneal lavage or positive wound exploration, or (2) tangential gunshot wounds without obvious peritoneal penetration. Laparoscopy was not considered in patients who required other emergent procedures such as exploration for peripheral vascular injury or those who were not hemodynamically stable. All procedures were performed in the operating room under general anesthesia and all patients consented to possible conversion to laparotomy. After induction of general anesthesia, a Foley catheter and orogastric or nasogastric tube were placed in each patient. An umbilical nondisposable trocar (Stortz, Charlton, MA) was placed by the Hasson technique and the abdomen was insufflated with CO2 to a pressure of 15 mmHg. A 0, 10-mm laparoscope was used initially in all patients. A 30 laparoscope was used as needed for evaluation of the diaphragm and upper abdominal organs. Additional 5-mm ports were placed under direct vision as necessary for manipulation of the bowel. Nondisposable, noncrushing bowel clamps (Olympus, Melville, NY; Marlow, Willoughby, OH) were used routinely to decrease the risk of bowel injury and to lower the variable costs incurred by laparoscopy. All quadrants were carefully inspected and the small bowel and colon were examined completely. All exploratory laparotomies were performed in standard fashion through a midline inci-

sion under general anesthesia. Variable and total costs were determined for each patient by the hospital billing office. All costs were adjusted for inflation to 1994 rates. The overall economic impact of laparoscopy was calculated by averaging the additional expense incurred by the conversion group to the DL group (overall laparoscopy costs DL costs + [CONV costs NL costs] / of DL + CONV patients). Statistical analysis was done by Students paired t-test, or z-test analysis.

Results Between January 1, 1992, and September 30, 1994, 131 patients were admitted to the trauma service with penetrating injuries to the flank or abdomen. There were 81 patients with gunshot wounds and 50 patients with stab wounds (Table 1). Ninety-five percent of patients with gunshot wounds underwent laparotomy (77/81). DL was performed on 3.7% of gunshot victims (3/81), and CONV on 1.2% of these patients (1/81). Seventy percent of patients with stab wounds underwent laparotomy (35/50), 24% underwent DL (12/50), and 6% had CONV (3/50). The combined negative or nontherapeutic laparotomy/laparoscopy rate was 27.5% (GSW 14.8%, STAB 48.0%), which was comparable to the negative or nontherapeutic laparotomy rate of 33.9% (p > 0.10, as determined by z-test) for a similar time period between March 1, 1989 and December 31, 1991, prior to the use of laparoscopy. The overall negative/nontherapeutic laparotomy rate was 19.1%, which was significantly lower than the negative or nontherapeutic exploration rate during the time period prior to the use of laparoscopy (p < 0.01, z 2.550) (Table 2). Four patients with gunshot wounds were evaluated by DL (Table 3). All four were negative and one was converted to laparotomy, which confirmed the negative findings at laparoscopy. Fifteen patients underwent diagnostic laparoscopy following stab wounds to the flank or abdomen. Three of these patients were converted to laparotomy. One patient had a colon injury identified during laparoscopic evaluation, and this was repaired primarily following conversion to formal laparotomy. The second patient had equivocal findings at DL and the third had definitive evidence of peritoneal penetration without an identifiable intraabdominal injury. Both of these last two patients underwent open laparotomy which confirmed the laparoscopic findings. One patient underwent a therapeutic laparoscopy with repair of a diaphragmatic defect. Overall, 14 patients underwent DL following a penetrating trauma to the abdomen or flank without an

274 Table 3. Comparison of preoperative and intraoperative findings in laparotomy and laparoscopy patients Laparotomy Number of patients (n) Gunshot/shotgun Stab wound Preoperative indications Positive wound exploration Positive DPL Positive CT scan Evisceration Suspicion Intraoperative findings: No peritoneal violation Peritoneal violation without intraabdominal organ injury Intraabdominal organ injury not requiring intervention Intraabdominal organ injury requiring intervention 19 8 11 7 0 1 2 9 8 5 6 0 Laparoscopya 18 4 14 13 0 0 1 4 13 4 0 1

a Includes conversion patients undergoing both laparoscopy and laparotomy

identifiable intraabdominal injury requiring therapeutic intervention or conversion to laparotomy between January 1, 1992, and September 30, 1994 (Table 3). There were 11 patients with stab wounds and three patients with tangential gunshot wounds. Nine of 11 patients with stab wounds had preoperative evidence of peritoneal penetration with either an omental evisceration or a positive wound exploration. None of these patients had intraabdominal injuries identified at DL. The three patients with gunshot wounds were without signs of peritonitis prior to DL and no intraabdominal injuries were discovered during laparoscopic evaluation. Four patients, described previously, underwent both laparoscopy and laparotomy and are included in the CONV group. Nineteen patients who underwent negative or nontherapeutic laparotomy (NL) were identified during the same time period (Table 3). There were 11 patients with stab wounds, seven patients with gunshot wounds, and one patient with a shotgun wound in the NL group. Nine of 11 patients with stab wounds had either an omental evisceration or positive wound exploration. Three of the gunshot wound patients had signs of peritoneal irritation preoperatively. None of the stab wound victims had evidence of peritonitis. The one patient with a shotgun wound underwent a laparotomy following a CT scan which revealed a single intraabdominal pellet. This pellet was located in the omentum and no therapeutic intervention was necessary. There were no intraoperative or postoperative complications that prolonged, hospitalization or increased costs. Also, there were no missed injuries in the three groups. There was no significant difference in age, operative times, or ISS between the DL and NL groups (Table 4). Mean ISS of CONV patients was significantly greater than DL patients, 5.75 1.97 vs 2.43 0.63 (p < 0.05). Mean operative time for CONV patients was also significantly greater than both DL and NL patients, 106.5 17.00 min vs 66.1 6.55 and 47.3 7.50 min, respectively (p < 0.05). The mean length of stay was significantly shorter in the DL group as compared to the NL and CONV groups, 1.43 0.20 days, vs 4.26 0.31 days and 5.00 0.82 days, respectively (p < 0.0001). Seventy-three percent (10/14) of

patients undergoing DL were discharged within 1 day of surgery. One patient required hospitalization for 3 days because of an associated chest injury. In comparison, 94.7% of NL patients (18/19) remained hospitalized for 3 days or more. Variable costs reflect the number and nature of procedures undergone by each patient and include the cost of supplies and labor specific to each patients care. The physicians fees for each patient were added to both the variable and total costs. The variable costs for the DL group were significantly lower than those incurred by patients in the NL and CONV groups, $2,917 175 vs $3,384 102 and $3,774 286, (p < 0.05). Variable costs were not significantly different between the NL and CONV groups. Total costs were also significantly lower in the DL group when compared to NL and CONV, $5,427 394 vs $7,026 251 and $7,855 750 (p < 0.005), but again were not statistically different between the NL and CONV groups. A detailed analysis comparing the average costs incurred by laparoscopy and laparotomy is provided in Table 5. The overall total costs for laparoscopy, including the costs incurred by conversion patients, were significantly less than the total costs for laparotomy patients, $5,664 394 vs $7,028.47 250 (p < 0.005). This resulted in an overall savings of $1,059.44 per laparoscopy performed (Table 6). When the costs were divided into preoperative, intraoperative, and postoperative expenditures, the greatest proportion of costs in both the DL and NL groups was intraoperative, both in the analysis of variable and of total costs. Postoperative expenditures in the NL group, however, were almost three times greater than those incurred by patients in the laparoscopy group, reflecting the increased cost of prolonged hospitalization.

Discussion The goal of our review was to assess the overall economic impact of diagnostic laparoscopy in the management of stable patients with penetrating abdominal trauma. The total costs calculated in this study, including the physicians fees, best represent the cost as perceived by a third-party payer. Alternatives to DL in the evaluation and management of hemodynamically stable patients with penetrating abdominal trauma include observation with selective laparotomy or mandatory laparotomy. We had no morbidity related to the laparoscopic procedures and demonstrated a significantly decreased length of stay. Several patients were discharged home immediately following DL without further observation, and over 70% were discharged by the 1 postoperative day. This would support the use of DL in the emergency department with local anesethesia and intravenous sedation [1, 2, 13]. Patients could be safely evaluated and discharged home without admission to the hospital. Salvino et al. have shown the safety and efficacy of emergency department laparoscopic evaluation [13]. In the ongoing evaluation of the use of DL in trauma patients, most protocols require taking the patient to the operating room, although not necessarily using general anesthesia [2, 3, 5, 8, 11, 15, 16]. It has been our practice, and it is our recommendation, to perform DL in the operating room setting. The use of the operating room affords the surgeon greater flexibility. Pa-

275 Table 4. Comparison of patients having undergone negative or nontherapeutic laparoscopy and/or laparotomy between January 1, 1992, and September 30, 1994 Negative or nontherapeutic laparoscopy Patients (n) Gunshot/shotgun victims (n) Stab wound victims (n) Mean age (years) Injury Severity Score Length of stay (days) Operative time (minutes) Variable costs (dollars) Total costs (dollars)
a b

Negative or nontherapeutic laparotomy

Laparoscopy and laparotomy (conversion) 4 1 3 of mean) 38.5 6.65 5.75 1.97 5.0 0.82 106.5 17.00 3774 286 7855 750

14 3 11

19 8 11 (Calculation of means standard error 31.2 2.23 3.21 0.66 4.26 0.31 66.1 6.55d 3384 102 7026 251

30.5 2.41 2.43 0.63a 1.43 0.20b 47.3 7.50c 2919 175e 5427 394f

p < 0.05 vs conversion p < 0.0001 vs laparotomy and conversion c p < 0.01 vs conversion d p < 0.05 vs conversion e p < 0.05 vs laparotomy and conversion f p < 0.001 vs laparotomy and conversion

tients may be easily repositioned and rotated on a standard operating room table to facilitate the complete evaluation of the abdomen. The ready availability of both 0 and 30 laparoscopes is also valuable. Lastly, in those patients who will require conversion to laparotomy due to inadequate laparoscopic evaluation or because of the need for therapeutic intervention, time will not be wasted transporting them to a different area of the hospital. The surgeon will be able to proceed directly to a laparotomy or to proceed with therapeutic laparoscopy. In addition, these procedures should be carried out by experienced laparoscopic surgeons who have familiarity with advanced laparoscopic techniques. Some of the reported studies in emergency department laparoscopy have also documented the use of smallersize laparoscopes [1, 2]. With the currently available 5-mm or 4-mm laparoscopes, however, visualization has been limited by a lack of brightness and a smaller field of vision. With newer technology and improvement in optics, the use of these smaller scopes may become more commonplace. There is no dispute regarding the efficacy and safety of DL in selected trauma patients. Carey et al. reported a decrease in their nontherapeutic laparotomy rate from 11% to 8.5% since initiating DL rather than diagnostic peritoneal lavage, computed tomography, or local wound exploration in the evaluation of patients with penetrating abdominal trauma [3]. The ability to decrease or even eliminate negative or nontherapeutic laparotomy could impact the overall morbidity and mortality of the trauma population [6, 10, 12]. Investigators at the University of Miami found that their 12.4% negative laparotomy rate following mandatory laparotomy was associated with a 22% morbidity rate in these patients and a mean hospital stay of 5.1 days [17]. By using diagnostic laparoscopy in a similar group of patients, they were able to decrease the morbidity rate to 3% and the mean hospital stay to 1.4 days. Both of these outcomes should significantly decrease the cost of caring for these patients. In our series, there were no missed injuries or complications. In addition, with the use of laparoscopy, we were able to significantly reduce our negative or nonthera-

peutic laparotomy rate from 33.9% to 19.1%. Laparoscopy proved to be not only safe, but it helped avoid an unnecessary laparotomy and the increased costs incurred by an extended hospitalization. Once it has been determined that a given procedure is safe and effective, the costs incurred by this procedure are commonly the next issue to be carefully evaluated. This study has demonstrated how decreased length of stay following DL leads to decreased total hospital costs, and we suspect that these costs will further decrease as the variable costs associated with laparoscopy decrease. The use of disposable instrumentation increases a hospitals variable costs. By eliminating the use of many disposable laparoscopic instruments, hospitals can reduce the costs of treating patients without affecting the standards of patient care. The use of gasless laparoscopy and conventional instruments has also been shown to be safe and cost effective [14]. This would not only lower the variable costs incurred by each patient but also lower the relative proportion of intraoperative costs associated with diagnostic or therapeutic laparoscopy. Also, by avoiding an unnecessary laparotomy and the increased costs incurred by this procedure, laparoscopy can provide further cost savings. Diagnostic laparoscopy is a cost-effective procedure for the evaluation of hemodynamically stable patients. It is vital that patients undergo a complete laparoscopic evaluation of the entire abdomen, preferentially in an operating room setting, where a surgeons resources are greatest. These procedures should all be carried out, or supervised, by experienced laparoscopic surgeons. In addition, all patients should consent to exploratory laparotomy, and there should be no hesitation to convert to an open procedure if the patient becomes unstable or if complete laparoscopic evaluation is not possible. In conclusion, hospital costs and length of stay were significantly lower in our population of patients when comparing negative or nontherapeutic laparoscopy with laparotomy in the evaluation of penetrating abdominal trauma. The overall economic impact of laparoscopy resulted in a $1,059 savings per laparoscopy performed.

276 Table 5. Detailed analysis of average costs incurred by laparoscopy and laparotomy Laparoscopy Variable Daily room and care IV therapy Emergency services Blood bank Chemistry lab Histology lab Hematology lab Electrocardiology lab Pharmacy Recovery room Operating room Central supply Respiratory care Anesthesia Microbiology lab Radiology Surgeons fee Anesthesiologists fee Total $244.67 6.92 94.67 26.00 46.42 0.00 13.17 2.00 76.25 51.33 171.25 634.67 9.50 253.83 1.75 28.08 950.00 308.77 $2,919.00 Total $843.42 13.67 246.17 66.50 110.08 0.00 40.42 5.42 142.33 120.33 869.58 1,122.67 24.33 435.25 3.83 124.67 950.00 308.77 $5,427.00 Variable $843.53 15.13 112.00 21.60 57.07 1.00 18.27 2.13 174.07 51.47 150.53 394.33 36.07 255.40 6.07 13.20 900.00 333.47 $3,384.00 Laparotomy Total $2,713.73 27.13 296.67 57.33 143.53 7.53 54.07 4.93 323.80 118.80 780.93 713.60 92.80 435.73 16.07 59.93 900.00 333.47 $7,026.00

Table 6. Comparison of overall costs incurred by laparoscopy during the study period vs the costs of negative or nontherapeutic laparotomy 6. Laparoscopya Variable costs (dollars) Total costs (dollars)
a

Laparotomy 7. 3,384.37 102 7,028.47 251

3,028.43 174b 5,664.57 394c

8.

overall DL costs ( DL costs + [CONV costs-NL costs])/DL + CONV patients (n) b p 0.07 c p < 0.005

9.

10. Acknowledgment. We wish to thank Leslie Brown and Jane Dostal for their excellent assistance in the preparation of this manuscript. 11. 12.

References
1. Berci G, Dunkelman D, Michel SL, Sanders G, Wahlstrom E, Morgenstern L (1983) Emergency minilaparoscopy in abdominal trauma. Am J Surg 146: 261265 2. Berci G, Sackier JM, Paz-Parlow M (1991) Emergency laparoscopy. Am J Surg 161: 332335 3. Carey JE, Koo R, Miller R, Stein M (1995) Laparoscopy and thoracoscopy in evaluation of abdominal trauma. Am Surg 61: 9295 4. Fabian TC, Croce MA, Stewart RM, Pritchard FE, Minard G, Kudsk KA (1993) A prospective analysis of diagnostic laparoscopy in trauma. Ann Surg 217: 557565 5. Fernando HC, Alle KM, Chen J, Davis I, Klein SR (1994) Triage by

13.

14.

15. 16. 17.

laparoscopy in patients with penetrating abdominal trauma. Br J Surg 81: 384385 Henderson VJ, Organ CH Jr, Smith RS (1993) Negative trauma celiotomy. Am Surg 59: 365370 Ivatury RR, Simon RJ, Stahl WM (1993) A critical evaluation of laparoscopy in penetrating abdominal trauma. J Trauma 34: 822828 Livingston DH, Tortella BJ, Blackwood J, Machiedo GW, Rush BF (1992) The role of laparoscopy in abdominal trauma. J Trauma 33: 471475 Marks Jm, Ramey R, Baringer DC, Aszodi A, Ponsly JL (1995) Laparoscopic repair of a diaphragmatic laceration. Surg Laparosc Endosc 5: 415418 Petersen SR, Sheldon GF (1979) Morbidity of a negative finding at laparotomy in abdominal trauma. Surg Gynecol Obstet 148: 2326 Ponsky JL, Marks JM (1995) Laparoscopic examination of the bowel in trauma patients. Gastrointest Endosc 143(2): 146148 Ryan M, Leighton T, Pianim N, Klein S, Bongard F (1993) Medical economic consequences of gang-related shootings. Am Surg 59: 831 833 Salvino CK, Esposito TJ, Marshall WJ, Dries DJ, Morris RC, Gamelli RL (1993) The role of diagnostic laparoscopy in the management of trauma patients: a preliminary assessment. J Trauma 34: 506515 Smith RS, Fry WR, Tsoi EKM, Henderson VJ, Hirvela ER, Koehler RH, Brams DM, Morabito DJ, Peskin GW (1993) Gasless laparoscopy and conventional instruments. Arch Surg 128: 11021107 Smith RS, Fry WR, Morabito DJ, Koehler RH, Organ CH Jr. (1995) Therapeutic laparoscopy in trauma. Am J Surg 170: 632636 Sosa JL, Sims D, Martin L, Zeppa R (1992) Laparoscopic evaluation of tangential abdominal gunshot wounds. Arch Surg 127: 109110 Sosa JL, Baker M, Peunte I, Sims D, Sleeman D, Ginzburg E, Martin L (1995) Negative laparotomy in abdominal gunshot wounds: potential impact of laparoscopy. J Trauma 38: 194197

Surg Endosc (1997) 11: 249252

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic salvage of malfunctioning peritoneal catheters

R. Amerling,1 D. Vande Maele,2 H. Spivak,2 A. Y. Lo,2 P. White,2 H. Beaton,2 J. Rudick2
1 2

Division of Nephrology and Hypertension, Beth Israel Medical Center, 17th Street and First Avenue, New York, NY 10003, USA Department of Surgery, Beth Israel Medical Center, 17th Street and First Avenue, New York, NY 10003, USA

Received: 27 March 1996/Accepted: 30 May 1996

Abstract Background: Malfunction of peritoneal catheters due to mechanical outflow problems is an annoying complication in patients undergoing chronic peritoneal dialysis (PD). Correction often involves catheter replacement or revision via laparotomy. Methods: Twenty-five patients undergoing PD who developed mechanical catheter flow restriction underwent 28 laparoscopic procedures. Preoperative diagnoses were made by contrast catheter radiography and were: catheter sequestration (36%), omental wrap (64%). Pneumoperitoneum was induced after general anesthesia and laparoscopy was performed using a Storz laparoscope. The catheter was then identified and manipulation was attempted using instruments placed percutaneously. Results: In 26 cases (93%), the catheter was freed and function restored. In two cases (7%), adhesions were so numerous and dense that the distal catheter could not be visualized. Four episodes of peritonitis occurred in the perioperative period. Four patients developed subcutaneous leakage of peritoneal fluid which responded to cessation of PD for 2 weeks. Four patients had recurrent occlusions; three of these were managed laparoscopically. Two patients developed late hernias at the site of insertion of the laparoscope. Catheter patency averaged 9.2 months postoperatively. Conclusions: Laparoscopic revision is a successful technique for salvage of occluded peritoneal catheters. Key words: Peritoneal catheters Malfunction Laparoscopy Repair Omentum Adhesions

outflow obstruction, though in our experience, this usually signifies omental wrapping. Simple displacement of a free catheter does not interfere with outflow [17]. Repositioning by external guidewire manipulation yields poor long-term results and is problematic in catheters with predefined bends and coiled intraperitoneal segments [13]. Successful treatment of omental wrapping or catheter sequestration has typically required laparotomy with omentectomy and manual repositioning of the catheter, or removal and replacement of the malfunctioning catheter. With the advent of high-resolution laparoscopic surgery, it has become possible to perform complex intra-abdominal procedures with a minimum of cutting and dissecting. Since 1991 we have employed laparoscopic techniques to repair malfunctioning peritoneal catheters. This approach offers significant advantages over traditional surgical methods.

