- Email: [email protected]
Wire-guided brush cytology: a new endoscopic method for diagnosis of bile duct cancer
judgment in regard to how much pressure to apply, knowing full well that applying too much is just as hazardous as applying too little. When utilizing stabilizing sutures, the risk of producing pressure necrosis is almost nil because the retaining device can be applied loosely. The push technique of PEG as described by Russell et al. I3 and others 14 is the preferred method in a patient with a partly obstructed esophagus. This method involves placing a Foley-type catheter through the abdominal wall and into the stomach with endoscopic monitoring. Retention of the gastrostomy tube is dependent upon a soft inflatable balloon. Partial dislodgement with intraperitoneal feeding and intra-abdominal abscess formation has occurred using this technique. I5 It appears that stabilizing sutures as described herein are indicated when using this technique and should be placed before the balloon is expanded. A method of preventing gastrostomy separation by placing four stabilizing sutures was first described by Hashiba et aUG They described a technique in which a specially designed notched needle was used to complete the second limb of the stabilizing suture. At last report they have used their technique successfully in 56 patients. I7 Our technique is basically the same except only two sutures are placed. More important, all equipment used in our technique should be readily available in all gastrointestinal endoscopy and surgical suites. We hope to make PEG safer by this refinement of the technique. REFERENCES 1. Gauderer MWL, Ponsky JL, Izant RJ Jr. Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr
Wire-guided brush cytology: a new endoscopic method for diagnosis of bile duct cancer P. G. Foutch, DO, FACP J. R. Harlan, MD D. Kerr, MD R. A. Sanowski, MD, FACP
Malignant strictures of the bile duct can be difficult to distinguish from benign conditions. Results of ERCP are suggestive but usually not diagnostic. DeReceived July 18, 1988. For revision August 29, 1988. Accepted i September 19, 1988. From the Departments of Gastroenterology ana Pathology, Carl T. Hayden Veterans Administration Medical Center, Phoenix, Arizona. Reprint requests: P. Gregory Foutch, DO, Carl T. Hayden Veterans Administration Medical Center, 7th Street and Indian School Road, Phoenix, Arizona 85012. VOLUME 35, NO.3, 1989
Surg 1980;15:872-5. 2. Ponsky JL, Gauderer MWL. Percutaneous endoscopic gastrostomy: a nonoperative technique for feeding gastrostomy. Gastrointest Endosc 1981;27:9-11. 3. Ponsky JL, Gauderer MWL, Stellato TA. Percutaneous endoscopic gastrostomy. Arch Surg 1983;118:913-4. 4. Foutch PG, Haynes WC, Bellapravalu S, Sanowski RA. Percutaneous endoscopic gastrostomy (PEG): a new procedure comes of age. J Clin Gastroenterol 1986;8:10-5. 5. Larson DE, Burton DO, Schroeder KW, DiMagno EP. Percutaneous endoscopic gastrostomy. Indications, success, complications and mortality in 314 consecutive patients. Gastroenterology 1987;93:48-52. 6. Greif JM, RaglandJJ, Ochsner MG, Riding R. Fatal necrotizing fasciitis complicating percutaneous endoscopic gastrostomy. Gastrointest Endosc 1986;32:292-3. 7. Cave DR, Robinson WR, Brotschi EA. Necrotizing fasciitis following percutaneous endoscopic gastrostomy. Gastrointest Endosc 1986;32:294-6. 8. Bronner MH. Percutaneous endoscopic gastrostomy and crepitus. Gastrointest Endosc 1987;33:270-1. 9. Schnall HA, Falkenstein DB, Raicht RF. Persistent pneumoperitoneum after percutaneous endoscopic gastrostomy. Gastrointest Endosc 1987;33:248-50. 10. Ciocon JO, Silverstone FA, Graver LM, Foley CJ. Tube feedings in elderly patients. Indications, benefits, and complications. Arch Intern Med 1988;148:429-33. 11. Solomon SM, Kirby OF. Percutaneous endoscopic gastrostomy: a matter of choice. Endosc Rev 1988;3:36-45. 12. Foutch PG, Woods CA, Talbert GA, Sanowski RA. A critical analysis of the Sacks-Vine gastrostomy tube: a review of 120 consecutive procedures. Am J GastroenteroI1988;83:812-5. 13. Russell TR, Brotman M, Norris F. Percutaneous gastrostomy: a new simplified and cost effective technique. Am J Surg 1984;148:132-7. 14. Negri G, Cosentino F, Spino GP. Fine-needle endoscopic percutaneous gastrostomy. Endoscopy 1984;16:223-5. 15. Kozarek RA, Ball TJ, Ryan JA Jr. When push comes to shove: a comparison between two methods of percutaneous endoscopic gastrostomy. Am J Gastroenterol 1986;81:642-6. 16. Hashiba K, Fabbri CE, Cappellanes CA, Branco PO, Birolini 0, Oliveira MR. Endoscopic percutaneous gastrostomy without laparotomy. Endoscopy 1984;16:219-22. 17. Hashiba K. Endoscopic gastrostomy. Endoscopy 1987;19:23-4.
