Estrogens and endometrial cancer
Horwitz, R.I.; Feinstein, A.R.
New England Journal of Medicine 300(23): 1333-1334
1979
ISSN/ISBN: 0028-4793 PMID: 440341 Document Number: 145129
The only active research performed by Jick and colleagues concerning the alleged causal relation between estrogens and endometrial cancer consisted of telephone interviews with 60 of 67 patients with endometrial cancer and a "control" group of 74 other women with intact uteri. In their case-control study, the authors describe no effort to deal with problems in surveillance or detection bis or to avoid bias in ascertainment of exposure to estrogens. The suggestion that 18,000 women were investigated is based on a statistical tour through secular trends in computerized accounts of estrogen prescriptions and separately reported diagnoses of endometrial cancer. Except for data in the small group selected for the case-control study, none of this information was checked or verified. On the basis of the statistical trends, the authors conclude that estrogens increase the risk of endometrial cancer to "even higher" than 20 times than that in nonusers but that the changes are stopped or reversed shortly after discontinuation of the estrogens. This capacity for de-carcinogenesis is explained by the conjecture that estrogen is a "promotor" rather than an "inducer" of endometrial cancer. An alternative hypothesis is that many uterine cancers are asymptomatic, never detected during life and 1st found (if at all) at necropsy.