Materials and methods

Twenty-five patients aged 2673 on peritoneal dialysis for 020 months developed mechanical catheter dysfunction over a 4-year period (Table 1). The cause of obstruction was diagnosed by contrast catheter radiography in all cases and was confirmed at operation (Fig. 1). The causes were omental wrap (18/28; Fig. 2) and sequestration by adhesions (10/28). All procedures were performed under general anesthesia. Prophylactic antibiotics were given routinely; either intravenous cefazolin (1 g) or intravenous vancomycin (1 g) and gentamicin (80100 mg). A 10/11-mm trocar was used at a periumbilical site for the laparoscope. Pneumoperitoneum was induced and the catheter was located and traced to the site of obstruction. In addition to the laparoscope trocar, two 5-mm trocars were used for seven procedures and three for 21. These were used to introduce instruments and were positioned to achieve maximum leverage relative to the catheter. When omental wrap was identified, simple stripping of the omentum usually released the catheter (Fig. 3). In four cases, the catheter was exteriorized for more meticulous debridement. Lysis of adhesions was performed using electrocautery or blunt dissection. Once freed, the catheter was repositioned within the pelvis. Partial omentectomy was carried out in five patients using the EndoGIA device. At this point, the catheter was irrigated with heparinized dialysate to assure hydraulic function and to prevent clotting within the catheter. Ports were removed and the sites were closed with nonabsorbable sutures using the Advanced Surgical trocar closing device. (The first 13 cases were performed before adequate fascial closure systems were available.) Catheter function was reconfirmed with heparinized dialysate; 250500 cc was left within the peritoneal cavity to prevent clot formation.

Mechanical obstruction of peritoneal dialysis catheters is a frustrating problem that occurs in 230% of patients treated by this modality [8, 15]. Severe obstruction often leads to catheter loss, interruption of peritoneal dialysis, and treatment failure. When it is due to fibrin plugging or adhesion, thrombolytic therapy with urokinase is frequently successful [4]. Migration of the catheter tip may be associated with
Correspondence to: R. Amerling

250 Table 1. Patient characteristics and utcome after laparoscopic catheter revisiona Proc. # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
a

Pt. # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 16 17 18 18 19 20 21 22 23 24 24 25

Age/sex 26 M 72 M 58 M 32 F 67 M 36 M 44 M 31 F 46 F 38 M 73 F 71 F 52 F 53 F 59 M 33 M 30 M 31 M 54 M 49 M 46 M 50 M 69 M 45 M 58 M

Cause ESRD RPGN Atheroemb HTN HTN HTN DM HIVAN HTN DM HIVAN DM HTN DM HTN DM HIVAN SLE HIVAN HTN HIVAN HIVAN DM HTN DM DM DM

Months on CPD 13 14 <1 <1 17 1 2 <1 20 <1 <1 <1 <1 <1 <1 1 1 <1 1.5 <1 12 <1 <1 6 <2

Cause obstruct Sequest Sequest Sequest Oment wrap Sequest Oment wrap Oment wrap Oment wrap Oment wrap Oment wrap Oment wrap Sequest Sequest Sequest Oment wrap Oment wrap Oment wrap Oment wrap Oment wrap Oment wrap Oment wrap Oment wrap Sequest Oment wrap Sequest Oment wrap Oment wrap Sequest

Cath type Cruz Cruz Tenck Cruz Crux Cruz Cruz Cruz Cruz Cruz Cruz Tenck Tenck Tenck Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz Cruz

Duration (mos) 4 4 0 22 4 5 2 36 36 1 17 11 17 15 17 3 8 12 1 3 7 5 0 4 1 8 3 12

Outcome Tx Hd Tx HD Expired Tx HD Tx HD Tx HD Tx HD On PD On PD Tx HD On PD On PD On PD On PD On PD Recurred On PD On PD Recurred Expired On PD On PC Tx HD Tx HD On PD Recurred On PD On PD

Complications Peritonitis Peritonitis Peritonitis Hernia Leak Leak Leak Leak Recurred Hernia

Recurred Cath replaced Recurred Cath replaced Failed Scrotal emphysema Recurred Peritonitis

RPGN rapidly progressive glomerulonephritis; atheroemb atheroembolic disease; HIVAN HIV-associated nephropathy, Cruz Cruz catheter; Tenck Tenckhoff catheter; DM Diabetes mellitus; HTN Hypertension; SLE systemic lupus erythematosus; Sequest catheter sequestered by adhesions; Oment wrap Omental wrap; OnPD Remains on peritoneal dialysis; TxHD Transferred to hemodialysis.

Results Operative time ranged from 40 to 120 min. In all but two cases, the catheter was freed and function was restored. In two cases, adhesions were so dense that the distal catheter could not be visualized. Lysis of adhesions in one patient permitted PD to be continued for 3 months with adequate volumes. In the other patient, the catheter was subsequently removed. Catheters remained patent for a mean of 9.2 months (range: 036). Ten catheters remain functional at the time of this writing. Four of 25 patients developed peritonitis in the perioperative period. In three of these the infection resolved with antibiotics within 1 week. In the other, the catheter again became occluded and the patient expired before the catheter could be removed. This was the only perioperative death. It was apparent that this patient had florid peritonitis at the time of laparoscopy. Three of the four episodes of peritonitis occurred during the early period of our experience, before routine antibiotic prophylaxis was used. Only one episode of perioperative peritonitis has been documented since adopting this measure. In four of 28 procedures, patients experienced subcutaneous leakage of peritoneal fluid. This manifested as either leakage at the exit site or as scrotal edema. All resolved with cessation of PD for 2 weeks. These episodes of leakage were presumed to be due to breeches of peritoneal integrity at the trocar sites. Once this was recognized as a complication, patients were either maintained on hemodialysis for 12 weeks or admitted for

low-volume, high-frequency PD. As we developed experience with port-site closure technique, fluid leakage disappeared. No leakage has been noted since using the newer endoscopic port closure devices, and we are able to resume full volume exchanges immediately. Two of 25 patients developed a significant hernia at the site of laparoscope insertion. Both underwent surgical correction of the hernia. With improved site closure we have not seen any hernia in the later series. One patient developed mild scrotal emphysema postoperatively which resolved spontaneously over 24 h. Not including the patient described above with intercurrent peritonitis who expired a week after the procedure, four patients developed recurrent occlusions. All had omental wraps. Two reoccluded within 4 weeks of revision, one after 12 weeks, and another after 8 months. Three of the four were successfully revised laparoscopically. The fourth elected to transfer to hemodialysis and the catheter was removed. When the procedure was performed on stable outpatients, they were able to go home the same day. There were no acute complications that necessitated hospitalization. There were no episodes of bleeding, and in no case was laparotomy required. Postoperative pain was mild to moderate and readily controlled with oral analgesics. Discussion Cunningham and Tucker used peritoneoscopy to evaluate two peritoneal dialysis patients with peritonitis in 1983 [7].

251

Fig. 1. Typical radiographic appearance of omental wrap. Catheter is displaced out of the pelvis. Contrast lines both the lumen and the exterior of the catheter, and intraluminal filling defects, representing ingrowth of omentum, are seen. Fig. 2. Omentum wrapped around catheter at point of entry into peritoneal cavity. Fig. 3. Omentum being stripped from catheter.

Wilson and Swartz were the first to describe the laparoscopic correction of a poorly functioning peritoneal catheter [18]. Since then, others have used laparoscopic techniques to implant [13, 16] and revise [5, 911, 14] peritoneal catheters. Laparoscopy is a relatively noninvasive approach to catheter salvage. It can be performed by any surgeon experienced with laparoscopic technique, and in most cases it can be done on an ambulatory basis, which saves costs. The traditional laparotomy with omentectomy and manual catheter repositioning is a major procedure with a prolonged recovery period. Removal and replacement of the catheter

necessitates the creation of a new exit site and tunnel which then must mature. The postoperative complications of subcutaneous fluid leakage and hernia formation have been eliminated by meticulous closure of the fascia. Recurrence of omental wrapping has been bothersome. Of the four instances, one had undergone omentectomy at the time of revision. One approach may be to suture the omentum to the anterior abdominal wall [12]. Residual devitalized omental tissue left within the catheter lumen has not caused obstruction. In conclusion, laparoscopic manipulation of occluded

252

peritoneal dialysis catheters is a useful technique for catheter salvage. It has the advantage of permitting early resumption of peritoneal dialysis, which greatly simplifies the management of these patients. References
1. Adamson AS, Kelleher JP, Snell ME, Hulme B (1992) Endoscopic placement of CAPD catheters: a review of one hundred procedures. Nephrol Dial Transplant 7: 855857 2. Amerling R, Cruz C (1993) A new laparoscopic method for implantation of peritoneal catheters. ASAIO J 39: M787M789 3. Ash SR, Wolf GC, Bloch R (1981) Placement of the Tenckhoff peritoneal catheter under peritoneoscopic visualization. Dial Transplant 10: 383385 4. Benevent D, Peyronnet P, Brignon P (1985) Urokinase infusion for obstructed catheters and peritonitis. Perit Dial Bull 5: 77 5. Chao SH, Tsai TJ (1993) Laparoscopic rescue of dysfunctional Tenckhoff catheters in continuous ambulatory peritoneal dialysis patients [letter]. Nephron 65: 157158 6. Cruz C (1988) The peritoneoscopic implantation of a polyurethane percutaneous access device for peritoneal dialysis: preliminary experience. Am Soc Artif Organs 34: 941944 7. Cunningham JT, Tucker CT (1983) Peritoneoscopy in chronic peritoneal dialysis: use in evaluation and management of complications. Gastrointest Endosc 29 (1): 4750 8. Diaz-Buxo J (1991) Mechanical complications of chronic peritoneal dialysis catheters. Semin Dial 4: 106110

9. Gibson DH, Heasley RN, Price JH, Doherty CC, Douglas JF (1990) Laparoscopic repositioning of blocked peritoneal dialysis catheters in patients on CAPD [letter]. Clin Nephrol 33: 208 10. Kimmelstiel FM, Miller R, Molinelli BM, Lorch JA (1993) Laparoscopic management of peritoneal dialysis catheters. Surg Gynecol Obstet 176: 565570 11. Kittur DS, Gazaway P, Abidin MR (1991) Laparoscopic repositioning of malfunctioning peritoneal dialysis catheters. Surg Laparosc Endosc 1: 179182 12. McIntosh G, Hurst P, Young A (1985) The omental hitch for the prevention of obstruction to peritoneal dialysis catheters. Br J Surg 72: 880 13. Moss JS, Minda SA, Newman GE, Dunnick NR, Vernon WB, Schwab SJ (1990) Malpositioned peritoneal dialysis catheters: a critical reappraisal of correction by wire manipulation. Am J Kidney Dis 15: 305308 14. Mutter D, Marichal JF, Heibel F, Marescaux J, Hannedouche T (1994) Laparoscopy: an alternative to surgery in patients treated with continuous ambulatory peritoneal dialysis. Nephron 68: 334337 15. Nolph KD (1993) Access problems plague both peritoneal dialysis and hemodialysis. Kidney Int 43: s8184 16. Pastan S, Gassensmith C, Manatunga AK, Copley JB, Smith EJ, Hamburger RJ (1991) Prospective comparison of peritoneoscopic and surgical implantation of CAPD catheters. ASAIO Trans 37: M154156 17. Twardowski Z (1990) Malposition and poor drainage of peritoneal catheters. Semin Dial 3: 57 18. Wilson JA, Swartz RD (1985) Peritoneoscopy in the management of catheter malfunction during continuous ambulatory peritoneal dialysis. Dig Dis Sci 30: 465467

Surg Endosc (1997) 11: 280282

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Thoracoscopic enucleation of a large leiomyoma located on the left side of the esophageal wall
E. Taniguchi,1 W. Kamiike,1 K. Iwase,1 T. Nishida,1 A. Akashi,2 S. Ohashi,2 H. Matsuda1
1 2

First Department of Surgery, Osaka University Medical School, 2-2 Yamada-Oka, Suita, 565, Japan Department of Surgery, Takarazuka City Hospital, 4-5-1 Kohama, Takarazuka, 665, Japan

Received: 23 February 1996/Accepted: 5 June 1996

Abstract Thoracoscopic enucleation through the right thorax was successfully performed for a large leiomyoma located on the left wall of the upper-middle esophagus. Using intraoperative esophagoscopy, the exact location of the tumor was identified by trans-illumination. Two sling tapes passed around the esophagus made the procedure as easy as that for tumor located on the right wall of the esophagus. We considered that the thoracoscopic approach through the right thorax can be applied to various types of esophageal leiomyoma, even a large one located on the left side of the esophagus. Key words: Thoracoscopic Esophageal leiomyoma Intraoperative esophagoscopy

Video-assisted thoracoscopic surgery (VATS) is widely used in intrathoracic surgery [8] and has been applied not only to partial resection of the lung for bulla or lung tumors [6, 11] but also to esophageal tumors [2, 5, 7, 9]. Using VATS, we successfully enucleated a large leiomyoma located on the left wall of the esophagus. Case report
A 24-year-old woman was hospitalized with dysphasia caused by a submucosal esophageal tumor. Barium swallow study showed a large defect with a smooth surface on the left wall of the upper-middle esophagus (Fig. 1). Chest MRI indicated that the tumor was located between the tracheal bifurcation and the descending aorta, and the maximal diameter was about 6.5 cm (Fig. 2). Endoscopic ultrasonography showed a hypoechoic and homogenous tumor which was consecutive to the proper muscular layer of the esophagus. The mass was diagnosed as leiomyoma of the esophagus. Surgery was performed under general anesthesia in the left lateral position, and the right lung was collapsed during the procedure using a double-lumen endotracheal tube. One 5-mm trocar was inserted at the second and two 10-mm trocars at the fourth and sixth intercostal spaces on

the right anterior axillary line. Another 10-mm trocar was inserted at the fifth intercostal space on the right middle axillary line (Fig. 3). No pleural adhesion were observed. Tumor was not identified over the mediastinal pleura. First, the azygos vein was dissected and divided. Then, esophagofiberoscopy was performed to identify the esophageal level at which the tumor was located. The esophagus was dissected at both the oral and the anal portion of the tumor using trans-illumination by the esophagofiberoscopy as a guide. Two sling tapes were passed around the esophagus. By retracting these, the esophagus was freed from the surrounding tissues. Then, the left side was rotated to the right using an Endo-Babcock grasper. The muscular layer was divided to expose the surface of the tumor, which was divided bluntly and sharply. Then the tumor was enucleated. After resection of the tumor, it was confirmed that the esophageal mucosal layer was intact using esophagofiberoscopy. The proper muscular layer was closed with three 3-0 silk interrupted sutures. Extraction was facilitated by sufficient dilation of the puncture (about 2 cm in the diameter) so that the bag could be withdrawn with the tumor intact. The tumor was elastic soft and measured 5.6 1.6 3.0 cm. Histological examination confirmed that the tumor was a leiomyoma. Postoperative pleural effusion was less than 70 ml/day. The patient required analgesics several times because of the pain from the wound in which the chest drain was inserted. After removal of the drain on the 5th postoperative day (POD), no analgesics were required. She discharged on 12 POD. To date, she has been followed for 16 months with no late complications.

Discussion Because of recent advances in video-assisted thoracoscopic surgery (VATS), it has been applied to a variety of diseases [2, 8]. Concerning esophageal leiomyoma, several authors have reported cases successfully managed by VATS [1, 3, 4]. The present case was the largest such tumor yet reported. Furthermore, the tumor in our case was located at the left wall of the upper-middle esophagus where thoracoscopic approach has been considered more difficult than on the right side. We easily enucleated the tumor using the following two techniques. First, to identify the exact location of the tumor, which could not be recognized through the right thoracic space, esophagofiberoscopy was performed. Transillumination of the fiberscope indicated the upper and lower ends of the tumor. The second technique, which was thought to be more important, was passing two slings around the esophagus at the oral and anal portions of the tumor. Retracting the slings provided a good surgical view,

Correspondence to: E. Taniguchi

281

Fig. 1. Barium swallow study. A defect with a smooth surface was observed on the left wall of the upper-middle esophagus. Fig. 2. Chest MRI. Left, coronal view; Right, axial view. The tumor was located between the tracheal bifurcation and the descending aorta, and the maximal diameter was about 6.5 cm. Fig. 3. Positions of the trocars. AAL, anterior axillary line; MAL, middle axillary line.