tection of metastases or a localized mass lesion by computed tomography (CT) is incriminating but these findings are frequently absent and percutaneous needle aspiration of bile duct strictures is extremely difficult when a mass is not present. 1. 2 Because of their location most cholangiocarcinomas are unresectable and palliative measures are indicated. 3.4 Endoscopic decompression is effective treatment for obstructive jaundice and some studies show enhanced survival for patients managed with radiation and chemotherapy.5-8 However, tumor-directed therapy is inappropriate in the absence of a tissue diagnosis. We have fashioned a cytology brush which can be pushed over a guide wire and precisely placed into tight strictures encountered during ERCP. Our data show that bile duct cancer can be reliably diagnosed by this method and thus specific antitumor treatment can be recommended with assurance. 243
PATIENTS AND METHODS
Eleven strictures (bile duct, nine; pancreatic, two) in nine male patients were brushed during ERCP in an effort to make a cytodiagnosis of cancer. Medical records from these individuals were reviewed to assess technical results of the procedure and eventual outcome for the patient. Blood test results and abnormal findings apparent on CT of the abdomen and those detected during surgery or postmortem examination were noted. All endoscopic procedures were performed by a member of our division using one of two standard side-viewing duodenoscopes (JFITlO, TJFI0; Olympus Corporation, Lake Success, NY) and informed consent was obtained in each case. When indicated 7,10, or 12 F stents or 7 F nasobiliary drainage catheters (Microvasive, Inc., Milford, Mass.) were pushed through the endoscope over a guide wire into the bile duct using standard techniques. 9 Because all patients had some degree of biliary obstruction, 1 g of cefoxitin (Merck, Sharp and Dohme, West Point, Pa.) was administered by vein 30 min prior to ERCP and continued in four daily doses for 24 to 48 hours. Bile duct strictures were considered to occur in the proximal, middle, or distal third of the extrahepatic portion of the duct when located above the cystic duct take-off, between Figure 2. The guide wire has been passed through the lumen of the sheath and out a side hole. The brush can be pushed out and retracted back into the sheath without resistance.
the cystic duct take-off and the pancreas, or within the substance of the pancreas, respectively. Mean age of patients was 75 years (range, 62 to 91 years). The pancreatic and common bile ducts were selectively cannulated for each patient. In two subjects concomitant pancreatic and bile duct strictures were brushed whereas seven individuals had specimens taken from biliary strictures alone. Initial mean values for serum bilirubin (normal <1.4mg/dl) and alkaline phosphatase (normal, 36 to 125 units/liter) were 18.2 mg/dl (range, 1.7 to 60.2 mg/dl) and 768 units/liter (range, 232 to 1583 units/liter), respectively. All patients have been followed up at intervals of 1 to 6 months or until their death. Follow-up has ranged from 2 weeks to 30 months (mean, 7.5 months). TECHNIQUE FOR ACQUISITION OF CYTOLOGICAL SPECIMENS
Figure 1. An endoscopic retrograde cholangiogram showing a distal common duct stricture due to cholangiocarcinoma.
244
After ERCP and radiographic confirmation of a stricture (Fig. 1), a sphincterotomy is made to facilitate insertion of the brush and placement of a drain or stent after the cytology specimen has been collected. The involved duct is cannulated and a 0.025-inch wire is inserted through the cannula and pushed into the duct beyond the stenosis. The cannula is removed and the wire is left in place. A standard cytology brush (Milrose Lab; 230 cm, 2.5 mm in diameter) is used to collect the specimen. The brush is retracted back into its sheath and a side hole is punched through one wall of the sheath with an 18 gauge needle, 3 mm from its distal end. The proximal end of the guide wire which exits the biopsy port of the endoscope is fed into the central lumen of the sheath at its tip and passed out the side hole (Fig. 2). The brush and sheath are pushed over the wire through the GASTROINTESTINAL ENDOSCOPY
endoscope and into the duct to the level of the stricture. The procedure is monitored by fluoroscopy. Difficulty in pushingthe brush out of the endoscope can be avoided by slightly straightening the tip of the instrument when the sheath makes contact with the elevator. The brush is pushed out beyond the sheath and into the stenosis (Fig. 3). The sheath is withdrawn slightly and the handle on the brush is moved in and out to create a to and fro motion on the brush against the stricture. After the specimen is collected, the brush is retracted back into the sheath which is removed from the endoscope over the wire and the sample is placed in a brush washer device (Microvasive Inc.). With the guide wire still in place, a drainage catheter or endoprosthesis can be placed when indicated. In the laboratory, the bile brush washer device is centrifuged for 5 min at 600 g. The supernatant is aspirated away and the sediment is smeared between two glass slides. These are sprayed with cytology fixative, oven dried at 60°C for at least 15 min, then stained by the standard Papanicolaou technique. Slides were interpreted as negative or positive for cancer (Fig. 4).
