Colorectal cancer screening can benefit patients through 2 mechanisms. First, a screening colonoscopy will increase the likelihood of early-stage cancer diagnosis among patients with prevalent disease (1). Because of the slow-growing nature of colorectal cancer, clinical benefits of this stage shift take years to accrue (1, 2). This is because the colorectal cancer survival benefit is due to the detection and treatment of early-stage cancer that otherwise would not have resulted in death for at least 5 years. Hence, further understanding of life expectancy after early-stage cancer diagnosis should inform decision making about screening by identifying patients who are unlikely to live long enough to receive the benefit. Second, as Dr. Burack emphasizes, colorectal cancer screening can also benefit patients by the removal of precancerous lesions and the decrease in the incidence of colorectal cancer. However, the interval between the onset of an adenoma and the clinical diagnosis of cancer has been estimated to be as long as 20 years (3).

Trial data support these estimates. Investigators of the Minnesota Colon Cancer Control Study randomly assigned more than 46 000 patients to a fecal occult blood test (FOBT) group or to a control group (1). Patients randomly assigned to the annual FOBT group had a significantly lower cumulative colorectal cancer mortality rate than that of the control group (0.59% vs.0.88%) after 13 years of follow-up. They were also more likely to receive a diagnosis of early-stage disease (1). Yet differences in incidence took longer to accrue. Whereas incidence rates at 13 years did not significantly differ, the incidence rate at 18 years was significantly lower in the annual FOBT group than in the control group (3.2% vs. 3.9%; P < 0.001) (4). Given that the clinical benefits of the stage shift may be observed in as little as 5 years, whereas those of adenoma removal may take 15 years or longer, we feel that the shorter-term interval should guide decision making. We agree that further work is needed to explore factors associated with life expectancy in the cancer-free population. While this analysis was beyond the scope of our study, the relative survival of early-stage colon cancer approaches 90% (5). Therefore, although substantial differences in life expectancy estimates between patients with early-stage colorectal cancer and those without cancer would be expected, empirical analyses should validate this assumption.