Appendectomy, Appendicitis, and Large Bowel Cancer1

(CANCER RESEARCH 50. 7549-7551. December I. 1990] Appendectomy, Appendicitis, and Large Bowel Cancer1

Gary D. Friedman2 and Bruce H. Fireman

Division of Research, Kaiser Permanente Medical Care Program, Northern California Region, Oakland, California 94611

ABSTRACT Subscribers are a socioeconomically and ethnically heterogeneous group and comprise about one fourth of the population in the San A cohort of 167,561 persons who received multiphasic health checkups Francisco Bay area. One of the self-administered questionnaires in the were followed up for cancer development. A history of appendectomy checkup included a question about having "an operation before one showed slightly negative nonsignificant associations with the development year ago." Subjects answering "yes" were asked to indicate on which, of cancer of the colon, rectum, and all sites combined. By inference, the in a list of ten organs, they had been operated. Persons who checked relation of appendicitis with these cancers was also inverse. Upper 95% "appendix" are classified as having had an appendectomy. Subjects confidence limits were compatible only with small positive associations who did not answer yes to the initial question but who nevertheless of appendectomy and appendicitis with these cancers. These data do not indicated a specific operation were considered to have had an operation support the view that removing the appendix increases cancer risk by and comprised 1.0% of the aforementioned subjects with appendectomy diminishing immunocompetency. A link between appendicitis and large and 0.5% of subjects without appendectomy. Subjects who did not bowel cancer has been noted in intersociety correlations and has been answer any question about past operations were excluded. If there was hypothesized to be due to prevention of both by a high-fiber diet. more than one checkup, only the first was used. However, appendicitis does not appear to be a useful predictor of large Follow-up and case identification have been described in detail in bowel cancer within a developed society. connection with a study of cholesterol levels and cancer incidence conducted in this cohort (10). In brief, cases of cancer were identified from two sourcesâ€”Kaiser Permanente hospitalization records for the INTRODUCTION years 1960 to 1980 and a local tumor registry operated by the California In 1969, Burkitt (1) proposed that large bowel cancer is one Resource for Cancer Epidemiology covering the years 1969 to 1980. of a group of diseases of the digestive system that shows large All cancers of the colon and rectum were verified by either noting that geographical variation in incidence because of differences in the the diagnosis was the same in both sources or manual review of medical charts. Follow-up of each subject ended on the date of diagnosis of consumption of dietary fiber. Epidemiological studies and clin ical observations had indicated that appendicitis, cancer, polyps, cancer, the actual or estimated date of termination of program mem bership, or December 31, 1980, whichever came first. Persons with the and diverticular disease of the large bowel were common in the cancer or cancers under study prior to their checkups were excluded Western world and rare in rural Africa. Proposed biological from follow-up. The numbers of subjects and person-years of follow-up mechanisms for fiber's preventive effect on colon cancer include are shown in Table 1. The age at entry ranged from 15 to 94 yr with a reduction of exposure to carcinogens due to more rapid fecal mean of 40.2 yr and interquartile range of 28 to 51 yr. "All cancer" transit, dilution of carcinogens, and other beneficial effects on here means all except nonmelanoma skin cancer. Of the initial cohort bowel contents, particularly bile acids and fecal flora (2). In of 177,729 persons, 1,604 had some prior cancer and 8,564 did not creased intraluminal pressures associated with a low-fiber diet answer the required questions. have been suggested to foster the development of appendicitis Incidence rates were age adjusted by the direct method, using the by contributing to the blockage of the appendiceal lumen and total study population as standard. Relative risks and 95% confidence intervals were calculated by the Mantel-Haenszel method applied to interfering with circulation to the mucosa (3). cohort study data (11). The strong correlation between appendicitis and large bowel cancer among societies (1) raises the question as to whether the occurrence of appendicitis is a useful predictor of this form of RESULTS cancer in individuals within a society. Studies to date have looked at appendectomy rather than appendicitis per se as a There was no evidence of an association between appendec possible risk indicator and have not provided a clear answer (4- tomy and cancer of the colon, rectum, or all sites combined 8). (Table 2). In fact, the incidence of these cancers was slightly There is also a possible causal link between appendectomy higher in those without, than in those with, a history of appen and cancer. The appendix contains significant lymphoid tissue, dectomy. The upper 95% confidence limits implied for men and it has been suggested that, rather than being a useless and women, respectively, possible excess risks no greater than vestigial organ, it may play some role in immunocompetency 8% and 16% for colon cancer, 8% and 33% for rectal cancer, (6). and 7% and 3% for all cancer. This study examined the relation of appendectomy and ap pendicitis to risk of subsequent cancer of the large bowel and DISCUSSION cancer of all sites combined. A history of appendectomy was not a risk indicator for subsequent cancer of the colon, rectum, or all sites combined SUBJECTS AND METHODS in this large cohort. The latency period studied ranged up to 16 The subjects were subscribers of the Kaiser Permanente Medical yr after completion of the questionnaire, with a mean of 10 yr. Care Program and underwent a multiphasic health checkup (9) in its Since the date of appendectomy was not recorded, only the Oakland or San Francisco medical center between 1964 and 1972. minimum interval from appendectomy to cancer is known. Received 6/5/90; accepted 8/30/90. Since the mean age at appendectomy in our program is 30.4 yr The costs of publication of this article were defrayed in part by the payment (1979 to 1987 data) and the mean age at entry of our study of page charges. This article must therefore be hereby marked advertisement in cohort was 40.2 yr, these data probably provide useful infor accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported by National Cancer Institute Grant R35 CA 49761. mation covering at least a few decades of latency. 2To whom requests for reprints should be addressed. The question that ascertained a history of appendectomy 7549

