CTIM-35. MULTI-OMIC AND SINGLE CELL ASSESSMENT OF SERIAL BIOPSIES REVEALS RESPONSES TO CAN-3110 ONCOLYTIC IMMUNOTHERAPY IN RECURRENT GLIOBLASTOMA THAT ARE NOT EVIDENT BY ROUTINE CLINICAL ANALYSES

Abstract Recurrent glioblastoma (rGBM) is incurable, and routine clinical monitoring is often insufficient to uncover mechanisms of treatment failure in clinical trials. To improve clinical monitoring and mechanistic insights from rGBM trials, we are conducting a Break Through Cancer collaborative clinical trial in which tumor biopsies are collected and subjected to multi-omic profiling at up to 6 timepoints in each patient over a 4 month period of treatment with intratumorally delivered oncolytic herpesvirus (HSV) (CAN-3110, aka rQNestin34.5v.2). Here we report findings from the first two patients in this study. Longitudinal sampling was feasible and well-tolerated, with 96 serial rGBM biopsy cores being obtained over a 4 month period from the two patients. Routine pathology confirmed infiltrating glioma with some necrosis and gliosis at the time of trial enrollment, and serial radiologic monitoring showed increased contrast enhancement consistent with continued tumor growth throughout the therapy period for both patients. In contrast, longitudinal multi-omic profiling revealed decreasing tumor cell density, changes in tumor cell states, increased immune infiltration/activation, and increased HLA protein expression following therapy. In addition to time-dependent changes, we also observed spatially driven changes. Biopsy sites further from HSV injection sites had higher tumor cell densities and HLA class II protein expression, while biopsies closer to HSV injection sites had higher immune activation and MHCI signatures. Functional profiling of TCR specificity also revealed increasing frequency of anti-HSV CD4/8 T cell clones following therapy, consistent with a treatment-induced intratumoral immune response. These results show that longitudinal tissue sampling during rGBM clinical trials is well tolerated and, even with a small number of patients, can reveal critical mechanistic insights into a therapy’s impact that would be missed via routine clinical monitoring. (clinicaltrials.gov NCT03152318)