Broad genomic testing is necessary to treat stage IV non-small cell lung cancer (NSCLC) patients. We describe a NSCLC precision medicine service at an academic-community practice and provide model-based estimates of the impact of a similar intervention. Will implementation of a precision medicine service increase NSCLC next-generation sequencing (NGS) rates, improve testing turnaround times (TAT), and increase rates of actionable genomic alterations (AGAs)? PREDICT (PREcision meDICine Thoracic) consisted of: 1) system-wide reflex testing of stage IV NSCLC patients by in-house solid tumor NGS focused assay and PD-L1 testing, 2) navigator, 3) molecular tumor board (MTB), 4) integrated information portal (OncoTracker) for real-time updates on samples processing, results, and treatment recommendations by the MTB. A decision analytic model compared four strategies: PREDICT versus standard care (send-out sequencing and PD-L1 testing), and sequencing of all stage IV NSCLC patients versus only non-squamous histologies. From 01/2016 to 12/2018, 626 retrospective patients with stage IV NSCLC were eligible. In the 17 months after PREDICT's launch, 290 prospective patients were identified. NGS testing rates increased significantly (91.3% vs 60.8%; p<0.0001) after PREDICT, whereas TAT from biopsy date were significantly shorter for both NGS (12 vs 18 days; p<0.0001) and PD-L1 (7 vs 10 days; p<0.0001). AGAs were identified in 29.3% of prospective patients, versus 22% of retrospective cohort (p=0.0172). Targeted therapy use increased from 6.8% to 10.6% (p=0.048). The decision analytic model predicted superior survival with the PREDICT initiative across all patients, resulting in the highest rates of AGA identification and lowest chemotherapy utilization rates in the first line setting. Implementation of a precision medicine service for stage IV NSCLC comprised of reflex testing, nurse navigation, molecular tumor board, and information portal at a combined academic-community practice led to higher NGS testing rates and shorter turnaround times resulting in identification of more actionable mutations and higher rates of first-line targeted therapy utilization. Decision analytic modeling suggests the superiority of this initiative across all patients.
Bruno et al. (Mon,) studied this question.