Annals of Oncology 0: 7, 999. 999 Kluwer Academic Publishers. Printed in the Netherlands. Original article Tumors of the papilla of Vater inadequate diagnostic impact of endoscopic forceps biopsies taken prior to and following sphincterotomy J. Menzel, C. Poremba, K.H. Dietl, W. Bocker & W. Domschke Departments of 'Medicine B, Pathology, ^General Surgery, University ofmunster, Munster, Germany Summary Background: It has been proposed that s of the papilla of Vater are precursors of adenocarcinomas. Duodenoscopy with ERCP and forceps biopsies have substantially improved the morphologic exploration of the major duodenal papilla. Yet there is little and contradictory information as to the diagnostic accuracy of endoscopic biopsies in tumors of the papilla. Moreover, after endoscopic sphincterotomy data on the diagnostic impact of endoscopic biopsies from the papilla are scarce and, in most cases, retrospectively obtained. Thus, the aim of the present prospective and histopathologically controlled study was to assess the diagnostic accuracy of endoscopic biopsies taken from tumors of the papilla before and after sphincterotomy. Patients and methods: Forty patients with tumors of the papilla of Vater were included in the study. In each case, a comparison was made between endoscopic forceps biopsy diagnoses prior to and following sphincterotomy and the definitive histological diagnosis after surgical tumor resection. Results: Resected tumors were diagnosed histomorphologically as follows: 9 adenocarcinomas (7%), 6 tubular s (%), 7 villous s (7%), 7 inflammatory nonneoplastic lesions (pseudotumors) (7%), and one adenomyoma (%). Overall accuracy for preoperative histopathological diagnosis was 6% ( of 0, 9% CI: 7%76%) prior to sphincterotomy while it was 70% (8 of 0, 9% CI: %8%) following the procedure. Regarding adenocarcinomas, sensitivity was found to be % ( of 9, 9% CI: 8%%) prior to and 7% (7 of 9, 9% CI: 9%8%) after sphincterotomy while specificity was 00% at both times. Conclusions: Endoscopic forceps biopsies do not allow for reliable preoperative diagnosis of tumors of the papilla of Vater. Key words: endoscopic biopsy, endoscopic sphincterotomy, tumors of the papilla of Vater Introduction Histomorphological studies support the hypothesis that invasive carcinomas of the papilla of Vater arise from preexisting mucosal lesions, such as or dysplasia [, ]. Despite the fact that duodenal and ampullary epithelial neoplasms are being recognized more frequently with increasing use of duodenoscopy, valid data on the incidence and prevalence of papilla of Vater s are still not available. However, an accurate preoperative diagnosis is essential to select the right patients and the most appropriate treatment. Endoscopically obtained forceps biopsies from suspicious papillas can establish an early and immediate preoperative diagnosis although even for skilled pathologists it is difficult, in some cases, to distinguish invasive carcinomas from noninvasive lesions on the basis of forceps biopsies. Moreover, forceps biopsy diagnosis of does not rule out the possibility of deeper carcinoma []. To improve the diagnostic accuracy of biopsies from suspicious papillas harvesting of specimens from the depth of the papilla following sphincterotomy is recommended [6]. However, surgically controlled data to support this recommendation are not available. Information regarding the diagnostic accuracy of endoscopic biopsies is limited and variable since the vast majority of studies is retrospective [7, 8]. To assess the preoperative diagnostic accuracy of biopsies taken from tumors of the papilla of Vater which, on endoscopy, are suspicious of neoplasia, we performed a prospective and histopathologically controlled study. The second goal was to evaluate whether or not sphincterotomy is able to improve diagnostic accuracy of endoscopic forceps biopsies. Patients and methods Patients Patients were admitted because of clinical symptoms such as