282

even when the left side of the esophagus extensively protruding into the left mediastinum was divided. Moreover, those slings made it easy to rotate the esophagus; then we could manage the case in just the same way as a tumor located on the right side of the esophagus. Bardini et al. reported that esophageal diverticulum developed after this procedure when reapproximation of the myotomy, and, using our technique, the procedure was not very difficult even on the left side. There are no papers on the thoracoscopic approach through the left thorax for esophageal submucosal tumor except for one involving a lipoma located at the lower esophagus [10]. For upper and middle esophagus, approach through the left thorax is considered difficult because of the anatomical relations. The success in our case showed that a thoracoscopic approach through the right thorax can be applied to various types of esophageal leiomyoma, even to a large tumor located on the left side of the esophagus. References
1. Bardini R, Segalin A, Ruol A, Pavanello M, Peracchia A (1992) Videothoracoscopic enucleation of esophageal leiomyoma. Ann Thorac Surg 54: 576577

2. Dallemagne B (1993) Endoscopic approaches to oesophageal disease. Baillieres Clin Gastroenterol 7: 795822 3. Evertt NJ, Glinatsis M, McMahon MJ (1992) Thoracoscopic enucleation of leiomyoma of the esophagus. Br J Surg 79: 643 4. Gossot D, Fourquier P, El Meteini M, Celerir M (1993) Technical aspects of endoscopic removal of benign tumors of the esophagus. Surg Endosc 7: 102103 5. Gossot D, Fourquier P, Celerier M (1993) Thoracoscopic esophagectomy: technique and initial results. Ann Thorac Surg 56: 667670 6. Jones DR, Tanguilig GG, Graeber GM (1994) Thoracoscopic resection of bilateral metastatic sarcomas causing spontaneous pneumothorax. Chest 106: 12741276 7. McAnena OJ, Rogers J, Williams NS (1994) Right thoracoscopically assisted oesophagectomy for cancer. Br J Surg 81:236238 8. Miller JI Jr (1993) The present role and future considerations of videoassisted thoracoscopy in general thoracic surgery. Ann Thorac Surg 56: 804806 9. Pellegrini CA, Leichter R, Patti M, Somberg K, Ostroff JW, Way L (1993) Thoracoscopic esophageal myotomy in the treatment of achalasi. Ann Thorac Surg 56: 680682 10. Salo JA, Kiviluoto T, Heikkila L, Perhoniemi V, Lamminen A, Kivilaakso E (1993) Enucleation of an intramural lipoma of the oesophagus by videothoracoscopy. Ann Chir Gynaecol 82: 6669 11. Saw EC, Ramachandra S, Franco M, Tapper DP (1994) Videothoracoscopic wedge resection for peripheral pulmonary nodules. J Am Coll Surg 179: 289294

Surg Endosc (1997) 11: 239244

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Ultrasound-guided alcoholization of celiac plexus for pain control in oncology

M. Caratozzolo,1 M. M. Lirici,1 M. Consalvo,2 F. Marzano,1 E. Fumarola,2 L. Angelini1
1 2

IV Clinica Chirurgica, Universita ` degli Studi di Roma La Sapienza, Policlinico Umberto I, Viale del Policlinico 155, 00161 Rome, Italy Palliative Care Unit, Department of Anesthesiology, Universita ` degli Studi di Roma La Sapienza, Policlinico Umberto I, Viale del Policlinico 155, 00161 Rome, Italy Received: 7 January 1996/Accepted: 1 June 1996

Abstract Background: Treatment of inoperable pancreatic cancer pain is of paramount importance. The ineffectiveness of pharmacological agents has led many investigators to recommend chemical neurolysis of the celiac ganglions for pain control. This procedure may be performed under either fluoroscopic or computed tomography (CT) guidance, or it may accompany laparotomy. The authors describe a modified sonographically (ultrasoundUS)-guided technique for alcoholization of the celiac ganglions. Methods: Twelve patients underwent the neurolytic procedure. Nine of 12 suffered from pancreatic cancer. The remaining three were affected by inoperable hepatic, gastric, or colon cancer, respectively, with multiple hepatic metastases. US-guided alcohol neurolysis was performed by an anterior approach. In the last four patients, PIA (percutaneous injection alcohol) needles, modified by the authors, replaced the spinal needles employed in the first eight patients to inject the alcohol. Pain and pain relief were rated according to a Simple Descriptive Scale (SDS), and treatment success was gauged by declining opiate doses and need for pharmacological therapy. Results after treatment performed using different needles were compared. Results: Procedure-related mortality was zero. Complications of the neurolytic procedure included left pleural effusion in one patient and mild diarrhea in two other patients. Positive, negative, and indeterminant results were noted in nine (75%, p < 0.001), two, and one patient(s), respectively. Conclusions: The neurolytic technique, although far from being considered a routine procedure, appears to provide patients with safe and effective pain relief for pain unresponsive to conventional medical treatment. Key words: Pain-relieving procedures Alcoholization of celiac plexus Ultrasound-guided Inoperable pancreatic cancer

Correspondence to: M. Caratozzolo

Malignant tumors of the pancreas, particularly those arising from the exocrine part, are extremely difficult to treat surgically, and 5-year mortality rates of up to 99% have been reported [22]. The incidence of pancreatic neoplasms has tripled over the last 40 years, and survival after demolitive procedures, when possible, is of very short duration [9, 22]. Extremely severe pain usually arises from invasion of nervous structures by the expanding neoplastic mass [8]. Pain relief, therefore, has become an issue of paramount importance, and the search for effective analgesic agents continues. Studies of variably increasing doses of different types of drugs, such as nonsteroidal antiinflammatory drugs (NSAIDs) and major and minor opioids, conducted along lines proposed by the World Health Organization (W.H.O.), have yielded poor or inconclusive results. Chemical neurolysis of celiac ganglions is currently an extremely effective procedure to block transmission and achieve significant or complete alleviation of deep visceral pain, which is responsible for a marked reduction of survival in 85% of patients with pancreatic neoplasms. The technique is common, has been described extensively [2, 4, 10, 11, 13, 15, 17], and may be performed by an anterior or posterior approach employing either laparotomy or fluoroscopic, computerized tomographic (CT), or sonographic (US) guidance. In 1990, Sharfman and Walsh [20] presented a 15-year retrospective study of 480 cases of successful chemical neurolysis. Nonetheless, widespread consensus for the procedure has not been forthcoming, and it has been performed in only a small number of patients affected by acute pain. Relative ignorance of its advantages, fear of therapeutic ineffectiveness, and possible complications due to erroneous performance of it have negatively influenced universal acceptance of the procedure among practitioners. Variability of results and difficulty in localizing the right celiac plexus have also been cited [13]. Recently, radiologic guidance has been shown to be fundamental in improving the quality and reproducibility of the neurolytic procedure and in making it safer and more effective. Serial computed tomography (CT) concomitant with chemical neurolysis has been proposed.

240

Fig. 1. The 3.5-MHz convex probe with mounted biopsy kit. Fig. 2. The 22-gauge spinal needle with Chiba tip (upper figure) and the modified 22-gauge PIA bore tip needle with multiple lateral holes (lower figure). PIA percutaneous injection alcohol. Fig. 3. Disposable parts composing the US-guided alcoholization kit. US ultrasound.

Montero Matamala et al. [17] have described both an anterior and posterior approach in which a 21- or 22-gauge needle is inserted laterally to the celiac trunk from the right or left side of the patient. The procedure is then repeated on the opposite side to produce the desired neurolysis of both celiac ganglions, and consequent maximal analgesia. Recently, Montero Matamala et al. employed ultrasound (US) guidance to perform celiac ganglion alcoholization by anterior approach [18]. Precise needle placement in the retroperitoneum posed a problem, however, as ultrasound reflection from the needle tip was poorly visible in the hyperechoic retroperitoneal tissue. For that reason, a variant of the original anteriorly executed, US-guided, celiac plexus alcoholization procedure first described by Montero Matamala et al. [18] has been proposed by the authors and described in this paper [3, 16].

Materials and methods

Between January 1991 and May 1995, a study population of 12 patients (eight women and four men) with a mean age of 57.2 years (range 4867) was evaluated at the 4th Department of Surgery and the Department of Anesthesiology (Palliative Care Unit) of the University of Rome (Italy) La Sapienza School of Medicine. Nine patients were affected by pancreatic cancer, one by gastric cancer with hepatic metastases, one by colon cancer with multiple hepatic metastases, and one by multifocal hepatocellular carcinoma. Three of nine pancreatic cancer patients had undergone prior resection of the pancreatic head by Whipple procedure and were subsequently

deemed candidates for celiac plexus alcoholization either 3, 4, or 6 months, respectively, after surgery. The remaining six pancreatic carcinoma patients had been considered inoperable due to vascular encasement or obstruction, distant metastases, or significant expansion of neoplastic mass. Also, prior total gastrectomy had been performed on the patient affected by gastric cancer. Moderate-to-severe dull, sore, or burning pain, localized in the abdomen, was noted in all patients, and became shooting and intolerable as it radiated to the back. Pain radiated to the right shoulder in the patient affected by hepatocellular carcinoma, and in one case of cancer of the pancreatic head with metastatic pleural effusion, radiation to the left hemithorax was observed. Pain was resistant to high oral doses of morphine in all patients and to continuous subcutaneous infusion of morphine in four patients. Administration of steroids or NSAIDs was without effect in all cases. All patients underwent US-guided celiac plexus neurolysis by anterior approach. Real-time two-dimensional US guidance was performed employing either an Aloka SSD 650 (Aloka, Mitaka-shi, Tokyo, Japan) scanner or an EsaOte 560 (EsaOte Biomedica, Genoa, Italy) scanner, both equipped with a 3.5-MHz convex transducer (Fig. 1). Percutaneous introduction of the needle by linear puncture probe, angled at 30, followed. Enhanced precision of plexus localization and rapidity of execution over the free-hand technique were characteristic of this procedure protocol. Two types of needles were used: (1) a 22-gauge spinal needle (Becton & Dickinson, Madrid, Spain) 17.8 cm long, with internal stylet incorporated, in eight patients; and (2) a modified 22-gauge PIA (Percutaneous Injection Alcohol) needle (Sterylab, Rho, Italy) 20 cm long, possessing both internal stylet and 0.5-mm-diameter lateral holes located 8 mm from the tip, in the remaining patients (Figs. 2, 3). The PIA needle was specifically modified by the authors to meet procedural needs, with a Chiba bore tip substituting for the usual conical tip. Although all patients were in the terminal phase of their diseases and extremely emaciated due to neoplastic cachexia and nutritional decline, when the alcoholization procedure was attempted, needle lengths >17 cm

241 Table 1. Score of the Simple Descriptive Scale (SDS) 5 4 3 2 1 Terrible pain Severe pain Moderate pain Mild pain No pain

were nonetheless required to reach the target area in the retroperitoneum; needle shafts had to traverse the entire abdomen in the anterior approach, and 56 cm of needle length remained incorporated in the puncture kit. In order to better evaluate treatment results, each patient was asked to judge the intensity of preprocedure pain, according to a Simple Descriptive Scale (SDS) of five categories, from grade 1 (no pain) up to grade 5 (terrible pain) (Table 1). The procedure was performed in an operating room with electrocardiogram (ECG) and blood pressure monitored by an anesthesiologist. Sedation was initiated by a 7 g/kg IV dose of fentanyl (Fentanest, Carlo Erba, Milan, Italy), followed by 2 mg/kg of propofol (Diprivan, Zeneca, Basiglio, Italy) to prevent unintentional movements during the procedure. A 10 ml/kg IV infusion of dextran (MW 40,000) in water was administered to expand blood volume and prevent a hypotensive reaction to sudden opening of splanchnic arterial shunts [3], a possible consequence of celiac plexus destruction. Patient position on the operating table was supine, and serial transverse sonographic scans were made to define the common celiac trunk at its origin from the aorta and at its division into splenic and gastrohepatic branches. Optimal ventilation was assured by the anesthesiologist, through hyperextension and forward displacement of the mandible and use of a ventilation bag. A spinal needle was inserted via anterior approach and advanced along a pathway lying perpendicular to the common celiac trunk. Contrary to the opinions expressed by other authors [18], the anterior lateral wall of the aorta was punctured to allow entrance of the needle tip into the vascular lumen (Fig. 4). As precise placement of the needle is crucial to the success of neurolysis, this variant of the usual procedure made that possible by permitting clear identification of the needle tip and its unequivocal demarcation from the surrounding hyperechoic retroperitoneal adipose tissue. Subsequently, the distance between needle tip and celiac ganglion was estimated visually on a video monitor, and the needle tip was then withdrawn from the aortic lumen and positioned exactly 1 cm from the anterior lateral aortic wall (Fig. 5). The stylet was then removed, and, after careful suction to exclude needle presence within a vascular lumen, 6 ml of 0.25% bupivacaine (Marcaine, Astra Farmaceutici, Milan, Italy) was injected. Injection of local anesthetic must always precede that of alcohol to reduce the pain of either ganglion neurolysis or retroperitoneal tissue necrosis. Shortly after bupivacaine injection and additional suction, 15 ml of 48% sterile alcohol was injected under US guidance (Fig. 6). The entire procedure was then repeated on the opposite side [3, 16] (Figs. 79), and careful monitorization of ECG, heart rate, blood pressure, serum amylase, and coagulation tests for the first 24 h postop followed in the Intensive Care Unit (ICU). Tolerance to pain classified by SDS was evaluated every 12 h for 2 days and then every 30 days. The results achieved after the treatments performed using spinal needles and specially modified PIA needles were compared.

ments/day), which resolved with medical treatment. Severe falls in blood pressure and orthostatic hypotension were not observed. Tolerance to pain, classified by SDS, was optimal in nine patients (75%, p < 0.001), who received only NSAIDs or cortisone until their deaths (26 months after the procedure). The puncture site in the patient who had succumbed on the 2nd postoperative day was examined at autopsy, and no procedure-related injury to the aortic wall was found. Although many intimal atherosclerotic plaques were evident in the aorta, with thrombi and ulcers at the aortic bifurcation, only a mild and relatively irrelevant adventitial hemorrhagic suffusion was found at the puncture site. The pharmacological therapy administered before and after celiac plexus alcohol neurolysis is reported in Table 2, with a net reduction in opiate administration evident. In only two patients was it necessary to prolong oral analgesic therapy with half doses (60 mg/day) of timed-delay morphine (MS Contin, Chinoin, Milan, Italy). When the modified PIA needle was employed in the last four cases to inject the alcohol, both radial distribution of alcohol in the retroperitoneum and patient reaction to treatment improved. Because of the limited number of patients treated, the success rate of alcoholization when the modified PIA needle was used (100%) was not significantly higher than that (75%) when a conventional spinal needle was employed (p 0.38) (Fig. 10). Discussion The reported positive results of celiac plexus chemical neurolysis have been rather variable, ranging from 60% to 85% of cases. The average postoperative pain-free interval has varied from 2 to 240 days [11, 14]. For those reasons, this procedure has been performed only when other medical treatments have failed to alleviate pain; and even then, universal consensus for the procedure has not been forthcoming [11]. Recent improvements in the procedure, such as its performance under fluoroscopic or CT guidance [17], have increased the percentages of positive results obtained in different studies, although nonuniformity of techniques has led to variability of those results [13]. Chemical neurolysis performed under US guidance offers many advantages over the other procedures proposed. First of all, it allows observation of the entire procedure on a video monitor in real time, with a clear and unobstructed view of needle puncture and needle insertion into the aortic lumen, precise localization of the needle on both sides of the celiac plexus, and demonstration of direction of alcohol diffusion in the retroperitoneal tissue. The latter permits correction of the diffusion pathway, when anatomic and structural alterations resulting from neoplastic expansion cause alcohol to diffuse in the wrong direction. Second, the US-guided procedure exposes neither patient nor physician to unnecessary radiation, and is also less time-consuming than either fluoroscopic or CT-guided procedures, which require serial scanning to precisely localize the needle before injection of alcohol can be initiated. Preparation, sedation, and alcoholization of both celiac gan-

Results Neither mortality nor major complications followed the transaortic approach to celiac ganglion block, although one patient did die 2 days after the procedure as a result of chronic restrictive cardiomyopathy and heart failure, as confirmed by autopsy. Minor complications included an increase of preexisting left pleural effusion of neoplastic origin in one patient, caused by alcohol dispersion along the left diaphragmatic crus, resulting in pleural inflammation. Evacuative thoracentesis was resolutive in that case. Two patients experienced mild diarrhea (48 bowel move-

242

Fig. 4. Transverse scan at celiac trunk level. The arrow indicates the needle tip within the aortic lumen. Left approach. Fig. 5. The needle is withdrawn 2 cm and repositioned anterior to the vessel. The arrow indicates the needle tip in the retroperitoneal space. Fig. 6. Arrowheads indicate alcohol diffusion along the aorta after injection. A aorta. Fig. 7. The white arrow indicates needle penetration to the aorta. Right approach. Fig. 8. Hyperechoic area (arrows) of alcohol diffusion anterior to the aorta. AO aorta; AE hepatic artery; AS splenic artery. Fig. 9. Sonographic control 10 h after treatment. Hyperechoic halo of alcohol-induced necrosis surrounds the celiac trunk. A aorta; AE hepatic artery; AS splenic artery.

243

Fig. 10. Diagram of pain decrease as median SDS value at 1-day and 1-month after treatments performed employing conventional spinal needles and specially modified PIA needles. The enclosed table shows the min and max SDS values before treatment, and at 1 day and 1 month after treatment. Table 2. Therapetucial results and variations of SDS Before alcoholization NSAID Y N Y Y N Y Y N Y Y Y Y Corticoids Y Y Y Y Y Y Y Y Y Y Y Y Morphine Y Y Y Y Y Y Y Y Y Y Y Y SDS 4 4 5 5 4 4 5 4 5 5 4 5 Patient 1 2 3 4 5 6 7 8 9 10 11 12 Needle Spinal Spinal Spinal Spinal Spinal Spinal Spinal Modified Modified Modified Spinal Modified After alcoholization NSAID N Y N Y N Y Y Y Y Y Y N Corticoids Y Y Y Y Y Y Y Y Y Y Y Y Morphine N N N Y N N Y N N N N N SDS 1 day 1 2 2 3 1 2 2 1 2 2 1 1 SDS 1 month 2 3 2 5 2 3 4 2 3 3 1

PIA PIA PIA PIA

glions under US guidance require a mere 3040 min to perform. Third, US-guided alcohol neurolysis is less expensive than the other techniques mentioned, and sonographic equipment is more readily available than CT units. Furthermore, the US-guided procedure is more cost-effective for an average hospital and frees the CT unit, which might otherwise be monopolized for up to 90 min per procedure. One final advantage of US-guided ganglion neurolysis is that it can be performed with the patient in a supine position, which is much more comfortable for pancreatic cancer patients than the prone position required for the posterior approach; this also eliminates the risk of accidents known to occur during a conventional posterior approach [1, 57, 12, 18, 21], such as cephalad diffusion of alcohol, posterior to sympathetic chain and lumbar plexus, along the aorta and diaphragmatic crura. Actually, the last few years have seen the posterior approach replaced by the anterior technique in the majority of cases, even in those performed under CT guidance [17]; and Montero Matamala chose the anterior approach for the US-guided technique he described [18]. The only side effect of alcohol neurolysis was an increase of preexisting left pleural effusion in one patient. Hiccups, a sign of chemical irritation of the phrenic nerve, accompanied the neurolytic procedure in the same patient.

