Abdom Imaging 22:307–310 (1997)

Abdominal Imaging q Springer-Verlag New York Inc. 1997

A percutaneous biopsy technique for patients with suspected biliary or pancreatic cancer without a radiographic mass R. E. Pelsang, F. C. Johlin Departments of Radiology and Internal Medicine, Division of Gastroenterology, University of Iowa College of Medicine, 200 Hawkins Drive, Iowa City, IA 52242, USA Received: 30 January 1996/Accepted: 28 March 1996

Abstract Background: Treatment of malignant pancreatic and/or biliary strictures requires tissue diagnosis. Since cytologic brushings at endoscopic retrograde cholangiopancreatography (ERCP) of these strictures has a poor sensitivity for malignancy (30–83%) (see ME Ryan. Gastrointestinal Endoscopy 1991;37(2):139–143; and MB Cohen, Wittchow RJ, Johlin FC, et al. Mod Pathol 1995;8:498–502), tissue diagnosis must be obtained by another route. We report our experience of percutaneous biopsy of malignant pancreatic and/or biliary strictures even when no radiographic mass is present. Methods: At ERCP, five patients demonstrated pancreatic and/or biliary duct strictures, had atypical cytological brushings, and had their strictures stented. No mass to account for the strictures could be identified on CT. These five patients underwent percutaneous biopsy of the peristent material by CT within 10 days of the ERCP. Results: Two patients had adenocarcinoma of the pancreas. One patient had malignant lymphoma and another had cholangiocarcinoma. One patient had inflammatory cells and was followed. Conclusions: If the diagnosis of malignancy cannot be made at the time of the ERCP sampling, then our experience suggests that a percutaneous biopsy should be performed even if a mass is not present using the stent as a target. Key words: Percutaneous biopsy—Biliary—Pancreas—Cancer—Computed tomography—ERCP.

Thirty-five % of patients ultimately proven to have pancreatic cancer have a ductal stricture but no identifiable Correspondence to: R. E. Pelsang

mass [1]. Because the only hope for curative resection depends on early diagnosis, it is these cases of small tumors that require precise tissue diagnosis. Yet in such cases readily obtaining a tissue diagnosis can be difficult as washings from ductal strictures are positive in only 30% of cases [2]. We present our experience in five patients with radiographic pancreatic and/or biliary duct strictures on endoscopic retrograde cholangiopancreatography (ERCP) whose cytologic brushings were atypical but not diagnostic for malignancy, and had no demonstrable mass. Biopsy in all five were performed percutaneously under CT guidance using the ductal stent placed at ERCP as the target.

Materials and methods Patients From February 1995 to June 1995, five patients (two women, three men) ages 45 to 74 years (mean: 65 years of age) who presented with jaundice (2), cholangitis (1), or relapsing pancreatitis (2) were referred for ERCP. Once a pancreatic and/or biliary stricture was demonstrated, cytologic brushings were performed at ERCP. At this institution, brushings are performed on all pancreatic and/or biliary strictures except on anastomotic strictures at the level of the duct to duct anastomosis. Atypical to suspicious cells were obtained from the cytologic brushings in these five patients who then received a CT-guided biopsy of the strictured area within 10 days of their ERCP.

Procedure All endoscopic procedures were performed by one of the authors (FCJ) using a standard diagnostic videoduodenalscope (Olympus, TJF100, Japan). Pancreatography and cholangiography were performed in a selective directed manner. Brush biopsies were performed after balloon dilatation. Brushings were fixed in 95% ethanol solution and stained with a modified Papanicolaou stain. An 11.5 French endoprosthesis was placed through the stricture.

4v0a 0014 Mp 307 Friday Feb 28 02:24 PM SV-Ab Image (v. 22, #4) 0014 (1383)

308

R. E. Pelsang and F. C. Johlin: Biopsy for biliary/pancreatic cancer

All percutaneous biopsies were performed by one of the authors (RP) under CT guidance within 10 days of the ERCP. All patients were scanned in a Siemens Somatom DRH CT scanner (Iselin, NJ). Initial 8-mm slice thickness and 10-mm intervals were use to localize the stent followed by 4-mm slice thickness and 4-mm intervals to determine a biopsy route. Stent localization bypassed the need for oral or intravenous contrast media. Biopsy needle-tip position was confirmed by CT guidance (Fig. 1, 2) and the needle was then attached to a 12-cc Luer lock syringe. Fine needle aspiration biopsy specimens were obtained with 3–4 cc of gentle suction and agitation until blood appeared at the needle hub or until approximately 15 s of agitation had occurred. Specimens were given to one of two cytologists who were present for all biopsies. The biopsies were interpreted by these two cytologists with knowledge of the patient’s history, radiographic findings, and ERCP results. The number of fine-needle biopsy passes ranged from two to five. Twenty-two- to 23-gauge Chiba brand needles (Cook, Bloomington, IN) of varying lengths were used. One patient returned for additional biopsies to obtain DNA sample for possible malignant lymphoma where a 22-gauge Franseen needle (Cook, Bloomington, IN) was used for fine-needle aspiration biopsies and a 20-gauge Bard Monopty biopsy gun (Bard Radiology, Covington, GA) was used for a core sample.

