Purpose: To determine the effect of elevation of se-rum HER-2/neu on response to hormone therapy.

Patients and Methods: Seven hundred nineteen met-astatic patients with estrogen receptor–positive (ER), progesterone receptor–positive, or both or ER status unknown breast cancer were randomized in three independent clinical trials to receive second-line hormone therapy with either megestrol acetate or an aromatase inhibitor (fadrozole or letrozole). An automated enzyme-linked immunosorbent assay specific for the extracellular domain of the HER-2/neu (c-erbB-2) oncoprotein product was used to detect serum levels.

Results: Two hundred nineteen patients (30%) had elevated serum HER-2/neu protein levels, using the mean 2 SD (15 ng/mL) from the serum of healthy women as an upper limit. Response to treatment was available for 711 patients. The response rate (complete responses plus partial responses plus stable disease) to endocrine therapy was 45% in 494 patients with nonelevated and 23% in 217 patients with elevated serum HER-2/neu levels (P<. 0001). Median duration of treatment response (using the time to progression [TTP] variable for patients who responded) was shorter in the group with elevated serum HER-2/neu levels (11.7 months) compared with the patient group with nonelevated levels (17.4 months). TTP, time to treatment failure, and median survival (17.2 months v 29.6 months) were also significantly shorter in the patients with elevated serum HER-2/neu levels (P<. 0001).

Conclusion: Patients with ER and serum HER-2/neu-positive metastatic breast cancer are less likely to respond to hormone treatment and have a shorter duration of response than ER and serum HER-2/neu–negative patients. Their survival duration is also shorter.