RESULTS
Seven strictures in six patients were confirmed to be malignant. Causes of malignant biliary obstruction included cholangiocarcinoma in four patients and metastatic lung cancer with nodal compression of the common duct in another individual. Adenocarcinoma of the pancreas caused an intrinsic stricture in one patient but also obstructed the distal common duct.
Both strictures were brushed. In five subjects cancer was proved by: laparotomy, one; cervical lymph node biopsy, one; cytological assessment of ascitic fluid, one; postmortem examination, two. A 91-year-old man had characteristic cholangiographic findings consistent with a proximal (Klatskin) bile duct cancer and a mass lesion in the porta hepatis on CT. The brush cytology specimen from this lesion was positive for cancer. The patient died 2 months after diagnosis and request for postmortem examination was denied. Four benign strictures were present in three patients. Chronic pancreatitis caused an intrinsic stricture in one subject but the inflammatory disease process also obstructed the distal common duct. Both lesions were brushed. A second patient had chronic pancreatitis and common duct obstruction but without stenosis of the main pancreatic duct. A third patient with chronic cholecystitis had an inflammatory mass around the gallbladder which also obstructed the common hepatic duct. Benignity of these lesions was confirmed at laparotomy in all cases. Each stricture (totalll) was brushed safely and all specimens were adequate for cytological examination. Of the seven malignant strictures, cytological specimens collected during ERCP were positive in four (sensitivity, 57%). There were no false positive results (specificity, 100%). All four patients with cholangiocarcinoma were diagnosed by this method (sensitivity for this lesion, 100%). Two cancers were located in
B Figure 3. A radiograph (A) and schematic drawing (B) showing the guide wire in an intrahepatic bile duct and the brush in the stricture. VOLUME 35, NO.3, 1989
245
Figure 4. Cluster of pleomorphic malignant cells (A) with high nuclear to cytoplasm ratios. A sheet of benign ductal cells (B)
recovered from the same patient is shown for comparison (Papanicolaou stain, original magnification x450).
the proximal, one in the middle, and one in the distal segment of the duct. A positive result in two patients permitted treatment with radiation and chemotherapy without the need for surgery. CT showed dilated bile ducts in all cases. In six patients with cancer, mass lesions responsible for strictures were detected in three (50%) but were absent in three (50%). Brush cytology specimens were positive for cancer in the latter three cases. Overall, eight of nine patients had a biliary stent or drain placed immediately after collection of the cytological specimen. Endoscopic management was sufficient in five cases. Our oldest patient (91 years) with cholangiocarcinoma had a nasobiliary drainage catheter placed as access for intracavitary irradiation. This treatment was combined with chemotherapy and external beam radiation but when the patient refused medical care, the catheter was removed. All three patients with benign disease eventually had definitive surgical treatment. DISCUSSION
It is logical to attempt to make a diagnosis of cancer in obstructed patients by assessment of exfoliated cells in bile and the concept is attractive because of its simplicity. Most large studies have evaluated subjects with pancreatic cancer. 1O- I2 Because cholangiocarcinoma is an uncommon neoplasm, there are relatively few reports which primarily assess the value of cytology in patients with this disease. Bile samples have 246
been obtained from percutaneous drains, T-tubes, duodenal drainage, and common duct aspiration performed at the time of surgery or ERCP, but the ability to diagnose cholangiocarcinoma by these methods has been disappointing. 13- I7 The desmoplastic nature of these tumors may be largely responsible for these results. It is possible that directly brushing the tumor can enhance the cancer detection rate and the endoscopic method which we describe has several important advantages. The procedure appears to be safe and relatively simple as long as the guide wire can be inserted above the strictured segment. Malignant lesions at all levels of the extrahepatic biliary system are accessible. In our series positive results were obtained in patients with tumors located in the proximal, middle, and distal thirds of the duct. The brush technique consistently retrieves large numbers of cells which usually cluster in groups or sheets. This facilitates assessment of nuclear pleomorphism and other features indicative of cancer. 18 A precise diagnosis can be made in a relatively short period of time which can guide the use of radiation and chemotherapy and in certain cases obviate the need for surgery. This should improve quality of patient care and lead to substantial savings in hospital costs. The procedure can be conveniently combined with endoscopic therapy, including placement of stents and drains. Consequently, diagnosis and treatment can be rendered in the same procedure. An endoscopic approach also permits accurate diagnosis GASTROINTESTINAL ENDOSCOPY
