Printer Friendly

Carcinomas of the pancreas, gallbladder, extrahepatic bile ducts, and ampulla of vater share a field for carcinogenesis: a population-based study.

The pancreas, biliary system, and ampulla of Vater share a common embryologic origin. Primarily ductal systems that facilitate digestion, these 3 anatomic sites are derived from the embryonic foregut. (1) Furthermore, the co-occurrence of carcinomas of the pancreas, extrahepatic bile ducts (EHBD), and ampulla and the coexistence of ampullary neoplasms with pancreatic intraepithelial neoplasia has been described. (2-6) This association suggests that tumors originating in the pancreas, gallbladder, EHBD, or ampulla are possibly related perhaps due to the common origin during embryogenesis. In this article, we provide epidemiologic and analytic evidence that malignant epithelial tumors of the pancreas, gallbladder, EHBD, and ampulla represent a common population that suggests a common pathogenesis, indicative of a potential field phenomenon for carcinogenesis. Specifically, the rate of cancer development by age is the same in all 4 sites even though the incidence of cancer varies among the sites, and the frequency of cancer may result from the relative surface area of the ductal system in each site.


Data Source

Data were obtained from the Surveillance Epidemiology and End Results (SEER) Program at the National Cancer Institute. (7) SEER has collected demographic, anatomic, histologic, extent of disease, mortality, and limited treatment information on cancer patients since 1973. Data were obtained for the years 1973 through 2005 from SEER's Registry 9, which covers approximately 12% of the US population. There were 3419 cases of cancer of the ampulla, 6838 of the extrahepatic bile ducts, 9744 of the gallbladder, and 81 042 cases of pancreatic cancer. Nonepithelial tumors including endocrine tumors, carcinoids, sarcomas, and leiomyosarcomas totaled 1806 (2.2%) in the pancreas, 73 (0.7%) in the gallbladder, 27 (0.4%) in the EHBD, and 103 (3.0%) cases in the ampulla. Because the fraction of nonepithelial cancers of the pancreas was very small, less than 4%, they were included in the analysis. The following code numbers were used for the non-epithelial tumors: 8013, 8150-8157, 8240-8249, 8800-8806, 8890-8896. All other International Classification of Diseases for Oncology (ICD-O) codes were used for the carcinomas. The Registry 9 Database includes SEER's original registries that covered Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco-Oakland, Seattle-Puget Sound, and Utah. For endocrine carcinomas, data were obtained only from cases recorded from 1992 to 2005 in SEER (Registry 13) because the 1992 edition of the ICD-O specifically included codes for endocrine pancreatic tumors. The validity of the morphologic data submitted to SEER has been considered previously. (8)

Histopathologic Codes

In SEER, the histopathologic types of neoplasia and grade assignment are coded from the pathology reports according to the ICD-O published by the World Health Organization (1992). Data are not available for benign tumors. Specific codes are also available for anatomic sites.

Data Analysis

Incidence trends and age-specific incidence rates (2000 US standard population) were calculated using SEER*Stat (6.1.4) and expressed as number of cases per 100 000 persons per year (100 000 person years). Only invasive cancers with a histologic diagnosis were included. Overall, 94% of cases reported to SEER have been histologically confirmed. Cases identified through death certificates only were excluded. All racial/ethnic groups were combined.

Incidence rates were age adjusted although age-specific rates represented crude rate. Age-specific incidence rates, which were calculated at 5-year intervals, were plotted on linear as well as on logarithmic scales. Age frequency density plots were constructed for each year from age 0 to 110 on linear scales. These plots represent the age of diagnosis for each year with a total probability of 1.0 for the population, which normalizes the age distribution of the diagnosis of cancer for each year for each site. (9)


Incidence Rates

Overall incidence rates from 1973 through 2005 for cancers of the pancreas, gallbladder, EHBD, and ampulla of Vater were 11.71, 1.43, 0.88, and 0.49 per 100 000 persons, respectively.

Incidence Rate Trends

From 1973 through 2005, the incidence rates of pancreatic EHBD and ampullary cancers remained relatively constant, although there was a decrease in the incidence of gallbladder cancer that was not statistically significant (Figure 1). Cancers of the pancreas are about 20 times more common than those in the ampulla. In view of the observation that rate patterns for these cancers have remained stable for at least 33 years, and given that the diagnostic criteria have not changed, the consistent rate patterns suggests that there has been no changes in disease etiology, in primary prevention, screening or detection rate, or in susceptibility to malignant transformation.

