Identification and validation of tissue or ctDNA PTPRD phosphatase domain deleterious mutations as prognostic and predictive biomarkers for immune checkpoint inhibitors in non-squamous NSCLC

Background With the revolutionary progress of immune checkpoint inhibitors (ICIs) achieved in non-small cell lung cancers (NSCLC), identifying patients benefiting from ICIs becomes critical and urgent. The associations of genomic alterations in protein tyrosine phosphatase receptor-type (PTPRs) and ICIs responses are unknown. Methods Whole-exome sequencing (WES) of 73 advanced NSCLC tumors sampled before anti-PD-(L)1 therapy was carried out with corresponding clinical data collected as a discovery cohort to find the associations of PTPR mutations and ICI responses. Three validation cohorts consolidated by 7 public cohorts of 1920 NSCLC patients with WES or target sequencing data of tumor tissue-derived DNA or circulating tumor DNA (ctDNA) and relevant clinical data were applied as validation cohorts. The lung adenocarcinoma (LUAD) cohort ( n =586) in The Cancer Genome Atlas (TCGA) database was used for analyzing the potential anti-tumor immunologic mechanisms. Results With the highest mutation frequency among all PTPRs, PTPRD mutations in non-squamous NSCLC (ns-NSCLC) were linked to longer progression-free survivals (PFS, 324 vs 63 days, hazard ratio (HR)=0.36, p = 0.0152) and higher objective response rate (ORR, p =0.0099). In validation cohort 1 ( n =377), ns-NSCLC patients with tissue PTPRD mutations had favorable PFS (9.10 vs 4.33 months, HR=0.62, p =0.0184) and ORR ( p =0.013). In validation cohort 2 ( n =406), ns-NSCLC patients with tissue PTPRD mutations had favorable overall survivals (OS, over 40 vs 11.94 months, HR=0.57, p =0.011). In validation cohort 3 ( n =1137), ns-NSCLC patients with ctDNA PTPRD mutations had longer PFS (6.97 vs 2.73 months, HR=0.63, p =0.028) and higher ORR ( p =0.047). Moreover, it was deleterious mutations in phosphatase domains (phosphatase-mut), rather than other mutations (other-mut), that were responsible of PTPRD’s prediction efficiency. In addition, in validation cohort 3, ctDNA phosphatase-mut also functioned as a predictive biomarker helping identify patients benefiting more from ICIs than chemotherapy (interaction P for PFS=0.0506, for OS=0.04). Univariate and multivariate regression analysis revealed that phosphatase-mut was independent on PD-L1 expression and tumor mutation burden (TMB) to predict. In silico analysis based on TCGA LUAD cohort discovered enhanced anti-tumor immunity in phosphatase-mut patients. Conclusions Tissue or ctDNA PTPRD phosphatase domain deleterious mutations might function as a both prognostic and predictive biomarker predicting clinical outcomes of ICIs in ns-NSCLC patients, independent on TMB or PD-L1 expression.