Characterization of TP53 mutations in actionable driver-negative non-small cell lung cancer (NSCLC).

Abstract

8621 Background: TP53 mutations are common in NSCLC and have been associated with poor prognosis in patients with actionable genomic drivers. Despite the increased prevalence in actionable driver-negative (DN) NSCLC, the role of TP53 mutations in this subtype is not fully explored. We characterize the association of TP53 mutations with the molecular and immune landscapes, as well as outcomes, in DN NSCLC histologic subtypes. Methods: 17,689 NSCLC specimens were analyzed using next-generation sequencing of DNA (DNA-592 panel or whole exome) or RNA (whole transcriptome) at Caris Life Sciences (Phoenix, AZ). Tumors with KRAS, EGFR, BRAF, MET, HER2, ALK, ROS1, NRG1, and NTRK 1-3 alterations were excluded. Specimens were grouped as TP53WT (wild type) or TP53MUT (pathogenic/likely pathogenic mutations). PD-L1 expression was analyzed by IHC (Dako 22c3; PD-L1 positive: TPS ≥1%). TMB was measured by counting all somatic mutations found per tumor (TMB-H: ≥10 mutations/Mb). Composition of the tumor microenvironment (TME) was estimated from bulk RNA sequencing using QuanTIseq method. Overall survival was extracted from insurance claims data and calculated from the time of tissue collection (OS) or initiation of Pembrolizumab treatment (Pembro-OS) to the last contact, using Kaplan-Meier estimates. Significance was determined using chi-square, Fisher’s Exact, Mann Whitney U and adjusted for multiple comparisons (p-value < 0.05). Results: Among DN tumors, the prevalence of TP53MUT was 86%. TP53MUT was more common in men (62% vs 38%), those with a smoking history (99% vs 1%), and those with squamous cell histology (46% vs33.6%, all p<0.05). Interestingly, histological stratification suggested that TP53MUT was associated with poor OS in adenocarcinoma (AC) (HR:1.3, p<0.05), and shorter Pembro-OS in squamous cell carcinoma (SCC) (HR:1.425, p<0.05). In AC, mutations in FAT1, ARID1A, SMARCA4, ARID2 and RBM10 were enriched, while in SCC, mutations in PIK3CA were less common with TP53MUT. Gene sets involved in cell cycle regulation such as G2M, MYC, and E2F targets were enriched in TP53MUT and to a greater extent in AC (1.22-1.29) vs SCC (1.06-1.08, both p<0.05). Looking at immunotherapy-related biomarkers, both AC and SCC TP53MUT tumors had a higher prevalence of TMB-H compared to TP53WT, while PD-L1 positivity was enriched in only AC TP53MUT tumors. Conversely, KEAP1 and STK11 mutations were enriched in AC TP53WT. M2 Macrophage (1.25-fold) and NK cell (1.3-fold) fractions were enriched and B cell fractions (1.25-fold) were less common in TP53MUT SCC TMEs (all p<0.05). Conclusions: TP53MUT were associated with poor OS and Pembro-OS exclusively in AC and SCC, respectively. The distinct molecular and immune landscapes associated with TP53MUT in AC and SCC potentially explain the differences in outcomes mentioned above. Further research is needed to better understand therapeutic implications of TP53 mutations in DN-NSCLC.