Hepatobiliar y Imaging Original Research

Pupulim et al. Cytologic Diagnosis of Hepatic Tumors Hepatobiliary Imaging Original Research

Algorithm for Immediate Cytologic Diagnosis of Hepatic Tumors

Lawrence F. Pupulim1 Michle Felce-Dachez 2 Valrie Paradis2 Marie-Pierre Vullierme1 Magaly Zappa1 Pierre Bedossa2 Valrie Vilgrain1
Pupulim LF, Felce-Dachez M, Paradis V, et al.

OBJECTIVE. Immediate cytologic assessment of hepatic lesions can help determine the adequacy of specimens and may yield a preliminary diagnosis. The purpose of this study was to analyze the diagnostic accuracy of on-site cytologic assessment compared with definitive cytologic examination in the detection and correct subtyping of malignant hepatic lesions. SuBJECTS AND METHODS. The study population included 472 consecutively registered patients with hepatic nodules who underwent sonographically guided fine-needle aspiration and core biopsies. During on-site cytologic analysis, the pathologist made a preliminary diagnosis of malignancy or negative for malignancy for each nodule. When a malignant lesion was diagnosed, immediate subtyping was attempted. RESuLTS. With immediate cytologic analysis, 280 (80.9%) of 346 malignant nodules were correctly identified without false-positive cases. With immediate subtyping, 113 (68.1%) of 166 cases of hepatocellular carcinoma, 28 (77.8%) of 36 cases of cholangiocarcinoma, and 113 (85.0%) of 133 cases of metastasis were detected. CONCLuSION. High diagnostic accuracy for malignancy can be achieved with onsite cytologic evaluation of hepatic tumors. Specific diagnoses of hepatocellular carcinoma, cholangiocarcinoma, and metastasis also can be made on-site in most cases. The absence of false-positive diagnosis of malignancy at on-site cytologic examination may make additional biopsy unnecessary. We propose an algorithm for the cytohistopathologic management of hepatic tumors. lthough clinical context associated with elevated levels of serum tumor markers and a typical imaging pattern may be sufficient for diagnosis of hepatic tumors [13], biopsy is mandatory in some clinical settings [3, 4]. Oncologists continue to recommend biopsy of most metastatic hepatic lesions to adapt clinical management, and biopsy is also used extensively if the lesion is atypical on imaging, whatever the clinical data. In one study [5], approximately 25% of patients who underwent transplantation because of an imaging-based diagnosis of hepatocellular carcinoma (HCC) were found not to have malignant findings at pathologic examination of the explanted liver. These results strengthen the importance of histologic diagnosis of malignancy before surgery. The two methods of obtaining diagnostic material are fine-needle aspiration biopsy (FNA) and core needle biopsy. These two methods have similar results in the literature, and their combination has had the best results

Keywords: core needle biopsy, fine-needle aspiration, immediate cytologic assessment, liver neoplasms DOI:10.2214/AJR.07.2549 Received May 11, 2007; accepted after revision September 11, 2007.
1

Service de Radiologie, Hpital Beaujon, Assistance Publique des Hpitaux de Paris, Universit Paris, 100, Blvd. du Gnral Leclerc, 92110 Clichy, France. Address correspondence to L. F. Pupulim (law_dp@hotmail.com).

2 Service de Pathologie, Hpital Beaujon, Assistance Publique des Hpitaux de Paris, Universit Paris, Clichy, France.

WEB This is a Web exclusive article. AJR 2008; 190:W208W212 0361803X/08/1903W208 American Roentgen Ray Society

[611]. For some experts, FNA is more effective than core needle biopsy on the basis of cost, procedure-associated morbidity, and overall time to diagnosis [12, 13], although the results of this procedure are highly cytopathologist dependent. Furthermore, immediate cytologic assessment during FNA can be used to determine the adequacy of the sample and may increase the diagnostic sensitivity of the technique. Although immediate cytologic assessment was proposed in the 1980s [14, 15], few centers have adopted this technique as routine, and the diagnostic accuracy of onsite assessment has been described mostly in small and retrospective series [1618]. The purposes of this study were to analyze prospectively with consecutive cases the diagnostic accuracy of immediate cytologic assessment in the detection of malignant hepatic lesions, to compare the results of immediate cytologic assessment with those of definitive cytologic examination, and to assess the performance of immediate cytologic assessment in the subtyping of malignant tumors.