Age-Specific Incidence Rates

The age-specific incidence rates of epithelial cancers arising in the pancreas, gallbladder, EHBD, and ampulla from 1973 through 2005 were calculated at 5-year intervals to age 85+ (Figure 2). Pancreatic cancer is clearly the most common and its age-specific rate rises rapidly in all age groups.

However, it is often more informative to explore age-specific incidence rates by logarithmic plots The age-specific rates were graphically analyzed on a logarithmic scale and plotted as the log age-specific rate versus the log age at diagnosis (Figure 3). The log-log plot revealed parallel linear rate patterns for the 4 sites. The mathematic transformation often reveals information that may not be obvious on linear graphic patterns. On a log-log plot, parallel rate patterns indicate that age-specific incidence rates are increasing by the same exponential rate. Therefore, parallel rate patterns in log-log plots indicate that the percentage change per unit time of each rate is the same even though the rates among the 4 tumor sites differ in absolute incidence.

Age Frequency Density Plots

The age frequency density plot for carcinomas of the pancreas, gallbladder, EHBD, and ampulla were identical for cancers originating in all 4 sites (Figure 4). The 4 isomorphous distributions have mean ages of 70.3, 71.2, 71.5, and 69.7 years, respectively.

Endocrine Carcinomas of the Pancreas

As a comparison, we used endocrine tumors of the pancreas. Endocrine pancreatic tumors, which have the same embryologic origin as the pancreatic ductal epithelium, follow different differentiation pathways than ductal cells. Endocrine pancreatic tumors are morphologically, functionally, and biologically different from tumors of the ductal epithelium. They also have a different age frequency density distribution (Figure 5) and their rate of development is not parallel with the rate of development of the ductal carcinomas in log-log plots (Figure 6). We conclude from this analysis that endocrine tumors constitute a different population than ductal carcinomas and, therefore, do not share in the field effect as seen in the malignant epithelial tumors.


We have considered the pathoepidemiology of carcinomas arising in the pancreas, gallbladder, EHBD, and ampulla of Vater listed in the SEER Program. In all 4 of the anatomic sites, carcinomas were found predominantly in older age groups, the mean age of diagnosis being around 70 years. Because it is often difficult to determine the exact origin of some ampullary carcinomas, it is possible that for some cases listed in SEER the carcinoma arose from the periampullary region of the duodenum and not from the mucosa of the ampulla proper. Nonetheless, carcinomas of the pancreas, EHBD, and ampulla share similar morphologic phenotypes. For example, carcinomas with intestinal and pancreatobiliary phenotypes have been described in these anatomic sites, complicating even further the identification of the primary site. (10,11) Endocrine carcinomas, on the other hand, are far less common than ductal carcinomas, representing less than 3% of all cancers (1765 cases) of the pancreas. Thus, by far most carcinomas in our study were ductal in origin.




Our results revealed a potential common pathogenesis for carcinomas arising in the pancreas, gallbladder, EHBD, and ampulla of Vater. As background to the log-log plots (Figure 3), Armitage and Doll (12,13) in 1954 demonstrated that the age-specific incidence rates of certain epithelial cancers have a linear slope when plotted against the age of diagnosis on logarithmic scales. This logarithmic transformation of population data confirmed a theoretical equation relating the sequential development of carcinoma to a series of molecular changes and cellular events, and thereby supporting the multistage process of carcinogenesis with a continuous increase in cancer rate as a function of age. (14) Many investigators have since exploited logarithmically scaled plots to investigate both the rates and origins of human cancer. (15-17) These log-log plots represent a mathematic transformation that provide insight into tumor development and comparisons among tumor populations. Similar graphical patterns imply that similar tumors arising in different organs might represent a single population with similar mechanisms of transformation or a similar carcinogenic process.

