Review: Meta-analysis shows that cholecystectomy is associated with a small increased risk for colorectal cancer of the proximal colon
ACP J Club. 1994 Jan-Feb;120:25. doi:10.7326/ACPJC-1994-120-1-025
Giovannucci E, Colditz GA, Stampfer MJ. A meta-analysis of cholecystectomy and risk of colorectal cancer. Gastroenterology. 1993 Jul;105:130-41.
To study the association between cholecystectomy and colorectal cancer.
Studies were identified using MEDLINE up to 1991 and bibliographies of relevant papers.
Studies were selected if they controlled for at least age and sex and if the confidence intervals (CIs) and P values were included or could be calculated. Study designs were cohort or follow-up studies (n = 5), case-control studies (n = 33), and autopsy studies (n = 7). Studies with no external controls (n = 9) that compared the prevalence of cholecystectomy for patients with proximal and distal colorectal cancer, studies that determined the incidence of colorectal adenomas (precursors of cancer) after cholecystectomy (n = 8), and studies that assessed the association between cholelithiasis and colorectal cancer (n = 9) were also included.
The proximal colon was defined as including up to the distal end of the transverse colon, and the distal colorectum as including the descending colon, sigmoid colon, and rectum. Information on how the control group was formed (hospital- or community-based controls) and data on relative risks (RRs) were extracted. Pooled RRs and 95% CIs were calculated. Heterogeneity was assessed.
Combining the 38 cohort and case-control studies, the pooled RR for colorectal cancer after cholecystectomy for men was 1.21 (95% CI 1.04 to 1.40, P = 0.01) and for women was 1.24 (CI 1.10 to 1.40, P < 0.001). The cohort studies alone and the community-based studies showed no association. The hospital-based studies had a pooled RR of 1.43 (CI 1.12 to 1.81). Data from case-control studies showed an increased risk for proximal cancer overall (pooled RR 1.88, CI 1.54 to 2.30) and for population-based studies (pooled RR 1.33, CI 1.09 to 1.62). Neither population-based nor community-based studies showed an increased risk for distal colorectal cancer. 6 of 7 autopsy studies and the 9 prevalence studies with no external controls showed that the ratio of proximal to distal colorectal tumors was increased in patients who had previous cholecystectomies. 5 case-control studies showed an increased risk for proximal cancer over time since cholecystectomy. Cholelithiasis was associated with colorectal cancer (pooled RR using data for 7 studies 1.24, CI 1.07 to 1.44) and with proximal cancer (pooled RR using data for 4 studies 1.55, CI 1.11 to 2.17), but the associations were not clearly independent from those of cholecystectomy with cancer.
Although the relative risks vary with study design, meta-analysis shows that cholecystectomy is associated with a small increased risk for cancer of the proximal colon.
Sources of funding: CIBA-GEIGY Corporation; American Cancer Society; Medical Foundation Incorporated.
For article reprint: Dr. E. Giovannucci, Channing Laboratory, 180 Longwood Avenue, Boston, MA 02115. USA. FAX 617-432-2435.
Approximately 20% of women and 10% of men have gallstones (1). Although most patients with gallstones are asymptomatic, approximately 600 000 cholecystectomies are done each year in the United States for symptomatic cholelithiasis (1). An association between cholecystectomy and colon cancer was first reported in 1978 (2), and since then more than 60 studies have attempted to confirm whether cholecystectomy increases the risk for colon cancer. The results have been inconsistent.
An association between cholecystectomy and colon cancer is biologically plausible. Without the gallbladder, continuous circulation of bile salts and increased bacterial degradation of primary to secondary bile salts occur (3, 4). Secondary bile salts may be carcinogenic because they can promote colonic epithelial cell proliferation in animal models (5). The key questions still remain: Is there truly an association, and, if a causal link exists, should we screen patients for colon cancer after cholecystectomy?
Ekbom and colleagues have essentially resolved the issue. Other cohort studies that also found no overall excess risk have been criticized for a potential lack of statistical power (6) or for insufficient follow-up (7). Because of the large sample size, a virtually complete ascertainment of patients, and extensive follow-up, bias is unlikely to explain the findings of the study by Ekbom and colleagues, and their estimates are probably precise. They did find that women had a small increased risk for right-sided colon cancer, particularly ≥ 15 years after surgery, but they note that only 23 more cases occurred than were expected in 41 870 women.
Moreover, these results were largely confirmed in the meta-analysis by Giovannucci and colleagues, although women and, to a lesser extent, men were found to be at a small but increased risk. Although studies that are weakest methodologically (hospital case-control studies) provided the strongest evidence for an increased risk for colon cancer after cholecystectomy, unrecognized bias (e.g., greater scrutiny of patients who had cholecystectomy compared with controls or selection of inappropriate controls) is likely to explain that observation. Cohort studies generally provide more convincing evidence than case-control studies. Why women are at a higher risk than men remains unexplained.
Although a small but definite association is now established for some patients with colon cancer, is this cause and effect? Other factors, such as weight, diet, or heredity, could predispose a person to have right-sided colon cancer and gallstones (and thus a cholecystectomy). On the other hand, confounding by lifestyle factors alone seems unlikely to entirely explain the association, in view of a positive association found in other study populations.
An increased risk for right-sided colorectal cancer appears to occur after cholecystectomy, particularly in women ≥ 15 years after surgery. Screening is, however, not justified because of the tiny attributable risk. Cholecystectomy should not be withheld for patients who need it because of any concern about the development of colon cancer.
Nicholas Talley, MD
Mayo ClinicRochester, Minnesota, USA
5. Narisawa T, Magadia NE, Weisburger JH, Wynder EL. Promoting effect of bile acids on colon carcinogenesis after intrarectal instillation of N-methyl-N'-nitro-N-nitrosoguanidine in rats. J Natl Cancer Inst. 1974;53:1093-7.
7. Maringhini A, Moreau JA, Melton LJ 3d, et al. Gallstones, gallbladder cancer, and other gastrointestinal malignancies. An epidemiologic study in Rochester, Minnesota. Ann Intern Med. 1987;107:30-5.