jaundice, abdominal discomfort, or dilated bile ducts on sonography. Distorted and polypoid enlarged papillas of Vater were detected endoscopically. From May 99 to September 998,0 consecutive patients (8 women, men; age range 07 years, mean 6 years) with polypoid tumors of the papilla of Vater were included in this study. Following surgical resection, tumors of the papilla of Vater were finally diagnosed to be adenocarcinomas, s, inflammatory nonneoplastic lesions and adenomyoma. Patients with pancreatic head carcinomas were excluded, since it was not the objective of this study to differentiate pancreatic head carcinomas from tumors of the papilla of Vater. Study protocol: none of the patients included had previously under Downloaded from https://academic.oup.com/annonc/articleabstract/0/0/7/0967 on 0 March 08
8 from the papilla without sphincterotomy: end of first examination (Figures la, b). Following staging with transabdominal ultrasonography, CT scan, EUS, usually within two to three days, a second ERCP was performed with sphincterotomy and four to six immediate forceps biopsies were taken from the depth of the papilla. Subsequently, surgery was performed in all study patients within two weeks. Techniques of surgery employed were papillectomy, pyloruspreserving resection of the pancreatic head, or Whipple's resection. All patients gave their written informed consent to participate in the study. The study protocol had previously been approved by the Local Ethics Committee of Muenster County and University Hospital. Methods Duodenoscopy was performed with fiberscopes (Olympus JFT0, Olympus Optical, Tokyo, Japan). Biopsies were taken directly with French forceps (MTW, Wesel, Germany), and sphincterotomies were performed employing. French guidewire sphincterotomes (Ultratome, Boston Scientific, Watertown, USA). The formalinfixed tissue was embedded in paraffin, and sections were stained with hematoxylineosin and PAS (periodic acid Schiff reaction). Assessment of biopsy specimens and resection specimens, respectively, was done by two experienced pathologists independently (C.P., W.B.). In the resection specimens the neoplastic tissue as well as adjacent surface and duct epithelium were examined extensively, to a maximum distance of. cm from the periphery of the tumor. Tumor typing and grading was performed according to the WHO classification of intestinal tumors [, 9]. Hyperplasia was diagnosed on the basis of the AlboresSaavedra criteria [0], s were defined as benign epithelial lesions consisting of tubular or villous components. In all of these lesions epithelial alterations were then classified according the WHO criteria for the presence or absence of dysplasia. Dysplasia, if identified, was diagnosed as lowgrade or highgrade (Figures ac). We calculated the 9% confidence interval (9% CI) to compare relative frequencies and the Fisher's exact test was used for statistical analysis. The statistical significance level was P < 0.0. Results After surgery the resected tumors of the papilla of Vater were definitively diagnosed as adenocarcinomas in 9 patients (7.%), tubular s in 6 (%), villous s in 7 (7.%) and adenomyoma in one patient (.%). The remaining seven patients were diagnosed as having inflammatory nonneoplastic lesions (pseudotumors) (7.%). There were no cases of familial polyposis coli among our patients. Sizes of carcinomas and benign tumors of the papilla of Vater were similar: In carcinomas, the mean size was 6. mm ( mm, ± 7 mm) and in benign tumors 6.0 mm (0 mm, ± 8 mm). Endoscopic biopsy diagnosis prior to sphincterotomy Figure. (a) ERCP of a patient presenting with elevated cholestatic enzymes. Dilated common bile duct (CBD) and pancreatic duct (PD). (b) Endoscopic appearance of the papilla of Vater in the respective patient. gone any manipulation of the papilla of Vater including biopsy or sphincterotomy. In all patients the order of procedures for diagnosis and tumor staging was as follows: Endoscopic