As demonstrated by the age-specific incidence rates (Figure 2), pancreatic cancer is significantly more common than cancers of the gallbladder, EHBD, or ampulla in all age groups. Because on a log-log plot a straight line indicates that tumor incidence increases at a constant exponential rate, the parallel lines (Figure 3), therefore, are indicative of a single cohort population among the tumors in the 4 organs that differ only in absolute incidence. Thus, according to the parallel graphic patterns in the log-log plots (Figure 3), the rate of carcinoma development is essentially the same in all 4 sites. Therefore, because of similar rates, the pathogenesis of carcinomas arising in the pancreas, gallbladder, EHBD, or ampulla may be similar or related. Armitage and Doll considered the log-log plots as a mathematical construct that relates to molecular/cellular events in carcinogenesis. Similar slopes therefore suggest similar mechanisms of carcinogenesis. Certainly, this conclusion is not unreasonable, because these anatomic sites have a common embryologic origin. The femalemale ratios for carcinomas of the pancreas, ampulla, and EHBD are, respectively, 0.74, 0.62, and 0.64, providing indirect but additional support for a similar carcinogenic development of these cancers. However, the ratio is 1.74 for gallbladder cancer, which is known to have a female predominance, owing to the additional risk factor of cholelithiasis that is more common in women.

These results seem to have a bearing on the etiology of cancer of these foregut derivatives. Gallstones along with inflammation are often considered the cause of most cancers of the gallbladder, but gallstones are not considered a cause of cancers of the EHBD, ampulla, or pancreas. Yet, the cancers arising in these sites have similar pathoepidemiologic characteristics. Thus, there must be something that is common in the carcinogenic process in the epithelium in these 4 sites, although the inciting environmental agent initiating and promoting the process may differ for each site.



The concept of field cancerization implies that cancers that develop at disparate sites with similar histopathologic patterns do not represent isolated carcinogenic pathways but rather that the sequential genetic changes and cellular events toward carcinogenesis share common pathways among the multiple isolated lesions. Clinically, a well-known example is the occurrence in the head and neck of synchronous and metachronous second primary squamous cell carcinomas, (18) although other investigators have also considered field changes. (19) We propose that in addition to similarities of tumor morphology among a common epithelium, the concept of field cancerization may be expanded by the commonality of the epidemiologic data derived from overlapping age frequency distributions and parallel log-log plots among the tumor types. The finding of similar tumor pathology from cells isolated in the same organ or analogous cells in different organs whose tumor have identical age frequency distributions and parallel log-log plots indicates that the tumors may represent a single population with similar if not identical carcinogenic pathways. Tumors having these pathologic and epidemiologic parameters may represent ideal models for molecular analysis of incipient neoplasia and tumor development.


On the assumption that the pathogenesis is similar or closely related among the tumors in these distinct sites, we may speculate that the incidence of pancreatic carcinoma is greatest because the total surface area for tumors to develop is much greater in the pancreas than in the gallbladder, EHBD, or ampulla. The suggestion is based on the results that carcinomas arising in these sites have a similar or related pathogenesis and equal epithelial susceptibility to cancer. The complex branching of the pancreatic ducts leads to a much greater surface area for tumors to develop.

In summary, our results suggest that carcinomas developing in tissues derived from the embryonic foregut are likely to be pathogenically related and share a common field. The potential for cancer is similar in each of the 4 sites even though the incidence rates are very different. These results are not considered an artifact because log-log plots of the endocrine carcinomas of the pancreas were not congruent with the epidemiologic patterns of the ductal carcinomas. An association between carcinomas arising in the pancreas, gallbladder, EHBD, and ampulla that has been documented in the literature (1-5) is only suggestive of a field effect, but the results presented in our study are more conclusive. Thus, these foregut derivatives share in a potential field phenomenon for carcinogenesis with the probability that the rates of clinical cancer in each site is most likely related to the relative surface area of the field within each site. Finally, we suggest that a definition for a field effect depends on similar epidemiologic parameters in addition to similar morphologic parameters.

Accepted for publication July 15, 2008.


(1.) Avisse C, Flament JB, Delattre JF. Ampulla of Vater: anatomic, embryologic, and surgical aspects. Surg Clin North Am. 2000;80:201-212.

(2.) Nishihara K, Tsuneyoshi M, Shimura H, Yasunami Y. Three synchronous carcinomas of the papilla of Vater, common bile duct and pancreas. Pathol Int. 1994; 44:325-332.

(3.) Fava M, Foradori G, Cruz F, Guzman S. Papillomatosis of the common bile duct associated with ampullary carcinoma. Am J Roentgenol. 1991;156:405-406.

(4.) Nordback IH, Hruban RH, Cameron JL. Second primary lesions in the biliary tree after successful resection of ampullary carcinoma. Surgery. 1992;112: 111-115.