of concomitant pancreatic disease which is helpful in the overall assessment of the patient. The sensitivity of our method may not be high for other lesions. Metastatic lung cancer and adenocarcinoma of the pancreas were not detected. Since individual strictures were brushed only once, it is unknown if additional samples would improve the diagnostic yield. However, tumors which extrinsically compress a duct or manifest predominant submucosal growth without intraductal exfoliation are unlikely to be diagnosed by cytology. It is important to note that false positive cytology results did not occur and this observation has been corroborated by other studies on bile specimens collected by different methods. 15 •17, 19 High specificity greatly enhances the clinical value of this test, particularly when supported by characteristic abnormalities detected by ERCP and other imaging modalities. Consequently, a positive cytological specimen is sufficient evidence for cancer and other diagnostic procedures are unnecessary. Wire-guided brush cytology is safe, relatively simple, and convenient and probably best suited for diagnosis of patients with primary malignant strictures of the bile duct. We recommend that all biliary strictures of unknown cause be routinely brushed during ERCP in an effort to make a diagnosis of cancer. ACKNOWLEDGMENTS
The authors thank Thomas Young, LPN, for his technical assistance and Joan Hillier for preparation of the manuscript.
REFERENCES 1. Hall-Craggs MA, Lees WR. Fine-needle aspiration biopsy: pancreatic and biliary tumors. Am J RoentgenoI1986;147:399-403. 2. Mendez G Jr, Russell E, Levi JU, et al. Percutaneous brush biopsy and internal drainage of biliary tree through endoprosthesis. Am J RoentgenoI1980;134:653-9. 3. Alexander F, Rossi RL, O'Bryan M, et al. Biliary carcinoma: a
VOLUME 35, NO.3, 1989
review of 109 cases. Am J Surg 1984;147:503-9. 4. Brandt-Rauf PW, Pincus MR, Adelson S. Carcinoma of the extrahepatic bile ducts: some new perspectives. Surv Dig Dis 1984;2:156-63. 5. Hishikawa Y, Shimada T, Miura T, et al. Radiation therapy of carcinoma of the extrahepatic bile ducts. Radiology 1983;146:787-9. 6. Wheeler PG, Dawson JL, Nunnerly H, et al. Newer techniques in the diagnosis and treatment of proximal bile duct carcinoma-an analysis of 41 consecutive patients. Quart J Med 1981;199:247-58. 7. Kopelson G, Harisiadis L, Tretter P, et al. The role of radiation therapy in cancer of the extrahepatic biliary system: an analysis of thirteen patients and a review of the literature of the effectiveness of surgery, chemotherapy and radiotherapy. Int J Radiat Oncol Bioi Phys 1977;2:883-94. 8. Hashmonai M. Long survival following combined treatment of inoperable cholangiocarcinoma: surgery, radiotherapy and chemotherapy. J Surg OncoI1980;13:231-9. 9. Cotton PB. Endoscopic methods for relief of malignant obstructive jaundice. World J Surg 1984;8:854-61. 10. Endo Y, Morii T, Tamura H, et al. Cytodiagnosis of pancreatic malignant tumors by aspiration under direct vision using a duodenal fiberscope. Gastroenterology 1974;67:944-51. 11. Hatfield ARW, Smithies A, Wilkins R, et al. Assessment of endoscopic retrograde cholangiopancreatography (ERCP) and pure pancreatic juice cytology in patients with pancreatic disease. Gut 1976;17:14-21. 12. Harada H, Sasaki T, Yamamoto N, et al. Assessment of endoscopic aspiration cytology and endoscopic retrograde cholangipancreatography (ERCP) in patients with cancer of the pancreas: Part I. Gastroenterol Jpn 1977;12:52-8. 13. Nishimura A, Den N, Sato H, et al. Exfoliative cytology of the biliary tract with the use of saline irrigation under choledochoscopic control. Ann Surg 1973;178:594-9. 14. Vilardell F. Cytological diagnosis of digestive cancer. Am J GastroenteroI1978;70:357-64. 15. Roberts-Thomson IC, Hobbs JB. Cytodiagnosis of pancreatic and biliary cancer by endoscopic duct aspiration. Med J Aust 1979;1:370-2. 16. Harell GS, Anderson MF, Berry PF. Cytologic bile examination in the diagnosis of biliary duct neoplastic strictures. Am J RoentgenoI1981;137:1123-6. 17. Cobb CJ, Floyd WN. Usefulness of bile cytology in the diagnostic management of patients with biliary tract obstruction. Acta CytoI1985;29:93-100. 18. Osnes M, Serck-Hanssen A, Myren J. Endoscopic retrograde brush cytology (ERBC) of the biliary and pancreatic ducts. Scand J Gastroenterol 1975;10:829-31. 19. Muro A, Mueller PR, Ferrucci JT, et al. Bile cytology: a routine addition to percutaneous biliary drainage. Radiology 1983;149:846-7.
