18P A translational project of the WSG-ADAPT-TN trial demonstrates immunomodulatory and anti-viral defense gene networks predicting pathological complete response (pCR) and survival after de-escalated neoadjuvant chemotherapy (NACT) in early triple-negative breast cancer (eTNBC)

NACT is the standard treatment for eTNBC, yet, molecular phenotypes driving chemotherapy sensitivity are poorly understood. Whole transcriptome profiling of baseline tumor biopsies from the neoadjuvant WSG-ADAPT-TN phase II trial (NCT01815242) was performed to identify gene networks predictive and prognostic for pCR and survival. Patients with cT1c-cT4c, cN+, centrally confirmed eTNBC were randomized to 12 weeks of nab-paclitaxel+gemcitabine (Arm A; n=182) or nab-paclitaxel+carboplatin (Arm B; n=154). The primary endpoint was pCR (ypT0/is, ypN0), secondary endpoints included survival and translational research. AmpliSeq RNA sequencing allowing simultaneous analysis of the expression of >20,000 genes was performed in 137 patients (Arm A/B: n=72/65). Differentially expressed genes were then evaluated in training (n=67) and validation (n=68) sets and two polygenic scores (PS) with a high diagnostic performance for prediction of pCR (PS:pCR) and invasive disease-free survival (PS:iDFS) were found. 49/135 (36.3%) patients had pCR; 30 iDFS events occurred during 60-month median follow-up. Genes and networks associated with immune recruitment, endogenous retroviral transcription, and viral defense were strongly associated with pCR and iDFS. PS:pCR achieved a high diagnostic accuracy (ROC AUC: 83%). In multivariate analysis (validation set) adjusted for clinical factors, pCR was associated with PS:pCR (OR 7.24; 95%CI 2.05, 25.58; p=0.002). Addition of PS:pCR to clinical factors increased ROC AUC for pCR from 0.707 to 0.887. iDFS was associated with PS:iDFS (HR 2.06; 95%CI 1.01, 4.20; p=0.046). Polygenic scores based on immunomodulatory and anti-viral defense gene networks may have utility in predicting pCR and survival, and could represent an opportunity for selecting patients for de-escalated NACT in eTNBC. This transcriptome network analysis also identifies potential new targets for personalized medicine approaches for patients without response to chemotherapy.