PFS Benefit in NSCLC Immunotherapy Predicted With Liquid Biopsy for Mutation Burden

According to a retrospective analysis of a randomized study presented at the 2017 ESMO Congress, a liquid biopsy that measures tumor mutational burden showed promise for predicting benefit in patients with non–small cell lung cancer treated with a checkpoint inhibitor. 

David R. Gandara, MD

According to a retrospective analysis of a randomized study presented at the 2017 ESMO Congress, a liquid biopsy that measures tumor mutational burden showed promise for predicting benefit in patients with non—small cell lung cancer (NSCLC) treated with a checkpoint inhibitor

Tumor mutation burden in blood (bTMB) ≥16 had a significant association with improved progression-free survival (PFS) in patients treated with the PD-L1 inhibitor atezolizumab (Tecentriq), according to findings reported at the 2017 ESMO Congress. The magnitude of improvement in PFS increased with bTMB. By contrast, patients with bTMB ≥16 or <16 bTMB, derived a similar survival benefit from treatment with atezolizumab.

“This is the first demonstration that tumor mutational burden can be measured in blood and that the blood tumor mutational burden is associated with progression-free survival from immune checkpoint inhibitor therapy,” said lead author David R. Gandara, MD, director of thoracic oncology at the University of California Davis Comprehensive Cancer Center in Sacramento. Gandara is a 2017 Giants of Cancer Care&reg;award winner.

“In this analysis, blood tumor mutational burden identified a unique patient population and was not significantly associated with PD-L1 status by immunohistochemistry. Prospective studies in first-line non—small cell lung cancer using this assay for blood tumor mutational burden are currently ongoing,” he said.

In the first-line setting of NSCLC, testing for PD-L1 expression has become standard of care and identifies a patient population that has improved PFS and overall survival (OS) when treated with a checkpoint inhibitor. Gandara and colleagues previously demonstrated that TMB in tumor tissue correlated with atezolizumab efficacy in patients with NSCLC.

Approximately 30% of patients have inadequate tumor tissue to perform molecular testing. Circulating blood-derived DNA might provide an alternative source of diagnostic material for patients with inadequate tissue, Gandara noted. Whether bTMB correlated with efficacy in NSCLC remained unknown.

To test and validate a bTMB assay and assess the association between bTMB and atezolizumab efficacy, investigators evaluated plasma samples from patients with NSCLC enrolled in 2 randomized trials comparing atezolizumab and docetaxel. The study involved 211 samples from the phase II POPLAR study and 583 samples from the phase III OAK trial.

The samples were tested for bTMB using a 394-gene next-generation sequencing assay. Specifically, the assay analyzes single nucleotide variants (SNVs) in the target genes and reports a score based on the number of high-confidence SNVs identified. The results are then grouped by cut points based on the minimum number of SNVs present.

Samples from the POPLAR trial served as the test set, and plasma samples from OAK represented the validation set. Based on results from the POPLAR samples, investigators chose bTMB ≥16 as the cutoff for assessing atezolizumab efficacy in the OAK validation set. In the POPLAR test set, bTMB ≥16 was associated with a PFS hazard ratio of 0.57 and an OS hazard of 0.56 for atezolizumab versus docetaxel.

In the OAK study, the bTMB ≥16 population accounted for 27% of all patients evaluable for assessment of TMB. The results showed that bTMB ≥16 was associated with a 35% reduction in the hazard for progression or death among patients treated with atezolizumab. In contrast, PFS did not differ significantly between patients with bTMB ≥16 or bTMB <16 in the docetaxel arm (HR, 0.98).

A significant PFS benefit of atezolizumab was observed across the bTMB range of 4 to ≥26, associated with hazard ratios of 0.73 to 0.51. The OS benefit ranged from 0.69 to 0.50 across the range of bTMB from 10 to ≥26.

An analysis of factors associated with increased tumor mutational identified smoking status as a significant predictor, consistent with previous studies, said Gandara. Additionally, the results suggested tumor volume and greater number of metastatic sites might correlate with higher bTMB.

An analysis of 69 paired tissue/blood specimens demonstrated a high concordance for assessments of TMB. Gandara said 41 of 46 positive results were true positives (TMB ≥ in tissue and blood) and all 23 negative results were true negatives.

In the OAK validation set, investigators found limited overlap between bTMB and PD-L1 expression, as 30 of 229 samples were positive by both assessments. PD-L1 positivity was associated with a PFS hazard of 0.62 and a survival hazard of 0.44 among patients treated with atezolizumab. Corresponding values for bTMB ≥16 were 0.64 and 0.64. Positivity by both measures produced the greatest benefit: HR 0.38 for PFS and HR 0.23 for OS.


Gandara DR, Kowanetz M, Mok T, et al. Blood-based biomarkers for cancer immunotherapy: tumor mutational burden in blood (bTMB) is associated with improved atezolizumab (atezo) efficacy in 2L+ NSCLC (POPLAR and OAK). Presented at: 2017 ESMO Congress; Madrid, Spain; September 9-12, 2017. Abstract 1295O.

Among patients treated with atezolizumab, the OS benefit did not differ between the bTMB ≥16 and bTMB <16 subgroups (HR, 0.64 and HR, 0.65, respectively). The lack of difference might reflect the impact of postprogression therapies, said Gandara. The bTMB ≥16 cutoff was associated with a median OS of 13.5 months among patients treated with atezolizumab versus 6.8 months for patients randomized to docetaxel.