Clinical Benefit of Comprehensive Genomic Profiling in a Mixed Cohort of Patients with Advanced Solid Tumors: a Predictive Model

INTRODUCTION: Despite achievements in oncology, metastatic forms of solid tumors still have a poor prognosis. One promising approach to personalizing treatment is comprehensive genomic profiling (CGP) permitting identification of molecular targets and selection of target therapy. However, the effectiveness of CGP in real clinical practice remains a subject of debate, especially in mixed cohorts and with limited resources. AIM: To assess the clinical benefit of prescribing CGP in patients with advanced solid tumors of different nosologies and to develop a predictive model to forecast the effectiveness of this approach. MATERIALS AND METHODS: The retrospective study included 169 patients with metastatic solid tumors who underwent CGP using targeted sequencing method with commercial panels (≥350 genes) in 2019–2024. Successive results of sequencing were obtained in 146 patients. The analysis included a comparison of the group with CGP and the control group without CGP by the ratio of the time to progression after two consecutive lines of therapy (PFS2/PFS1) and median progression-free survival (PFS). A predictive model was constructed on the basis of clinical and molecular factors. RESULTS: In breast cancer and non-small cell lung cancer, the use of CGP was associated with a statistically significant increase in the proportion of patients with PFS2/PFS1 >1.3 (p=0.039) and an improvement in PFS median (p <0.05). No such dependence was found in the groups with colorectal cancer, gastric cancer, and pancreatic cancer (p >0.1). The predictive model demonstrated high diagnostic value (AUC ROC=0.781; p=0.021): in patients with predicted benefit from CGP, the PFS2/PFS1 median was 1.41 versus 0.79 (p=0.001), and the PFS median was 15 weeks versus 9 weeks (p <0.001). CONCLUSION: Comprehensive genomic profiling has selective clinical benefit in patients with advanced solid tumors, particularly breast cancer and non-small cell lung cancer. The developed predictive model allows for identifying a subgroup of patients that are most likely to benefit from comprehensive genomic profiling, which is important for rational use of resources in oncology.