Results Radiographic interpretation of ERCP

Fig. 1. Patient no. 1 with a stricture of the common bile duct and pancreatic duct on ERCP. CT-guided percutaneous biopsy of peristent material in the pancreas diagnosed adenocarcinoma. Black arrows identify pancreatic duct stent. White arrowhead denotes percutaneous biopsy needle with tip adjacent to pancreatic duct stent.

Two patients had biliary strictures and one had a pancreatic stricture. Two patients had both biliary and pancreatic strictures. CT interpretation There was no identifiable focal mass on CT to account for the pancreatic and/or biliary strictures. One patient had ascites and small, õ1-cm retroperitoneal nodes. Two patients had signs of pancreatitis with diffusely enlarged pancreas and mesenteric fat inflammatory changes. Biopsy results All had atypical to suspicious cells on cytologic brushings, but none were diagnostic for malignancy. Percutaneous biopsy diagnosed adenocarcinoma of the pancreas, cholangiocarcinoma, inflammatory disease, and malignant lymphoma of the pancreas. A preliminary diagnosis of adenocarcinoma of the pancreas was made at the time of the biopsy for patient no. 5, but the final pathologic report did not confirm malignancy.

Fig. 2. Patient no. 2 with a stricture of the common hepatic duct on ERCP. CT-guided percutaneous biopsy of peri-stent material in the liver diagnosed cholangiocarcinoma. Black curved arrow identifies biliary duct stent. Black on white arrow denotes percutaneous biopsy needle with tip anterior to biliary duct stent.

Complications One patient was admitted and monitored for bleeding after ERCP’s sphincterotomy. No transfusions were necessary and the bleeding was controlled with an injection of 1:10,000 epinephrine at the bleeding site. No

complications resulted from the percutaneous CTguided biopsy procedure itself. However, the lack of a final malignant diagnosis for patient no. 5 from the fineneedle aspirated specimens could be included as a complication. This patient’s CT biopsy procedure was ter-

4v0a 0014 Mp 308 Friday Feb 28 02:24 PM SV-Ab Image (v. 22, #4) 0014 (1383)

R. E. Pelsang and F. C. Johlin: Biopsy for biliary/pancreatic cancer

309

minated after only two samples, at which time the cytologist gave the preliminary diagnosis of adenocarcinoma. However, the cytologist was unable to confirm malignancy in the final report. This patient subsequently went to surgery for a definitive diagnosis where adenocarcinoma of the pancreas was confirmed. However, he was unresectable at the time of surgery due to vascular encasement.

moplastic reaction by the tumor or obstruction of the main pancreatic duct causing acute or chronic pancreatitis. Repeat percutaneous biopsy is advocated to evaluate negative percutaneous biopsies. Teplick et al. repeated biopsies in 27 patient with previously negative biopsies. Of these 27 patients, 21 became true positives for malignancy [11]. No complications resulted from our percutaneous biopsy procedure itself. The reported complication rate of percutaneous pancreatic biopsies under CT, ultrasound, or fluoroscopic guidance ranges from 3.3 to 8.6% [11, 13]. Minor complications include vasovagal reactions, hematoma, pain, and fever, which rarely require hospitalization. Major complications (1.1%) include pancreatitis requiring hospitalization and a report of a pancreatic duct leak requiring surgery [13]. A few anecdotal cases in literature have described seeding after pancreatic biopsies [14, 15] and after endoscopic sphincterotomy [16]. Treatment protocols require tissue diagnoses. If this diagnosis cannot be obtained at the time of ERCP sampling, and malignancy remains in the differential diagnosis, then percutaneous biopsy should be performed. Our study population was limited to patients with atypical cells on cytologic brushings. Since this technique is relatively easy, less invasive than surgery, and has relatively few complications, percutaneous biopsy of peristent material could also be performed in those patients with a negative or inadequate cytologic sampling but who have a high likelihood of underlying malignancy.

