Tchou et al.1 recently published a retrospective chart review of women choosing to receive tamoxifen for chemoprevention of breast carcinoma. One hundred thirty-seven women seeking risk counseling for breast carcinoma were offered tamoxifen for chemoprevention, and 57 (42%) chose to receive this agent. The authors found that a history of atypical hyperplasia or lobular carcinoma in situ (LCIS) and older age were significant predictors of being offered and accepting treatment with tamoxifen. The figures reported by Tchou et al. are in sharp contrast to the results reported by Port et al,2 who found that 43 women with a 5-year risk of developing breast carcinoma > 1.7% (as predicted by the Gail model) were given information regarding the benefits and risks of tamoxifen and that only 2 of these women (4.7%) chose to receive tamoxifen therapy.
These discrepancies in terms of acceptance rates may be attributable to the way in which information was provided. Despite the best intentions of physicians, it may be impossible to present information to patients in a truly unbiased fashion. A physician's description of a treatment and the strength of the recommendation,3 in addition to the framing of information,4 may influence patients. Although in these two previous studies, the authors state that they provided ‘neutral risk counseling’1 or ‘neutral education sessions and literature’,2 this neutrality was not explicit, and some bias in favor or against tamoxifen use may have been present. The Tchou et al. study did not include the physician as a variable having a possible effect on the offering or acceptance of tamoxifen therapy.
In our own prospective study, which attempted to remove physician-related biases through the use of a standardized decision guide, only 6 of 41 women (14.6%) who had high-risk status (mean 5-year Gail risk, 3.4%; atypia rate, 61%) were willing to receive tamoxifen for chemoprevention, a figure that was intermediate relative to the Tchou et al. and Port et al. studies. Although information cannot be tailored specifically to the patient using the decision guide, generic information can be conveyed to introduce the concept of chemoprevention, and the issue can be explored further at the request of the patient.
We commend Tchou et al.1 for offering more insight into the proportion of women expressing an interest in tamoxifen chemoprevention. There remain several challenges related to how to best identify and approach women who have an elevated risk of developing breast carcinoma (as routine screening using the Gail model is not recommended5) and how to best provide information on the small absolute benefits and the small but significant risks associated with tamoxifen use. The balance between the benefits and harms of tamoxifen use is a delicate one, and decisions regarding tamoxifen use are best made by informed patients who are free from the influence of physician-related biases. The existing literature suggests that the issue of physician-related bias has not yet been addressed.