W208

AJR:190, March 2008

Cytologic Diagnosis of Hepatic Tumors Subjects and Methods Patient Selection

During the 6-year period April 2000July 2006 all patients at our institution who had an indication for percutaneous hepatic biopsy for hepatic lesions underwent a combined core needle biopsy and FNA procedure with immediate cytologic assessment. We routinely perform both FNA and core needle biopsy because studies [611] have shown that combining the two techniques yields the best results. Biopsy of hepatic lesions was indicated when the diagnosis was not established after imaging and when biopsy results might have changed patient care according to the most recent guidelines [24]. Contraindications were carefully sought and included abnormal coagulation indexes [1, 4], the presence of cholangitis, lack of a safe access route, and lack of cooperation on the part of the patient. Institutional review board approval was obtained, and all patients gave written informed consent before procedure. Five hundred seven consecutively registered patients were prospectively enrolled. We excluded patients with nonsolid lesions or lesions secondarily invading the liver by contiguity (n = 15) and patients with an uncertain definitive diagnosis (n = 20). The study population included 472 adult patients. The 269 (57%) men and 203 (43%) women had a mean age of 58 years (range, 1987 years). A total of 345 (73%) of the patients had one or two liver nodules, and 127 (27%) had multiple nodules. The mean size of the biopsied nodules was 3.9 3.5 (SD) cm; median, 3 cm; range, 0.720 cm. All patients stayed in the hospital for at least 1224 hours for postbiopsy observation, which is common practice at our institution. No major complications occurred. part of the samples in the radiology department. Two to 10 slides were prepared for each procedure depending on the amount of material obtained. The cytopathologist decided whether material was adequate after rapid microscopic evaluation. If not, FNA was repeated a maximum of three times during the same session in an attempt to obtain an adequate sample. After spreading the aspirated material on slides, the cytopathologist prepared air-dried direct smears that were fixed on alcohol and prepared with eosin and Giemsa stains for immediate cytologic assessment. A number of slides were separated out for definitive cytologic analysis, and these slides were prepared in a different manner from the immediately evaluated slides. They were prepared in the pathology department with May-Grnwald and Giemsa stains after air-drying. Alcohol fixation was not used for these slides. Immediately after the FNA procedure, core needle biopsy samples were obtained with an 18-gauge needle loaded into a semiautomatic biopsy system. Up to three core samples were obtained, and the adequacy of core needle biopsy was judged on the basis of the position of the needle in the target lesion and the size and color of the specimens. Biopsy of the tumor-free portion of the liver was performed on patients with chronic liver disease and when a diagnosis of benign tumor was suspected. After formalin fixation, the core needle biopsy samples were routinely processed and prepared with H and E, Masson trichrome, picrosirius red, Perls, and reticulin stains. The number of passes for both FNA and core needle biopsy was recorded for each patient. All slides subjected to immediate and definitive cytologic assessment were analyzed by the same cytopathologist. The histologic analysis was performed by another pathologist. The interpretations were performed independently, and the findings were documented in separate reports.

Immediate Cytologic Diagnosis

On the basis of the results of rapid on-site FNA assessment in the radiology department, the cytopathologist made a preliminary diagnosis of malignant or negative for malignancy for each aspirated nodule. Doubtful cases with no clear evidence of malignancy were considered negative for malignancy. Cytologic features defining malignancy during rapid assessment were basically the same as those used in the definitive cytologic analysis and included the presence of high cellularity and marked cellular atypia. These features have been well defined and extensively reported in the medical literature [12, 1923]. When a diagnosis of malignancy was made on site, the cytopathologist subtyped the lesions as HCC, metastasis, or cholangiocarcinoma according to the following specific features: HCC,

TABLE 1: Final Diagnosis According to Method of Determination (n = 472)