Discussion Pancreas or biliary neoplasms produce strictures of the pancreas or biliary ducts or both. Patient may present with jaundice in the case of biliary obstruction or with symptoms of pancreatitis in the case of pancreatic duct obstruction. Since neoplasm of the pancreas and bile ducts arise from ductal epithelium, tissue from the abnormal ductal segment seems appropriate for pathologic examination. Cytologic brushing at the time of ERCP has a sensitivity of 30% but a specificity of 100% for malignant pancreatic strictures and 48.8–83% sensitivity and 98–100% specificity for malignant biliary strictures [2, 3]. False-positive results are exceptionally rare. False-negative results, however, need continued scrutiny. Reasons for false-negative results include extensive fibrosis or ulceration of the neoplasm, high degree of tumor differentiation, low-grade or benign epithelium overlying malignant strictures, difficult anatomic sites, or poor visualization [4]. At some centers, the radiographic interpretation of a CT scan of the pancreas for change consistent with malignancy has a sensitivity of 76% and a specificity of 50–58% [5]. If a mass is seen, then percutaneous biopsy can be done under CT, ultrasound, or fluoroscopy. Accuracy of 50–94% has been reported for CT-guided pancreas biopsies [6, 7] and 67–97% for ultrasoundguided pancreatic biopsies [8–10]. Teplick et al. [11] performed 64% of his 201 pancreas and/or extrahepatic bile duct biopsies under fluoroscopy at a time other than the ERCP. He reported a combined accuracy of 82.4% for biopsies performed with ultrasound or fluoroscopy. Gagnon et al. [12] biopsied 14 patients using bi-plane fluoroscopy during pancreatography with a sensitivity of 90% and specificity of 100%. His series had no complications from this technique. Radiation exposure during this biplane technique, however, would be significant. Also, in most centers biplane fluoroscopy is not available in the ERCP suite. How are negative percutaneous biopsy results handled? Similar to our series, a minimum of four sampling attempts is advocated [11]. When to stop sampling varies with the patient, the types of samples obtained during the current procedure, and clinical judgement. Even large masses can produce false-negative results. This is usually secondary to sampling errors due to the des-

References 1. Ferrari AP, Lichtenstein DR, Slivka A, Chang C, Carr-Locke DL. Brush cytology during ERCP for the diagnosis of biliary and pancreatic malignancy. Gastrointest Endosc 1994;40:140–145 2. Ryan ME. Cytologic brushings of ductal lesions during ERCP. Gastrointest Endosc 1991;37:139–143 3. Cohen MB, Wittchow RJ, Johlin FC, Bottles K, Raab SS. Brush cytology of the extra hepatic biliary tract: comparison of cytologic features of adenocarcinoma and benign biliary strictures. Mod Pathol 1995;8:498–502 4. Layfield LJ, Wax TD, Lee JG, Cotton PB. Accuracy and morphologic aspects of pancreatic and biliary duct brushings. Acta Cytol 1995;39:11–18 5. Rodriguez J, Kasberg C, Nipper M, Schoolar J, Riggs MW, Dyck WP. CT-guided needle biopsy of the pancreas: a retrospective analysis of diagnostic accuracy. Am J Gastroenterol 1992; 87:1610–1613 6. Fekete PS, Nunez C, Pitlik VA. Fine needle aspiration biopsy of the pancreas: a study of 61 cases. Diagn Cytopathol 1986;2:301–306 7. Del Maschio A, Vanzulli A, Sironi S, Castrucci M, Mellone R, Staudocher C, Carlucci M, Zerbi A, Parolini D, Faravelli A, Cantaboni A, Garancini P, DiCarlo V. Pancreatic cancer vs. chronic pancreatitis: diagnosis of CA 19-9 assessment. Ultrasound, CT, and CT-guided fine needle biopsy. Radiology 1991;178:95–98 8. Hall-Craggs MA, Lees WR. Fine needle aspiration biopsy: pancreatic and biliary tumors. AJR 1986;147:399–403 9. Hajud AD, Kamari-Subaiya S, Phillips G. Ultrasonically-guided

4v0a 0014 Mp 309 Friday Feb 28 02:24 PM SV-Ab Image (v. 22, #4) 0014 (1383)

310

R. E. Pelsang and F. C. Johlin: Biopsy for biliary/pancreatic cancer

percutaneous aspiration biopsy of the pancreas. Semin Diagn Pathol 1986;3:166–175 10. Bret PM, Nicolet B, Labadie M. Percutaneous fine needle aspiration biopsy of the pancreas. Diagn Cytopathol 1986;2:221–222 11. Teplick SK, Haskin PH, Kline TS, Sammon JK, Laffey PA. Percutaneous pancreaticobiliary biopsies in 173 patients using primary ultrasound or fluoroscopic guidance. Cardiovasc Interv Radiol 1988;11:26–28 12. Gagnon P, Boustiere C, Ponchon T, Valette P-J, Genin G, Labadie M. Percutaneous fine-needle aspiration cytologic study of main pancreatic duct stenosis under pancreatographic guidance. Cancer 1991;67:2395–2400

13. Brandt KR, Charboneau JW, Stephens DH, Welch TJ, Goellner JR. CT and US guided biopsy of the pancreas. Radiology 1993; 187:99–104 14. Rashleigh-Belcher HJC, Russell RCG, Lees WR. Cutaneous seeding of pancreatic carcinoma by fine-needle aspiration biopsy. Br J Radiol 1986;59:183–185 15. Ferrucci JT, Wittenberg J, Margolies MN, Carey RW. Malignant seeding of the tract after thin-needle aspiration biopsy. Radiology 1979;130:345–346 16. Studley JGN, Sami AR, Williamson RCN. Dissemination of pancreatic carcinoma following endoscopic sphincterotomy. Endoscopy 1993;25:301–302

4v0a 0014 Mp 310 Friday Feb 28 02:24 PM SV-Ab Image (v. 22, #4) 0014 (1383)