Diarrhea of 3 days duration, with multiple bowel movements, was noted in two other patients; it resolved with medical treatment and should be considered a consequence of celiac plexus block, rather than a true complication [19]. Patient sedation, to avoid inconvenient and dangerous reflex reactions, and aortic perforation, to provide a constant and unequivocal landmark even in an anatomic field altered by neoplastic expansion or surgery, may be considered disadvantages of the neurolytic procedure. Aortic perforation, however, is not a limitation of the procedure, as demonstrated in the autopsied patient who died 2 days after celiac ganglion neurolysis for reasons unrelated to aortic perforation with a 22-gauge needle, and subsequent needle withdrawal. Conventional PIA needles have a conical tip which excludes their use, when major arterial walls must be traversed, due to the risk of hemorrhage. For that reason, a new open-ended Chiba-like tip has been designed in order to guarantee safe introduction into the aortic lumen. When the newly designed, multiperforated, Chiba bore tip needle was employed in the last four patients treated by chemical neurolysis, significantly better alcohol diffusion in the retroperitoneum (as monitored by US guidance) seemed to lead to improved therapeutic results. Although definitive conclusions cannot be drawn due to the small size of the study population in which the multiperforated needle was em-

244

ployed, and the improvement of therapeutic results is not statistically significant, the observed relative improvement of the medium-term results in the last four patients can be theoretically attributed to radial alcohol diffusion from the multiperforated needle vs a less-effective 90 diffusion pattern produced by spinal needles in the other patients. The anterior approach to US-guided celiac ganglion neurolysis has been presented as a cheap, easy to perform, time-saving, safe procedure whose therapeutic efficacy equals, if not exceeds, that of the posterior approach and those of the fluoroscopic and CT-guided procedures. Far from a routine technique to relieve upper abdominal pain, alcoholization of the celiac plexus must, on the contrary, be considered an extraordinary measure to manage pain unresponsive to conventional medical treatment. References
1. Benzon HT (1979) Convulsions secondary to intravascular phenol: a hazard of celiac plexus block. Anesth Analg 58: 150152 2. Bonica JJ (1953) Management in pain. Lea & Febinger, Philadelphia, pp 208212, 13851390, 14861492 3. Caratozzolo M, Mattia C, Scardella L, Laurenzi L, Carassiti M, Angelini L (1991) Lalcolizzazione del plesso celiaco nel trattamento del dolore da cancro del pancreas: proposta di un nuovo approccio sotto guida ecografica. Proceedings 93 Congress Societa ` Italiana di Chirurgia, Florence, vol 1: 325329 4. Das KM, Chapman AH (1992) Sonographically guided coeliac plexus block. Clin Radiol 45: 401403 5. Fine PG, Bubela C (1985) Chylothorax following celiac plexus block. Anesthesiology 63: 454456 6. Fujita Y, Takaori M (1987) Pleural effusion after CT-guided alcohol celiac plexus block. Anesth Analg 66: 911912 7. Galizia EJ, Lahiri SK (1974) Paraplegia following coeliac plexus block with phenol. Br J Anaesth 46:539540

8. Greenwald HP, Bonica JJ, Berger M (1987) The prevalence of pain in four cancers. Cancer 60: 25632569 9. Gudjonsson B (1987) Cancer of the pancreas. 50 years of surgery. Cancer 60: 22842303 10. Ischia S, Luzzani A, Ischia A, Faggion S (1983) A new approach to the neurolytic block of the coeliac plexus: the transaortic technique. Pain 16: 333341 11. Jones J, Gough D (1977) Celiac plexus block with alcohol for relief of upper abdominal pain due to cancer. Ann R Coll Surg Engl 59: 4649 12. Learned LO, Calhoon RF (1951) Retroperitoneal hemorrhage as a complication of lumbar paravertebral injection: report of three cases. Anesthesiology 12: 391393 13. Lebovits AH, Lefkowitz M (1989) Pain management of pancreatic carcinoma: a review. Pain 36: 111 14. Leung JW, Bowen Wright M, Aveling W, Shorvon PJ, Cotton PB (1983) Coeliac plexus block for pain in pancreatic cancer and chronic pancreatitis. Br J Surg 70: 730732 15. Lieberman RP, Waldman SD (1990) Celiac plexus neurolysis with the modified transaortic approach. Radiology 175: 274276 16. Mattia C, Laurenzi L, Caratozzolo M, Carassiti M, Scardella L, Pinto G (1993) Alcolizzazione percutanea del plesso celiaco con approccio per via anteriore. Minerva Anestesiol 59: 193199 17. Montero Matamala A, Vidal Lopez F, Aguilar Sanchez JL, Donoso Bach L (1989) Percutaneous anterior approach to the celiac plexus using ultrasound. Br J Anaesth 62: 637640 18. Montero Matamala A, Vidal Lopez F, Inaraja Martinez L (1988) The percutaneous anterior approach to the celiac plexus using CT guidance. Pain 34: 285288 19. Owitz S, Koppolu S (1983) Celiac plexus block: an overview. Mt Sinai J Med 50: 486490 20. Sharfman WH, Walsh TD (1990) Has the analgesic efficacy of neurolytic celiac plexus block been demonstrated in pancreatic cancer pain? Pain 41: 267271 21. Thompson GE, Moore DC, Brinderbaugh LD, Artin RY (1977) Abdominal pain and alcohol celiac plexus nerve block. Anesth Analg 56: 15 22. Warshaw AL, Swanson RS (1988) Pancreatic cancer in 1988. Possibilities and probabilities. Ann Surg 208: 541553

Technique
Surg Endosc (1997) 11: 295298

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

A four-point fixation method for the resection of early gastric cancer, with particular reference to the analysis of cases of incomplete resection
M. Tanaka,1 S. Inatsuchi2
1 2

School of Nursing, Faculty of Medicine, Toyama Medical and Pharmaceutical University, 2630 Sugitani, Toyama 930-01, Japan Department of Gastroenterology, Toyama Red Cross Hospital, 1-5-25 Higashidenjigata, Toyama 930, Japan

Received: 23 February 1996/Accepted: 28 May 1996

Abstract. We developed a four-point fixation method for the resection of gastric lesions. The aim of the present study was to compare the characteristics of early gastric cancer between the complete-resection group (56 patients) and the incomplete-resection group (10 patients). The results showed that the incomplete-resection group included a significantly higher proportion of patients with large lesions and patients in whom the lesion was located at the body or lesser curvature of the stomach. These results were similar to those previously reported for treatment with conventional strip biopsy. However, in contrast to conventional strip biopsy, relatively few patients with depressed-type gastric cancer were included in the incomplete-resection group. This is considered one of the most important features of the four-point fixation method. Key words: Four-point fixation method Endoscopic mucosal resection (EMR) Early gastric cancer

With EMR, complete resection of the target lesion on initial treatment is crucial to the outcome of therapy [9]. We analyzed the characteristics of gastric cancer and gastric adenomas able to be resected completely with the initial procedure of the four-point fixation method. We compared those to cancers and adenomas that underwent incomplete resection. This, we feel, will likely contribute to refinement of the method and, in a broader sense, promote the progress of endoscopic therapy for gastric cancer. A discussion of our results and their clinical implications follows.

Materials and methods

Sixty-six patients with gastric cancer limited to the mucosa were studied. Before endoscopic therapy, the histologic diagnosis was established by histologic examination of biopsy specimens. The depth of cancer invasion was determined on the basis of the results obtained by conventional endoscopy and endoscopic ultrasonography of the stomach before endoscopic therapy. Endoscopic features of gastric cancer were classified into type I, type IIa, type IIb, type IIc, type III, and type IIa + IIc [8]. Premedication similar to that used for conventional endoscopy was administered. A two-channel direct-viewing endoscope (model GIF-2T20 or GIF-2T200, Olympus, Tokyo, Japan) was used. The endoscope was inserted into the stomach, and the lesion was observed. The four-point fixation procedure is illustrated in Fig. 1. The region to be resected endoscopically was decided upon, and two different-colored clips were applied to the mucosa, one on the oral side and the other on the anal side of the resection line. The clips had an arm length of only 5.5 mm so that the tips would not reach the deep part of the mucosa. They were covered with an insulated coating so that the surrounding tissue would not be cauterized by the transmission of high-frequency current. Two to five milliliters of physiological saline was injected into the submucosa adjacent to the lesion to induce separation of the mucosa from the underlying tissue. This caused protrusion of the lesion and its surrounding mucosa. The anterior side and posterior side of the lesion were then grasped between the blades of a jumbo grasping forceps. The mucosa around the lesion was thus secured by the two clips and the two blades of the jumbo grasping forceps. Next, a snare forceps (SD-7P, Olympus, Tokyo, Japan) was placed around these four designated regions. The snare forceps was tightened, and a coagulation current was applied to the snare. The lesions with surrounding mucosa were resected from the stomach wall and retrieved, with the two colored clips in place (Fig. 2).

Endoscopic mucosal resection (EMR) is now used extensively for radical therapy in patients with early gastric cancer [4]. However, EMR is not without its drawbacks, and several modifications of strip biopsy [14], the underlying principle of EMR, have been attempted [1, 6, 7, 11]. We developed a modification of strip biopsy, referred to as the four-point fixation method and reported our initial results in 1994 [5]. This new method not only had a significantly higher complete-resection rate for early gastric cancer and gastric adenoma; it also permitted the correct identification of the oral and anal sides of resected materials after retrieval. Subsequent clinical experience has contributed to improved understanding of the features of the fourpoint fixation method.
Correspondence to: M. Tanaka

296

Fig. 1. The four-point fixation method. (a) Indication of the region to be resected; (b) saline injection into submucosal layer; (c) lifting the lesion by grasping forceps; (d) snaring and resection, the snare is fixated at four points; (e) removal of the specimen.

Table 1. Clinical characteristics of the patients Group Complete resection (n 56) Age (years) Sex (M/F) 68.5 9.1 40/16 Incomplete resection (n 10) 72.0 7.6 7/3 p value NS NS

Fig. 2. A resected specimen with the two colored clips. After retrieval, the length of the major axis of the lesion was measured and the tissue was fixed in formalin. The lesion was then sliced into serial sections (thickness, 2.0 mm), and all sections were stained with hematoxylin and eosin. These tissue sections were evaluated histologically to establish the final diagnosis of the lesion, the depth of cancer invasion, and whether or not the lesion was resected completely. This latter evaluation was made according to the histologic criteria proposed by Tada et al. [10]. The 66 patients were divided into a group undergoing complete resection and a group undergoing incomplete resection on the basis of the outcome of initial EMR using the four-point fixation method. These two groups were compared with respect to treatmentrelated factors (single-step resection or piecemeal resection, and the size, endoscopic features, histological type, and site of the lesion). The results were analyzed using 2 test; p < 0.05 was considered to indicate statistical significance.

were similar in terms of the age and sex of the patients. The groups also did not differ with regard to the proportion of patients who underwent single-step resection and the proportion who underwent piecemeal resection. The length of the major axis of the lesion was >20 mm in a significantly higher proportion of patients in the incomplete-resection group. There was no difference between the groups with respect to the endoscopic classification of lesions (type I, type IIa, type IIa + IIc, type IIc) or the histological type (differentiated carcinoma, undifferentiated carcinoma). The percentage of patients in whom the lesion was located at the body of the stomach was significantly higher in the incomplete-resection group, while that of patients in whom the lesion was located at the antrum was significantly higher in the complete-resection group. In addition, the proportion of patients in whom the lesion was located at the lesser curvature was significantly higher in the incomplete-resection group. Discussion The present study demonstrated that there are similarities and differences between patients in whom lesions are resected incompletely with the four-point fixation method and those in whom lesions are resected incompletely with conventional strip biopsy or modified EMR technique. With regard to the resection technique, there was no difference between the complete-resection group and the

Results The results are summarized in Tables 1 and 2. The complete-resection group and the incomplete-resection group

297 Table 2. Characterization of gastric cancer Group Parameter Resection Single step Piecemeal Size 20 mm >20 mm Endoscopic feature Type I Type IIa Type IIc or IIa + IIc Histology Differentiated cancer Undifferentiated cancer Location Cardia Body Angle Antrum Lesser curvature Greater curvature Anterior wall Posterior wall Complete resection (%) 63 (35/56) 37 (21/56) 91 (51/56) 9 (5/56) 13 (7/56) 45 (25/56) 42 (24/56) 91 (51/56) 9 (5/56) 4 (2/56) 28 (16/56) 14 (8/56) 54 (30/56) 30 (17/56) 18 (10/56) 20 (11/56) 32 (18/56) Incomplete resection (%) 60 (6/10) 40 (4/10) 30 (3/10) 70 (7/10) 10 (1/10) 50 (5/10) 40 (4/10) 90 (9/10) 10 (1/10) 10 (1/10) 70 (7/10) 20 (2/10) 0 (0/10) 80 (8/10) 0 (0/10) 0 (0/10) 20 (2/10) p value NS NS <0.0001 <0.0001 NS NS NS NS NS NS <0.05 NS <0.01 <0.01 NS NS NS

incomplete-resection group in the proportion of patients who underwent single-step resection and the proportion who underwent piecemeal resection. Few studies have addressed this issue previously. With the HSE (hypotonic saline-epinephrine solution) technique [7], a modified procedure for conventional strip biopsy, there was similarly no significant difference between these groups. Since the introduction of the four-point fixation method, the frequency of piecemeal resection has increased. With conventional strip biopsy, we performed piecemeal resection of gastric cancers or adenomas in only 12% (2/17) of patients [5]. Since we began to use the four-point fixation method, this figure has increased significantly to 38% (23/66 patients). There are two main reasons for this increase. First, clipping of the mucosa around the lesion with the four-point fixation method has facilitated the performance of piecemeal resection, thereby eliminating the need for single-step resection of difficult-to-manage lesions. Second, the clips remain fixed to the resected materials after retrieval, permitting one to identify the oral and anal sides and the anterior and posterior wall side of the stomach. In the past, piecemeal resection has been discouraged [12]. However, with the advent of the four-point fixation method, piecemeal resection can be recommended in selected cases. An analysis of the morphologic characteristics of the lesions indicated that there were significantly more patients who had large lesions (>20 mm) in the incomplete-resection group. A similar trend has been noted with conventional strip biopsy [13, 15]. These results imply limitations with regard to the size of lesions resectable with the four-point fixation method. There was no difference in the endoscopic classification of lesions between the complete-resection group and the incomplete-resection group. With conventional strip biopsy, the resection rate of depressed-type gastric cancer (type IIc, type IIa + IIc) was reported to be low [3, 9]. Since the height of this type of gastric cancer is

nearly the same as that of the surrounding mucosa, the snare forceps may slip across the mucosal surface when it is placed and tightened at the resection line circumscribing the lesion. The designated area may therefore not be resected as planned. This is the major cause of the low resection rate with conventional strip biopsy. However, with the fourpoint fixation method, the clips attached around the lesion prevent slippage of the snare forceps, which facilitates the complete resection of even depressed-type gastric cancer of virtually the same height as the surrounding mucosa. This mechanism is considered responsible for the higher resection rate of depressed-type gastric carcinomas with the fourpoint fixation method than with conventional strip biopsy. With regard to histological type, there was no difference between the complete-resection group and the incompleteresection group. The resection rate of undifferentiated carcinoma has been reported to be low with conventional strip biopsy [12]. Accurate evaluation of the depth of invasion of undifferentiated carcinoma is often difficult on endoscopic examination. This also applies when using a dissecting microscope to evaluate the depth of invasion of resected lesions [2]. The low resection rate for undifferentiated carcinoma with conventional strip biopsy may therefore result from over- or underestimation of the extent of invasion on endoscopic evaluation before endoscopic therapy. Analysis of the lesion location in the incompleteresection group revealed a significantly higher proportion of patients in whom the lesion was located in the body of the stomach. A low rate of complete resection for lesions by conventional strip biopsy occurring in the body has been reported in numerous studies [3, 9, 13, 15]. We have also found similar trends with conventional strip biopsy [5]. This problem arises because a lesion in the body cannot be viewed in entirety and resected unless the scope is turned 180 (J- or U-turn), or because it is often difficult to obtain a complete view of a lesion even when the scope is turned;

298

in many cases, only a lateral view is attainable. Another potential problem is that grasping and lifting of a lesion with a grasping forceps may be insufficient when the scope tip is too close to or far from the lesion. Since these problems are also encountered with the four-point fixation method, the clinical outcome was similar to that for conventional strip biopsy. In contrast to lesions in the body, the incompleteresection group had a significantly lower incidence of lesions in the antrum. This implies that it is easier to completely resect lesions in the antrum with the four-point fixation method. Another characteristic of the incompleteresection group was a significantly higher incidence of lesions in the lesser curvature. A low complete-resection rate of lesions in the lesser curvature has been reported with conventional strip biopsy in many studies [3, 9, 15]. Similar to the outcome when lesions in the body underwent EMR, this low rate may reflect insufficient grasping and lifting of a lesion with a grasping forceps, which is liable to occur when the scope tip approaches a lesion too closely. This problem cannot be eliminated with the use of the four-point fixation method. Consequently, similar clinical results were obtained with conventional strip biopsy and with the fourpoint fixation method. In summary, patients in whom lesions were resected incompletely with the four-point fixation method tended to have (1) a large lesion size and (2) lesion occurrence in the body and lesser curvature. The former suggests that only lesions up to a certain size can be resected with the fourpoint fixation method. The latter finding implies that it is difficult to correctly orient the scope at the body and lesser curvature, irrespective of the procedure used for EMR, because the therapeutic outcome was similar to that obtained with conventional strip biopsy. It should be noted that a high incidence of depressed-type gastric cancer among patients in whom resection was incompletea problem with conventional strip biopsydid not occur with the four-point fixation method. This is one of the most important features of this procedure. With the use of the four-point fixation method, further enhancement of the complete-resection rate of lesions located at the body and lesser curvature of the stomach may be difficult to achieve because of the problems involved in correct scope placement for EMR. During treatment the endoscopist should attempt to change the spatial relations between the scope tip and the lesion by adjusting the insufflation of the stomach, altering the patients position, and using an oblique- or lateral-viewing endoscope instead of a direct-viewing scope.

