Immune checkpoint inhibitors (ICIs) are highly concerned in the treatment of non-small cell lung cancer (NSCLC), represented by inhibitors of programmed death protein 1 (PD-1) and its ligand (PD-L1), and inhibitors of cytotoxic T lymphocyte-associated antigen-4 (CTLA-4). The introduction of immunotherapy in the treatment of perioperative NSCLC has improved the prognosis to a great extent, as demonstrated by several phase II and III clinical trials. The target population for immunotherapy in early-stage NSCLC is still under discussion, and the biomarkers for neoadjuvant immunotherapy population selection are the next pending problem. The predictive efficacy of many potential makers is still being explored, including PD-L1 expression levels, tumor mutation burden, circulating tumor DNA, components of the tumor microenvironment, and several clinical factors. We summarize key findings on the utility of ICIs in clinical trials of preoperative NSCLC patients and conclude analyses of relevant biomarkers to provide a better understanding of potentially predictive biomarkers in neoadjuvant immunotherapy.

Introduction: Recent studies exhibited the unstable prediction ability of blood-based tumor mutational burden (bTMB) when predicting the response of immune checkpoint inhibitors (ICIs) therapy in patients with non-small cell lung cancer (NSCLC). Circulating tumor DNA (ctDNA) abundance, usually represented by maximum somatic allele frequency (MSAF), was one possible confounding factor influencing bTMB ability in ICIs response prediction. Methods: MSAF-adjusted bTMB (Ma-bTMB) was established and validated in patients with advanced NSCLC among Geneplus Cancer Genome Database (GCGD, n = 1679), Zhuo (n = 35), Wang (n = 45), POPLAR (NCT01903993, n = 211) and OAK (NCT02008227, n = 642) cohorts. Results: MSAF demonstrated a modest positive correlation with bTMB and a negative one with survival benefit. Improved survival outcomes of ICIs therapy have been observed among patients with high-Ma-bTMB compared to those with low-Ma-bTMB in Zhuo and Wang cohorts. In addition, compared to low-Ma-bTMB, high-Ma-bTMB was associated with more positive clinical benefits from ICIs therapy than chemotherapy both in POPLAR and OAK cohorts. Further exploration suggested that Ma-bTMB could precisely identify more potential ICIs beneficiaries compared to bTMB and LAF-bTMB, complementary to PD-L1 expression. Conclusions: We developed Ma-bTMB, a convenient, readily available, non-invasive predictive biomarker effectively differentiates beneficiaries of ICIs therapy in advanced NSCLC, warranting future clinical trials.