(5.) Agoff SN, Crispin DA, Bronner MP, Dail DH, Hawes SE, Haggitt RC. Neoplasms of the ampulla of Vater with concurrent pancreatic intraductal neoplasia: a histological and molecular study. Mod Pathol. 2001;14:139-146.

(6.) Liu TH, Chen J, Zeng XJ. Histogenesis of pancreatic head and ampullary region carcinoma. Chin Med J (Engl). 1983;96:167-174.

(7.) Surveillance, Epidemiology, and End Results (SEER) Program. SEER*Stat Database: Incidence--SEER 17 Regs Limited-Use, Nov 2006 Sub (1973-2005 varying)--Linked To County Attributes--Total US, 1969-2005 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2008, based on the November 2006 submission. Available at:

(8.) Henson DE, Schwartz AM, Tilara A, Grimley PM, Anderson WF. Population-based analysis of pathologic data: a new approach to the investigation of uterine endometrial and ovarian endometrioid carcinomas. Arch Pathol Lab Med. 2007; 131:1337-1342.

(9.) Anderson WF, Chatterjee N, Ershler WB, Brawley OW. Estrogen receptor breast cancer phenotypes in the Surveillance, Epidemiology, and End Results database. Breast Cancer Res Treat. 2002;76:27-36.

(10.) Albores-Saavedra J, Henson DE, Klimstra DS. Tumors of the Gallbladder, Extrahepatic Bile Ducts, and Ampulla of Vater. Washington, DC: Armed Forces Institute of Pathology; 2000. Atlas of Tumor Pathology; 3rd series, fascicle 27.

(11.) Albores-Saavedra J, Simpson K, Dancer YJ, Hruban R. Intestinal adenocarcinoma of the pancreas: a previously unrecognized variant of ductal carcinoma of the pancreas. Ann Diag Pathol. 2007;11:3-9.

(12.) Armitage P, Doll R. The age distribution of cancer and a multi-stage theory of carcinogenesis. 1954. Int J Epidemiol. 2004;33:1174-1179.

(13.) Armitage P, Doll R. A two-stage theory of carcinogenesis in relation to the age distribution of human cancer. Br J Cancer. 1957;11:161-169.

(14.) Knudson AG. Two genetic hits (more or less) to cancer. Nature. 2001;1: 157-162.

(15.) Pierce DA, Mendelsohn ML. A model for radiation-related cancer suggested by atomic bomb survivor data. Radiat Res. 1999;152:642-654.

(16.) Kaldor JM, Day NE. Mathematical models in cancer epidemiology. In: Schottenfeld JF, Fraumeni J, eds. Cancer Epidemiology and Prevention. 2nd ed. New York, NY: Oxford University Press; 1996:127-137.

(17.) Moolgavkar SH. Carcinogenesis models: an overview. Basic Life Sci. 1991; 58:387-396, discussion 396-399.

(18.) Slaughter DL, Southwick HW, Smejkal W. "Field cancerization" in oral squamous epithelium: clinical implications of multicentric origin. Cancer. 1953; 6:963-968.

(19.) Willis FA. The Pathology of Tumours. 4th ed. London, England: Butterworths; 1967:107-126.

Donald Earl Henson, MD; Arnold M. Schwartz, MD, PhD; Hala Nsouli, MPH; Jorge Albores-Saavedra, MD

From the George Washington University Cancer Institute, Office of Cancer Prevention and Control (Dr Henson), the Department of Pathology, The George Washington University Medical Center (Dr Schwartz), and the Department of Epidemiology and Biostatistics, School of Public Health and Health Services, The George Washington University (Ms Nsouli), Washington, DC; and the Instituto Nacional de Ciencias Medicas y Nutricion, Salvador Zubiran, Mexico City, Mexico (Dr Albores-Saavedra).

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Donald Earl Henson, MD, George Washington University Cancer Institute, Ross Hall, Room 502, 2300 "I" Street, NW, Washington, DC 20037 (e-mail:
COPYRIGHT 2009 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Henson, Donald Earl; Schwartz, Arnold M.; Nsouli, Hala; Albores-Saavedra, Jorge
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2009
Previous Article:Micropapillary variant of urothelial carcinoma in the upper urinary tract: a clinicopathologic study of 11 cases.
Next Article:Improvement of coagulation laboratory practice in Thailand: the first-year experience of the National External Quality Assessment Scheme for Blood...

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters