Biomarker data from KATHERINE: A phase III study of adjuvant trastuzumab emtansine (T-DM1) versus trastuzumab (H) in patients with residual invasive disease after neoadjuvant therapy for HER2-positive breast cancer.

Abstract

502 Background: The phase 3 KATHERINE study (NCT01772472) compared adjuvant T-DM1 versus H in patients with residual invasive breast cancer after neoadjuvant chemotherapy plus HER2-targeted therapy. Here we report exploratory analyses of the relationship between invasive disease-free survival (IDFS) and biomarkers potentially related to response. Methods: Formalin fixed paraffin-embedded tissue samples were collected before neoadjuvant treatment and/or at surgery. Surgical samples were used for analyses, except when only pre-treatment samples were available (~20% of cases). DNA was derived to identify PIK3CA hotspot mutations and gene expression (RNA) analysis was used to detect HER2, PD-L1, CD8 and predefined immune signatures including 3-gene, 5-gene, Teffector, chemokine signaling, and checkpoint inhibitor signatures. RNA analysis was adjusted for tumor content and expression levels were dichotomized at the median into low (≤) and high (>) groups. The effect of treatment and biomarkers on IDFS was assessed. Results: PIK3CA mutation (mut) status was available from 1363 (91.7%) patients. T-DM1 IDFS benefit was independent of PIK3CA mut status (mut: HR 0.54; 95%CI 0.23–0.90; non-mut: HR 0.48; 95%CI 0.35–0.65) and no impact of PIK3CA mut was observed within either treatment arm. Gene expression data were available from 1059 (71.3%) patients. Similar gene expression levels were observed between treatment arms, but, unlike the surgical samples (n = 815), the pre-treatment samples (n = 244) were not representative of the ITT population. Thus, subsequent analyses were based on surgical samples (H n = 398; T-DM1 n = 417). Consistent treatment benefit with T-DM1 vs H was observed across the single-gene and immune gene-signature subgroups as in the ITT population. High vs low HER2 expression was associated with worse outcome (HR 2.02; 95% CI 1.32–3.11) within the H arm, but not within the T-DM1 arm (HR 1.01; 95% CI 0.56–1.83). High vs low PD-L1 expression was associated with better outcome within the H arm (HR 0.66; 95% CI 0.44–1.00) but not within the T-DM1 arm (HR 1.05; 95% CI 0.59–1.87). Similar trends were observed in the checkpoint inhibitor subgroups. Conclusions: These exploratory analyses provide the first data on the relationship between biomarker expression in residual disease after HER2-targeted therapy and outcomes. PIK3CA mut status did not influence outcomes with H or T-DM1. T-DM1 benefit appeared to be independent of all biomarkers assessed. Clinical trial information: NCT01772472 .