Retrospective Comparison of Targeted Anticancer Drugs Predicted by the CNS-TAP Tool Versus Those Selected by a Molecularly Driven Tumor Board in Children With DIPG.

Abstract

The recent trial Pediatric Neuro-Oncology Consortium 003 (PNOC003) utilized a molecular tumor board to recommend personalized treatment regimens based on tumor sequencing results in children with DIPG. We separately developed the Central Nervous System Targeted Agent Prediction (CNS-TAP) tool, which numerically scores targeted anticancer agents using preclinical, clinical, and patient-specific data. We hypothesized that highly scored agents from CNS-TAP would overlap with the PNOC003 tumor board's recommendations. For each of the 28 participants, actionable genetic alterations were derived from PNOC003 genomic reports and input to CNS-TAP to identify the highest scoring agents. These agents were then compared with PNOC003 recommendations, with a resultant concordance percentage calculated. Overall, 38% of the total agents recommended by the tumor board were also selected by CNS-TAP, with higher concordance (63%) in a subanalysis including only targeted anticancer agents. Furthermore, nearly all patients (93%) had at least 1 drug chosen by both methods. We demonstrate overlap between agents recommended by CNS-TAP and PNOC003 tumor board, though this does not appear to improve survival. We do observe some discordance, highlighting strengths and limitations of each method. We propose that a combination of expert opinion and data-driven tools may improve targeted treatment recommendations for children with DIPG.