inspection of the papilla. ERCP and harvesting of up to 6 superficial forceps biopsies Downloaded from https://academic.oup.com/annonc/articleabstract/0/0/7/0967 on 0 March 08 Endoscopic biopsy specimens obtained from tumors of the papilla prior to sphincterotomy allowed for diagnoses of inflammatory nonneoplastic pseudotumors, tubular s, 0 villous s and carcinomas. Two endoscopically obtained specimens were largely acellular or with severe artifacts and therefore not sufficient for pathological diagnosis (Table ). Five of the eleven tumors which had endoscopically been diagnosed to be inflammatory nonneoplastic
9 Table. Comparison of presphincterotomy and postsurgery diagnoses. Presphincterotomy Postsurgery Carci AdenoSpecimen Pseudo Tubular insufficient tumor noma myoma for I ft. ; Tubular Table. Comparison of postsphincterotomy and postsurgery diagnoses. Postsphincterotomy Postsurgery Specimen Pseudo Tubular insufficient tumor for Tubular Figure. Specimens from the patient presented in Figure. (a) Biopsy specimen prior to sphincterotomy: chronic inflammation and stromal fibrosis. HE stain (x 00). (b) Biopsy specimen after sphincterotomy: tubular with moderate nuclear atypia (arrow). HE stain (x 00). (c) Resection specimen revealing carcinoma of the papilla of Vater (pt N 0 M x ). Stroma invasion of atypical tubules (arrows). HE stain (x 00). pseudotumors finally turned out to be carcinomas, while only three of the eleven tumors were confirmed to be inflammatory nonneoplastic pseudotumors. Among the remaining tumors, two were eventually diagnosed as tubular s and one as adenomyoma. In patients, endoscopic biopsies revealed tubular s presenting with mild (6), moderate (), and severe () epithelial dysplasia. After surgical resection, three tubular s, five carcinomas, three inflammatory nonneoplastic pseudotumors, and two villous s were diagnosed. Cardnoma Adenomyoma 6 7 7 Preoperatively, 0 villous s were encountered presenting with mild (), moderate (), and severe dysplasia (). Five of these ten tumors were finally confirmed to be villous s while the remaining specimens revealed four carcinomas and one inflammatory nonneoplastic pseudotumor. In two cases, endoscopic biopsies obtained from the papilla prior to sphincterotomy were not suitable for pathological examination (Table ). These tumors were eventually diagnosed as tubular and carcinoma, respectively. Endoscopic biopsy diagnosis after sphincterotomy Endoscopic biopsy specimens obtained from tumors of the papilla after sphincterotomy allowed for diagnosis of 7 inflammatory nonneoplastic pseudotumors, tubular s, villous s, and 7 carcinomas. Two endoscopically obtained specimens were not suitable for pathological diagnosis (Table ). Three of the seven endoscopically diagnosed nonneoplastic pseudotumors were postoperatively found to be carcinomas. Regarding tubular s, distribution of detected dysplasias did not differ from presphincterotomy diag Downloaded from https://academic.oup.com/annonc/articleabstract/0/0/7/0967 on 0 March 08
0 Table. Nineteen carcinomas of the papilla of Vater: endoscopic biopsy diagnoses prior to and after sphincterotomy. Tubular Endoscopic diagnoses Prior to sphincterotomy After sphincterotomy noses. Four out of twelve endoscopically diagnosed tubular s were finally confirmed to be tubular s. However, resection specimens revealed six carcinomas and two nonneoplastic pseudotumors. Postsphincterotomy biopsies revealed villous s which in 7 cases were confirmed postsurgically. In the remaining five cases, however, resection specimens allowed for diagnosis of three carcinomas, one inflammatory nonneoplastic pseudotumor, and one tubular, respectively. s of the papilla of Vater Following surgical tumor resection, 9 carcinomas of the papilla of Vater were diagnosed: TiN 0 M x, T N 0 M x, T N,M X, T N O M X, TaN.M,, T N,M X. Biopsy specimens prior to sphincterotomy truly revealed of these 9 carcinomas while in the remaining carcinomas inflammatory