Serum Human Epidermal Growth Factor 2 Extracellular Domain as a Predictive Biomarker for Lapatinib Treatment Efficacy in Patients With Advanced Breast Cancer

Abstract

Purpose We examined the prognostic and predictive value of serum human epidermal growth factor 2 (HER2) extracellular domain (sHER2) in patients with advanced breast cancer treated with lapatinib using data from three randomized trials. Patients and Methods We analyzed sHER2 and tissue HER2 (tHER2) data from 1,902 patients (84%) who were randomly assigned to receive lapatinib or control in the trials EGF30001, EGF30008, and EGF100151. Cox regression analyses were performed to correlate both biomarkers with progression-free survival (PFS) and overall survival (OS). Results Median sHER2 levels were 25.1 and 10.1 ng/mL in tHER2-amplified (tHER-positive) and nonamplified (tHER-negative) populations, respectively (r = 0.42 for sHER2-tHER2 correlation). Lapatinib had significant PFS benefit over control (hazard ratio [HR], 0.855; P = .004), but not OS (HR, 0.941; P = .33). Lapatinib PFS benefit is independently predicted by higher sHER2 values (HR per 10-ng/mL increase in sHER2: lapatinib-containing therapies, 1.009 v nonlapatinib-containing therapies, 1.044; Pinteraction < .001) and by positive tHER2 (HR [lapatinib v nonlapatinib]: tHER2 positive, 0.638 v tHER2 negative, 0.940; Pinteraction = .001). Within the tHER2-positive subpopulation (n = 515), higher sHER2 values still independently predicted lapatinib PFS benefit (HR per 10-ng/mL increase in sHER2: lapatinib-containing therapies, 1.017 v nonlapatinib-containing therapies, 1.041; Pinteraction = .008). In control arms (n = 936), higher sHER2 was associated with worse prognosis in multivariable analyses (PFS HR per 10 ng/mL: PFS, 1.024; P < .001; and OS, 1.018; P < .001). Conclusion Higher sHER2 predicts greater PFS benefit with lapatinib independent of tHER2 status. High sHER2 is also independently prognostic for worse survival in patients who received nonlapatinib-containing therapies. The predictive role of sHER2 for other anti-HER2 agents requires further research.