Until more biomarkers are available, the use of tumor mutational burden with PD-L1 expression could help oncologists further personalize immunotherapy choices for patients with non–small cell lung cancer.
Christina S. Baik, MD, MPH
The search continues for a biomarker for directing patients to immunotherapy treatments; however, until more biomarkers are available, the use of tumor mutational burden (TMB) with PD-L1 expression could help oncologists further personalize immunotherapy choices for patients, according to Christina S. Baik, MD, MPH.
Baik, an associate professor at the University of Washington School of Medicine, and assistant member of the Clinical Research Division at Fred Hutchinson Cancer Research Center, discussed the potential role of TMB and PD-L1 expression as markers for treatment with immunotherapy and chemotherapy in lung cancer, as well as the current challenges with TMB, during a presentation at the 16th Annual Winter Lung Cancer ConferenceTM.
TMB, she explained, is a measurement of somatic mutations that are present in the tumor, and it estimates the tumor neoantigenic load that is presented by the tumor cells and are recognized by T cells.1
“It’s really an estimate of how immunogenic a tumor is,” she noted.
Currently, there is only 1 FDA-approved tissue-based next-generation sequencing (NGS) assay for TMB testing: Foundation Medicine. The Foundation Medicine assay looks at 324 genes and calculates both synonymous and nonsynonymous mutations and then subtracts germline mutations using bioinformatics. With this assay, 10 mutations per megabase is considered to be “high” TMB.
Although the FDA has cleared the NGS assay MSK-IMPACT for testing for single-nucleotide variants, indels, and microsatellite instability, it has not yet been approved for TMB. The MSK-IMPACT assay looks at 468 genes and calculates nonsynonymous mutations and subtracts germline mutations using patient-matched blood samples. Per this assay, approximately 7 mutations per megabase is considered “high” TMB.
An initial clinical study that showed the importance of TMB found that in 2 small independent cohorts of patients treated with pembrolizumab (Keytruda) monotherapy, durable clinical benefit was associated with a high degree of nonsynonymous mutational burden and high neoantigen burden.2
This has since been confirmed by larger studies, such as CheckMate-026, which looked at first-line nivolumab (Opdivo) compared with chemotherapy in patients with stage IV or recurrent NSCLC and a PD-L1 expression of ≥5%. The results in the overall population were negative, showing a median progression-free survival (PFS) of 4.2 months (95% CI, 3.0-5.6) with nivolumab versus 5.9 months (95% CI, 5.4-6.9) with chemotherapy (HR, 1.15; 95% CI, 0.91-1.45; P = .25). Yet, among patients with a high TMB, the median PFS was 9.7 months with nivolumab (95% CI, 5.1-not reached [NR]) compared with 5.8 months with chemotherapy (95% CI, 4.2-8.5), leading to a 38% reduction in the risk of disease progression (HR, 0.62; 95% CI, 0.38-1.00).3
“One question you may ask is, ‘Is TMB really providing additional information to PD-L1 or are these just interchangeable biomarkers?’”
The CheckMate-026 trial, however, did not show a significant association between PD-L1 and TMB. But with regards to response rates, patients with high TMB and PD-L1 expression (≥50%) had the highest objective response rate (ORR) at 75%, and those with a low/median TMB and lower PD-L1 expression (1%-49%) had the lowest ORR at 16%. This was compared with the ORR in the overall population of 26%, “suggesting that these are complimentary biomarkers,” Baik said.
The CheckMate-227 trial brought TMB to the forefront, according to Baik. The study randomized patients with stage IV or recurrent nonsquamous or squamous NSCLC who had not previously received chemotherapy according to their PD-L1 status (≥1% or <1%). Patients in the 2 cohorts then received nivolumab plus ipilimumab (Yervoy), nivolumab with or without chemotherapy, or histology-based chemotherapy. A co-primary analysis from the study also looked at TMB in patients treated with nivolumab plus ipilimumab or chemotherapy alone.
In patients with high TMB, there was a PFS benefit with the immunotherapy doublet compared with chemotherapy for both patients with positive and negative PD-L1 expression. For patients with PD-L1 ≥1%, the hazard ratio for disease progression or death was 0.62 (95% CI, 0.44-0.88), and for patients with PD-L1 <1%, the hazard ratio was 0.48 (95% CI, 0.27-0.85).4
When PFS was analyzed for patients with negative PD-L1 expression according to TMB, PFS was prolonged in patients with high TMB (≥10 mutations per megabase) who received the immunotherapy combination versus in those who received nivolumab/chemotherapy and compared with those who received chemotherapy (7.7, 6.2, and 5.3 months, respectively). In the lower TMB group (<10 mutations per megabase), patients receiving nivolumab and chemotherapy and those receiving chemotherapy alone both had a median PFS of 4.7 months compared with 3.1 months in those receiving nivolumab and ipilimumab.5
“At least the PFS data are very compelling for TMB being a predictive marker…especially for nivolumab/ipilimumab,” Baik said.
However, Bristol-Myers Squibb recently withdrew their supplemental biologics license application to the FDA for the combination of nivolumab and ipilimumab for the treatment of patients with advanced NSCLC with high TMB.6 While she said that not much information is known yet, as several pieces of data from the trial are still pending, she noted that more recent data did not show a different in overall survival between patients according to TMB status—and not all patients had TMB-evaluable tumors.7
“I think we need the full data set to really make sense of the data.”
Overall, Baik suggested that literature is consistent in showing that high TMB is associated with a clinical response to immune checkpoint inhibition therapy. Additionally, she suggested that TMB and PD-L1 are likely complementary biomarkers, reflecting activity at different steps of the immunity cycle.
“With all of this, the million-dollar question is, ‘Is TMB ready to be used in the clinic?’ There are varying opinions on this out there, but my personal opinion is: not yet,” she commented.
Challenges also exist for implementing TMB in clinical use, Baik noted, as there is not a great degree of standardization between assay results.
Baik explained how she currently stratifies her patients with lung cancer for chemotherapy and/or immunotherapy based on current information with TMB and PD-L1. For patients with “hot” or inflamed tumors, which demonstrate both high PD-L1 expression and TMB, she gives anti—PD-1 monotherapy. For patients with high PD-L1 expression but low TMB, she gives chemoimmunotherapy. For those with high TMB but low or negative PD-L1 expression, she gives chemoimmunotherapy or anti–PD-1/CTLA-4 therapy. Moreover, for patients with “cold” or noninflamed tumors that demonstrate both low TMB and low or negative PD-L1 expression, she administers chemotherapy with or without immunotherapy, or possibly cellular immunotherapy.
Concluding, Baik said that what is most important with regard to the future use of biomarkers for immunotherapy, is rigorous validation before they are to be used in the clinic.