247
Wire-guided brush cytology: a new endoscopic method for diagnosis of bile duct cancer P. G. Foutch, DO, FACP J. R. Harlan, MD D. Kerr, MD R. A. Sanowski, MD, FACP
Malignant strictures of the bile duct can be difficult to distinguish from benign conditions. Results of ERCP are suggestive but usually not diagnostic. DeReceived July 18, 1988. For revision August 29, 1988. Accepted i September 19, 1988. From the Departments of Gastroenterology ana Pathology, Carl T. Hayden Veterans Administration Medical Center, Phoenix, Arizona. Reprint requests: P. Gregory Foutch, DO, Carl T. Hayden Veterans Administration Medical Center, 7th Street and Indian School Road, Phoenix, Arizona 85012. VOLUME 35, NO.3, 1989
Surg 1980;15:872-5. 2. Ponsky JL, Gauderer MWL. Percutaneous endoscopic gastrostomy: a nonoperative technique for feeding gastrostomy. Gastrointest Endosc 1981;27:9-11. 3. Ponsky JL, Gauderer MWL, Stellato TA. Percutaneous endoscopic gastrostomy. Arch Surg 1983;118:913-4. 4. Foutch PG, Haynes WC, Bellapravalu S, Sanowski RA. Percutaneous endoscopic gastrostomy (PEG): a new procedure comes of age. J Clin Gastroenterol 1986;8:10-5. 5. Larson DE, Burton DO, Schroeder KW, DiMagno EP. Percutaneous endoscopic gastrostomy. Indications, success, complications and mortality in 314 consecutive patients. Gastroenterology 1987;93:48-52. 6. Greif JM, RaglandJJ, Ochsner MG, Riding R. Fatal necrotizing fasciitis complicating percutaneous endoscopic gastrostomy. Gastrointest Endosc 1986;32:292-3. 7. Cave DR, Robinson WR, Brotschi EA. Necrotizing fasciitis following percutaneous endoscopic gastrostomy. Gastrointest Endosc 1986;32:294-6. 8. Bronner MH. Percutaneous endoscopic gastrostomy and crepitus. Gastrointest Endosc 1987;33:270-1. 9. Schnall HA, Falkenstein DB, Raicht RF. Persistent pneumoperitoneum after percutaneous endoscopic gastrostomy. Gastrointest Endosc 1987;33:248-50. 10. Ciocon JO, Silverstone FA, Graver LM, Foley CJ. Tube feedings in elderly patients. Indications, benefits, and complications. Arch Intern Med 1988;148:429-33. 11. Solomon SM, Kirby OF. Percutaneous endoscopic gastrostomy: a matter of choice. Endosc Rev 1988;3:36-45. 12. Foutch PG, Woods CA, Talbert GA, Sanowski RA. A critical analysis of the Sacks-Vine gastrostomy tube: a review of 120 consecutive procedures. Am J GastroenteroI1988;83:812-5. 13. Russell TR, Brotman M, Norris F. Percutaneous gastrostomy: a new simplified and cost effective technique. Am J Surg 1984;148:132-7. 14. Negri G, Cosentino F, Spino GP. Fine-needle endoscopic percutaneous gastrostomy. Endoscopy 1984;16:223-5. 15. Kozarek RA, Ball TJ, Ryan JA Jr. When push comes to shove: a comparison between two methods of percutaneous endoscopic gastrostomy. Am J Gastroenterol 1986;81:642-6. 16. Hashiba K, Fabbri CE, Cappellanes CA, Branco PO, Birolini 0, Oliveira MR. Endoscopic percutaneous gastrostomy without laparotomy. Endoscopy 1984;16:219-22. 17. Hashiba K. Endoscopic gastrostomy. Endoscopy 1987;19:23-4.
