Patients with ulcerative colitis had increased mortality in the year of diagnosis, but no increased risk for colorectal cancer
ACP J Club. 1993 Mar-April;118:53. doi:10.7326/ACPJC-1993-118-2-053
Langholz E, Munkholm P, Davidsen M, Binder V. Colorectal cancer risk and mortality in patients with ulcerative colitis. Gastroenterology. 1992 Nov;103:1444-51.
To ascertain survival and the risk for colorectal cancer and colectomy associated with ulcerative colitis.
Inception cohort of patients diagnosed between 1962 and 1987 and followed for a mean of 12 years.
Copenhagen County, Denmark.
1161 adults with ulcerative colitis confirmed by sigmoidoscopy and double-contrast barium enema. At diagnosis 515 patients (44%) had ulcerative proctosigmoiditis, 422 (36%) had substantial colitis (left-sided and extensive), and 207 (18%) had pancolitis (involving the cecum). Follow-up was 99.9%.
Assessment of prognostic factors
Annual assessments were done to collect information on disease activity, treatments, endoscopic and radiologic findings, and work capacity.
Main outcome measures
National data bases provided cancer diagnoses and mortality. Information on colectomies was obtained from yearly assessments.
Of the 141 deaths that occurred, 26 were related to ulcerative colitis and 115 to other causes. Survival analysis from 1 year after diagnosis showed mortality did not differ for patients with colitis compared with the general population. In the year of diagnosis, 20 patients died (9 deaths related to ulcerative colitis and 11 related to other causes). The ratio of observed-to-expected deaths was 2.4, indicating increased mortality within the year of diagnosis of colitis. Risk factors associated with increased mortality were short interval from onset to diagnosis, weight loss or fever at diagnosis, and total colitis (P < 0.02 for all comparisons). 235 patients had a colectomy. In the year of diagnosis, the colectomy rate was 9%. The cumulative colectomy rates were 24% (95% CI 20% to 27%); 30% (CI 26% to 34%); and 32% (CI 28% to 37%) at 10, 15, and 25 years, respectively, after diagnosis. The probability of colectomy within 5 years was 35% for patients with pancolitis.
6 patients developed colorectal cancer. The median duration of ulcerative colitis until cancer diagnosis was 12 years. The relative risk for cancer was 0.9 (CI 0.1 to 1.4) overall and 0.8 (CI 0.3 to 2.0) excluding patients with proctitis. The calculated lifetime risk for development of colorectal cancer was 3.5% for patients and 3.7% for the total Danish population.
Patients with ulcerative colitis had increased mortality only in the year of diagnosis. The risk for colorectal cancer did not increase.
Sources of funding: Aage Louis Hansen's Mindefond; Direktør Emil C Hertz og hustru Inger Hertz's Fond; Fonden til fremme af klinisk Forskning.
For article reprint: Dr. E. Langholz, Rigs Hospitalet, National University Hospital, Department of Medical Gastroenterology, CA 3129, DK 2100, Copenhagen, Denmark. FAX 45-3545-2452.
Patients with long-standing ulcerative colitis are widely acknowledged to be at higher risk for colorectal cancer. Although initial reports from referral centers were compromised by various forms of bias, more careful studies have generally, but not universally, confirmed an increased risk. It is important for clinicians and patients to understand the true magnitude of risk in order to make decisions about surveillance and colectomy.
The design of the study by Langholz and colleagues has much to recommend it. They have assembled and meticulously followed a large, population-based cohort of patients. The excellent survival after the first year is good news for patients with colitis. The conclusion that patients have no increased risk for malignancy, however, should be viewed cautiously. The authors attribute this finding to colectomy for unresponsive disease and continuous use of sulfasalazine. Indeed, the rate of colectomy was high overall and especially high in those with pancolitis (39%). Because pancolitis elevates the risk for cancer, colectomies may have eliminated those at greatest risk. Further, the authors combined patients with left-sided colitis (a lower risk group) with patients with pancolitis. It would be interesting to have more specific information on the outcome of patients with pancolitis. The number of cancers observed from 1962 to 1987 was compared with the expected number in Denmark in 1985. Because colon cancer incidence has been increasing in Denmark, the expected number may be inflated, and this could lower the relative risk estimates. Finally, despite the large size of the cohort, few patients were followed for more than 20 years.
Although the results of this study are intriguing, they should not cause physicians to abandon careful follow-up of patients with long-standing pancolitis. For patients with limited colitis, the risk for cancer may be modest.
Robert S. Sandler, MD, MPH
University of North CarolinaChapel Hill, North Carolina, USA
Anders Ekbom, MD, PhD
University HospitalUppsala, Sweden