Table 1 Subjects and person-years in this study tomies in women did not represent appendicitis. If we assume Numbers of subjects and person-years were slightly larger for colon and rectal cancer because subjects with other cancers at entry were not excluded. further that appendectomy without appendicitis is not associ ated with risk, the dilution of true appendicitis by using appen Men Women dectomy as its proxy implies that, for appendicitis, the negative No. Person-years No. Person-years associations observed should be more negative; that is, the At risk of any cancer (except skin) relative risks should be 31%/(1-31%) or 45% farther from 1.0 Appendectomy 10,585 103,865 18,033 181,655 No appendectomy 66,447 647,998 72,496 707,231 in men (i.e., 0.71 for colon cancer, 0.45 for rectal cancer, and 0.96 for all cancer) and 51%/(1-51%) or 104% farther from 1.0 in women (i.e., 0.88 for colon cancer, 0.79 for rectal cancer, referred to the period before 1 yr prior to the checkup. Some and 0.91 for all cancer). The upper 95% confidence limits of people classified as having no appendectomy might have had these relative risks, estimated by computer simulation, were one during the prior year. There was another question that similar to, or lower than, those for appendectomy. It would, of asked about "any operations ... in the past year," but the course, be preferable to draw conclusions about the risks asso specific operations were written on the form and not transferred ciated with appendicitis by following a cohort with pathologi to computer storage. Altogether, 5.6% of men and 6.7% of cally proven appendicitis. However, this study appears to rule women of the study group indicated that they had had some out all but small increases in the risk of large bowel cancer operation in the past year. Since only a small fraction of these associated with a history of appendicitis as well as with a history would have been appendectomies, the misclassification intro of appendectomy. duced by using only the "before one year ago" question was Does the absence of an association of appendectomy (or negligible. Although some errors and misclassification inevita appendicitis) with large bowel cancer within our society cast bly occur in questionnaire responses, it is likely that the vast doubt on the fiber hypothesis? Not necessarily. It may be that majority of subjects knew whether they had an appendectomy the vast majority of persons within our society consume much and recorded the information accurately. The likely low rate of less fiber than rural Africans and that, at this low level, differ appendectomy during the previous year and the aging of the ences in characteristics other than fiber intake become more cohort away from the time of life when most appendectomies important in determining who will get appendicitis or large are done (70.5% below age 40, 1979 to 1987 Kaiser Permanente bowel cancer. Intersociety variation is sometimes more reveal data) both suggest that few of the persons without appendec ing than interindividual variation (13). tomy would have had this operation during follow-up, and that Appendectomy itself has been suggested as a risk factor for misclassification for this reason would have had little effect on cancer of several sites, quite independent of the postulated our findings. relation of appendicitis and colon cancer to a low fiber diet. In Of more concern is the fact that not all appendectomies a literature review covering papers until 1982, Howson (6) represent appendicitis, since some are performed incidentally observed that epidemiological data on this subject were incon during other abdominal operations. Additionally good surgical clusive but noted that the appendix of humans and rabbits practice results in the removal of some normal appendices, in contains significant amounts of lymphoid tissue and reiterated perhaps 20% to 25% of operations for suspected appendicitis the question of whether its removal could raise cancer risk by (12), in order that no irreversibly inflamed ones be missed. decreasing immunocompetency. More recent investigations What can be concluded about appendicitis, given our results for continue to give conflicting results. Silingardi et al. (14) found appendectomy? To estimate roughly the proportion of appen that appendectomy did not increase the risk of subsequent dectomies that were incidental to other abdominal surgery, we malignant lymphoma. In a study with few appendectomized noted how many patients with them also listed operations on subjects, Vobecki et al. (5) found appendectomy associated with the stomach, gallbladder, and colon, and in women also removal an elevated risk of colonie but not rectal cancer in men and of the uterus. These additional operations were recorded by with risk of neither in women. Fan and Zhang (7) found a 8.2% of the men and 34.6% of the women with appendectomy. history of appendectomy to be more common in hospitalized If we make the extreme assumption that the appendectomies patients with cancers of the stomach, large bowel, and breast in all of these persons were incidental and assume that appen than among those with cerebral hemorrhage. It was not noted dicitis was not present in 25% of the remaining appendectomies, whether the patients with cerebral hemorrhage might have been then 31% of the appendectomies in men and 51% of appendec less capable of giving such a history or, conversely, whether