tection of metastases or a localized mass lesion by computed tomography (CT) is incriminating but these findings are frequently absent and percutaneous needle aspiration of bile duct strictures is extremely difficult when a mass is not present. 1. 2 Because of their location most cholangiocarcinomas are unresectable and palliative measures are indicated. 3.4 Endoscopic decompression is effective treatment for obstructive jaundice and some studies show enhanced survival for patients managed with radiation and chemotherapy.5-8 However, tumor-directed therapy is inappropriate in the absence of a tissue diagnosis. We have fashioned a cytology brush which can be pushed over a guide wire and precisely placed into tight strictures encountered during ERCP. Our data show that bile duct cancer can be reliably diagnosed by this method and thus specific antitumor treatment can be recommended with assurance. 243
PATIENTS AND METHODS
Eleven strictures (bile duct, nine; pancreatic, two) in nine male patients were brushed during ERCP in an effort to make a cytodiagnosis of cancer. Medical records from these individuals were reviewed to assess technical results of the procedure and eventual outcome for the patient. Blood test results and abnormal findings apparent on CT of the abdomen and those detected during surgery or postmortem examination were noted. All endoscopic procedures were performed by a member of our division using one of two standard side-viewing duodenoscopes (JFITlO, TJFI0; Olympus Corporation, Lake Success, NY) and informed consent was obtained in each case. When indicated 7,10, or 12 F stents or 7 F nasobiliary drainage catheters (Microvasive, Inc., Milford, Mass.) were pushed through the endoscope over a guide wire into the bile duct using standard techniques. 9 Because all patients had some degree of biliary obstruction, 1 g of cefoxitin (Merck, Sharp and Dohme, West Point, Pa.) was administered by vein 30 min prior to ERCP and continued in four daily doses for 24 to 48 hours. Bile duct strictures were considered to occur in the proximal, middle, or distal third of the extrahepatic portion of the duct when located above the cystic duct take-off, between Figure 2. The guide wire has been passed through the lumen of the sheath and out a side hole. The brush can be pushed out and retracted back into the sheath without resistance.
the cystic duct take-off and the pancreas, or within the substance of the pancreas, respectively. Mean age of patients was 75 years (range, 62 to 91 years). The pancreatic and common bile ducts were selectively cannulated for each patient. In two subjects concomitant pancreatic and bile duct strictures were brushed whereas seven individuals had specimens taken from biliary strictures alone. Initial mean values for serum bilirubin (normal <1.4mg/dl) and alkaline phosphatase (normal, 36 to 125 units/liter) were 18.2 mg/dl (range, 1.7 to 60.2 mg/dl) and 768 units/liter (range, 232 to 1583 units/liter), respectively. All patients have been followed up at intervals of 1 to 6 months or until their death. Follow-up has ranged from 2 weeks to 30 months (mean, 7.5 months). TECHNIQUE FOR ACQUISITION OF CYTOLOGICAL SPECIMENS
Figure 1. An endoscopic retrograde cholangiogram showing a distal common duct stricture due to cholangiocarcinoma.
244
After ERCP and radiographic confirmation of a stricture (Fig. 1), a sphincterotomy is made to facilitate insertion of the brush and placement of a drain or stent after the cytology specimen has been collected. The involved duct is cannulated and a 0.025-inch wire is inserted through the cannula and pushed into the duct beyond the stenosis. The cannula is removed and the wire is left in place. A standard cytology brush (Milrose Lab; 230 cm, 2.5 mm in diameter) is used to collect the specimen. The brush is retracted back into its sheath and a side hole is punched through one wall of the sheath with an 18 gauge needle, 3 mm from its distal end. The proximal end of the guide wire which exits the biopsy port of the endoscope is fed into the central lumen of the sheath at its tip and passed out the side hole (Fig. 2). The brush and sheath are pushed over the wire through the GASTROINTESTINAL ENDOSCOPY
endoscope and into the duct to the level of the stricture. The procedure is monitored by fluoroscopy. Difficulty in pushingthe brush out of the endoscope can be avoided by slightly straightening the tip of the instrument when the sheath makes contact with the elevator. The brush is pushed out beyond the sheath and into the stenosis (Fig. 3). The sheath is withdrawn slightly and the handle on the brush is moved in and out to create a to and fro motion on the brush against the stricture. After the specimen is collected, the brush is retracted back into the sheath which is removed from the endoscope over the wire and the sample is placed in a brush washer device (Microvasive Inc.). With the guide wire still in place, a drainage catheter or endoprosthesis can be placed when indicated. In the laboratory, the bile brush washer device is centrifuged for 5 min at 600 g. The supernatant is aspirated away and the sediment is smeared between two glass slides. These are sprayed with cytology fixative, oven dried at 60°C for at least 15 min, then stained by the standard Papanicolaou technique. Slides were interpreted as negative or positive for cancer (Fig. 4).
RESULTS
Seven strictures in six patients were confirmed to be malignant. Causes of malignant biliary obstruction included cholangiocarcinoma in four patients and metastatic lung cancer with nodal compression of the common duct in another individual. Adenocarcinoma of the pancreas caused an intrinsic stricture in one patient but also obstructed the distal common duct.
Both strictures were brushed. In five subjects cancer was proved by: laparotomy, one; cervical lymph node biopsy, one; cytological assessment of ascitic fluid, one; postmortem examination, two. A 91-year-old man had characteristic cholangiographic findings consistent with a proximal (Klatskin) bile duct cancer and a mass lesion in the porta hepatis on CT. The brush cytology specimen from this lesion was positive for cancer. The patient died 2 months after diagnosis and request for postmortem examination was denied. Four benign strictures were present in three patients. Chronic pancreatitis caused an intrinsic stricture in one subject but the inflammatory disease process also obstructed the distal common duct. Both lesions were brushed. A second patient had chronic pancreatitis and common duct obstruction but without stenosis of the main pancreatic duct. A third patient with chronic cholecystitis had an inflammatory mass around the gallbladder which also obstructed the common hepatic duct. Benignity of these lesions was confirmed at laparotomy in all cases. Each stricture (totalll) was brushed safely and all specimens were adequate for cytological examination. Of the seven malignant strictures, cytological specimens collected during ERCP were positive in four (sensitivity, 57%). There were no false positive results (specificity, 100%). All four patients with cholangiocarcinoma were diagnosed by this method (sensitivity for this lesion, 100%). Two cancers were located in
B Figure 3. A radiograph (A) and schematic drawing (B) showing the guide wire in an intrahepatic bile duct and the brush in the stricture. VOLUME 35, NO.3, 1989
245
Figure 4. Cluster of pleomorphic malignant cells (A) with high nuclear to cytoplasm ratios. A sheet of benign ductal cells (B)
recovered from the same patient is shown for comparison (Papanicolaou stain, original magnification x450).
the proximal, one in the middle, and one in the distal segment of the duct. A positive result in two patients permitted treatment with radiation and chemotherapy without the need for surgery. CT showed dilated bile ducts in all cases. In six patients with cancer, mass lesions responsible for strictures were detected in three (50%) but were absent in three (50%). Brush cytology specimens were positive for cancer in the latter three cases. Overall, eight of nine patients had a biliary stent or drain placed immediately after collection of the cytological specimen. Endoscopic management was sufficient in five cases. Our oldest patient (91 years) with cholangiocarcinoma had a nasobiliary drainage catheter placed as access for intracavitary irradiation. This treatment was combined with chemotherapy and external beam radiation but when the patient refused medical care, the catheter was removed. All three patients with benign disease eventually had definitive surgical treatment. DISCUSSION
It is logical to attempt to make a diagnosis of cancer in obstructed patients by assessment of exfoliated cells in bile and the concept is attractive because of its simplicity. Most large studies have evaluated subjects with pancreatic cancer. 1O- I2 Because cholangiocarcinoma is an uncommon neoplasm, there are relatively few reports which primarily assess the value of cytology in patients with this disease. Bile samples have 246
been obtained from percutaneous drains, T-tubes, duodenal drainage, and common duct aspiration performed at the time of surgery or ERCP, but the ability to diagnose cholangiocarcinoma by these methods has been disappointing. 13- I7 The desmoplastic nature of these tumors may be largely responsible for these results. It is possible that directly brushing the tumor can enhance the cancer detection rate and the endoscopic method which we describe has several important advantages. The procedure appears to be safe and relatively simple as long as the guide wire can be inserted above the strictured segment. Malignant lesions at all levels of the extrahepatic biliary system are accessible. In our series positive results were obtained in patients with tumors located in the proximal, middle, and distal thirds of the duct. The brush technique consistently retrieves large numbers of cells which usually cluster in groups or sheets. This facilitates assessment of nuclear pleomorphism and other features indicative of cancer. 18 A precise diagnosis can be made in a relatively short period of time which can guide the use of radiation and chemotherapy and in certain cases obviate the need for surgery. This should improve quality of patient care and lead to substantial savings in hospital costs. The procedure can be conveniently combined with endoscopic therapy, including placement of stents and drains. Consequently, diagnosis and treatment can be rendered in the same procedure. An endoscopic approach also permits accurate diagnosis GASTROINTESTINAL ENDOSCOPY
of concomitant pancreatic disease which is helpful in the overall assessment of the patient. The sensitivity of our method may not be high for other lesions. Metastatic lung cancer and adenocarcinoma of the pancreas were not detected. Since individual strictures were brushed only once, it is unknown if additional samples would improve the diagnostic yield. However, tumors which extrinsically compress a duct or manifest predominant submucosal growth without intraductal exfoliation are unlikely to be diagnosed by cytology. It is important to note that false positive cytology results did not occur and this observation has been corroborated by other studies on bile specimens collected by different methods. 15 •17, 19 High specificity greatly enhances the clinical value of this test, particularly when supported by characteristic abnormalities detected by ERCP and other imaging modalities. Consequently, a positive cytological specimen is sufficient evidence for cancer and other diagnostic procedures are unnecessary. Wire-guided brush cytology is safe, relatively simple, and convenient and probably best suited for diagnosis of patients with primary malignant strictures of the bile duct. We recommend that all biliary strictures of unknown cause be routinely brushed during ERCP in an effort to make a diagnosis of cancer. ACKNOWLEDGMENTS
The authors thank Thomas Young, LPN, for his technical assistance and Joan Hillier for preparation of the manuscript.