Diagnosis Malignant lesions Hepatocellular carcinoma Metastasis Cholangiocarcinoma Lymphoma Other malignant tumorsa Benign lesions Focal nodular hyperplasia Hemangioma Adenoma Regenerative macronodule Low-grade dysplastic nodule Abscess Pseudonodular steatosis Other benign hepatic lesionb Total Surgery 50 (10.6) 31 7 8 1 3 14 (2.9) 4 1 6 1 1 1 64 (13.5) Core Needle Biopsy 261 (55.3) 117 113 24 5 2 105 (22.2) 48 6 10 17 2 8 6 8 366 (77.5) Follow-Up 35 (7.4) 18 13 4 7 (1.5) 2 5 42 (8.9) Total 346 (73.3) 166 (35.2) 133 (28.2) 36 (7.6) 6 (1.3) 5 (1.0) 126 (26.7) 52 (11.0) 9 (1.9) 16 (3.4) 18 (3.8) 2 (0.4) 8 (1.7) 7 (1.5) 14 (3.0) 472 (100)

FNA and Core Needle Biopsy Procedure

Four senior radiologists with more than 10 years of experience in abdominal imaging performed liver biopsy at our institution. FNA and core needle biopsy were always performed during the same session; FNA was performed first. Patient sedation (10 mg of diazepam) was administered 12 hours before the procedure. After induction of local anesthesia, the FNA samples were obtained by an attending radiologist using a 22-gauge spinal needle (Spinocan, Braun) or a 22-gauge Westcott biopsy needle under sonographic freehand guidance (Sonoline Elegra sonography unit, Siemens Medical Solutions, or Aplio sonography unit, Toshiba). Coaxial needles were not used. In patients with multiple nodules, the most accessible nodule was targeted. After nodule localization, the needle was introduced and passed gently through the lesion without suction, and the samples were given to the cytopathologist. Direct smears were prepared from

NoteValues are number with percentages in parentheses. Dash [] indicates no case diagnosed by corresponding method. Follow-up includes imaging, clinical, and biologic findings. aIncludes the rare malignant tumors angiosarcoma and leiomyosarcoma. bIncludes the rare benign liver lesions angiomyolipoma, inflammatory pseudotumor of the liver, hydatid disease, granuloma, intrahepatic splenosis, and confluent fibrosis.

AJR:190, March 2008

W209

Pupulim et al.
presence of tridimensional cell groups, basketing pattern (endothelial cells wrapped around groups of hepatocytes), tumoral cells with prominent nucleoli and atypical naked hepatocytic nuclei; metastasis, malignant pleomorphic cells in isolated or loosely cohesive groups with or without an acinar pattern, rare traversing capillaries through cell groups; cholangiocarcinoma, features of adenocarcinoma with tumoral round cells arranged in a papillary or glandular pattern. The results of immediate cytologic analysis were recorded to be compared with those of off-site cytologic examination and with the definitive diagnosis. of benign tumor by the number of proved cases of benign tumor and multiplication of the quotient by 100; positive predictive value, division of the number of correct diagnoses of malignant cases by the total number, correct and incorrect, of diagnoses of malignant tumor and multiplication of the quotient by 100; negative predictive value, division of the total number of correct diagnoses of benign cases by the total number, correct and incorrect, of diagnoses of benign tumors and multiplication of the quotient by 100. The differences in proportion of diagnoses of malignancy in the two groups (immediate cytologic vs definitive cytologic) were compared. The differences were evaluated with the McNemar test because the groups included the same sample of subjects. Differences were considered significant at p < 0.05. All statistical analyses were performed with a statistical software package (StatView version 5.0, SAS Institute).

nant disease. The overall results of the immediate and definitive cytologic examinations are summarized in Table 2. There was no statistical difference in the sensitivity or specificity of the two procedures for the diagnosis of malignancy. Nevertheless, 11 (2.3%) more cases of malignant tumors were detected at definitive cytologic examination than at onsite assessment; this difference was significant according to the McNemar test. Immediate Cytologic Subtyping Subtyping was correct in 254 (90.7%) of the 280 cases diagnosed as malignant on-site. Immediate cytologic diagnosis of malignant subtypes was correct in 113 (68.0%) of 166 cases of HCC, 113 (85.0%) of 133 cases of metastasis, and 28 (77.8%) of 36 cases of cholangiocarcinoma. There were no false-positive diagnoses of cholangiocarcinoma, two false-positive diagnoses of HCC (one cholangiocarcinoma, one metastatic lesion), and six false-positive diagnoses of metastasis (six cholangiocarcinomas). In all of these false-positive cases, the final diagnosis was another malignant tumor, and in all cases of false-positive diagnosis of metastasis, there was no suspected primary site at the time of biopsy. Algorithm An algorithm for the cytohistopathologic management of hepatic tumors was developed on the basis of the results of this study (Fig. 1). The absence of false-positive diagnosis of malignancy at on-site cytologic examination suggests that in well-defined cases, the additional core biopsy procedure is unnecessary, especially for cholangiocarcinoma and metastatic lesions with known primary sites. Use of this algorithm would have allowed 120 (25%) of the patients in this study to avoid additional biopsy (28 patients with cholangiocarcinoma, 92 patients with hepatic metastasis and known primary cancer site). Discussion Our results show that immediate cytologic assessment is a highly accurate procedure in the diagnosis of malignant hepatic tumors, having a sensitivity of 81% and specificity of 100%. Our study also showed that correct subtyping of most malignant lesions (90%) is achieved at immediate cytologic assessment. Moreover, on the basis of these results, we propose an algorithm for a cytohistopathologic approach to hepatic nodules. This study was performed prospectively with consecutive cases of a large number of patients,

Definitive Diagnosis
Sixty-four (13.5%) of the patients underwent surgical treatment, and the definitive diagnosis was determined at histologic analysis of the surgical specimen, which confirmed the core needle biopsy results in 46 cases. For the other 18 patients, the decision for surgery was made even though core needle biopsy results were inconclusive, because the cytologic results or follow-up features indicated the need for surgery. For 366 (77.5%) of the patients, biopsy analysis was used as standard of reference. The other 42 (8.9%) patients had inconclusive histologic findings, and the tumors were not resected. For these patients, the definitive diagnosis was based on imaging, clinical, and biologic followup findings. Lack of tumor growth over 12 years defined a hepatic tumor as benign. Elevation of levels of serum tumor markers and evidence of tumor growth on images or survey of metastatic lesions during follow-up were used as criteria for defining a tumor as malignant. Final diagnoses and the techniques of determination of the diagnoses are summarized in Table 1.

Statistical Analysis
The sensitivity, specificity, positive predictive value, and negative predictive value of immediate and definitive cytologic examinations were calculated as follows: sensitivity, division of the number of correct diagnoses of malignancy by the number of proved malignant tumors and multiplication of the quotient by 100; specificity, division of the total number of correct diagnoses

Results The definitive final diagnoses for the 472 patients are shown in Table 1. Three hundred forty-six (73.3%) of the patients had malignant disease, and 126 (26.7%) had benign lesions. The overall sensitivity and specificity of core needle biopsy were 85% and 100%. The mean numbers of passes for FNA and core needle biopsy for each patient were 1.2 (range, 13 passes) and 1.4 (range, 13 passes), respectively. In nine patients, on-site cytologic analysis was impossible because the material was inadequate, even after three aspirations. The cytopathologist classified these samples as hemorrhagic sample or acellular material, and they were considered negative for malignancy. The final diagnoses for these nine patients were HCC (n = 3), metastasis (n = 2), hemangioma (n = 1), focal nodular hyperplasia (n = 1), pseudonodular steatosis (n = 1), and hepatic abscess (n = 1). Immediate Cytologic Diagnosis of Malignancy A correct diagnosis of malignancy was made at immediate cytologic assessment for 280 (81.0%) of the 346 patients with malig-

TABLE 2: Comparison of the Accuracy Rates of Immediate and Definitive Cytologic Examination in the Diagnosis of Malignant Nodules
Characteristic Sensitivity (%) Specificity (%) Positive predictive value (%) Negative predictive value (%) Immediate Cytologic Examination Definitive Cytologic Examination 81.0 (280/346) 100.0 (126/126) 100.0 (280/280) 65.6 (126/192) 84.1 (291/346) 100.0 (126/126) 100.0 (291/291) 69.6 (126/181)

NoteNumbers in parentheses are the data used to calculate the percentages.

