List of 301 genes included in all versions of MSK-IMPACT and DFCI OncoPanel NGS platforms used for all analyses.

Kaplan-Meier curves of overall survival of patients with long-term response (LTR), short-term response (STR), and progressive disease (PD) from individual sites.

Time on treatment among patients with long-term response (LTR).

Kaplan-Meier curve of progression-free survival (PFS) among patients with long-term response (LTR) in the combined cohort.

Univariable and multivariable analyses of association of clinical characteristics with patients achieving complete response (CR) compared to patients not achieving complete response (non-CR).

PURPOSE:Among patients with advanced non-small cell lung cancer (NSCLC) who discontinue immune checkpoint inhibitors (ICI) because of immune-related adverse events (irAE), post-discontinuation clinical outcomes and factors associated with disease progression after discontinuation are largely unknown. EXPERIMENTAL DESIGN:Clinicopathologic data were abstracted from patients with advanced NSCLC who received ICI and discontinued treatment because of irAE. Factors associated with post-discontinuation progression-free survival (PFS) and post-discontinuation overall survival (OS) were evaluated. RESULTS:Of 2,794 patients, 10% (N = 271) discontinued ICI because of irAE, and the median duration of ICI treatment before discontinuation for irAE was 5.9 months (range, 0.03-73.5). A longer treatment duration before discontinuation for irAE was associated with improved post-discontinuation outcomes: for patients on ICI for <3 months (N = 89), 3 to 6 months (N = 49), and >6 months (N = 133) before discontinuing for irAE, the median post-discontinuation PFS was 6.2, 13.9, and 25.8 months (P < 0.001), respectively, and the median post-discontinuation OS was 21.7, 42.7, and 86.9 months (P < 0.001), respectively. At multivariable analyses, predictors of longer post-discontinuation PFS were PD-L1 ≥ 50%, complete response/partial response (CR/PR) to treatment, and treatment duration before discontinuation between 3 to 6 months and >6 months; predictors of longer post-discontinuation OS were nonsquamous histology, CR/PR, and treatment duration before discontinuation >6 months. The use of immunosuppressive agents for toxicity management did not affect post-discontinuation outcomes. CONCLUSIONS:A longer treatment duration before discontinuation, a best objective response of CR/PR, PD-L1 ≥50%, and nonsquamous histology may help clinicians identify patients who may experience long-term disease control after discontinuation of ICI for irAE.

Distribution of tumor mutational burden (TMB) z-scores among patients with long-term response (LTR), short-term response (STR), and progressive disease (PD).

Frequency of long-term response (LTR) and short-term response (STR) according to composite PD-L1 and tumor mutational burden (TMB) status.

Genomic summary metrics and HLA features of patients with long-term response (LTR), short-term response (STR), and progressive disease (PD) among patients with next-generation sequencing by MSK-IMPACT.

List of PD-L1 antibodies used at each site and in the combined cohort.

Univariable analyses of association of clinical characteristics with long-term response (LTR) compared to short-term response (STR), non-LTR, and progressive disease (PD).

Genomic features of patients with long-term response (LTR) compared to short term response (STR), non-LTR, and progressive disease (PD) among patients with TMB z-score ≥ 0.

Supplemental Figure 1: OS and PFS for KRAS-mutated NSCLC treated with first-line platinum-based chemotherapy.

Frequencies of alterations in individual genes among non-squamous patient groups stratified by TMB z-score.

Whole exome sequencing (WES) features of patients with long-term response (LTR), short-term response (STR), and progressive disease (PD).

Frequencies of alterations in individual genes among squamous and non-squamous patient groups.

AbstractBackground: The genetic factors that modulate risk for developing lung cancer have not been fully defined. Here, we sought to determine the prevalence and clinical significance of germline pathogenic/likely pathogenic variants (PV) in patients with advanced lung cancer. Methods: We studied clinical and tumor characteristics of germline PV in 5,118 patients who underwent prospective genomic profiling using paired tumor–normal tissue samples in 468 cancer genes. Results: Germline PV in high/moderate-penetrance genes were observed in 222 (4.3%) patients; of these, 193 patients had PV in DNA damage repair (DDR) pathway genes including BRCA2 (n = 54), CHEK2 (n = 30), and ATM (n = 26) that showed high rate of biallelic inactivation in tumors. BRCA2 heterozygotes with lung adenocarcinoma were more likely to be never smokers and had improved survival compared with noncarriers. Fourteen patients with germline PV in lung cancer predisposing genes (TP53, EGFR, BAP1, and MEN1) were diagnosed at younger age compared with noncarriers, and of tumor suppressors, 75% demonstrated biallelic inactivation in tumors. A significantly higher proportion of germline PV in high/moderate-penetrance genes were detected in high-risk patients who had either a family history of any cancer, multiple primary tumors, or early age at diagnosis compared with unselected patients (10.5% vs. 4.1%; P = 1.7e−04). Conclusions: These data underscore the biological and clinical importance of germline mutations in highly penetrant DDR genes as a risk factor for lung cancer. Impact: The family members of lung cancer patients harboring PV in cancer predisposing genes should be referred for genetic counseling and may benefit from proactive surveillance.

Genomic features of patients with long-term response (LTR) compared to short term response (STR), non-LTR, and progressive disease (PD) among patients with TMB z-score < 0.

Supplemental Figure 3: PFS estimates for patients with NSCLC treated with standard first- and second-line therapies.