In an updated analysis of CheckMate-586, investigators found that tumor mutational burden identified patients with non–small cell lung cancer who were more likely to respond to first-line combination immunotherapy with nivolumab and ipilimumab.
Suresh S. Ramalingam, MD
In an updated analysis of CheckMate-586, investigators found that tumor mutational burden (TMB) identified patients with nonsmall cell lung cancer (NSCLC) who were more likely to respond to first-line combination immunotherapy with nivolumab (Opdivo) and ipilimumab (Yervoy).1
According to results of the analysis, a cutoff of 10 mutations per megabase of DNA (mut/Mb) was associated with an objective response rate (ORR) of 44%. Increasing the cutoff to 15 mut/Mb did not identify patients with a greater likelihood of response, and a lower cutoff was associated with significantly lower response rates. The 10 mut/Mb cutoff also identified a patient subgroup with better 6-month progression-free survival (PFS).
A comparison of performance metrics with PD-L1 assessment confirmed previous evidence showing that TMB and PD-L1 are independent biomarkers of response to immunotherapy, Suresh S. Ramalingam, MD, director of medical oncology at Emory University’s Winship Cancer Institute in Atlanta, Georgia, said at the American Association for Cancer Research (AACR) meeting in Chicago, Illinois.
“A tumor mutational burden greater than or equal to 10 mutations per megabase was identified as the optimal cutoff for tratment4 benefit for first-line nivolumab-ipilimumab,” said Ramalingam. “In this study, 49% of first-line nonsmall cell lung cancer patients had a tumor mutational burden of 10 mutations per megabase or greater.”
“PD-L1 and TMB identify distinct and independent populations of non-small cell lung cancer,” he added. “PD-L1 and TMB independently are associated with enhanced objective response rate and progression-free survival.”
Although anti-PD-1 immunotherapy has several approved indications in NSCLC, the need has persisted for a biomarker to identify patients more likely to benefit from the treatment. PD-L1 expression has been evaluated most extensively, and approved approaches to first-line therapy for NSCLC include anti-PD-1 monotherapy for patients with PD-L1 expression ≥50% by a required assay, as well as immuno-oncology combination therapy.1
More recently, TMB has emerged as a potential predictive biomarker of response to immuno-oncology therapy in multiple tumor types, including lung cancer. In the CheckMate 012 trial of first-line combined immunotherapy for NSCLC, TMB was shown to be predictive of response, independent of PD-L1 expression status.2
Ramalingam reported findings from an analysis of CheckMate 568, which evaluated first line nivolumab-ipilimumab immunotherapy in advanced NSCLC. The trial had a primary endpoint of ORR in PD-L1-stratified populations (<1% and ≥1%). Secondary endpoints included ORR, PFS, and overall survival (OS) by TMB.
For assessment of TMB, investigators used the FoundationOne CDx test, which uses next generation sequencing to detect aberrations in 324 genes. They compared results with TMB assessment and PD-L1 expression. Ramalingam said all 288 patients in CheckMate 568 had tumor tissue for assessment of PD-L1 expression and 120 for TMB assessment; 95 patients had data for both assessments.
The primary objective was to determine the TMB cutoff associated with improved outcomes in patients who received the nivolumab-ipilimumab combination as first-line treatment for advanced NSCLC. Investigators excluded patients with EGFR/ALK alterations, and the combination therapy continued for as long as 2 years.
The combination resulted in an ORR of 30%, including 29% in PDL-1 evaluable patients, and 28% in TMB-evaluable patients. Patients with PD-L1 expression ≥1% had an ORR of 41%, whereas those with lower levels of PD-L1 expression had an ORR of 15%.
Analysis of ORR by TMB showed response rates of 9% for TMB <5, 15% for TMB of 5-10, 44% for TMB ≥10, and 39% for TMB ≥15.
A comparison of receiver operating characteristic curves (ROC) showed an area under the curve (AUC) of 0.70 for PD-L1 expression and 0.73 for a TMB of 9 to 10 mut/Mb.
Patients with a TMB ≥10 had a 6-month PFS of 55% as compared with 31% for patients who had a TMB <10. Median PFS was 7.1 months in association with a TMB ≥10 as compared with 2.6 months for a TMB <10.
Consistent results were observed in a TMB analysis of CheckMate 227, which evaluated the nivolumab-ipilimumab combination as first-line therapy for advanced NSCLC. Data from CheckMate 227, which included 330 TMB-evaluable patients, showed a 6-month PFS of 55% and 1-year PFS of 43% in association with a TMB ≥10 as compared with 36% and 25% among patients with a TMB <10. Median PFS was 7.2 months for patients with TMB ≥10 and 3.2 months for patients with TMB <10.3
CheckMate 227 also was reported at the AACR meeting, and was published simultaneously inThe New England Journal of Medicine.3
Ramalingam reported that response by TMB value was independent of PD-L1 expression status. In patients with PD-L1 expression <1%, the response rate was 47% for patients with TMB ≥10 and 5% for those with TMB <10, although the analysis included only 41 patients. For the subgroup of patients with PD-L1 ≥1% (N = 54), response rates were 42% with TMB ≥10 and 18% for TMB <10.1
ROC analysis for TMB and ORR by PD-L1 expression showed an AUC of 0.90 for patients with PD-L1 <1% and 0.64 for those with PD-L1 ≥1%.
“TMB was an informative classifier of response with nivolumab and ipilimumab in patients with less than 1% tumor PD-L1 expression and 1% or greater PD-L1 expression,” said Ramalingam.