W210

AJR:190, March 2008

Cytologic Diagnosis of Hepatic Tumors

Liver nodule

FNA with immediate cytologic analysis

MALIGNANT

NEGATIVE for malignancy

Cholangiocarcinoma

Metastasis

HCC

No subtyping

Fig. 1Proposed diagnostic algorithm for cytohistopathologic approach to liver nodules. Dotted line indicates that although biopsy is recommended for histologic grading of hepatocellular carcinoma (HCC), therapeutic approach can be initiated after diagnosis on basis of on-site cytologic findings. FNA = fineneedle aspiration.

Primary site Primary site known unknown

THERAPEUTIC STRATEGIES

Core needle biopsy

which strengthens the results and prevents most types of bias. To our knowledge, this series is the largest of patients undergoing immediate cytologic assessment of hepatic lesions. Although FNA is often performed for hepatic tumors, on-site cytologic evaluation is not generally performed at most centers, and published results on this approach are limited [24]. On-site evaluation of hepatic nodules with FNA is often performed only to determine the adequacy of material obtained but not to discern an immediate diagnosis [15, 25]. As in most published reports [2628], the specificity in our series reached 100% because only tumors that had clear evidence of a malignant pattern were classified in the malignant group. To our knowledge, few studies have been conducted to evaluate on-site cytologic diagnosis of hepatic lesions. In a series of evaluations of various sites, Silverman et al. [18] found a sensitivity of 96% and a specificity of 100% for immediate cytologic FNA diagnosis of malignancy in 31 patients with hepatic lesions. Stewart et al. [16] also emphasized the importance of rapid assessment of cytologic preparations. In that series of 141 patients with abdominal lesions (105 with hepatic lesions), a provisional diagnosis was made during the procedure for 103 (73%) FNA specimens (92 malignant cases, 11 reactive or inflammatory lesions). The sensitivity of immediate cytologic evaluation in the diagnosis of malignancy was 71.3%. In that experience, there was no significant difference between provisional and final cytologic diagnoses. We found that on-site cytologic assessment can be used for accurate diagnosis of

the main subtypes of malignant hepatic tumors with a sensitivity of 68% for HCC, 78% for cholangiocarcinoma, and 85% for metastasis. These results confirm those of a study in which the investigators [29] found good performance of immediate cytologic assessment of hepatic tumors with correct subtyping of 71% of lesions. To our knowledge, the diagnosis of cholangiocarcinoma with onsite cytologic assessment has not been studied. Not surprisingly, as with conventional cytologic examination, the diagnosis of HCC with immediate assessment was more difficult than that of metastasis, especially for well-differentiated HCC [3032]. We propose a diagnostic algorithm based on the results of on-site cytologic evaluation (Fig. 1). For instance, patients with a diagnosis of metastasis at immediate cytologic assessment and a known extrahepatic site of primary cancer can avoid further biopsy. In contrast, when the primary site is unknown, biopsy is warranted, especially to obtain additional morphologic and phenotypical features of the malignant proliferation. For primary hepatic cancer, cytologic diagnosis of cholangiocarcinoma appears to be reliable. There were no false-positive cases in our series, so additional core needle biopsy could have been avoided. Finally, in cases of HCC, we still recommend core needle biopsy to evaluate additional prognostic histopathologic features of the tumor (especially the grade of differentiation), although treatment strategies such as radiofrequency ablation can be initiated immediately after diagnosis with immediate cytologic assessment. The proposed algorithm may have an important

effect on the cytohistopathologic management of hepatic tumors. In our series, use of the algorithm would have avoided additional core needle biopsy on 25% of patients. Our study had limitations. First, both immediate and definitive cytologic evaluations were performed by the same cytopathologist. Therefore the definitive cytologic diagnostic accuracy was probably related to or influenced by the immediate assessment, especially when malignancy was diagnosed during the on-site evaluation. Second, only one cytopathologist prospectively interpreted all the findings. We acknowledge that diagnostic accuracy is related to the experience of the cytopathologist, and we did not evaluate interobserver variability. On the other hand, cytologic diagnosis, which can be considered an art, is based on well-defined cytologic features that have been extensively described in the medical literature [12, 1923]. Third, core needle biopsy results were used to determine the definitive diagnosis in a great number of cases, as in most published series. Therefore, we did not compare on-site cytologic results with those from core needle biopsy because such a comparison has inherent bias. Results of some studies have suggested that the caliber of the needles [13, 26] and the small number of passes needed to obtain adequate material [7, 33] may be responsible for the lower rate of complications of FNA compared with core needle biopsy. Thus we can hypothesize that optimization of this technique during the procedure with on-site cytologic evaluation may decrease the risk of complications. Another useful technique for percutaneous biopsy of the liver is the coaxial needle system. In this technique, a needle introducer remains in position while multiple needle passes are made. The technique has been shown [34] to prevent tumor seeding in the needle track. In small tumors, however, adequate needle positioning with coaxial technique can be challenging [35]. The assistance of a pathologist in the radiology department is time-consuming [36], but it may be useful for reducing the number of repeated procedures. Cost-effectiveness, which is another important issue, was beyond the scope of our study. This study showed that at FNA of hepatic tumors, immediate cytologic assessment by a pathologist in the radiology department has high accuracy in the diagnosis of malignancy and can be used to assess tumor subtype in most cases. This practice not only increases diagnostic yield but also may prevent core

