D.F . Ransohoff and C.A. Lang concluded in an article published earlier this year in the “New England Journal of Medicine” that screening for colorectal cancer is not justified. Jerome DeCosse, professor of surgery at New York Hospital-Cornell Medical Center, took exception. “NEJM” has accepted a letter from DeCosse, coauthored with Judith Jacobson, responding to the article.
The letter follows:
“Drs. Ransohoff and Lang conclude that colorectal cancer screening of asymptomatic persons is not justified because it has not been shown to reduce mortality. This benefit of breast cancer screening required several decades to become apparent. The issue now is whether, until the results of screening trials are available, the indirect evidence is strong enough to support a public policy of screening for colorectal cancer.
“The authors speculate that three clinical trials of fecal occult blood test screening now in progress in Europe will demonstrate a 10 to 30 percent reduction in mortality from screening. The methodology of the three trials is similar enough to permit their interim findings to be examined collectively. Between 1982 and 1989, 98,193 persons were screened in the three trials; colorectal cancers detected in this group were compared with those in an unscreened group of 99,803 persons. A total of 312 invasive colorectal cancers were found in the screened group, while 180 invasive colorectal cancers were diagnosed in the comparison group.
“Of the cancers diagnosed in the screened group, 60 percent were Dukes stage A or B, compared to 45 percent in the unscreened group. If five year survival rates in these cancer patients are similar to those found in other populations, the diagnosed cancer patients in the screened group have the prospect of an 11 percent advantage over the unscreened patients. It also seems reasonable to assume that additional late stage cancers remain to be diagnosed in the unscreened group. As Ransohoff and Lang point out, lead time bias has little relevance to colorectal cancer. Late recurrence of a Dukes stage A cancer is particularly rare.
“Drs. Ransohoff and Lang describe a 10 percent gain as small. However, in the United States a 10 percent reduction in colorectal cancer mortality amounts to preventing 6,000 deaths a year. We believe that the effort to save these lives is justified.”