References
1. Fujimori T, Nakamura T, Hirayama D, Satonaka K, Ajiki T, Kitazawa S, Maeda S, Nagasako K, Yamaguchi H, Yoshida S (1992) Endoscopic mucosectomy for early gastric cancer using modified strip biopsy. Endoscopy 24: 187189 2. Fujimori T, Hirayama D, Satonaka K, Nakamura T, Yukawa M, Tabata T, Teramoto T, Kitazawa S, Maeda M, Ishikawa Y, Kaji T, Tamada F, Nishitani K, Tsukamoto T, Tabata B, Kuroda D, Kashiwagi K, Kato M, Saito Y (1991) Limitation of dissecting microscopy for the diagnosis of the extent of the spread of cancer cellsa fundamental study on endoscopic complete resection of gastric carcinoma. Stomach Intest 26: 301310 3. Hamada T, Kondo K, Ota H, Takazoe M, Ukai T, Nishida J, Itagaki Y, Inoue N, Kitayama S, Watanabe H, Murai T, Shirakabe H (1993) Cooperation of endoscopic resection of early gastric cancer and surgery. Prog Dig Endosc 42: 2225 4. Hiki Y, Shimao H, Mieno H, Sakakibara Y, Kobayashi N, Saigenji K (1995) Modified treatment of early gastric cancer: evaluation of endoscopic treatment of early gastric cancers with resect to treatment indication groups. World J Surg 19: 517522 5. Inatsuchi S, Tanaka M (1994) Clinical evaluation of improved technique in strip biopsy for gastric lesion. Gastroenterol Endosc 36: 939948 6. Inoue H, Takeshita K, Hori H, Muraoka Y, Yoneshima H, Endo M (1993) Endoscopic mucosal resection with a cap-fitted panendoscope for esophagus, stomach, and colon mucosal lesions. Gastrointest Endosc 39: 5862 7. Nakamura M, Ishibashi T, Taniguchi H, Yamamoto H, Nakajima Y, Umezu H, Chiba K, Hoshino K, Kimura K, Ishii K, Nakamura Y, Umetani K (1991) Evaluation of endoscopic resection of gastric tumors following local injection of HSE (hypotonic saline-epinephrine solution). Dig Endosc 3: 475484 8. Okajima K (1993) Macroscopic classification of gastric cancer. In: Japanese research society for gastric cancer (ed) The general rules for the gastric cancer study. Kanehara-syuppan, Tokyo, pp 1821 9. Tada M, Matsumoto Y, Murakami A, Nishizaki Y, Karita M, Yanai H, Okita K (1993) Problems and their solution in curative endoscopic resection of early gastric carcinomas. Endosc Dig 5: 11691174 10. Tada M, Karita M, Yanai H, Takemoto T (1988) Evaluation of endoscopic strip biopsy therapeutically used for early gastric cancer. Stomach Intest 23: 373385 11. Takechi K, Mihara M, Saito Y, Endo J, Maekawa H, Usui T, Moriwaki H, Muto Y (1992) A modified technique for endoscopic mucosal resection of small early gastric carcinomas. Endoscopy 24: 215217 12. Takekoshi T, Fujii A, Baba Y, Takemoto N, Kaku Y, Shimizu H, Tomimatsu H, Koizumi K, Kamei A, Hase Y, Akamatsu M, Ogawa T, Yonemura H, Ogata E, Ohta H, Nishi M, Kato H, Yanagisawa A (1993) Endoscopic resection of early gastric cancerspossibilities and limitations based on long term follow-up. Clin Gastroenterol 8: 649667 13. Takekoshi T, Baba Y, Ohta H, Kato Y, Yanagisawa A, Takagi K, Noguchi Y (1994) Endoscopic resection of early gastric carcinoma: Results of a retrospective analysis of 308 cases. Endoscopy 26: 352358 14. Takemoto T, Tada M, Yanai H, Karita M, Okita K (1989) Significance of strip biopsy, with particular reference to endoscopic Mucosectomy. Dig Endosc 1: 49 15. Yokoyama Y, Ishii K, Tanabe S, Koizumi W, Mitsuhashi T, Ooida M, Saigennji K, Shimao H, Mieno H, Hiki Y, Kan K, Atari E (1992) Course of endoscopic resection for early gastric cancer. Prog Dig Endosc 41: 119128

Surg Endosc (1997) 11: 230234

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

The safety of helium for abdominal insufflation

R. Y. Declan Fleming,1 T. B. Dougherty,2 B. W. Feig1
1 2

Department of Surgical Oncology, The University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA Department of Anesthesiology, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA

Received: 20 March 1996/Accepted: 19 July 1996

Abstract Background: A search for alternative methods of abdominal insufflation has been prompted by the fact that CO2 insufflation may cause acidosis, decreased cardiac output, increased systemic vascular resistance, and increased cardiac filling pressures. This study evaluates the safety and the cardiopulmonary effects of helium abdominal insufflation (HAI). Methods: Thirteen ASA class III and IV patients undergoing laparoscopic procedures were studied in a prospective, nonrandomized protocol using HAI. Cardiopulmonary parameters were measured before and after anesthetic induction and every 30 min during HAI. Abdominal insufflation pressure was initially 10 mmHg and was increased to 15 mmHg after 30 min. All measurements were repeated 15 min after deflation of the abdomen. Changes were evaluated by ANOVA. Results: No significant cardiopulmonary complications were observed. No patient developed hypercarbia or acidosis. Peak inspiratory pressure increased with HAI from 20 1 to 34 2 cm H2O (p < 0.0001). Cardiac index decreased (3.35 0.19 vs 2.37 0.19 l/min/m2; p 0.0303) and systemic vascular resistance increased (1,123 66 vs 1,406 126 dyne s/cm5; p 0.0512) while cardiac filling pressures increased with insufflation to 15 mmHg. Conclusions: Minimal cardiac and pulmonary aberrations were observed. Helium was safe for abdominal insufflation and may be the insufflating agent of choice in patients with significant cardiopulmonary disease. Key words: Helium insufflation Laparoscopy Cardiopulmonary changes

Laparoscopic surgical techniques have become important diagnostic and therapeutic tools of the general surgeon in the 1990s. As many as 80% of the cholecystectomies perCorrespondence to: B. W. Feig

formed in the United States today are accomplished laparoscopically. Many surgeons have begun to apply laparoscopic techniques to as broad a range of procedures as possible, citing potential benefits of smaller incisions, decreased postoperative pain, shortened periods of hospitalization, and increased cost effectiveness [20]. Public acceptance of and demand for minimally invasive surgery has caused the clinical application of laparoscopy to advance far beyond any scientific evaluation of the physiologic effects (either deleterious or beneficial) of this technique. To date, the vast majority of laparoscopic surgeries have been low-risk, elective procedures performed in young, healthy patients. However, advances in laparoscopy-related technology and surgeons experience with laparoscopy have led to the use of laparoscopy for increasingly more complex and lengthy procedures in a broader patient population. Advanced laparoscopic procedures frequently last 1 or more h and these patients not infrequently may have compromised cardiopulmonary function. Several reports have documented the development of hypercarbia and subsequent acidosis in patients undergoing carbon dioxide (CO2) pneumoperitoneum [12, 22]. Unfortunately there has been little scientific evaluation of the perioperative and/or long-term consequences of these physiologic changes. As information has increased about potential risks of CO2 abdominal insufflation (CAI) during laparoscopic surgery, a search for other methods of obtaining visualization of the peritoneal cavity has been initiated. Both gasless laparoscopy employing mechanical lifting devices and use of other insufflating gases have been examined. Recently, three reports have examined the use of helium as the insufflating gas for laparoscopic procedures [2, 14, 17]. These studies have demonstrated some diminution of the hypercarbia associated with CAI. The majority of the patients studied, however, have been healthy patients undergoing uncomplicated laparoscopic cholecystectomy. The purpose of this study was to examine the respiratory and hemodynamic effects of helium abdominal insufflation (HAI) for laparoscopic surgery in a group of higher-risk surgical patients, all of whom possess significant concomitant cardiac and/or pulmonary disease.

231 Table 1. Patient characteristics and laparoscopic procedures Patient no. 1 2 3 4 5 6 7 8 9 10 11 12 13 Gender M M M M M M F M M F F M F Age (years) 65 74 71 67 53 73 73 72 73 70 81 39 71 Significant medical problems Lymphoma s/p CHOP/Bleomycin chemo; cardiomyopathyEF 51%; 1 AV block NIDDM; COPD; sigmoid colon cancer Parkinsons disease; HTN; carotid arterial stenosis; sigmoid colon cancer NIDDM; HTN; colon cancer and multiple colonic polyposis COPD; IDDM; CADMI 1 yr prior; 100% occlusion of RCA; 50% EF; LAFB gastric cancer Rectal carcinoma s/p preop Chemo/XRT; HTN; LAFB; anemia; smoking > 50 pk-yr Colon cancer; viral cardiomyopathyEF 45%; angina at rest; heart block with V-demand pacing; Goldman class III Rectal cancer s/p preop chemo/XRT; seizure disorder; NIDDM; 25 pk-yr smoking; PVCs Colon cancer; CAD s/p CABG; NIDDM; HTN Colonic polyposis; esophageal leiomyoma; poorly controlled IDDM; HTN; CAD with stable angina; s/p MI & PTCA; EF 45% Transverse colon carcinoma; multiple descending and sigmoid colon polyps; 30+ pk-yr smoking; LVH AML s/p BMT with GVH disease; pulmonary fibrosis; septicemia from Salmonella cholecystitis; Colon cancer American Society of Anesthesiology class IV III III III IV III IV III III IV III III III Laparoscopic procedure Mesenteric lymph node biopsy Sigmoid colectomy Low anterior resection Total abdominal colectomy Staging laparoscopy for gastric cancer Low anterior resection Right hemicolectomy Low anterior resection Right hemocolectomy Left hemicolectomy Diagnostic laparoscopy with hepatic ultrasound Cholecystectomy Sigmoid colectomy

Methods
All patients classified as American Society of Anesthesiology (ASA) class III or IV scheduled to undergo a laparoscopic procedure were considered as candidates in this prospective, nonrandomized protocol employing helium as the insufflating agent. Prior to the initiation of the study, the proposal was reviewed and approved by the Institutional Review Board of the University of Texas M.D. Anderson Cancer Center. Informed consent was obtained from all participants. All patients underwent placement of pulmonary artery and radial arterial catheters prior to the initiation of anesthesia. Hemodynamic parameters were measured and cardiac parameters were normalized to achieve a systemic vascular resistance (SVR) of less than 1200 dyne s/cm5, cardiac index (CI) greater than 2.1 l/min/m2, and pulmonary capillary wedge pressure between 8 and 12 mmHg. Normalization of these parameters was achieved with the intravenous administration of fluid with or without nitrates and/or -adrenergic agonists. Baseline hemodynamic and blood gas values were then obtained just prior to induction of general anesthesia. General anesthesia was induced with etomidate and maintained with isoflurane, sufentanil, and vecuronium bromide. Paralysis was monitored by twitch response and by observation of the end-tidal CO2 (ETCO2) waveform to ensure that the patients were not breathing spontaneously between mechanical ventilations. Five minutes after induction of anesthesia, hemodynamic and blood gas values were again obtained. Pneumoperitoneum by helium abdominal insufflation (HAI) to an intraabdominal pressure of 10 mmHg was accomplished and maintained with an automated servoinsufflator with intraabdominal pressure being monitored continuously. Hemodynamic and blood gas measurements were obtained at 5 and 30 min after HAI to 10 mmHg; intraabdominal pressure was then increased to 15 mmHg and measurements were again taken at 5 and 30 min following the pressure change. Recordings were then made at 30-min intervals until the end of HAI. Once HAI was complete, a final set of data was obtained 15 min after deflation of the abdomen. Hemodynamic variables that were measured included heart rate (HR); systolic, diastolic, and mean arterial blood pressure (SBP, DBP, MBP); SVR; CI; pulmonary capillary wedge pressure (PCWP); and central venous

pressure (CVP). Pulmonary data that were recorded included respiratory rate (RR); tidal volume (TV); minute ventilation (MV); peak inspiratory pressure (PIP); ETCO2; and oxygen content, oxygen saturation, carbon dioxide content, bicarbonate content, and pH of both arterial and mixedvenous blood (PaO2, PvO2, PaCO2, PvCO2, aHCO3, vHCO3, apH, vpH, SaO2, SvO2). Statistical analysis was performed by analysis of variance (ANOVA) for repeated measures with significance determined at a 95% confidence interval. Where significant differences were discovered by ANOVA, comparisons between measurement points were performed using the Fishers protected least significant difference test.

Results Thirteen patients were studied and there were no deviations from the protocol. Average age of the patients was 68 years (range 39 to 81 years), and all of the patients were ASA class III or IV. The characteristics of the patients and the laparoscopic procedures are presented in Table 1. All patients tolerated HAI without significant perioperative cardiac or pulmonary complications. There was no operative or perioperative mortality. Average length of HAI was 102 min. No patients required inotropes; however, three patients received intravenous nitroglycerin (NTG) infusions during their operative procedures. In all cases the NTG infusion was initiated prior to the creation of pneumoperitoneum and continued, without interruption or alteration of dose, through the operation and into the postoperative period. Tables 2 and 3 record the maximal physiologic changes (representing the greatest deviation from baseline readings) that were observed during HAI.

232 Table 2. Hemodynamic parameters in 13 patients with helium insufflation for laparoscopic surgerya Preanesthetic Pulse BP (mean) CVP (cm H2O) PCWP (cm H2O) SVR (dyne s/cm5) CI (l/min/m2)
a

Induction 66 3* 83 3*** 8.2 0.8 11.4 0.9 1,217 104 2.69 0.23

10 mmHg 70 2* 88 4 13.5 1.0*1 15.6 1.1*2 1,259 88 2.81 0.29

15 mmHg 70 2** 79 4*** 13.6 2.0*1 16.0 2.0*1 1,406 126* 2.37 0.19*

Deflation 80 3 83 3*** 10.3 0.9 14.2 0.9 992 95 3.28 0.30

78 3 98 5 8.2 0.8 12.2 0.6 1,123 66 3.35 0.19

Values listed are average SEM; 15-mmHg value includes the extreme value recorded during the 15-mmHg insufflation period (minimum or maximum value depending on the direction of deviation from the baseline value). * p < 0.05 vs Preanesthetic and Deflated by Fishers PLSD. ** p < 0.05 vs Deflated by Fishers PLSD. *** p < 0.05 vs Preanesthetic by Fishers PLSD. *1 p < 0.05 vs Preanesthetic and induction by Fishers PLSD. *2 p < 0.05 vs Induction by Fishers PLSD. Table 3. Respiratory parameters in 13 patients with helium insufflation for laparoscopic surgerya Induction 10 mmHg 28.5 2.0* 5.04 0.27 35.5 1.0 31.4 0.8 200 25 25.0 0.5 7.46 0.01 77.4 1.1 15 mmHg 33.8 1.9* 4.96 0.29 36.5 0.9 30.7 0.6 175 23 24.3 0.5 7.44 0.01 78.0 1.7 Deflation 23.5 1.2 5.19 0.28 36.5 1.0 31.7 0.8 162 21 24.6 0.5 7.43 0.01 80.1 1.3 p <0.0001 0.7694 0.8593 0.2368 0.0036 0.4447 0.2669 0.2377

PIP (cm H2O) MV (l/min) PaCO2 (mmHg) ETCO2 (mmHg) PaO2 (mmHg) HCO3 (mg/dl) pH SvO2 (%)
a

20.0 1.4 5.36 0.31 36.1 1.1 32.8 0.6 309 43** 25.5 0.5 7.46 0.01 81.0 1.4

Values listed are average SEM; p value by ANOVA; 15-mmHg value includes the extreme value recorded during the 15-mmHg insufflation period (minimum or maximum value depending on the direction of deviation from the baseline value). * p < 0.05 vs Induction and Deflated by Fishers PLSD. ** p < 0.05 vs 10 mmHg, 15 mmHg, and Deflated by Fishers PLSD.

Cardiac variables Table 2 summarizes the mean hemodynamic changes that were observed for the entire group of 13 patients. The values expressed for the 15 mmHg helium insufflation period represent those values that deviated to the greatest degree from the baseline measurement. Introduction of anesthesia resulted in significant decreases in pulse and blood pressure with only slight alteration in SVR and CI. (These changes did not reach significance). As intraabdominal pressure was increased by HAI to 10 mmHg, the CVP and PCWP increased significantly. Additionally, there was an increase in the SVR and a concomitant decrease in the CI. This trend, however, did not reach statistical significance until the intraabdominal pressure was increased to 15 mmHg. At that point CI fell to 71% of baseline (p 0.03) and SVR increased 25% above baseline readings. Upon desufflation, values for CI and SVR returned to near-baseline level. CVP and PCWP levels reduced with desufflation but remained significantly elevated compared to baseline readings. No patient experienced cardiac arrhythmia or EKG changes that might have necessitated a reduction of insufflation pressure or the termination of laparoscopy with conversion to an open operation. Pulmonary variables No preanesthetic respiratory parameters were included as many of the measurements required that the patient be in-

tubated and on a ventilator to obtain reproducible recording of these values. PIP was observed to increase 65% over baseline values (p < 0.0001 by ANOVA) with HAI to 15 mmHg. HAI did not result in significant hypercarbia or acidosis at any measured time point. This was without any alteration in minute ventilation to maintain these normal arterial CO2, HCO3, and pH values. PaO2 was reduced from 300 torr just following induction of anesthesia to levels of 200 torr as helium insufflation was accomplished. This reduction in PaO2 likely represents an adjustment of the mixture of the inhalation gases by the anesthesiologist after completion of the anesthetic induction. Deflation of the abdomen did not significantly alter arterial oxygenation, indicating that the reduction in PaO2 likely was not caused by HAI but rather reflects the equilibration of blood oxygenation after the patient is no longer breathing 100% oxygen. Mixed venous oxygen saturation, an indicator of oxygen transport and extraction, did not change significantly at any point during the procedures, implying that the patients had no oxygen delivery or extraction difficulties. Results of the effect of HAI on PIP, MV, PaCO2, ETCO2, PaO2, HCO3, pH, and SvO2 are summarized in Table 3. Discussion Minimally invasive procedures have become increasingly popular. Laparoscopy is currently being applied to a variety of procedures ranging from herniorrhaphy [15] to pancreatic

