TBCRC030: A randomized, phase II study of preoperative cisplatin versus paclitaxel in patients (pts) with BRCA1/2-proficient triple-negative breast cancer (TNBC)—Evaluating the homologous recombination deficiency (HRD) biomarker.

TPS1145 Background: Both platinum and taxane chemotherapy have activity in TNBC, however biomarkers predictive for activity of either agent are lacking. The HRD assay detects impaired double strand DNA break repair, and may identify BRCA1/2-proficient tumors with a ‘BRCA-like’ phenotype suitable for treatment with DNA repair targeted therapies. A significant correlation between HRD score and response to platinum was reported in a preoperative study of gemcitabine, carboplatin and iniparib for TNBC. The current trial will prospectively determine the association between HRD score and response to platinum or taxane preoperative chemotherapy in TNBC, as well as explore other potential novel biomarkers of response. Methods: This is a phase II multicenter study randomizing pts with BRCA1/2-proficient, stage I (T1 > 1.5 cm)-III TNBC to preoperative cisplatin 75 mg/m2 q3 wks x 4 or paclitaxel 80 mg/m2 weekly x 12 wks, followed by surgery. Mandatory tissue collection will occur at baseline and surgery. Pts with significant residual disease after 12 weeks will have a tumor biopsy, be classified as a non-responder (RCB>1), and can cross-over to alternative treatment. The primary objective is to determine the association of HRD score with pathologic response, defined as RCB 0-1, to neoadjuvant platinum or taxane in TNBC. Secondary objectives include overall clinical and pathologic responses and the positive predictive value of HRD for platinum therapy. Correlative analyses include evaluation of markers: NtAI, LST, HRD combined score, Hereditary Cancer Profile, PMS2 and CHEK2 HCP, RAD17/RAD50 expression values, PAM50, Vanderbilt TNBC subgroups, tumor infiltrating lymphocyte predictor, and signatures of taxane response; whole exome and RNA sequencing will be used to identify novel markers of response. Target accrual is 160 pts randomized 1:1. Assuming response rates of 40% to cisplatin and 30% to taxane and a drop-out rate of 12%, there will be 91% and 87% power, respectively, to detect a difference of 0.75 standard deviations in HRD score by RCB score. Clinical trial information: NCT01982448.