nonneoplastic pseudotumors (), tubular s (), or villous s () were diagnosed. One forceps biopsy specimen did not allow for pathological diagnosis. After sphincterotomy, 7 carcinomas of the papilla of Vater were truly diagnosed by endoscopic biopsies while in the remaining carcinomas inflammatory nonneoplastic pseudotumors (), tubular s (6) and villous s () were determined (Table ). Accuracy of endoscopic biopsy diagnosis Endoscopic biopsy specimens of forty tumors of the papilla of Vater obtained prior to sphincterotomy allowed for correct preoperative histopathological diagnosis in cases, resulting in an overall accuracy of 6% (P < 0.0 to differentiate benign from malignant specimen). Regarding benign papillary tumors specificity rated at 00% while in carcinomas sensitivity was only % ( of 9). After endoscopic sphincterotomy, the overall accuracy for biopsy specimen diagnosis insignificantly increased to 70% (8 of 0) while specificity remained at 00%. In the subset of carcinomas sensitivity increased from % (9% CI: 8%%) to 7% (9% CI: 9% 8%). Discussion Results of this histopathologically controlled, prospective study indicate that endoscopic biopsies obtained from tumors of the papilla of Vater prior to and after sphincterotomy do not allow for adequate preoperative diagnosis. Particularly the sensitivity rate of % as found in carcinomas is definitely unsatisfactory. To improve the diagnostic accuracy of endoscopic biopsies several authors have recommended harvesting of biopsies following endoscopic sphincterotomy [ 6, ]. Yet in our series proceeding along these lines, as many as out of 9 carcinomas were missed. These data confirm earlier reports of retrospective studies [,,] suggesting that diagnostic accuracy of papillary forceps biopsies is essentially limited. This failure may be due to some sampling error inherent in biopsy technique and to the fact that overlying mucosal lesions like s or dysplasias may disguise deeper located carcinomas. The limited accuracy of preoperative diagnoses of tumors of the papilla of Vater, however, gives rise to much controversy regarding the appropriate diagnostic and therapeutical management of patients with suspected, but yet unproven neoplasms of the papilla of Vater [6]. Owing to this diagnostic dilemma of endoscopic forceps biopsies, further prospective studies should be designed to elucidate whether or not endoscopic snare excisions yielding more sample tissue material might improve preoperative diagnostic accuracy in tumors of the papilla of Vater. Combining ERCP with miniprobe ultrasonography promises a new diagnostic modality that has some potential advantages for local staging of small tumors of the papilla of Vater. Intraductal ultrasonography (IDUS) using miniprobes is superior to conventional endoscopic ultrasonography (EUS) in detection and staging of tumors of the papilla of Vater. Especially the high accuracy of IDUS in assessment of tumor size and staging might have considerable impact on therapeutic procedures. In cases of periampullary local tumor resection such as transduodenal ampullectomy can be performed instead of Whipple's procedure [7,, ]. Based on data from the literature and results of the present study, we should like to recommend the following sequence of diagnostic procedures: In case of enlarged or suspicious papillas, endoscopic biopsies from the surface should be obtained as well as biopsies from the depth immediately after sphincterotomy. Once a carcinoma or an with dysplasia has been diagnosed, the tumor should surgically be resected whenever possible. If, however, biopsies are negative while the tumor remains evident endoscopically, more sophisticated imaging techniques including CT and intraductal ultrasonography using miniprobes should be employed to advance diagnostics []. Downloaded from https://academic.oup.com/annonc/articleabstract/0/0/7/0967 on 0 March 08