233

resection [7]. All this has taken place rapidly, with relatively little study of the physiologic effects of CAI preceding the broad application of the techniques. Even in early studies which involved young, healthy women undergoing elective gynecologic procedures which lasted for a short period of time, there were indications that CAI might cause significant derangement from normal physiology [12, 14, 17]. Laparoscopy requires some method of expanding the peritoneal cavity to provide adequate visualization and exposure. Air insufflation was used initially to create pneumoperitoneum but was discontinued in favor of CAI following the introduction of incandescent illumination [8]. Carbon dioxide was ultimately selected because it was chemically inert, colorless, inexpensive, readily available, and did not support combustion. Carbon dioxide instilled in the abdominal cavity is rapidly absorbed into the bloodstream across the peritoneum [18]. The majority of the CO2 is then eliminated by the lungs, but the sustained elevation of blood CO2 levels associated with CAI likely cannot all be eliminated by hyperventilation alone, and thus visceral storage sites for the excess CO2 are recruited [3]. Concomitant to the increased absorption of CO2 during CAI is an increase in the pulmonary dead space [4]. Both of these effects may lead to the development of hypercarbia and subsequently to acidosis. The physiologic effects of hypercarbia and acidosis are well documented. Hypercarbia stimulates the sympathetic nervous system, stimulating tachycardia and vasoconstriction. Acidosis causes vasodilation, which may counteract some of the vasoconstriction. The combination of acidosis and hypercarbia, however, results in a decrease in myocardial contractility and a lowering of the arrhythmia threshold [19]. Studies in both experimental animals and in patients with preexisting cardiopulmonary disease undergoing laparoscopic cholecystectomy have revealed significant alterations in hemodynamic parameters with CAI. Ishizaki et al. correlated adverse hemodynamic effects in dogs with increasing intraabdominal pressure [9]. They demonstrated a significant reduction in cardiac output with a corresponding increase in systemic vascular resistance as intraabdominal pressure was increased from 12 to 16 mmHg. Safran et al. likewise found significant increases in SVR and decreases in CI upon insufflation of the abdomen to 15 mmHg [21]. Our group has previously demonstrated that these changes could be overcome by the use of intravenous nitroglycerine and that longer, complex laparoscopic surgical procedures could be completed safely even in high-risk patients [4]. Concerns about hypercarbia and acidosis from CO2 insufflation have led to an examination of the use of alternate gases to create pneumoperitoneum. The ideal insufflating agent would be inert, nonflammable, colorless, easily handled, and readily available, as well as inexpensive. Nitrous oxide (N2O) meets many of these criteria; however, it can support combustion and would therefore be unsuitable for use with electrocautery. Initial results testing helium insufflation in animals was encouraging [6, 10] in that pneumoperitoneum by HAI did not produce the respiratory acidosis associated with CAI and helium would not allow combustion in the presence of electrocautery or laser coagulation. Subsequently, HAI was used in clinical trials. Studies by Bongard et al. [2], McMahon et al. [14], and Neuberger

et al. [17] found that the hypercarbia and acidosis associated with CAI was ameliorated by the use of helium as the insufflating agent. Only Bongard et al. examined any hemodynamic parametersthey found that pulse and blood pressure increased regardless of the gas used for insufflation and that there was no significant change in cardiac output for either group. None of the patients in Bongards study and only a minority of patients in the studies of McMahon and Neuberger had significant comorbid conditions. This study represents only the fourth report of the use of HAI for laparoscopy in humans and the only study to exclusively examine high-risk patients to assess the safety and cardiopulmonary effects of HAI. We found that HAI could be safely employed for laparoscopic surgery even in high-risk surgical patients. No patient suffered any significant intraoperative or postoperative complication related to helium insufflation. Furthermore, all patients tolerated insufflation at 15 mmHg for at least 30 min without requiring reduction in the intraabdominal pressure due to the development of hypercarbia, acidosis, or arrhythmia. This contrasts with a previous study from our institution in which four of 15 patients required a reduction in CO2 insufflation pressure from 15 to 10 mmHg for persistent hypercarbia and acidosis. Despite a significant increase in PIP, PaCO2 and pH remained constant in this group of patients. No alterations in MV were necessary to maintain normocarbia and a normal pH, indicating that helium insufflation for laparoscopy alleviated the hypercarbia and acidosis in this group of patients that may be associated with CAI. This finding is in concurrence with the findings of the previous studies examining HAI in humans [2, 14, 16]. Further, this indicates that increased intraabdominal pressure along is not sufficient to cause the respiratory alterations sometimes seen in CAI. Significant alterations in hemodynamic parameters in this group of patients included elevations in CVP and PCWP (5 and 4 mmHg, respectively) above their baseline preanesthetic values. These increases persisted upon deflation of the abdomen and appear to be a function of the intraoperative fluid status of these patients. All patients (n 3) who received intravenous NTG were started on the NTG infusion prior to the initiation of helium pneumoperitoneum. All of these patients had preexisting cardiac disease, and prophylactic NTG infusion was a part of the preoperative anesthetic plan. None of these patients required a change in the dose of their NTG with either the initiation of helium pneumoperitoneum or with stepwise increases in HAI pressures. A major concern of many authors regarding the use of HAI has been the risk of gas embolization during laparoscopy. In particular, helium is highly diffusible, but it has an extremely low solubility in water. Therefore, embolization of helium could have potentially greater side effects than embolization of a gas with greater water solubility, like CO2. Although the reports of the use of helium in humans are limited in number (45 patients in the English literature, including this report), there are currently no reports documenting helium embolization during laparoscopy [1, 5, 14, 17]. In laparoscopic procedures with a high risk of embolization (e.g., laparoscopic liver resection), helium may not be the insufflating agent of choice. Our study indicates that helium insufflation for laparo-

234

scopic surgery is a safe alternative to CO2 insufflation for the creation of pneumoperitoneum. Adverse blood-gas changes that may be associated with CAI were eliminated by the use of helium and no patient suffered an adverse consequence from helium insufflation. Helium may be the agent of choice for abdominal insufflation in high-risk surgical patients with significant cardiopulmonary disease owing to its decreased perturbation of cardiopulmonary and acidbase status as compared to conventional CO2 insufflation. Further prospective, randomized studies are needed to examine the use of other sources to visualize and explore the peritoneal cavity (e.g., argon or other gas insufflation or mechanical lifting devices). References
1. Bongard FS, Pianim NA, Se-Yuan, Lippmann M, Davis I, Klein S (1991) Using helium for insufflation during laparoscopy. JAMA 266: 3131 2. Bongard FS, Pianim NA, Leighton TA, Dubecz S, Davis IP, Lippmann M, Klein S, Liu SY (1993) Helium insufflation for laparoscopic operation. Surg Gynecol Obstet 177: 140146 3. Farhi LE, Rahn H (1960) Dynamics of changes in carbon dioxide stores. Anesthesiology 21: 604614 4. Feig BW, Berger DH, Dougherty TB, Dupuis JF, Bartholomew H, Hickey RC, Ota DM (1994) Pharmacologic intervention can reestablish baseline hemodynamic parameters during laparoscopy. Surgery 116: 733741 5. Fernandez-Cruz L, Saenz A, Taura P, Benarroch G, Nies C, Austudillo E (1994) Pheochromocytoma: laparoscopic approach with CO2 and helium pneumoperitoneum. End Surg 2:300304 6. Fitzgerald SD, Andrus CH, Baudendistel LJ, Dahms TE, Kaminski DL (1992) Hypercarbia during carbon dioxide pneumoperitoneum. Am J Surg 163: 186190 7. Gaigner M (1994) Laparoscopic pancreatic resection. Presented at the Society of Surgical Oncology, Houston, TX 1720 March 8. Gunning JE, Rosenzeig BA (1991) Evolution of endoscopic surgery.

9.

10.

11. 12.

13.

14.

15.

16.

17.

18.

19. 20. 21. 22.

In: White RA, Klein SR (eds) Endoscopic surgery. Mosby Year Book Medical, St Louis, pp 159169 Ishizaki Y, Bandai Y, Shimomura K, Abe H, Ohtomo Y, Idezuki Y (1993) Safe intraabdominal pressure of carbon dioxide pneumoperitoneum during laparoscopic surgery. Surgery 114: 549545 Leighton TA, Bongard FS, Liu SY, et al. (1991) Comparative cardiopulmonary effects of carbon dioxide versus helium pneumoperitoneum. Surg Forum 62: 485487 Lenz RJ, Thomas TA, Wilkins DG (1976) Cardiovascular changes during laparoscopy. Anaesthesia 31: 412 Liu SY, Leighton T, Davis I, Stanley K, Lippmann M, Bongard F (1991) Prospective analysis of cardiopulmonary responses to laparoscopic cholecystectomy. J Laparoendosc Surg 1: 241246 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 McMahon AJ, Baxter JN, Murray W, Imrie CW, Kenny G, ODwyer PJ (1994) Helium pneumoperitoneum for laparoscopic cholecystectomy: ventilatory and blood gas changes. Br J Surg 81: 10331036 Winchester DJ, Dawes LG, Modeleski DD, et al. (1993) Laparoscopic inguinal hernia repair. A preliminary experience. Arch Surg 128: 781 786 Motew M, Ivankovic AD, Bienearz J, Zahed B, Scommegna A (1973) Cardiovascular effects and acid-base and blood gas changes during laparoscopy. Am J Obstet Gynecol 115: 10021012 Neuberger TJ, Andrus CH, Wittgen CM, Wade TP, Kaminski DL 91996) Prospective comparison of helium versus carbon dioxide pneumoperitoneum. Gastrointest Endosc 43: 3841 Piiper J (1965) Physiologic equilibria of gas cavities in the body. In: Fenn O, Rahn SM (eds) Handbook of physiology, vol II, respiration. American Physiological Society, Washington, DC, pp 12051218 Rasmussen JP, Dauchot PJ, DePalma RG, et al. (1978) Cardiac function and hypercarbia. Arch Surg 113: 11961200 Reddick EJ, Olsen DO (1990) Outpatient laparoscopic laser cholecystectomy. Am J Surg 160: 485488 Safran D, Sgambati S, Orlando R (1993) Laparoscopy in high-risk cardiac patients. Surg Gynecol Obstet 176: 548554 Wittgen CM, Andrus CH, Fitzgerald SD, Baudenstel LJ, Dahms TE, Kaminski DL (1991) Analysis of the hemodynamic and ventilatory effects of laparoscopic cholecystectomy. Arch Surg 126: 9971001

Surg Endosc (1997) 11: 303307

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic tension-free repair of large paraesophageal hernias

M. G. Paul, R. P. DeRosa, P. E. Petrucci, M. L. Palmer, S. H. Danovitch

Abstract The paraesophageal hernia is an unusual disorder of the esophageal hiatus that may be associated with lifethreatening mechanical problems. Elective repair is recommended at the time the condition is diagnosed, and open surgery can be accomplished with a low incidence of complications. The option of performing these repairs through a laparoscopic approach may further reduce morbidity and recovery time associated with surgical intervention. The purpose of this report was to review available options for laparoscopic repair and to present our experience with a tension-free technique for large paraesophageal hernias. Three patients with large diaphragmatic defects had laparoscopic repairs using an expanded polytetrafluorethylene (PTFE) patch secured with intracorporeal suturing techniques. One of these patients also underwent laparoscopic Toupet fundoplication in conjunction with repair of the hernia. In the other two patients, the fundus was secured to the right diaphragmatic crus to reduce the potential for recurrence and minimize postoperative reflux symptoms. All patients underwent successful repair without perioperative complications and had excellent long-term results. Laparoscopic repair of paraesophageal hernias can be accomplished by a number of different reported techniques. The use of a tension-free repair with PTFE may be particularly suitable for large diaphragmatic defects. An antireflux operation may be added selectively depending on clinical circumstances. Key words: Paraesophageal hernia Laparoscopic surgery Fundoplication

Paraesophageal hernias represent 5% of all hernias through the esophageal hiatus [21]. The more common sliding hiatal hernia (type I) involves herniation of the cardia and gastroesophageal junction into the chest and is often associated with reflux esophagitis. Paraesophageal hernias (type II)

Correspondence to: M. G. Paul, Present address: 3301 New Mexico Ave, N.W., Washington, DC 20016, USA

occur when the fundus is the lead point of herniation, with the gastroesophageal junction maintained below the diaphragm by the phrenoesophageal ligament. Reflux is distinctly less common in these pure paraesophageal hernias, although combined forms (type III) may occur. Postprandial discomfort with substernal fullness and belching are the most common symptoms [17]. Thirty percent of patients present with gastric ulceration and hemorrhage in the herniated segment [12]. Bleeding complications are a consequence primarily of impaired venous drainage as well as a direct constricting effect of the rigid hiatal margins on the wall of the stomach. The most serious complication is organoaxial gastric volvulus, which may lead to gangrene and perforation. Patients who present with this latter complication have an operative mortality as high as 100% [23]. Expectant management of paraesophageal hernias was unsuccessful due to progression of symptoms in 45% of patients [25]. In another study of 21 patients followed longitudinally, 27% of patients treated with observation eventually died from complication of their hernia [23]. Surgical treatment is therefore indicated in all cases unless there are overwhelming medical contraindications [27]. Ellis et al. [11] reported a series of 116 patients undergoing open surgical repair of paraesophageal hernias. Anterior crural repair using the Collis technique was performed in all cases. A gastrostomy was also performed to help fix the stomach below the diaphragm and protect against recurrence. Symptoms were relieved in over 80% of patients. Despite unanimity on the need for surgical repair and the fact that open operation may be accomplished with a high rate of success, the pain and disability associated with recovering from a large upper abdominal incision continued to be a deterrent to the referral of patients for surgical treatment. Associated medical conditions often lead to delays in recommending surgical therapy, until life-threatening mechanical complications occur. The advent of laparoscopic surgery now seems to be changing the perception of patients and referring physicians toward surgery. The same principles of open surgical repair may be applied through a minimal access technique, allowing the laparoscopic approach to become the preferred technique for gastroesophageal surgery. This report reviews techniques currently available for the laparoscopic repair of paraesophageal hernias and examines some of the persistent controversies in this field. We also describe our experience

304

with a tension-free procedure that may be suitable for the repair of particularly large paraesophageal hernias. Materials and methods Patient 1
A 65-year old dentist with a history of recurrent upper gastrointestinal bleeding was diagnosed as having a large paraesophageal hernia. He also had marked reflux symptoms in the recumbent position. Esophageal manometry revealed a decreased lower esophageal sphincter (LES) tone and decreased peristalsis in the esophageal body. Laparoscopic reduction of the hernia was performed, along with a Toupet fundoplication and expanded polytetrafluorethylene (PTFE) repair of the 10-cm diaphragmatic defect. The postoperative course was uneventful, and he was discharged on the 4th postoperative day, tolerating a regular diet. He was asymptomatic at 14month follow-up exam. Upper endoscopy and contrast studies done at that time showed no evidence of any recurrence.

Patient 2
A 79-year-old female on steroids for Crohns disease was found on upper gastrointestinal contrast series to have a large paraesophageal hernia (Fig. 1). Symptoms included intermittent chest pain and upper gastrointestinal bleeding, necessitating several emergency admissions. Elective laparoscopic repair was performed, consisting of reduction of the stomach and expanded PTFE patch repair of a 10-cm diaphragmatic defect. She was discharged on the 2nd postoperative day on a regular diet. There were no postoperative complications, and, at 13-month follow-up, the patient was entirely asymptomatic with a normal hematocrit.

then carried out at this stage. Depending on the adequacy of esophageal peristalsis as determined by preoperative manometric studies, either a 270 Toupet fundoplication or a conventional Nissen fundoplication is performed. Laparoscopic techniques for these procedures are well described elsewhere [24]. A 5 10 cm piece of expanded polytetrafluoroethylene (Gore-Tex soft tissue patch, W. L. Gore and Assoc., Flagstaff, AZ), 1 mm in thickness, is further cut to conform to the size of the diaphragmatic defect. The patch is then inserted through an 11-mm port and positioned over the defect. Two 25-cm lengths of CV-2 PTFE suture are used to suture the patch to the diaphragmatic crura. The patch is first secured at its lower edges, and then the sutures are advanced in a running fashion and tied together at the upper portion of the patch (Fig. 3). All sutures are tied using intracorporeal knotting techniques. PTFE is ideally suited for a running continuous suture as it resists the fraying often seen with synthetic braided or silk sutures when handled repeatedly by grasping instruments. The fundus is at this point extremely mobile, theoretically still prone to reherniation, and must be secured below the diaphragm. This is accomplished by securing the fundus to the right crus using two to three interrupted sutures of 2-0 silk (Fig. 4). This maneuver also serves to recreate the angle of His, similar to a Dor fundoplication [24], reducing any reflux that may theoretically occur as a consequence of division of the phrenoesophageal membrane. This step is omitted if either a Toupet or Nissen fundoplication has been performed. A nasogastric tube is kept in place until the morning after surgery. The diet is advanced as tolerated, and patients are given a 10-day course of prokinetic agents such as cisapride (Propulsid, Janssen Pharmaceutica Inc., Titusville, NJ) to temporarily aid in gastric emptying.

Discussion Cuschieri et al. reported one of the first series of patients with large paraesophageal hernias treated through a laparoscopic approach [8]. He described eight patients whose diaphragmatic defects were closed with a running suture of 2-0 silk. A Nissen fundoplication was added in all eight cases. Follow-up at three months showed excellent functional results. Koger and Stone [14] performed a laparoscopic reduction alone in acutely ill patients with no attempt to repair the defect. Congrave [6] reported the laparoscopic repair of two paraesophageal hernias with primary suture closure of the diaphragmatic defect in conjunction with anterior gastropexy and gastrostomy. Oddsdottir et al. [20] reported ten patients undergoing laparoscopic repair of paraesophageal hernia with excellent results. In all cases the diaphragmatic defect was closed with pledgeted, non absorbable horizontal mattress sutures. A short, floppy Nissen fundoplication was performed in nine patients. Postoperative complications were few and patient satisfaction with the operation at 3 months was near universal. One patient was noted to have a herniated wrap on barium swallow with mild dysphagia and heartburn. The simplicity and apparent durability of primary suture repair of the hiatus would certainly seem to make this the technique of choice in most instances of small to moderatesize diaphragmatic defects. Larger defects, however, may not be suitable for primary repair, as the tension generated during suture approximation of widely spaced diaphragmatic edges may predispose to recurrence. Ellis et al. found the recurrence rate of primary repair to be 10% at median follow-up of 61 months [11]. These occurred despite gastropexy, suggesting that excessive tension on the repair, rather than lack of intraabdominal fixation, contributed to the recurrence. Pressure gradients between the abdominal and thoracic cavities also may contribute to reherniation.

Patient 3
The patient was an 87-year-old male who had numerous Emergency Room visits because of chest pain and upper abdominal pain. Chest X-ray and subsequent barium contrast studies revealed a paraesophageal hernia. Abnormal liver enzymes also prompted an ultrasound exam, revealing the presence of numerous gallstones. He underwent combined laparoscopic cholecystectomy and paraesophageal hernia repair. The latter was accomplished by laparoscopic reduction and expanded PTFE patch repair of the 8-cm diaphragmatic defect. He was discharged on the 2nd postoperative day and remained symptoms free at 9-month follow-up.