Table 2 Incidence and relative risk (with 95% confidence interval) of cancer of the colon, rectum, and all sites combined (except skin) in relation to a history of appendectomy

Cancer rale/ 1000 rate/ 1000 siteAppendectomyColon person-years0.41 Cases1180.80(0.59-1.08)Â°risk person-years0.47 risk0.94(0.76-1.16)0.90(0.61-1.33)0.96(0.89-1.03) Yes

NoRectum 317141205012755MenAge-adjusted0.510.12 31834 0.500.14

Yes 0.62(0.36-1.08) NoAll 0.194.53 991089 0.154.94

YesNoCases48 0.97 (0.89-1.07) 4.62Relative 3280WomenAge-adjusted5.17Relative 1Numbers in parentheses, 95% confidence interval. 7550

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1990 American Association for Cancer Research. APPENDECTOMY AND CANCER greater efforts to elicit a history of past surgery would be made REFERENCES with patients about to undergo surgery for cancer. Linet and 1. HuÃkin.D. P. Related diseaseâ€”related cause? Lancet, 2: 1229-1231, 1969. Cartwright (15) found a diminished risk of chronic lymphocytic 2. Council on Scientific Affairs. Dietary fiber and health. JAMA, 262: 542- 546, 1989. leukemia in patients with appendectomy. Jarebinski et al. (8) 3. Unikm. D. P. The aetiology of appendicitis. Br. J. Surg., 5Â«:695-699, 1971. found appendectomy to be associated with an elevated risk of 4. Abcatici, S., Raso, A. M., and Locatelli, L. Appendectomy and colo-rectal carcinoma. Clinico-stalistical and experimental studies. Minerva Chir.. 32: rectal cancer when cases were compared to both hospital con 333-340, 1977. trols and neighborhood controls. 5. Vobccki, J., Caro, J., and Devrocde, G. A case-control study of risk factors In conclusion, this study provides evidence against the hy for large bowel carcinoma. Cancer (Phila.), SI: 1958-1963, 1983. 6. Howson, C. P. Appendectomy and subsequent cancer risk. J. Chronic Dis., potheses that a history of appendectomy or appendicitis indi 36: 391-396, 1983. cates an increased risk of cancer of the colon, rectum, or all 7. Fan. Y. K., and Zhang, C. C. Appendectomy and cancer. An cpidcmiological evaluation. Chin. Med. J. (Engl. Ed.), 99: 523-526. 1986. sites combined. These data provide no support to the view that 8. Jarebinski, M., Adanja. B., and Vlajinac, H. Case-control study of relation a missing appendix lessens immunocompetency and thereby ship of some biosocial correlates to rectal cancer patients in Belgrade, Yugoslavia. School of Medicine, Belgrade University, Yugoslavia. Neo increases the likelihood of cancer. Our findings do not support plasma (Bralisl.). 36: 369-374, 1989. the hypothesis that there is an appendicitis-colon cancer asso 9. Collen, M. F.. and Davis, L. F. The multitest laboratory in health care. J. Occup. Med., //: 355-360, 1969. ciation due to low dietary fiber. However, an association be 10. Hiatt, R. A., and Fireman, B. H. Serum cholesterol and the incidence of tween these two diseases may be easier to demonstrate in cancer in a large cohort. J. Chronic Dis., 39: 861-870, 1986. 11. Breslow. N. E.. and Dav. N. E. Statistical Methods in Cancer Research, Vol. comparisons among, rather than within, societies. II. The Design and Analysis of Cohort Studies, pp. 109-110. Lyon: Inter national Agency for Research on Cancer. 1987. 12. Sleisenger. M. H. Miscellaneous inflammatory diseases of the intestine. In: J. B. Wyngaardcn and L. H. Smith, Jr. (cds.), Cecil Textbook of Medicine, ACKNOWLEDGMENTS Ed. 18, pp. 800-807. Philadelphia: W. B. Saunders. 1988. 13. Elliott, P., and Marmot, M. International studies of salt and blood pressure. Ann. Clin. Res.. 16 (Suppl. 43): 67-71, 1984. The authors thank Donna Wells for computer programming. Dr. 14. Silingardi, V., Venezia, L., Tampieri, A., and Gramolini, C. Tonsillectomy, Robert Hiatt for assembling the cohort follow-up data and for helpful appendectomy, and malignant lymphomas. Scand. J. Haematol., 28: 59-64, 1982. comments on the manuscript, and Professor Nick Jewell for statistical 15. Line!. M. S., and Cartwright, R. A. Chronic lymphocytic leukemia: epide consultation. miology and ctiologic findings. Nouv. Rev. Fr. Hematol., 30:353-357, 1988.

7551

Gary D. Friedman and Bruce H. Fireman

Cancer Res 1990;50:7549-7551.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/50/23/7549

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/50/23/7549. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.