Examining and predicting drug preferences of patients with metastatic breast cancer: using conjoint analysis to examine attributes of paclitaxel and capecitabine.

The purpose of the study was to examine variables that influence patients' decisions to undergo treatment for metastatic breast cancer as part of a larger effort to understand the utility of biomarkers in treatment decisions. An online survey containing treatment scenarios for a conjoint analysis was e-mailed to members of breast cancer support organizations. The survey contained 14 different scenarios in which survey respondents with a history of metastatic breast cancer were asked to choose between two treatments and whether or not they would undergo the treatment. The scenarios were designed based on paclitaxel and capecitabine profiles related to medication format, likelihood of benefit, and side effects. The likelihood of benefit and likelihood/severity of side effects associated with these two drugs were varied based on the range of predictability afforded by current biomarkers. Most of the 641 respondents indicated that they would opt to undergo a treatment with 27 or 33 % likelihood of benefit regardless of the toxicity scenario presented. In general, a high percentage of respondents indicated that they would choose treatment in all of the scenarios (e.g., for an intravenous medication with a 20-50 % likelihood of benefit and a 20-60 % likelihood of moderate peripheral neuropathy for 1 year, 65-94 % of respondents indicated that they would choose treatment). The conclusion is that this study found that likelihood of benefit was more important than toxicity when considering particular benefit and toxicity ranges associated with common treatment options for metastatic breast cancer.