Genomic Features of Response to Combination Immunotherapy in Patients with Advanced Non-Small-Cell Lung Cancer

Abstract

Combination immune checkpoint blockade has demonstrated promising benefit in lung cancer, but predictors of response to combination therapy are unknown. Using whole-exome sequencing to examine non-small-cell lung cancer (NSCLC) treated with PD-1 plus CTLA-4 blockade, we found that high tumor mutation burden (TMB) predicted improved objective response, durable benefit, and progression-free survival. TMB was independent of PD-L1 expression and the strongest feature associated with efficacy in multivariable analysis. The low response rate in TMB low NSCLCs demonstrates that combination immunotherapy does not overcome the negative predictive impact of low TMB. This study demonstrates the association between TMB and benefit to combination immunotherapy in NSCLC. TMB should be incorporated in future trials examining PD-(L)1 with CTLA-4 blockade in NSCLC.

Keywords: CTLA-4; PD-1; TMB; immunotherapy; lung cancer; mutation burden.

Figure 1

TMB Correlates with Efficacy in Patients with NSCLC Treated with Nivolumab Plus Ipilimumab (A) TMB in patients with complete response (CR)/partial response (PR) (n = 24, blue) versus stable disease (SD)/progressive disease (PD) (n = 51, red) (median 273 versus 114 mutations, Mann-Whitney p = 0.0004) and TMB in patients with DCB (green, n = 37) versus those with NDB (purple, n = 38) (median 210 versus 113 mutations, Mann-Whitney p = 0.0071). Medians, interquartile ranges, and minimum/maximum shown in boxplots. (B) Objective response and durable clinical benefit in patients with high TMB (>median, 158 mutations) versus low TMB (≤median) (ORR 51% versus 13%, odds ratio 6.97 [95% confidence interval (CI) 2.19–19.0], Fisher's exact p = 0.0005; DCB 65% versus 34%, odds ratio 3.55 [95% CI 1.3–8.64], Fisher's exact p = 0.011). Proportion of CR/PR or DCB, respectively, are colored on histograms with rate (n/N) shown above each bar. (C) PFS in patients with high TMB versus low TMB (median 17.1 versus 3.7 months, Mantel-Haenszel hazard ratio 0.41 [95% CI 0.23–0.73], log rank p = 0.0024). (D) Receiver operating characteristic (ROC) curves for correlation of TMB with objective response (CR/PR; blue line) (AUC 0.75 [95% CI 0.62–0.88], p = 0.0006) and DCB (green line) (AUC 0.68 [95% CI 0.56–0.8], p = 0.0076). (E) PFS in cohorts of patients defined by quartiles of TMB percentile rank among NSCLC tumors profiled by TCGA (log rank for trend p = 0.01). See also Figure S2 and Table S3.

Figure 2

Summary of Clinical and Molecular Features Associated with Response or Non-response in Patients with NSCLC Treated with Nivolumab Plus Ipilimumab Individual patients are represented in each column, organized by those with objective response on the left (blue) and those with no objective response on the right (red). Categories of histology (squamous or non-squamous) and smoking status (never or ever) are characterized. PD-L1 expression is stratified as 0%, 1%–49%, or ≥50%. PFS is shown in months, with the color of each bar representing those who are censored (dark blue) or have progressed (light blue). The NSCLC TCGA percentile rank for each case is described from 0% to 100% in light to dark purple. Nonsynonymous TMB and mutation burden quantified using genes including in the MSK-IMPACT targeted next-generation sequencing panel are shown in histograms. The percent of transitions (light green) and transversions (dark green) are shown. Candidate neoantigen burden is quantified in histograms, stratified by predicted patient-specific HLA binding affinity 0–50 nM (orange) or 50–500 nM (light yellow). The occurrences of selected genes in each case are represented in the oncoprint, with the percent frequency in responders or non-responders shown. See also Figures S2 and S3; Tables S3 and S4.

Figure 3

Association between TMB and Efficacy in Multivariate Context (A) Correlation between TMB and PD-L1 expression (Spearman ρ −0.087 [95% CI −0.32 to 0.16], p = 0.48). Patients with CR/PR (n = 24) are colored in blue circles; those with SD/PD (n = 51) are colored in gray squares. (B) ROC curves for multivariate model correlation with objective response (CR/PR), with model including TMB (continuous variable), PD-L1 (continuous), histology (binary, squamous versus non-squamous), smoking status (binary, ever versus never), performance status (Eastern Cooperative Oncology Group [ECOG] 0 versus 1), and tumor burden (binary, > versus ≤ median) (plain line, AUC 0.869). Univariate correlation of TMB with objective response is shown again for reference (dotted line). (C) ROC curves for univariate correlation of TMB (continuous) with progression-free survival (dotted line) at 6 months (purple, AUC = 0.585) or 12 months (yellow, AUC = 0.558). ROC curves for multivariate correlation of model including TMB (continuous), PD-L1 (continuous), histology (squamous versus non-squamous), smoking status (ever versus never), performance status (ECOG 0 versus 1), and tumor burden (binary, > versus ≤ median) also shown (plain lines; at 6 months AUC = 0.764, at 12 months AUC = 0.831). (D) Histogram of objective response (CR/PR) to nivolumab plus ipilimumab in patients characterized by high mutation burden (>median TMB) and PD-L1 expression (≥1%), high mutation burden or PD-L1 expression, or neither. Response rates (n/N) are shown above each bar, with proportion of those with PR/CR colored in blue. Chi-square for trend p < 0.0001. See also Figure S4.