Operative technique
The patient is placed in the lithotomy position in padded Allen stirrups. We have found it also helpful to use a bean-bag cushion molded around the patient, as this helps keep the patient immobilized during periods of steep reverse Trendelenburg position. Five trocars are used in a Chevron distribution along the upper abdomen, as originally described by Dallemagne in his Nissen fundoplication technique [9]. We consider a 30 or 45 scope essential to permit maximum different viewing angles, particularly when it is necessary to perform posterior esophageal mobilization for a fundoplication. A reusable fan-shaped retractor is used to displace the medial segment of the liver. Securing this retractor with an automated arm (Mediflex Holder, Islandia, NY) is helpful in maintaining uninterrupted exposure. Division of the triangular ligament is necessary to fully expose the edges of the esophageal hiatus. The stomach is then carefully reduced from its intrathoracic position using atraumatic grasping forceps (Ethicon, Summerville, NJ). A nasogastric tube is then placed to fully decompress the stomach.. Adhesions between the stomach and intrathoracic structures are divided using bipolar cautery, and the short gastric vessels are divided between laparoscopic clips. The excess hernia sac is excised to fully expose the edges of the diaphragmatic defect (Fig. 2) and the gastroesophageal junction is carefully identified at the level of the hiatus. If preoperative tests or clinical circumstances dictate the need for a concomitant fundoplication, the posterior dissection of the esophagus is

305

Fig. 1. Upper gastrointestinal contrast study demonstrating a paraesophageal hernia with an intrathoracic stomach. The gastroesophageal junction is maintained below the diaphragm by the phrenoesophageal ligament.

Fig. 2. Laparoscopic inspection revealing a large paraesophageal hernia with an intrathoracic stomach.

Ackerman et al. found open primary repair to have a 50% recurrence rate radiographically [1]. These findings suggest that it may be necessary to consider the use of a synthetic material to accomplish a tension-free repair, applying the same principles used in the laparoscopic repair of inguinal hernias. The first descriptions of prosthetic repairs of a paraesophageal hernia involved the use of large pieces of synthetic mesh as an onlay patch. Kuster and Gilroy [15] reported successful laparoscopic repairs in five of six patients using Mersilene mesh (Ethicon, Summerville, NJ). The hernia sacs were not excised. A laparoscopic hernia stapler was used to secure the mesh to the edge of the diaphragm. A gastropexy was also added by stapling the seromuscular layer of the fundus of the stomach to cover the mesh in its entirety. No significant complications occurred and there were no recurrences noted at 822 months following surgery. In Pitchers et al. report of 12 patients undergoing laparoscopic paraesophageal hernia repair, two patients had large defects successfully repaired with polypropylene mesh [22]. Similar techniques and results have been published as case reports by others [4, 5, 13]. Edelman [10] reported a series of five patients undergoing laparoscopic mesh repair of paraesophageal hernias. A large piece of polypropylene mesh (Surgipro Mesh, Autosuture, Norwalk, CN) was stapled in place to cover the defect and a laparoscopic gastropexy was accomplished by fixing the cardia of the stomach to the mesh. A temporary gastrostomy was performed for additional fixation. Three patients had temporary episodes of heartburn postoperatively. One patient required laparoscopic reoperation because of a stricture at the gastroesophageal junction due to adhesions of the mesh to the stomach. One patient died because of a ruptured gastrostomy balloon which resulted in intraperitoneal tube feedings and peritonitis. The apparent disadvantage of polypropylene mesh would seem to be its tendency to cause an undesirable degree of visceral adhesions to adjacent organs. Indeed, the clinical experience with polypropylene mesh has been replete with a variety of complications, such as wound sepsis, erosion into intraabdominal organs, bowel fistula, and mesh

extrusion [2]. The same concerns have been raised over its use as an onlay patch in laparoscopic inguinal hernia repairs [19]. If used in the abdominal cavity, it should ideally be covered with a layer of peritoneum to prevent undesirable visceral adhesions. We feel that while the concept of a prosthetic, tensionfree repair is sound, materials other than polypropylene may be more suited for laparoscopic repair of paraesophageal hernias. Expanded PTFE, in particular, may be a better choice of prosthetic material in this location. Laboratory studies have indicated that expanded PTFE causes considerably less visceral adhesions than other materials tested and allowed normal diaphragmatic motion on fluoroscopy [18]. It allows tissue ingrowth without stimulating an extensive inflammatory response. Expanded PTFE is considered the preferred material for the repair of congenital diaphragmatic hernias, and excellent long-term results have been reported without evidence of an undesirable degree of visceral adhesions [7]. It elicits a low foreign body reaction, has low infectability, exhibits a low rate of adhesion formation, yet supports fibrous tissue incorporation [2, 16]. The disadvantage of working with expanded PTFE in this setting is that fixation with the customary hernia stapling devices is impractical. Not only is the material more difficult to penetrate than polypropylene mesh, but its opacity makes precise application of the staples difficult. For this reason, proper fixation must involved the sue of intracorporeal suturing techniques as described in this article. A further issue is the technique of fixation of the stomach below the diaphragm in order to prevent hernia recurrences. Most reports have described depending on either visceral adhesions to the mesh or the addition of a gastrostomy. We have chosen either to perform a fundoplication in those patients in whom it is clinically indicated or to suture the fundus to the right crus. This latter technique not only should protect adequately against reherniation, but, by a mechanism similar to a Dor fundoplication, may also have some protective effect against reflux. An additional argument presented in favor of a gastrostomy is the need for temporary decompression of a stomach that, as a consequence of chronic incarceration, may be prone to delay in emptying. None of our patients had post-

306

Fig. 4. Completed laparoscopic repair of paraesophageal hernia. The fundus has been secured to the right diaphragmatic crus using interrupted silk sutures.

Fig. 3. Intracorporeal suturing techniques are used to secure the expanded PTFE patch to the edges of the diaphragm. The patch is first secured at its lower edges with two separate PTFE sutures. The sutures are then advanced in a running fashion and tied together at the upper portion of the patch.

operative nausea or vomiting, and all were taking solid food by the 2nd postoperative day. Newer-generation prokinetic agents, such as cisapride, seem in our experience to overcome the stasis that may occur in the incarcerated segment of the stomach. A gastrostomy, when not physiologically desirable, is an unappealing means of fixation from the patients perspective. Even if only a temporary inconvenience, it may detract from an otherwise minimally invasive operation when other forms of internal gastric fixation are possible. A still-unresolved controversy is whether to perform an antireflux operation in conjunction with closure of the diaphragmatic defect. Elegant studies by DeMeester have documented that 60% of patients with type III hernias had hypotensive lower esophageal sphincters and an abnormal 24-h pH test [26]. These data were felt to substantiate the need for routine antireflux surgery as a component of the repair of paraesophageal hernias. In contrast, Williamson et al. have long questioned the value of 24-h pH monitoring in patients with paraesophageal hernia [27]. In the presence of gastric outlet obstruction, a normal LES may become incompetent at times and may result in reflux symptoms and abnormal reflux by 24-h pH studies. They found an antireflux operation to be necessary on only 15% of patients, and all of these were identified by preoperative screening with a combination of endoscopy, manometry, and 24-h pH monitoring. In their series, 98% of the remaining patients with preoperative reflux symptoms had complete resolution of the heartburn after reduction and hiatal closure alone. Ellis believes that as long as the posterior attachments of the esophagus at the hiatus are not disturbed, postoperative reflux symptoms are unlikely after anatomic repair of a paraesophageal hernia. Hill and Tobias also feel it is a technical mistake to divide these

in the repair of a paraesophageal hernia [12], and he expresses a concern that extensive mobilization of inferior esophagus unnecessarily destroys the angle of His. While 24-h pH monitoring is clearly the gold standard for diagnosing reflux disease in the setting of a sliding (type I) hernia, we share Elliss reservations about the accuracy of 24-h pH monitoring in the setting of a paraesophageal hernia. Reflux can occur in individuals with a normally functioning sphincter when gastric obstruction is present, as would seem to be the case in most patients with a paraesophageal hiatal hernia [11]. Our own approach has been to add a conventional antireflux operation if contrast studies indicate a combined type III paraesophageal hernia in which the phrenoesophageal ligaments, by definition, have already been disrupted. We would also advocate a formal antireflux operation if the patient has a decreased lower esophageal sphincter pressure as identified by preoperative manometry. In all other cases, we perform closure of the diaphragmatic defect with suture of the fundus to the right crus of the diaphragm. This latter step is a modified form of anterior Dor fundoplication that we believe helps protect against postoperative reflux. As this report has indicated, numerous techniques are available for the laparoscopic repair of paraesophageal hernias. Smaller defects are best closed using primary suture repair, while large defects may require a tension-free repair using prosthetic patches. We have demonstrated the successful use of expanded PTFE in the repair of large diaphragmatic defects and presented our criteria for inclusion of an antireflux operation. Disorders of the esophageal hiatus encompass a wide spectrum of clinical disease, and the operative approach should be individualized to the patients anatomic and pathophysiologic condition. References
1. Ackermann C, Bally H, Rothenbuehler JM, Harder F (1989) The surgery of paraesophageal hernias: technique and results (translated from German). Schweiz Med Wochenschr 119: 723725 2. Bauer JJ, Salky BA, Gelernt IM, Kreel I (1987) Repair of large ab-

307 dominal wall defects with expanded polytetrafluorethylene (PTFE). Ann Surg 206: 765769 Brown GL, Richardson JD, Malangoni MA, et al. (1985) Comparison of prosthetic materials for abdominal wall reconstruction in the presence of contamination and infection. Ann Surg 81: 705711 Bueno R, Brooks DC (1993) Laparoscopic repair of paraesophageal hernias: preliminary results (abstract). Surg Endosc 7: 128 Cloyd DW (1994) Laparoscopic repair of incarcerated paraesophageal hernias. Surg Endosc 8: 893897 Congrave DP 91992) Laparoscopic paraesophageal hernia repair. J Laparoendosc Surg 2: 4548 Cullen ML, Klein MD, Philippart AI (1985) Congenital diaphragmatic hernia. Surg Clin North Am 65: 11151138 Cuschieri A, Shimi S, Nathanson LK (1992) Laparoscopic reduction, crural repair, and fundoplication of large hiatal hernia. Am J Surg 163: 425430 Dallemagne B, Weerts JM, Jehaes C (1991) Laparoscopic Nissen fundoplication: preliminary report. Surg Laparosc Endosc 1: 138143 Edelman DS (1995) Laparoscopic paraesophageal hernia repair with mesh. Surg Laparosc Endosc 5: 3237 Ellis FH, Crozier RE, Shea J (1986) Paraesophageal hiatus hernia. Arch Surg 121: 416420 Hill LD, Tobias J (1968) Paraesophageal hernia. Arch Surg 96: 735 744 Johnson PE, Persuad M, Mitchell T (1994) Laparoscopic anterior gastropexy for the treatment of paraesophageal hernias. Surg Laparosc Endosc 4: 152154 Koger KE, Stone JM (1993) Laparoscopic reduction of acute gastric volvulus. Am Surg 59: 325328 Kuster GG, Gilroy S (1993) Laparoscopic technique for repair of paraesophageal hiatal hernias. J Laparoendosc Surg 3: 331338 Lamb JP, Vitale T, Kaminski DL (1983) Comparative evaluation of synthetic meshes used for abdominal wall replacement. Surgery 93: 643648 Landreneau RJ, Johnson JA, Marshall JB, Hazelrigg SR, Boley TM, Curtis JJ (1992) Clinical spectrum of paraesophageal herniation. Dig Dis Sci 37: 537544 Newman BM, Jewett TC, Lewis A, Cerny F, et al. (1985) Prosthetic materials and muscle flaps in the repair of extensive diaphragmatic defects: an experimental study. J Pediatr Surg 20: 362367 Nguyen NX, Camps J, Fitzgibbons RJ (1994) Laparoscopic intraperitoneal inguinal hernia repair. Semin Laparosc Surg 1: 106115 Oddsdottir M, Franco AL, Laycock WS, Waring JP, Hunter JG (1995) Laparoscopic repair of paraesophageal hernia. Surg Endosc 9: 164 168 Ozdemir L, Burke W, Ikins P (1973) Paraesophageal herniaa life threatening disease. Ann Thorac Surg 16: 547552 Pitcher DE, Curet MJ, Martin DT, Vogt DM, et al. (1995) Successful laparoscopic repair of paraesophageal hernia. Arch Surg 130: 590596 Skinner D, Belsey RH (1967) Surgical management of esophageal reflux and hiatus hernia: Long term results with 1030 patients. J Thorac Cardiovasc Surg 53: 3334 Swanstrom LL, Hunter JG (1995) Laparoscopic partial fundoplication. In: Peters JH, DeMeester TR (eds) Minimally invasive surgery of the foregut. Quality Medical, St Louis, MO, pp 159177 Treacy PJ, Jamieson G (1987) An approach to the management of para-oesopohageal hiatus hernia. Aust NZ J Surg 57: 813817 Walther B, DeMeester TR, Lafontaine E, Courtney JV, Little AG, Skinner DB (1984) Effect of paraesophageal hernia on sphincter function and its implication on surgical therapy. Am J Surg 147: 111116 Williamson WA, Ellis FH, Streitz JM, Shahian DM (1993) Paraesophageal hiatal hernia: is an antireflux procedure necessary? Ann Thorac Surg 56: 447452

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

17.

18.

19. 20.

21. 22. 23.

24.

25. 26.

27.

Surg Endosc (1997) 11: 257260

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic lumbar sympathectomy

N. Kathouda,1 S. Wattanasirichaigoon,2 E. Tang,1 P. Yassini,1 U. Ngaorungsri2
1 2

Department of Surgery, School of Medicine, University of Southern California, 1510 San Pablo Street, Suite 514, Los Angeles, CA 90033, USA Department of Surgery, Faculty of Medicine, Vajira Hospital, 681 Samsen Road, Dusit, Srinakharinwirot University, Bangkok 10300, Thailand

Abstract Background: Lumbar sympathectomy retains a role in the treatment of patients with causalgia, Symptomatic vasospasm, and nonreconstructable arterial occlusive disease. Open surgical sympathectomy, with its attendant morbidities, remains the standard. Chemical sympathectomy has been introduced as a less invasive means of achieving sympatholysis. However, this has been associated with incomplete and transient denervation. Methods: We present a series of five lumbar sympathectomies performed laparoscopically. Results: All patients sustained symptomatic relief and no postoperative complications were noted. Postoperative skin thermometry and resistance measurements confirmed adequacy of sympatholysis. Conclusion: We conclude that lumbar sympathectomy can be performed laparoscopically. Our preferred technique is now the extraperitoneal approach. Such an approach combines the durability and reliability of standard open sympathectomy with the minimal invasiveness of laparoscopic surgery. Key words: Laparoscopic Lumbar Sympathectomy

used with some success. However, incomplete sympathectomy and return of sympathetic tone remain the significant limitations with these modalities [4]. Laparoscopically performed lumbar sympathectomy promises all the advantages of a minimally invasive approach without the inconsistent therapeutic results of percutaneous ablation techniques [10]. However, there has as of yet been no description of a laparoscopic approach to this particular operation. Here we describe our technique of laparoscopic lumbar sympathectomy and present five patients who have undergone this operation.

Methods and materials

Five patients underwent laparoscopic lumbar sympathectomy. Four of the patients were diabetic males aged 2954 years with known unreconstructable peripheral arterial disease. The predominant presenting symptoms were unilateral pain and lower extremity coldness. The fifth patient was a 27-year-old man with severe right foot rest pain and dry gangrene of the big toe. He also experienced intermittent vasospastic-like symptoms including right leg coldness. He had a 20-year history of smoking two packs of cigarettes per day and 7-year heroin addiction. Physical examination of the fifth patient revealed a blood pressure of 122/80 mmHg, absent pulses below the right knee, and marked hyperhidrosis below the right ankle. His right great toe had dry gangrene. The admission hemoglobin, white blood count, coagulation profile, and serum chemistry were all within normal limits. A right femoral angiogram revealed occlusion of the popliteral artery with reconstitution of the distal anterior tibial artery. A femoral-to-anterior-tibial-artery reverse saphenousvein bypass graft was performed without relief of the foot pain or the hyperhidrosis. This initial experience with laparoscopic sympathectomy was through an anterior transperitoneal approach. In this approach, the patient was placed in a lateral position, with the table broken between the ribs and the iliac crest. The surgeon stood in front of the patient, with a first assistant on the opposite side. A Veress needle (Ethicon Endosurgery, Cincinnati, Ohio) was inserted at the edge of the rectus sheath in line with the umbilicus. The abdomen was then insufflated with carbon dioxide and a 10-mm port was inserted at the site of the Veress needle and used for the laparoscope. Under direct vision, two secondary ports, one 5 mm and one 10 mm, were then inserted in the midclavicular line. If required, a fourth port (5 mm) was placed halfway between the umbilicus and the symphysis pubis. This port was used for a fan retractor to displace the kidney and colon medially. The lateral peritoneal attachments of the right or left colon were then incised from the hepatic or splenic flexure down to the pelvic brim. The colon was then reflected medially by virtue of gravity. The kidney, including the perirenal fat and adrenal gland, was then dissected from the retroperitoneum and then medially rotated to expose the anterolateral sur-

Lumbar sympathectomy was first popularized for symptomatic vasospasms by Adson and Brown in the United States and Diez in South America in the 1920s. Subsequently, it enjoyed a period of popularity during which time it was applied to arterial occlusive disease as well. With the development of arterial reconstructive techniques in the 1960s, its use has diminished. However, it still remains a role in the treatment of patients with unreconstructable arterial occlusive disease, symptomatic vasospasm unresponsive to medications, and causalgia [1, 4]. Percutaneous lumbar sympathectomy by radiologically guided injections of phenol or alcohol has been increasingly popular in Europe. Radiofrequency ablation has also been

Correspondence to: N. Katkhouda

258

Fig. 1. Patient positioning: A: Supine position with the right side supported 30 upward. B The thighs are flexed 20 and the table is flexed 20.

