Positive PD-L1 staining was more frequent in patients with
KRAS mutation compared with wild-type
KRAS patients in the KEYNOTE-001 study
[56]. Consistently, monotherapy with anti-PD-1 antibodies, such as nivolumab or pembrolizumab, initially showed a greater clinical benefit in patients with
KRAS mutation compared with
KRAS wild-type patients
[56]. However, a multi-omics analysis uncovered the heterogeneity of
KRAS-mutant lung adenocarcinomas based on co-occurring genetic alterations including inactivation of TP53 or LKB1 and low expression of the thyroid transcription factor-1 (TTF-1)
[23][24]. The integrative analysis with clinical data indicated that these distinct subsets affect PD-L1 expression and the response to PD-1/PD-L1 inhibitors (
Figure 1). Among them,
KRAS-mutant lung adenocarcinomas with TP53 inactivation is characterized as high PD-L1 expression together with high TMB and marked T-cell infiltration, and showing favorable responses to monotherapy with anti-PD-1/PD-L1 antibodies
[23][24].
In contrast to TP53 inactivation, lung adenocarcinomas with LKB1 inactivation, which is encoded by
serine/threonine kinase 11, is associated with the downregulation of PD-L1 expression and reduced T-cell infiltration. Somatic mutation of
LKB1 occurs in approximately 20% of lung adenocarcinomas and 30% of
KRAS-mutant lung adenocarcinomas, whereas LKB1 inactivation is present as a germline mutation of the autosomal dominant disorder, Peutz–Jeghers syndrome
[23][24][57]. The loss of LKB1 function affects tumor initiation though the dysregulation of cell polarity and the reprograming of energy metabolism, including glucose uptake and pyrimidine/purine balance
[58][59][60][61]. These drastic intracellular transformations can affect the secretion of proinflammatory cytokines, such as interleukin-6 (IL-6) and chemokine (C-X-C motif) ligand 7 (CXCL7), resulting in the accumulation of immunosuppressive neutrophils and exhausted or suppressed infiltrated T cells
[62]. Consistent with these basic molecular analyses, a pan-cancer T-cell-inflamed gene expression profile (GEP) consisting of 18 genes, which represent the T-cell-activated tumor microenvironment (TME), revealed that somatic mutation of LKB1 was one of the most prevalent driver alterations in immunosuppressed phenotypes in NSCLC known as “cold tumor”
[19]. In fact, anti-PD-1/PD-L1 monotherapy is ineffective in NSCLC with LKB1 inactivation, which exhibits primary resistance to PD-1/PD-L1 blockade with PD-L1 negativity and intermediate or high TMB
[63].
The
KEAP1 inactivating mutation is associated with the immunosuppressive phenotype and is frequently involved in TTF-1-negative lung adenocarcinoma, which was reported as deficient T-cell infiltration from the analysis of a pan-cancer T-cell-inflamed GEP
[64]. KEAP1 is a redox-regulated substrate for the cullin-3 dependent E3 ubiquitin ligase complex, which facilitates the ubiquitination and subsequent proteolysis of nuclear factor erythroid 2-related factor 2 (NRF2), a master regulator of detoxification, antioxidant response, and anti-inflammatory activity
[65]. KEAP1 inactivation results in persistent NRF2 activation; therefore, the tumors are highly resistant to radiotherapy and cytotoxic chemotherapy
[65][66][67]. KEAP1 inactivation is also involved in reprograming to an immunosuppressive TME through Srglycin (SRGN) secretion, which is a chondroitin sulfate proteoglycan that plays an intricate role in inflammation by regulating several inflammatory mediators
[68][69][70]. SRGN expression is transcriptionally upregulated by NRF2 activation and epigenetically induced through nicotinamide N-methyltransferase-induced perturbation of methionine metabolism in TTF-1–negative lung adenocarcinoma
[70]. Cancer cell-derived SRGN upregulates PD-L1 expression on the cancer cell itself and increases the secretion of proinflammatory cytokines, including IL-6, interleukin-8 (IL-8), and chemokine (C-X-C motif) ligand 1 (CXCL1), indicating that it contributes to reprogramming into an aggressive and immunosuppressed phenotype
[70][71]. Similar to NSCLC with LKB1 inactivation, NSCLC with disruption of the KEAP1-NRF2 pathway is widely known respond poorly to monotherapy with anti-PD-1/PD-L1 antibodies. Arbour et al. analyzed co-occurring the genetic alterations of 330 patients with
KRAS-mutant NSCLC by NGS and found that
KEAP1-NRF2 alterations occurred in 27% of the patients that had shorter OS from the initiation of immunotherapy
[23]. Furthermore, a subset of NSCLC harbors inactivating mutations of both
LKB1 and
KEAP1/NRF2 and demonstrate a further aggressive clinical course with strong resistance to ICIs treatment
[72]. Papillon-Cavanagh et al. analyzed the clinical efficacy of PD1/PD-L1 inhibitors or platinum-based chemotherapy against NSCLC with the double-mutational status in a real world-setting. Patient outcome for both treatments was worse progression-free survival (PFS) and OS compared with patients harboring only an
LKB1 alteration, only
KEAP1/NRF2 alterations, or a negative status for both
[73]. These results indicate that co-occurring genetic alterations of
LKB1 and
KEAP1/NRF2 have an additive effect for tumor aggressiveness even with combined ICI regimens containing cytotoxic chemotherapy. To improve the outcome of NSCLC with these aggressive phenotypes, new therapeutic developments are needed.