Abstract 2757: Correlates of peptide-HLA manufacturing success, TCR capture and neoTCR trafficking from patients using the PACTImmune࣪ Database

Abstract

Abstract PACT Pharma uses a state-of-the-art approach to predict and validate neoepitopes (neoEs) and their cognate T cell receptors (neoTCRs) by capturing neoE-specific T cells from peripheral blood. This neoTCR discovery and validation process is being applied in a clinical trial (NCT03970382) evaluating personalized neoTCR-T cell therapy in solid tumors. The PACTImmune Database (PIDB) accumulates extensive pre-, on- and post-treatment data related to this trial. Here we present PIDB data on the prediction of peptide-HLA manufacturing success, as well as correlates of success of neoTCR capture and neoTCR T cell tumor trafficking, all of which have direct applications for prospective patient enrollment and treatment. The PIDB contains distinct data sets on our neoTCR platform that we examined for predictive capabilities for three distinct applications. The first application was predicting manufacturing success rates of our single chain trimer (comPACT) which consists of the predicted neoE peptide together with B2M and the HLA heavy chain. Using multiple linear regression (MLR), we built a model predicting success rates of comPACT protein manufacturing (defined as comPACT protein yields >detection threshold) and compared it to prediction based solely on expressed tumor mutation burden (eTMB). Using univariate analysis, we compared variables between patients with 3 neoTCRs selected versus those with 0 neoTCRs. Finally, using principal component analysis (PCA) we evaluated if the PIDB neoantigen and TCR characteristics correlate with neoTCR T cell trafficking in patients (n=4) dosed with neoTCR T cells. Trafficking was determined by sequencing (neoTCR specific reads) and in some cases, IHC/RNAScope imaging of post-dosing biopsies. MLR analysis of comPACT protein expression data resulted in significant improvement in comPACT manufacturing success rate predictions (R2=0.66, P<<0.001). These predictions were better than using eTMB alone (R2=0.33, P<<0.001). Patients with 3 neoTCRs selected had a significantly higher number of comPACTs and mutations covered by those comPACTs as compared to 0 neoTCR patients (P=0.004 and 0.002, respectively). All four patients demonstrated trafficking of neoTCR products into the tumor in post-dosing biopsies. Based on TCR characteristics PCA showed separation between tumor trafficking and non-trafficking TCRs (data pending). Application of PIDB data improves predictions of peptide-HLA manufacturing success, identifies differences between patients with 3 and no neoTCR products and provides insights into trafficking TCR characteristics, which in turn enables us to optimize our neoTCR selection strategy for patients with solid cancers. PIDB thus represents a significant & maturing dataset for patient-specific tumor immunogenicity in solid cancers and provides opportunities to optimize personalized neoTCR T cell treatment. Citation Format: Eric W. Stawiski, Vinnu Bhardwaj, Jyoti Mathur, Eva Huang, Olivier Dalmas, Zheng Pan, Cliff Wang, James Heath, Barbara Sennino, Mark Frohlich, Arati Rao, Stefanie Mandl. Correlates of peptide-HLA manufacturing success, TCR capture and neoTCR trafficking from patients using the PACTImmune࣪ Database [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2757.