Fig. 2. The operative setup: A shows position of primary surgeon (S), camera operator (C), assistant (A), and scrub nurse (N). B Port placements for camera (C), fan retractor (F), and working instruments (W).

face of the vertebral column. The second lumbar ganlion, lying below the renal pedicle, was then identified and lifted away from the vertebra. The upper end of the sympathetic chain was clipped and transected, and the rami communicantes were dissected free and secured with clips. The dissection then proceeded inferiorly until the iliac arteries were encountered. The chain was then clipped, transected, and removed. A closed suction drain was placed at the end of the procedure. After initial experience with the anterior transperitoneal approach to lumbar sympathectomy, the approach was modified to remain completely extraperitoneal. For a right extraperitoneal lumbar sympathectomy, the patient was placed on a bean bag with the right side supported to 30 (Fig. 1A). The table was then flexed 20, and the thighs were flexed to 20 to relieve tension on the psoas muscle (Fig. 1B). The surgeon stood on the right of the patient with an assistant on the left (Fig. 2A). An 11-mm transverse incision was made above the anterior superor iliac spine in line with the umbilicus. This was deepened in a muscle-splitting manner until a finger could be gently pushed over the peritoneum into the retroperitoneal space. The space was then further developed by digital dissection alternating with the use of a standard open peanut dissector. A blunt-tip 10 11-mm port was then inserted through the incision and a pursestring suture was used to secure it into position. The retroperitoneal space was then insufflated with carbon dioxide to a pressure of 13 mmHg, and this pneu-

modissection completed the development of the retroperitoneal space. A 30 laparoscope (Karl Storz, Tuttlingten, Germany) was introduced to inspect the retroperitoneal space, and under direct vision to avoid peritoneal violation, two secondary ports were inserted at the edge of the rectus sheath in a triangulated fashion to allow the surgeon to operate with two hands. A fourth port was used to place a nontraumatic retractor (Ethicon Inc., Cincinnati, Ohio) (Fig. 2B). The anatomic landmarks for the adequate exposure of the right side include the inferior vena cava medially as it runs over the right side of the vertebral bodies. The right renal vein forms the superior extent of the dissection, and the right psoas muscle and the inferior pole of the kidney form the lateral extent. The sympathetic chain is located between the inferior vena cava and the psoas muscle. Dissection of the L2 sympathetic ganglion was again started below the renal pedicle (Fig. 3). The upper end of the sympathetic chain was then identified, clipped, and transected. The rami communicantes were likewise dissected, clipped, and transected (Fig. 4). One lumbar vein was encountered and was carefully ligated with two clips and divided. The sympathetic chain was divided at the level of the common iliac artery with the specimen sent for pathologic examination. Great caution was taken to preserve the first lumbar ganglion in order to preserve sexual function. No drains were used at the end of the procedure.

259

Fig. 3. Intraoperative exposure of the right sympathetic chain (S) completely free from its retroperitoneal attachments. Inferior vena cava (I) is medially retracted by an atraumatic retractor. The psoas muscle (P), lower pole of right kidney (K), and ureter (U) are seen.

Fig. 4. Division of sympathetic chain between clips.

Discussion Lumbar sympathectomy has been performed for a variety of indications although its role has diminished with the success of infrapopliteal arterial reconstruction and newer pharmacologic management of vasospasm and caulsalgia. Nevertheless, a significant number of patients still present with nonreconstructable arterial occlusive disease, symptomatic vasospasm, or caulsalgia not responsive to medical therapy. Open lumbar sympathectomy, with all the inherent drawbacks of open surgery, remains the standard of care for these patients. Chemical sympathectomy by means of percutaneous radiologically guided injections of phenol or alcohol has been performed with some success. However, these techniques have been plagued with inconsistent results, particularly concerning the duration of sympathectomy block. There is also a significant incidence of incomplete block and injection-site pain. Furthermore, multiple injections can cause enough inflammation to preclude subsequent surgical sympathectomy. The use of laparoscopic techniques to perform a standard sympathectomy brings together the advantages of minimally invasive surgery and the reliability of an established open procedure. Our initial experience was with the transperitoneal approach since newly described laparoscopic procedures on retroperitoneal structures have primarily used this approach [2, 6, 79]. Furthermore, early attempts at direct retroperitoneal laparoscopy by Wickham and Mille were unsuccessful [11]. Although the transperitoneal approach to this operation has been successful in our hands, a retroperitoneal approach would be preferable due to fewer long-term complications resulting from intraabdominal adhesion formation. An effective minimally invasive retroperitoneal route has recently been described by Gaur [5]. Other authors have

Results A total of five patients underwent laparoscopic lumbar sympathectomy. An anterior transperitoneal approach was used for the first four patients, and the last patient underwent an extraperitoneal procedure. The mean operating time for the transperitoneal approach was 126 min (range 75140 min), and the operating time for the extraperitoneal approach was 145 min. Pathologic examination of all of the specimens demonstrated removal of the sympathectic chain. No parenteral analgesics were either requested or administered postoperatively. The patients resumed oral intake on the following morning, and all had resumed normal activity by the end of the 1st postoperative week. Postoperative radiographs demonstrated the positions of clips placed during the procedure (Fig. 5). Skin thermometry and resistance measurements were done on the last patient. Preoperatively, his skin temperature on the affected right side was 35.6C, compared to 37.2 on the left side. Ten days postoperatively, the temperatures were 37.2 and 37.0, respectively. Skin resistance measurements were made on the medial aspect of the foot using electrodes spaced 5 cm apart. Five measurements were taken at each site. The mean resistance on the right foot preoperatively was 0.581 M, compared to 15.9 on the left. Postoperatively, the values were 12.1 and 8.31, respectively. At 1 month postoperatively, no patients complained of neuralgia, and no male patients reported sexual dysfunction. All patients reported relief of their rest pain and improvement of trophic changes in the affected extremity. The one patient with a nonhealing ulcer demonstrated the ulcer to be healing.

260

laparoscopic dissection is essentially identical. Retraction of the large vessels such as the inferior vena cava on the right side and the aorta on the left side should be performed with extreme caution using atraumatic retractors handled by an experienced assistant. In conclusion, we believe that laparoscopic retroperitoneal lumbar sympathectomy is a safe and technically achievable operation for the experienced laparoscopic surgeon. Excellent clinical outcome appears to be comparable to that of the open procedure with all the advantages of a minimally invasive approach. A larger clinical experience and long-term follow-up will ultimately determine if this will become the procedure of choice.

References
1. Baker DM, Lamerton AJ (1994) Operative lumbar sympathectomy for severe lower limb ischemia: still a valuable treatment option. Ann R Coll Surg Engl 76: 5053 2. Clayman R, Kavoussi L, Soper NJ (1991) Laparoscopic nephrectomy: initial case report. J Urol 146: 278282 3. Coptcot MJ, Eden CG (1994) Laparoscopic retroperitoneal surgery. In: Coptcoat MJ, Joyee AD (eds) Laparoscopy in urology. Blackwell, Oxford 4. Cotton LT, Cross FW (1985) Lumbar sympathectomy for arterial disease. Br J Surg 72: 678683 5. Gaur DD (1992) Laparoscopic operative retroperitoneoscopy: use of a new device. J Urol 148: 11371139 6. Gurshman A, Daykhovsky L, Chanda M (1990) Laparoscopic pelvic lymphadenectomy. J Laparoendosc Surg 1: 6368 7. Janetschek G, Reissigl A, Peschel R, Bartsch G (1992) Laparoscopic retroperitoneal lymphadenectomy for testicular tumor: animal studies and first clinical experience. Minimally Invasive Ther 1: 68 (b-31) 8. Lipskey H, Wuernschinnel E (1993) Laparoscopic lithotomy for ureteral stones. Minimally Invasive Ther 2: 1922 9. Nezhat C, Nezhat F, Green B (1992) Laparoscopic treatment of obstructed ureter due to endometriosis by resection and ureteroureterostomy: a case report. J Urol 148: 865868 10. Reddick DJ, Olsen DO (1989) Laparoscopic laser cholecystectomya comparison with mini-lap cholecystectomy. Surg Endosc 3: 131313 11. Wicham JEA, Miller RA (1983) Percutaneous renal access. In: Percutaneous renal surgery. Chapter 2, Churchill Livingstone, New York, pp 3339

Fig. 5. Postoperative X-ray demonstrating position of clips and adequate levels of transection of the ganglionic chain.

also used a dissecting balloon placed through a small incision into the retroperitoneal space [3]. We have not used this balloon in order to limit the cost of the procedure, as blunt finger dissection and insufflation created an adequate working space. As the technology evolves and the cost of a dissecting balloon becomes lower, its use can facilitate and speed the creation of the retroperitoneal space. One pitfall of the retroperitoneal approach is the potential violation of the peritoneum, which can occur particularly upon insertion of trocars. This can be avoided by excellent direct visualization during introduction of trocars. Following exposure of the retroperioneum, the remainder of the operation proceeds as in the open surgery, and the

Surg Endosc (1997) 11: 287289

Surgical Endoscopy
Springer-Verlag New York Inc. 1997

Laparoscopic intragastric surgery for gastric leiomyoma

E. Taniguchi,1 W. Kamiike,1 H. Yamanishi,1 T. Ito,1 R. Nezu,1 T. Nishida,1 T. Momiyama,1 S. Ohashi,2 T. Okada,2 H. Matsuda1
1 2

First Department of Surgery, Osaka University Medical School, 2-2 Yamada-Oka, Suita, 565, Japan Department of Surgery, Takarazuka City Hospital, 4-5-1 Kohama, Takarazuka, 665, Japan

Abstract. Laparoscopic intragastric surgery (LIGS) was performed on a 63-year-old man with a gastric leiomyoma adjacent to the cardia. Because the tumor was about 5 cm in maximum diameter and showed ulceration, the possibility that the tumor was a leiomyosarcoma could not be ruled out preoperatively. Conventionally, major surgery has been performed on patients with a tumor located near the cardia, although it was not always malignant. Enucleation by LIGS enabled us to avoid excessive invasiveness and provided a favorable result. LIGS may be an appropriate new, minimally-invasive operation for gastric myogenic tumors and should be considered for such cases. Key words: Laparoscopic intragastric surgery (LIGS) Gastric leiomyoma Leiomyosarcoma

Case Report
A 63-year-old man was admitted to our hospital because of an asymptomatic gastric submucosal tumor. Barium examination and gastrofiberscopy showed a submucosal tumor protruding into the gastric lumen near the cardia. Ultrasound examination and X-ray CT scan showed no extragastric growth. Because the tumor had a central ulceration and was about 5 cm in diameter, malignancy could not be excluded although pathological examination of an endoscopically biopsied specimen suggested a leiomyoma. Due to its location, partial resection with sufficient margins was thought to be difficult. Accordingly, we decided to perform enucleation by LIGS. Then, if the postoperative pathological examination showed it to be a leiomyosarcoma, we would perform total gastrectomy with lymph node dissection. Under general anesthesia, a gastrofiberscope was inserted perorally into the stomach for observation of the gastric lesion. Then, a nasogastric tube with a balloon was inserted into the duodenum, and the balloon was inflated to prevent air-flow from the stomach to the intestine. Guided by the gastrofiberscope, three trocars, one 12-mm trocar and two 5-mm ones, with a balloon (Laparo SAC, Marlow Inc., USA), were inserted into the stomach through the abdominal wall (Fig. 1). Enucleation was performed using a laparoscopic dissector and electrocautery. First, a dilute epinephrine solution was injected into the submucosal layer. The mucosal layer around the ulceration was cut by electrocautery. The surface of the tumor was exposed, and the dissection along it was performed sharply and bluntly (Fig. 2). A small amount of bleeding was observed, but irrigation and suction made it easy to identify the surface of the tumor. The mucosal layer around the ulceration was removed together with the tumor. The tumor, which appeared to be arising from the lamina muscularis, did not invade into the muscularis propria (Fig. 3). The defect of the mucosal layer was laid open. The enucleated tumor was put into a plastic bag (Catch Purse, Hakko Trading Co., Japan) and brought out through the esophagus using a gastrofiberscopic grasper. The wounds of the gastric wall were pulled out through the wounds of the abdominal wall and closed by 10 silk continuous sutures. The tumor, with multinodular expansion, was elastic, soft, and 5.4 3.4 2.2 cm in size. Histological examination revealed that the tumor was a benign leiomyoma. The patients postoperative course was uneventful.

It is sometimes difficult to diagnose whether gastric myogenic tumors are malignant or benign preoperatively [12]. Accordingly, it is not easy to decide the operative procedure for such cases. Because gastric myogenic tumors are often located in the cardia, major surgery, such as total gastrectomy or proximal gastrectomy, should be selected if they are suspected to be malignant. Even if they are not malignant, those procedures have been often performed when malignancy cannot be ruled out. And they result in lowering the patients quality of life. Laparoscopic surgery has come into widespread use in intraabdominal operations. Recently, a new concept, laparoscopic intragastric surgery (LIGS), has been proposed for early gastric cancer [9]. We successfully performed LIGS in a patient with a gastric leiomyoma for which the possibility of being a leiomyosarcoma could not be excluded preoperatively.

Discussion Enucleation is thought to be a satisfactory procedure for gastric leiomyomas. However, the surgical strategies for gastric leiomyosarcomas have been controversial. Some authors have advocated the need for gastrectomy with system-

Correspondence to: E. Taniguchi

288

Fig. 1. Intraoperative laparoscopic view. At the beginning of the operation. T, tumor; NG, nasogastric tube. Fig. 2. Intraoperative laparoscopic view. The surface of the tumor was exposed, and the dissection along it was performed sharply and bluntly. T, the surface of the tumor; E, electrocautery with scissors; S, irrigation and suction tube. Fig. 3. Intraoperative laparoscopic view. The tumor (T) was not attached to the muscularis propria (M).

atic lymph node dissection [13], while others have reported that wedge resection was adequate for gastric leiomyosarcomas [3, 11, 12]. Recently, there have been some papers reporting that the operative procedure for gastric leiomyosarcoma should be selected according to the grade of malignancy [2, 4]. At least, resection of the full thickness of the gastric wall with adequate margins is required for leiomyosarcomas. As partial resection of the stomach with margins cannot be applied to tumors located close to the cardia, those have been subjected to major surgery such as total gastrectomy or proximal gastrectomy. It is sometimes difficult to determine whether the tumors are benign or malignant preoperatively or even from intraoperative frozen sections [12]. Major surgery would be excessive if the tumor is proven to be benign by postoperative pathological examination. Therefore, enucleation is considered to be a better procedure for patients whose tumors are suspected to be benign but in which sarcoma cannot be ruled out; such cases are occasionally encountered. If the tumor is proved to be benign, as had been thought preoperatively, enucleation enables us to avoid excessive invasiveness and will provide a less invasive outcome. If the tumor is diagnosed to be malignant, a second-look operation will be necessary. Even in such cases, pathological findings of the whole tumor can be obtained, which is available for determination of the procedure for the next operation. Concerning the tumor seedings, the risk is thought to be negligible because almost the tu-

mors to which this procedure will be applied are considered to be low-grade malignancy even if they are malignant. Recent advances in laparoscopic surgical techniques have enabled us to perform various operations for intraabdominal diseases under laparoscopy [15], including surgery on the stomach [1]. Several papers reported that laparoscopic partial gastrectomy had been successfully performed for early gastric cancer or peptic ulcer [5, 8, 10]. More recently, laparoscopic operation has also been applied to gastric leiomyomas. Formerly, endoscopic resection using a gastrofiberscope was undertaken only in patients with gastric leiomyomas of small size and located where operation by gastrofiberscopy could be easily performed [14]. Llorente reported a case of gastric leiomyoma subjected to laparoscopic gastric resection [7]. Yamashita et al. reported two techniques for resection of gastric leiomyomas: laparoscopy-guided extracorporeal enucleation and laparoscopic intracorporeal enucleation [16]. These were ordinary extragastric laparoscopic operations, and it was impossible to apply those procedures to our case because of its size and location and because it had an ulceration. Ohashi developed LIGS for early gastric cancer and enucleated a small gastric leiomyoma using this technique [9]. There had been a report on a trial of intragastric manipulation using laparoscopy for bleeding from a gastric ulcer [6]. However, Ohashi was the first to apply LIGS for neoplasms of the stomach and established it as an operative technique [9]. The present case is considered to be the first one in which LIGS was applied for a large leiomyoma located close to the esophagogastric junction. Enucleation of gastric leiomyomas by LIGS has some advantages over laparoscopic wedge resection: LIGS can be applied to large tumors and those located near the cardia and pylorus as well as at the posterior wall of the stomach, where laparoscopic wedge resection is difficult to apply. However, LIGS cannot apply to tumors existing at the anterior wall or showing extragastric growth. The op-

289

erative procedure for gastric submucosal tumor should be selected based on the features of the tumor. LIGS should be considered in cases whose tumors are located near the cardia and are thought to be benign, although malignancy cannot be ruled out. References
1. Abercrombie JF, McAnena OJ, Rogers J, Williams NS (1993) Laparoscopic resection of a bleeding gastric tumour. Br J Surg 80: 373 2. Bandoh T, Isoyama T, Toyoshima H (1993) Submucosal tumors of stomach: a study of 100 operative cases. Surgery 113: 498506 3. Crocker DW (1969) Smooth muscle tumors of the stomach. Ann Surg 170: 239243 4. Eng-Hen NG, Pollock RE, Munsell MF, Atkinson EN, Romsdahl MM (1992) Prognostic factors influencing survival in gastrointestinal leiomyosarcomas. Ann Surg 215: 6877 5. Goh P, Kum CK (1993) Laparoscopic Billroth II gastrectomy: a review. Surg Oncol 2 (Suppl) 1: 1318 6. Kitano S, Kawanaka H, Tomikawa M, Hirabayashi H, Hashizume M, Sugimachi K (1994) Bleeding from gastric ulcer halted by laparoscopic suture ligation. Surg Endosc 8: 405407 7. Llorente J (1994) Laparoscopic gastric resection for gastric leiomyoma. Surg Endosc 8: 887889

8. Lointier P, Leroux S, Ferrier C, Dapoigny M (1993) A technique of laparoscopic gastrectomy and Billroth II gastrojejunostomy. J Laparoendosc Surg 3: 353364 9. Ohashi S (1995) Laparoscopic intraluminal (intragastric) surgery for early gastric cancer. A new concept in laparoscopic surgery. Surg Endosc 9: 169171 10. Ohgami M, Kumai K, Otani Y, Wakabayashi G, Kubota T, Kitajima M (1994) Laparoscopic wedge resection of the stomach for early gastric cancer using a lesion-lifting method. Dig Surg 11: 6467 11. Ripstein CB, Flint GW (1952) Leiomyosarcoma of the gastrointestinal tract. Gastroenterology 20: 315326 12. Schiu MH, Farr GH, Papachristou DN, Hajdu SI (1982) Myosarcoma of the stomach: natural history, prognostic factors and management. Cancer 49: 177187 13. Skandalakis JE, Gray SW, Shepard D (1960) Smooth muscle tumor of the stomach. Int Abstr Surg 110: 209226 14. Spinelli P, Cerrai FG, Cambareri AR, Meroni E, Pizzetti P (1993) Two-step endoscopic resection of gastric leiomyoma. Surg Endosc 7: 9092 15. Taniguchi E, Kamiike W, Iwase K, Nishida T, Miyata M, Inoue M, Ohashi S, Okada T, Matsuda H (1995) Laparoscopic extramucosal myectomy with anterior fundoplication (Dor) for esophageal achalasia using intraoperative manometry. Surg Endosc 9: 817819 16. Yamashita Y, Bekki F, Kakegawa T, Umetani H, Yatsuka K (1995) Two laparoscopic techniques for resection of leiomyoma in stomach. Surg Laparosc Endosc 5: 3842