Next-generation sequencing-based evaluation of the actionable landscape of genomic alterations in solid tumors: the "MOZART" prospective observational study

Abstract

Background: The identification of the most appropriate targeted therapies for advanced cancers is challenging. We performed a molecular profiling of metastatic solid tumors utilizing a comprehensive next-generation sequencing (NGS) assay to determine genomic alterations' type, frequency, actionability, and potential correlations with PD-L1 expression.

Methods: A total of 304 adult patients with heavily pretreated metastatic cancers treated between January 2019 and March 2021 were recruited. The CLIA-/UKAS-accredit Oncofocus assay targeting 505 genes was used on newly obtained or archived biopsies. Chi-square, Kruskal-Wallis, and Wilcoxon rank-sum tests were used where appropriate. Results were significant for P < .05.

Results: A total of 237 tumors (78%) harbored potentially actionable genomic alterations. Tumors were positive for PD-L1 in 68.9% of cases. The median number of mutant genes/tumor was 2.0 (IQR: 1.0-3.0). Only 34.5% were actionable ESCAT Tier I-II with different prevalence according to cancer type. The DNA damage repair (14%), the PI3K/AKT/mTOR (14%), and the RAS/RAF/MAPK (12%) pathways were the most frequently altered. No association was found among PD-L1, ESCAT, age, sex, and tumor mutational status. Overall, 62 patients underwent targeted treatment, with 37.1% obtaining objective responses. The same molecular-driven treatment for different cancer types could be associated with opposite clinical outcomes.

Conclusions: We highlight the clinical value of molecular profiling in metastatic solid tumors using comprehensive NGS-based panels to improve treatment algorithms in situations of uncertainty and facilitate clinical trial recruitment. However, interpreting genomic alterations in a tumor type-specific manner is critical.

Keywords: ESCAT; clinical actionability; metastatic; molecular profiling; next-generation sequencing; solid tumors.

Figure 1.

PD-L1 status according to tumor groups. External circle’s % are referred to the proportion of PD-L1+ and —patients within each tumor type/group. Internal circle’s % are referred to the frequency of a specific tumor type/group with respect to the overall patients cohort. Abbreviations: +, positive; −, negative; BC, breast cancer; CNS, central nervous system primary tumors; GI, gastrointestinal; GU, genitourinary (including prostate); Gyneco, gynecological malignancies; H&N, head and neck; HCC, hepatocellular carcinoma; PA, pancreatic adenocarcinoma; BT, biliary tract; skin, includes melanoma and non-melanoma skin cancers.

Figure 2.

Mutational landscape. In the central pie plot, OTHERS includes all mutant genes with a prevalence <0.8% and values are rounded. Abbreviation: TKR, tyrosine kinase receptor.

Figure 3.

Actionability of detected mutations according to the ESCAT scale. A: Global proportion of different ESCAT tiers in the mutant cohort. Here, the best ESCAT tier for each single mutation detected was considered, independently of the tumor where it was detected. B: Prevalence of ESCAT I-II and III-V or X mutations according to tumor. Abbreviations: ESMO, European Society for Medical Oncology; ESCAT, ESMO Scale for Clinical Actionability of Molecular Targets; CNS, central nervous system primary tumors; H&N, head and neck; HCC, hepatocellular carcinoma; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; CUP, cancer of unknown primary.