Purpose: Preclinical data indicate that expression of the ErbB family of receptors, such as HER-2 and HER-1 (EGFR) may be involved in endocrine resistance. Evidence of resistance from clinical studies has been inconsistent. The present study examined whether HER-2 gene amplification or HER-1 expression predicted response to tamoxifen.
Patients and methods: Three hundred and forty nine patients had estrogen receptor (ER)-positive breast cancer and received daily tamoxifen as initial therapy for advanced disease. HER-2 gene amplification, detected by fluorescence in situ hybridization, and HER-1 expression, evaluated by immunohistochemistry, was determined on 136 and 204 patients, respectively.
Results: HER-2 amplification was correlated with lower ER (P = 0.02), HER-1 positivity (P = 0.004), and HER-2 protein overexpression (P < 0.00001). The response rate was 56% for HER-2 non-amplified versus 47% for HER-2 amplified tumors (P = 0.38), and 58% for HER-1-negative versus 36% for HER-1-positive (P = 0.05). Time to treatment failure (TTF) was 7 months for non-amplified HER-2 tumors and 5 months (P = 0.007) for amplified HER-2 tumors, and there was a trend toward a better overall survival (OS) in patients with non-amplified HER-2 tumors (median 31 versus 25 months, respectively, P = 0.07). For positive versus negative HER-1 tumors, TTF was 4 versus 8 months (P = 0.08) and median survival was 24 versus 31 months (P = 0.41). Combining HER-1 expression and HER-2 gene status, patients with both negative HER-1 expression and non-amplified HER-2 had longer TTF (P = 0.001) and OS (P = 0.03) than if either were positive. In multivariate analysis, HER-2 was not an independent factor for TTF and OS, although HER-1 was significant for TTF only (P </= 0.001).
Conclusion: Patients with HER-2 amplification and HER-1 expression had lower ER levels and were modestly less responsive to tamoxifen, suggesting that molecular events in addition to those involving the ErbB receptors are important in determining the endocrine-resistant phenotype.