PREDICTIVE SIGNIFICANCE OF IDH1/2 MUTATION AND 1P/19Q CO-DELETION STATUS IN A POST-HOC ANALYSIS OF NRG ONCOLOGY/RTOG 9802: A PHASE III TRIAL OF RT VS RT plus PCV IN HIGH RISK LOW-GRADE GLIOMAS

This study investigated the predictive significance of IDH1/2 mutations and 1p/19q co-deletion in patients with high risk (age≥40 or subtotal resection/biopsy) low-grade gliomas (LGGs), randomized to receive RT with or without PCV after biopsy/surgical resection in NRG Oncology/RTOG 9802. Importantly, this is the very first phase III study to analyze the predictive value of these sub-groups in LGGs using prospectively-collected, well-annotated clinical data. Immunohistochemistry and/or IonTorrent sequencing were used to determine IDH1/2 status. Oncoscan and/or 450K methylation data were used to determine 1p/19q status. To estimate by marker status the treatment effect on overall survival (OS) and progression-free survival (PFS), hazard ratios (HRs) were calculated using the Cox proportional hazard model and tested using the log-rank test in a post-hoc and exploratory analysis. Interaction effects between marker status and treatment were tested. Of all the randomized eligible patients in NRG Oncology/RTOG 9802, 97(39%) had sufficient DNA for profiling. Of these, 36(37%) were IDHmut/non-co-deleted, 33(34%) were IDHmut/co-deleted, and 28(29%) were IDHnon-mut. Upon univariate analyses, the IDHmut/non-co-deleted sub-group was significantly correlated with better PFS with the addition of PCV (HR=0.31; p=0.005) and OS (HR=0.36; p=0.02). The IDHmut/co-deleted sub-group was significantly correlated with better PFS with the addition of PCV (HR=0.16; p=0.002), but not OS (HR=0.31; p=0.13). There was no significant difference observed with the addition of PCV in the IDHnon-mut sub-group for either OS or PFS. Tests on interaction effects revealed no statistical significance. Our analyses suggest a putative predictive value of 1p/19q co-deletion and IDH1/2 mutations for high-risk LGGs, based on the observed differences in the treatment effect by marker status. Our results also support the hypothesis that IDHnon-mut high-risk LGG patients do not benefit from the addition of PCV to RT. FUNDING: U10CA21661, U10CA180868, U10CA180822, and U10CA37422. Also, R01CA108633, R01CA169368, RC2CA148190, U10CA180850-01, BTFC, OSU-CCC (all to AC).