BACKGROUND – Molecular tumour boards (MTB) are the de facto standard for deriving consensus therapy recommendations for patients with advanced cancer using genomic biomarker testing. However, finding relevant treatment options remains challenging, demanding new informatics strategies.
METHODS – This implementation study evaluates the impact of a decision support system (DSS) in the MTB of the Molecular Screening and Therapeutics (MoST) program. The DSS includes the Gentian variant interpretation pipeline for identifying druggable genomic alterations, the TOPOGRAPH knowledge base linked with an annotated trials registry through a symbolic reasoning system (POTTR), and a web platform for recommendation reporting, to streamline the matching of patients' clinicogenomic profile onto potential therapies and trials. The primary endpoint is the proportion of patients participating in biomarker-matched trials (outside the MoST Program) after molecular profiling. The secondary endpoints are the proportion of patients that participated in any trials and overall survival (OS) from subsequent therapies stratified by the receipt of MTB recommended therapy. This study was pre-planned on 7 April, 2020. The data from the first 298 (of 500 planned) patients, up to December 2021, are presented.
RESULTS – 1,141 patients received systemic treatments after MTB (excluding MoST-related trials) with 843 and 298 patients profiled before and after the DSS implementation (July 2020), respectively. Before the implementation, 43 patients (5.1%) participated in biomarker-matched trials after molecular screening. A trend towards a higher rate of participation (24 patients, 8.1%) was observed during the post-implementation period (p=0.08). Overall, a significantly higher proportion of any trial participation following molecular screening was also seen after DSS implementation (61, 20.5% v 126, 14.9%, p=0.03). Notably, while MTB recommendations were not associated with a difference in OS before DSS implementation (received recommended v other therapy, n=144 v 699, median 11.5 v 9.8 months, HR 0.95, p=0.65), OS was significantly longer in patients who received an MTB recommended therapy after implementation (n=59 v 239, median 13.2 v 8.3 months, HR 0.54, 95% CI 0.36 to 0.80, p=0.024). The DSS was instrumental in scaling this research program, enabling a ten-fold increase in the weekly case numbers over a 6-year period, ensuring contemporaneous recommendations are provided to the referring oncologists.
CONCLUSIONS – Algorithmic, evidence-based decision support may be associated with observable improvement in trial participation and survival outcomes in patients with refractory cancer referred for genomic biomarker testing. Further analysis and prospective studies will ascertain the benefits of integrating digital DSS into precision cancer medicine.