Biases in study design compromise estimates of cancer risk reduction

Unrecognized biases in study design are probably causing researchers to over- or underestimate the value of prophylactic surgery for preventing breast and ovarian cancer in high-risk women.

In a commentary in the July 2nd issue of the Journal of the National Cancer Institute, Dr. Hester M. Klaren, and colleagues at The Netherlands Cancer Institute in Amsterdam, Holland, discuss these potential biases.One such bias is confounding by indication. This type of bias occurs, for example, when surgery and nonsurgery patients from families with BRCA 1/2 mutations and different baseline risks of ovarian cancer are compared. Such a comparison could lead to an underestimation of risk reduction after prophylactic oophorectomy.Familial-event bias might occur when a woman’s decision to undergo prophylactic surgery is influenced by a family member’s diagnosis or death from cancer. By including this familial event in the study, cancer risk among women in the nonsurgery group may be overestimated, which could lead to overestimation of the risk reduction after prophylactic surgery, according to a JNCI statement.Furthermore, if only women with proven BRCA1 or BRCA2 mutations are included in risk reduction analyses, a cancer-induced testing bias might result. “Most women without cancer would generally not consider prophylactic surgery unless they are certain about their BRCA mutation carrier status,” the authors point out. “Consequently, the date of testing generally precedes the date of the prophylactic surgery,” they write. “Many women who do not consider prophylactic surgery and remain cancer-free may not undergo genetic testing at all.”They point out further that if such women are not included in risk reduction analyses, “the differential selection of identified mutation carriers with cancer from the total group of mutation carriers may lead to an overestimation of cancer incidence in the nonsurgery group.” Therefore, “cancer risk reduction might be overestimated.”Other potential biases include survival bias, detection bias, and confounding by competing risk factors for breast and ovarian cancer.The authors caution that it may “not always be possible to resolve all of the methodological problems of efficacy studies, or resolution would be only at the expense of a considerable loss of power.” When this is the case, they add, more critical consideration of potential biases is necessary.The authors warn that “valid estimate of risk reduction [from prophylactic surgery] may become even more crucial…when data become available regarding the efficacy of new surveillance methods.”(Source: J Natl Cancer Inst 2003;95:941-947: Reuters Health: July 11, 2003: Oncolink)