REFERENCES 1. Hall-Craggs MA, Lees WR. Fine-needle aspiration biopsy: pancreatic and biliary tumors. Am J RoentgenoI1986;147:399-403. 2. Mendez G Jr, Russell E, Levi JU, et al. Percutaneous brush biopsy and internal drainage of biliary tree through endoprosthesis. Am J RoentgenoI1980;134:653-9. 3. Alexander F, Rossi RL, O'Bryan M, et al. Biliary carcinoma: a
VOLUME 35, NO.3, 1989
review of 109 cases. Am J Surg 1984;147:503-9. 4. Brandt-Rauf PW, Pincus MR, Adelson S. Carcinoma of the extrahepatic bile ducts: some new perspectives. Surv Dig Dis 1984;2:156-63. 5. Hishikawa Y, Shimada T, Miura T, et al. Radiation therapy of carcinoma of the extrahepatic bile ducts. Radiology 1983;146:787-9. 6. Wheeler PG, Dawson JL, Nunnerly H, et al. Newer techniques in the diagnosis and treatment of proximal bile duct carcinoma-an analysis of 41 consecutive patients. Quart J Med 1981;199:247-58. 7. Kopelson G, Harisiadis L, Tretter P, et al. The role of radiation therapy in cancer of the extrahepatic biliary system: an analysis of thirteen patients and a review of the literature of the effectiveness of surgery, chemotherapy and radiotherapy. Int J Radiat Oncol Bioi Phys 1977;2:883-94. 8. Hashmonai M. Long survival following combined treatment of inoperable cholangiocarcinoma: surgery, radiotherapy and chemotherapy. J Surg OncoI1980;13:231-9. 9. Cotton PB. Endoscopic methods for relief of malignant obstructive jaundice. World J Surg 1984;8:854-61. 10. Endo Y, Morii T, Tamura H, et al. Cytodiagnosis of pancreatic malignant tumors by aspiration under direct vision using a duodenal fiberscope. Gastroenterology 1974;67:944-51. 11. Hatfield ARW, Smithies A, Wilkins R, et al. Assessment of endoscopic retrograde cholangiopancreatography (ERCP) and pure pancreatic juice cytology in patients with pancreatic disease. Gut 1976;17:14-21. 12. Harada H, Sasaki T, Yamamoto N, et al. Assessment of endoscopic aspiration cytology and endoscopic retrograde cholangipancreatography (ERCP) in patients with cancer of the pancreas: Part I. Gastroenterol Jpn 1977;12:52-8. 13. Nishimura A, Den N, Sato H, et al. Exfoliative cytology of the biliary tract with the use of saline irrigation under choledochoscopic control. Ann Surg 1973;178:594-9. 14. Vilardell F. Cytological diagnosis of digestive cancer. Am J GastroenteroI1978;70:357-64. 15. Roberts-Thomson IC, Hobbs JB. Cytodiagnosis of pancreatic and biliary cancer by endoscopic duct aspiration. Med J Aust 1979;1:370-2. 16. Harell GS, Anderson MF, Berry PF. Cytologic bile examination in the diagnosis of biliary duct neoplastic strictures. Am J RoentgenoI1981;137:1123-6. 17. Cobb CJ, Floyd WN. Usefulness of bile cytology in the diagnostic management of patients with biliary tract obstruction. Acta CytoI1985;29:93-100. 18. Osnes M, Serck-Hanssen A, Myren J. Endoscopic retrograde brush cytology (ERBC) of the biliary and pancreatic ducts. Scand J Gastroenterol 1975;10:829-31. 19. Muro A, Mueller PR, Ferrucci JT, et al. Bile cytology: a routine addition to percutaneous biliary drainage. Radiology 1983;149:846-7.
247