AJR:190, March 2008

W211

Pupulim et al. biopsy procedures on selected patients. We recommend that a pathologist routinely perform on-site evaluation of FNA specimens of hepatic tumors. References
1. Siegel CA, Silas AM, Suriawinata AA, van Leeuwen DJ. Liver biopsy 2005: when and how? Cleve Clin J Med 2005; 72:199224 2. Bruix J, Sherman M, Llovet JM, et al. Clinical management of hepatocellular carcinoma: conclusions of the Barcelona 2000 EASL Conference. J Hepatol 2001; 35:421430 3. Bruix J, Sherman M. Management of hepatocellular carcinoma. Hepatology 2005; 42:12081236 4. Grant A, Neuberger J. Guidelines on the use of liver biopsy in clinical practice. Gut 1999; 45[suppl IV]: IV1IV11 5. Wiesner RH, Freeman RB, Mulligan DC. Liver transplantation for hepatocellular cancer: the impact of the MELD allocation policy. Gastroenterology 2004; 127:S261S267 6. Farnum JB, Patel PH, Thomas E. The value of Chiba fine-needle aspiration biopsy in the diagnosis of hepatic malignancy: a comparison with Menghini needle biopsy. J Clin Gastroenterol 1989; 11:101109 7. Tikkakoski T, Paivansalo M, Siniluoto T, et al. Percutaneous ultrasound-guided biopsy: fine needle biopsy, cutting needle biopsy, or both? Acta Radiol 1993; 34:3034 8. Borzio M, Borzio F, Macchi R, et al. The evaluation of fine-needle procedures for the diagnosis of focal liver lesions in cirrhosis. J Hepatol 1994; 20:117121 9. Glenthoj A, Sehested M, Torp-Pedersen S. Diagnostic reliability of histological and cytological fine needle biopsies from focal liver lesions. Histopathology 1989; 15:375383 10. Moulton JS, Moore T. Coaxial percutaneous biopsy technique with automated biopsy devices: value in improving accuracy and negative predictive value. Radiology 1993; 186:515522 11. Franca AVC, Valerio HM, Trevisan M, et al. Fine needle aspiration biopsy for improving the diagnostic accuracy of cut needle biopsy of focal liver lesions. Acta Cytol 2003; 47:332336 12. Dodd L, Mooney E, Layfield L, Nelson R. Fineneedle aspiration of the liver and pancreas: a cytology primer for radiologists. Radiology 1997; 203:19 13. Guo Z, Kurtycz D, Salem R, De Las Casas L, Caya J, Hoerl HD. Radiologically guided percutaneous fine-needle aspiration biopsy of the liver: a retrospective study of 119 cases evaluating diagnostic effectiveness and clinical complications. Diagn Cytopathol 2002; 26:283289 14. Civardi G, Fornari F, Cavanna L, Di Stasi M, Sbolli G, Buscarini L. Value of rapid staining and assessment of ultrasound-guided fine needle aspiration biopsies. Acta Cytol 1988; 32:552554 15. Miller DA, Carrasco CH, Katz RL, Cramer FM, Wallace S, Charnsangavej C. Fine needle aspiration biopsy: the role of immediate cytologic assessment. AJR 1986; 147:155158 16. Stewart CJR, Coldewey J, Stewart IS. Comparison of fine needle aspiration cytology and needle core biopsy in the diagnosis of radiologically detected lesions. J Clin Pathol 2002; 55:9397 17. Wee A. Fine needle aspiration biopsy of the liver: algorithmic approach and current issues in the diagnosis of hepatocellular carcinoma. Cytojournal 2005; 2:7 18. Silverman JF, Finley JL, OBrien KF, et al. Diagnostic accuracy and role of immediate interpretation of fine needle aspiration biopsy specimens from various