Acknowledgement We are indebted to C. Sauerland, Department of Medical Statistics, for her valuable support. References. Baczako K, Biichler M, Beger HG et al. Morphogenesis and possible precursor lesions of invasive carcinoma of the papilla of Vater: Epithelial dysplasia and. Hum Pathol 98; 6: 00.. Kozuka S, Tsubone M, Yamaguchi A et al. Adenomatous residue in cancerous papilla of Vater. Gut 98; : 0.. Seifert E, Schulte F, Stolte M. Adenoma and carcinoma of the duodenum and papilla of Vater: A clinicopathologic study. Am J Gastroenterol 99; 87: 7^.. Ponchon T, Berger F, Chavaillon A et al. Contribution of endoscopy to diagnosis and treatment of tumors of the ampulla of Vater. Cancer 989; 6: 67.. Huibregtse K, Tytgat GNJ. of the ampulla of Vater: The endoscopic approach. Endoscopy 988; 0: 6. 6. Bourgeois N, Dunham F, Verhest A et al. Endoscopic biopsies of the papilla of Vater at the time of endoscopic sphincterotomy: Difficulties in interpretation. Gastrointest Endosc 98; 0: 66. 7. Nakao NL, Siegel JH, Stenger RJ et al. Tumors of the ampulla of Vater: Early diagnosis by intraampullary biopsy during endoscopic cannulation. Gastroenterology 98; 8: 96. 8. Sauvanet A, Chapuis O, Hammel P et al. Are endoscopic procedures able to predict the benignity of ampullary tumors? Am J Surg 997; 7: 8. 9. Kimura W, Ohtsubo K.. Incidence, sites of origin, and immunohistochemical and histochemical characteristics of atypical epithelium and minute carcinoma of the papilla of Vater. Cancer 988; 6: 90. 0. AlboresSaavedra J, AlcantraVazquez A, CruzOrtiz H et al. The precursor lesions of invasive gallbladder carcinoma. Hyperplasia, atypical hyperplasia and carcinoma in situ. Cancer 980; : 99 7.. Topazian M, Salem RR. Sphincter of Oddi dysfunction caused by an ampullary neoplasm. Gastroenterology 99; 08: 6.. Leese T, Neoptolemos JP, West K.P et al. Tumours and pseudotumours of the region of the ampulla of Vater: An endoscopic, clinical and pathological study. Gut 986; 7: 869.. Martin FM, Rossi RL, Dorrucci Vet al. Clinical and pathologic correlations in patients with periampullary tumors. Arch Surg 990; : 76.. Yamaguchi K, Enjoji M. of the ampulla of vater. A clinicopathologic study and pathologic staging of 09 cases of carcinoma and cases of. Cancer 987; 9: 06.. Kimchi NA, Mindrul V, Broide E et al. The contribution of endoscopy and biopsy to the diagnosis of periampullary tumors. Endoscopy 998; 0: 8. 6. Asbun HJ, Rossi RL, Munson JL. Local resection for ampullary tumors. Is there a place for it? Arch Surg 99; 8: 0. 7. Rosenberg J, Welch JP, Pyrtek LJ et al. Benign villous s of the papilla of Vater. Cancer 986; 8: 68. 8. Goldberg M, Zamir O, Hadary A et al. Wide local excision as an alternative treatment for periampullary carcinoma. Am J Gastroenterol 987; 8: 697. 9. Grace PA, Pitt HA, Longmire WP. Pylorus preserving pancreatoduodenectomy: An overview. Br J Surg 990; 77: 9687. 0. HenneBruns D, Kremer B, Meyer Pannwitt U et al. Partial duodenopancreatectomy with radical lymphadenectomy in patients with pancreatic and periampullary carcinomas: Initial results. Hepatogastroenterology 99; 0: 9.. Chijiiwa K, Yamashita H, Kuroki S. Wide ampullectomy for patients with villous of duodenal papilla and followup results of pancreaticobiliary tract. Int Surg 99; 79: 788.. Binmoeller KF, Boaventura S, Ramsperger K et al. Endoscopic snare excision of benign s of the papilla of Vater. Gastrointest Endosc 99; 9: 7.. Gouma DJ, Obertop H, Vismans J et al. Progression of a benign epithelial ampullary tumor to adenocarcinoma. Surgery 987; 0:0.. Menzel J, Hoepffner N, Sulkowski U et al. Polypoid tumors of the major duodenal papilla: Preoperative staging with intraductal US, EUS, and CT a prospective, histopathologically controlled study. Gastrointest Endosc 999; 9: 97. Received June 999; accepted August 999. Correspondence to: J. Menzel, MD Department of Medicine B University of Miinster AlbertSchweitzerStr. D89 Munster Germany Email: jmenzel@unimuenster.de Downloaded from https://academic.oup.com/annonc/articleabstract/0/0/7/0967 on 0 March 08