OBJECTIVES: This study aims to evaluate the impact of shifting diagnostic strategies from current practice to next generation sequencing (NGS) based comprehensive genomic profiling (CGP) on the treatment guidance for patients with advanced or metastatic cancer in Germany while accounting for variability in real-world diagnostic practices and correlation between multiple actionable biomarkers.

METHODS: A decision model was developed in Excel® that examined multiple diagnostic strategies, including common single gene tests, small and CGP panels, and PD-L1 testing, in parallel or in sequence, with frequencies based on Ipsos Healthcare chart review data. CGP panels detected rare mutations including known and novel fusions with NTRK, and measured tumor mutation burden (TMB), an emerging biomarker for immune-oncology agents. The model factored sensitivity and specificity of tests in each strategy to predict observed diagnoses versus actual patient characteristics. Patients can have a genetic biomarker for targeted therapy along with high TMB and/or PD-L1 expression, and approximately 24% of the modeled population fell into this category. For patients positive for multiple actionable biomarkers, biomarkers for targeted therapy took precedence over TMB and PD-L1 in driving treatment. The base case analysis evaluated the impact of CGP panels on the fraction of non-small cell lung cancer (NSCLC) patients receiving optimal treatments. Scenarios of alternative cancer indications were also evaluated.

RESULTS: The model predicted that higher use of CGP panels in NSCLC can improve detection of genetic biomarkers for targeted therapy by as much as 3.4% of standard care and biomarkers for immune-oncology agents by as much as 17.0%. Detection of NTRK can improve by as much as 52.7%.

CONCLUSIONS: CGP panels can lead to more patients being matched with optimal treatments. Assessing the impact of NGS in oncology benefits from a modeling approach that captures genetic heterogeneity and treatment decisions when dealing with multiple actionable biomarkers.