sites. Acta Cytol 1989; 33:791796 19. Longchampt E, Patriarche C, Fabre M. Accuracy of cytology vs. microbiopsy for the diagnosis of well-differentiated hepatocellular carcinoma and macroregenerative nodule: definition of standardized criteria from a study of 100 cases. Acta Cytol 2000; 44:515523 20. Boer WB, Segal A, Frost F, Sterrett G. Cytodiagnosis of well differentiated hepatocellular carcinoma: can indeterminate diagnosis be reduced? Cancer Cytopathol 1999; 87:270277 21. Bottles K, Cohen MB. An approach to fine-needle aspiration biopsy diagnosis of hepatic masses. Diagn Cytopathol 1991; 7:204210 22. Sbolli G, Fornari F, Civardi G, et al. Role of ultrasound guided fine needle aspiration biopsy in the diagnosis of hepatocellular carcinoma. Gut 1990; 31:13031305 23. Fornari F, Civardi G, Cavanna L, et al. Ultrasonically guided fine-needle aspiration biopsy: a highly diagnostic procedure for hepatic tumors. Am J Gastroenterol 1990; 85:10091013 24. Kocjan G, Feichter G, Hagmar B, et al. Fine needle aspiration cytology: a survey of current European practice. Cytopathology 2006; 17:219226 25. Saleh HA, Khatib G. Positive economic and diagnostic accuracy impacts of on-site evaluation of fine needle aspiration biopsies by pathologists. Acta Cytol 1996; 40:12271230 26. Buscarini L, Fornari F, Bolondi L, et al. Ultrasound-guided fine-needle biopsy of focal liver lesions: techniques, diagnostic accuracy and complicationsa retrospective study of 2091 biopsies. J Hepatol 1990; 11:344348 27. Logrono R, Kurtycz D, Sproat IA, Shalkham JE, Stewart JA, Inhorn SL. Multidisciplinary approach to deep-seated lesions requiring radiologically guided fine-needle aspiration. Diagn Cytopathol 1998; 18:338342 28. Hertz G, Reddy VB, Green L, et al. Fine-needle aspiration biopsy of the liver: a multicenter study of 602 radiologically guided FNA. Diagn Cytopathol 2000; 23:326328 29. Ceyhan K, Kupana S, Bektas M, et al. The diagnostic value of on-site cytopathological evaluation and cell block preparation in fine-needle aspiration cytology of liver masses. Cytopathology 2006; 17:267274 30. Wee A, Nilsson B, Tan LK, Yap I. Fine needle aspiration biopsy of hepatocellular carcinoma: diagnostic dilemma at the ends of the spectrum. Acta Cytol 1994; 38:347354 31. Jain D. Diagnosis of hepatocellular carcinoma: fine needle aspiration cytology or needle core biopsy. J Clin Gastroenterol 2002; 35:S101S108 32. Chhieng DC. Fine needle aspiration biopsy of liver: an update. World J Surg Oncol 2004; 2:5 33. Plecha DM, Goodwin DW, Rowland DY, Varnes M, Haaga JR. Liver biopsy: effects of biopsy needle caliber on bleeding and tissue recovery. Radiology 1997; 204:101104 34. Maturen KE, Nghiem HV, Marrero JA, et al. Lack of tumor seeding of hepatocellular carcinoma after percutaneous needle biopsy using coaxial cutting needle technique. AJR 2006; 187: 11841187 35. Zins M, Vilgrain V, Gayno S, et al. US-guided percutaneous liver biopsy with plugging of the needle track: a prospective study in 72 high-risk patients. Radiology 1992; 184:841842 36. Layfield LJ, Bentz JS, Gopez EV. Immediate on-site interpretation of fine-needle aspiration smears: a cost and compensation analysis. Cancer Cytopathol 2001; 93:319322

W212

AJR:190, March 2008