Impact on costs and outcomes of multi-gene panel testing for advanced solid malignancies: A cost consequence analysis using linked administrative data.

6650 Background: To date, economic analyses of tissue-based next generation sequencing genomic profiling (NGS) have required models with multiple assumptions, with little real-world evidence on overall survival (OS), clinical trial enrollment or end-of-life (EOL) quality of care. The OCTANE clinical trial (NCT02906943) is a prospective study evaluating the role of NGS for advanced solid tumors in Ontario, Canada. We performed a cost consequence analysis of OCTANE. Methods: We undertook a longitudinal, propensity score-matched retrospective cohort study using linked administrative data. OCTANE patients (pts) at Princess Margaret Cancer Centre from August 2016 until March 2019 undergoing NGS panel testing (555 or 163-gene panels) were matched with contemporary controls from across Ontario not enrolled in OCTANE. Patients were matched according to 19 variables including age, sex, place of residence, tumor site, symptom burden, income quintile, comorbidities and prior lines of systemic therapy. Primary outcomes were mean per capita health care costs (2019 Canadian dollars [CAD]) from the public payer’s perspective, OS, clinical trial enrollment and EOL quality metrics. Full 2-year follow-up data was available. Sensitivity analyses considered alternative matched cohort specifications. Results: There were 782 OCTANE pts with 782 matched controls. Variables were balanced after matching (standardized difference [std. diff.]<0.10). Most common tumor sites were: Ovary (30.4%), endometrium (15.0%) breast (12.3%) and colon (8.6%). OCTANE pts had higher mean healthcare costs than controls ($79,702 vs. $59,550), mainly due to costs of oncology visits ($33,165 vs. $26,197), outpatient clinic visits ($8,696 vs. $5,114) and emergency visits ($1,723 vs. $1,373) (all p<0.05). Publicly funded drug costs were less for OCTANE pts ($20,015 vs. $24,465). Overall, OCTANE enrollment was not associated with improved OS (restricted mean survival time (RMST) [standard error]: 1.50 (±0.03) vs. 1.44 (±0.03) years, log-rank p=0.153), but OCTANE was associated with longer OS in ovarian cancer (RMST: 1.69 (±0.05) vs. 1.45 (±0.06) years, p=0.011) and biliary tract tumors (RMST: 1.16 (±0.13) vs. 0.80 (±0.11) years, p=0.02). Importantly, OCTANE correlated with increased clinical trial enrollment (25.5% vs. 9.5%, p<0.001) and better EOL quality due to fewer deaths in hospital (10.2% vs 16.4%, p=0.003). Results were robust in sensitivity analysis. Conclusions: There was an increase in healthcare costs associated with NGS testing for advanced cancers. The impact on OS was not significant in the overall population, but varied across tumor types. OCTANE was associated with greater trial enrollment, lower publicly funded drug costs and fewer in hospital deaths suggesting important considerations in determining the value of NGS for advanced cancers. Clinical trial information: NCT02906943 .