OBJECTIVES: To examine physicians' assessments of tamoxifen effectiveness in breast cancer patients, identify predictors of these assessments, and estimate the relationship between these assessments and receipt of tamoxifen prescription. DESIGN: A cohort of breast cancer patients aged 65 and older at diagnosis and their physicians were surveyed using mailed questionnaires and telephone interviews. SETTING: Community and academic hospitals in Rhode Island; North Carolina; Minnesota; and Los Angeles, California between 1996 and 1998. PARTICIPANTS: Physicians completed treatment recommendation forms for 496 of 865 Stage Ic to IIIa breast cancer patients. MEASUREMENTS: Visual scales measured physicians' assessments of the risk that individual patients would have a breast cancer recurrence or die of breast cancer with, and without, tamoxifen therapy. RESULTS: The mean risk ratio+/-standard deviation comparing risk of recurrence without tamoxifen with the risk with tamoxifen was 1.8+/-1.0 and for breast cancer mortality was 1.8+/-1.2. Only estrogen-receptor status and enrollment site emerged as significant predictors of recurrence and mortality risk ratios in regression models. Patients for whom the physician estimated that the recurrence or mortality risk doubled without tamoxifen were more likely to receive a tamoxifen prescription than patients for whom the physician estimated that tamoxifen would have no effect (odds ratio (OR)=1.4, 95% confidence interval (CI)=0.98-2.1 for recurrence risk, OR=1.8; 95% CI=1.2-2.6 for mortality risk). CONCLUSION: Estrogen receptor status most strongly influenced physicians' assessments of tamoxifen's effectiveness in individual patients; this effectiveness was not found to be associated with advancing patient age. Estrogen receptor status and enrollment site were related to receipt of tamoxifen prescription, but advancing age was not after accounting for physician's individualized assessment of tamoxifen's effectiveness. These findings suggest that an evidence-based approach for hormonal therapy has been widely adopted for care of older patients with breast cancer.

The role of androgen receptor (AR) mutations in androgen-independent prostate cancer (PCa) was determined by examining AR transcripts and genes from a large series of bone marrow metastases. Mutations were found in 5 of 16 patients who received combined androgen blockade with the AR antagonist flutamide, and these mutant ARs were strongly stimulated by flutamide. In contrast, the single mutant AR found among 17 patients treated with androgen ablation monotherapy was not flutamide stimulated. Patients with flutamide-stimulated AR mutations responded to subsequent treatment with bicalutamide, an AR antagonist that blocks the mutant ARs. These findings demonstrate that AR mutations occur in response to strong selective pressure from flutamide treatment.