ABSTRACT
This study is an analysis of the relative survival of patients diagnosed with breast cancer in the state of Connecticut between 1950 and 1973. The data were collected by the End Results Section, Biometry Branch, US National Cancer Institute during a time when the incidence of breast cancer was increasing (an 18% increase from 1936 to 1965 and 50% from 1965 to 1975) but mortality remained constant. The relative annual mortality of all patients included in the study dropped from about 12% in the first year following diagnosis to 2.5% at 10 years. It remained constant at about 2.5% per year thereafter. Because of the relatively sharp break in the curve at 10 years, the author hypothesized that there are two distinct patient populations. One died at a rate of 25% per year. These patients were all dead by year 10. The second group died at a rate of 2.5% per year. Half of these patients would have survived 30 years in the absence of other causes, and they constituted about 60% of the population of breast cancer patients in the registry. The relative mortality of this second group was calculated to be nearly equal to that of cigarette smokers. Among those patients with localized or stage I disease, 85% fell in the group with an annual mortality rate of 2.5%, compared with only 40% of those with regional or stage II breast cancer. Fox applied the same analysis to untreated patients in the Middlesex Hospital series (related reference 1), and found that those patients died at an annual rate of 25%!
