Non-small cell lung cancer (NSCLC) represents one of the most multifaceted tumor based on its complex molecular features. The identification of driver mutations, predominantly identified in the adenocarcinoma histotype, have opened the era of ‘biomarker-driven therapy’ (1). The ‘oncogene addicted’ disease, characterized by a targetable gene alteration, represents a subgroup of NSCLC in which targeted agents have radically changed prognosis and quality of life of these patients (2). With the advent of tumor genotyping, upfront routine testing for targetable biomarker alterations is strongly recommended in all patients with diagnosis of NSCLC to personalize treatment (3). Additional molecular analysis by next generation sequencing (NGS) may be considered as part of broader testing panels, with the final aim of detecting rare druggable alterations to enroll patients in clinical trials. Among gene alterations involved in cancer development, kirsten rat sarcoma (KRAS) represents the most common oncogene driver in human cancer. It has been demonstrated that the wild-type KRAS is a proto-oncogene that is frequently activated during many types of cancer progression (4). Once the KRAS gene mutates, it acquires oncogenic properties that are causally involved in the development of human cancers (5, 6). Such oncogenic forms of the KRAS gene are prevalent in pancreatic carcinomas (>80%), colon carcinomas (40–50%), and lung carcinomas (30–50%), but are also present in biliary tract malignancies, endometrial cancer, cervical cancer, bladder cancer, liver cancer, myeloid leukemia and breast cancer (7). KRAS protein is a small guanine triphosphatase (GTPase) that serves as a binary switch in signal transduction for most receptor tyrosine kinases and plays a key role in regulating cell growth, differentiation and apoptosis by interacting with multiple effectors including mitogen-activated protein kinase (MAPK) (8). Oncogenic KRAS mutations mostly occur in exon 2 at codon 12, less frequently at codon 13 (3-5%) and rarely at exon 3 codon 61 (less than 1%) (9). KRAS signaling begins with the stimulation of a vast array of upstream receptors including receptor tyrosine kinases (RTKs). Adaptor proteins interact with the intracellular domain of activated EGFR and in turn recruit guanine nucleotide exchange factors (GEFs), such as Son of Sevenless (SOS), to the cellular membrane where they can associate with KRAS to promote the switch from inactive GDP to active GTP. KRAS signaling terminates upon the hydrolysis of GTP to GDP by the intrinsic GTPase activity of RAS. Cancer-causing mutations in KRAS drastically impair the GTPase activity, resulting in KRAS proteins that are locked in the active GTP-bound conformation, regardless of the upstream signal. As consequence, these missense mutations cause a constitutive activation of KRAS downstream pathways and promote cell proliferation and tumor development (10). In this review we aim to give an overview on KRAS role in lung cancer pathogenesis and targeting with a special focus on new potential treatments based on epigenetic modulation.

In NSCLC, KRAS mutations occur in 30% of adenocarcinoma and less frequently (about 7%) in squamous-cell lung carcinoma (11). Nevertheless, more recent results suggest that, in most cases, KRAS mutations do not occur in pure squamous-cell lung carcinoma but in mixed histology of adenosquamous or neuroendocrine carcinomas (12, 13). Although most of KRAS-mutant NSCLC are diagnosed in former or active smokers, KRAS mutations can also be detected in never smoker patients with early onset of cancer, thus its mutational state cannot be predicted on the basis of smoking history alone (14). Because of its high frequency in lung adenocarcinoma, several preclinical and clinical investigations have been conducted seeking effective therapeutic approaches targeting KRAS mutations but, for many years, these efforts have been unsuccessful. The deeper knowledge of genomic alterations has led to the identification of different KRAS subtype mutations. Indeed, it has been demonstrated that in NSCLC the most frequent KRAS allelic variations are the p.G12C (39%) and p.G12V (17-18%), resulting from a G:C to T:A transversion as a classical smoking-induced alteration (15). Remarkably, smokers and never smokers have a different spectrum of mutations and codon variants in KRAS. Thus, transition mutations (G12D) are more common in never smokers, whereas transversion mutations (G12C and G12V) are more common in former or current smokers (16). Moreover, KRAS-mutant NSCLC in smoker patients are genomically more complex, with a higher mutational burden and higher frequency of additional mutations in TP53 or STK11 genes compared to never smoker tumors, as result of antigenic exposure and oxidative stress in epithelial cells (17). From a clinical point of view, KRAS-mutant cancers have generally been associated with poorer overall survival (OS) compared to KRAS wild type tumors, especially in the advanced stages (18–20); however, other studies in early (where the benefit of adjuvant chemotherapy is minimal) or advanced stage of KRAS-mutant lung cancer have provided conflicting results, thus the prognostic value of KRAS alteration is still debated (21–23). A systematic review and meta-analysis including 3,620 patients has shown that KRAS mutations confers a significantly worse prognosis in patients with lung adenocarcinoma (19). Furthermore, KRAS mutations also have a predictive value in lung cancer. Indeed, as already outlined, KRAS represents a critical downstream effector on the EGFR pathway. Thus, there is a biologic rationale supporting the hypothesis that KRAS-mutant NSCLC are intrinsically resistant to EGFR TKIs because of a constitutive activation of the kinase that regulates EGFR signaling (24–27). With this regard, several studied have evaluated the efficacy of erlotinib or gefitinib according to KRAS mutational status confirming an association between KRAS mutations and limited activity of EGFR TKIs (28, 29). Several investigators have also hypothesized that KRAS may be a predictive biomarker of response to chemotherapy. Indeed, early preclinical data have shown that KRAS mutations correlate with a greater sensitivity to pemetrexed in NSCLC models (30) probably due to the pemetrexed ability to alter KRAS RNA expression. From a biologic point of view, this result is explained by a strong dependency of KRAS-mutant cells on folate metabolism pathways. Unfortunately, these preclinical results did not translate into an increased clinical activity (31, 32); indeed, a retrospective study conducted on 1,190 KRAS-mutant NSCLC patients treated with upfront platinum-based doublet chemotherapy has demonstrated that pemetrexed was associated with a shorter progression-free interval compared to other agents, such as taxanes (32). Nevertheless, at the ESMO 2021 congress, in the NVALT22 study cisplatin-pemetrexed chemotherapy plus bevacizumab did not improve PFS compared with carboplatin-paclitaxel plus bevacizumab in KRAS mutant patients (33). Concerning other potential targets, Mainardi et al. (34) have found that Src homology region 2 (SHP2), a cytoplasmic Src homology 2 domain containing protein tyrosine phosphatase that regulates several cellular processes, is necessary for KRAS-mutant tumor cell growth in in vivo models of NSCLC. Accordingly, the authors have highlighted that SHP2 inhibition promotes a senescence response in KRAS-mutant NSCLC cells, suggesting a rationale for targeting SHP2 in NSCLC harboring KRAS mutations. Indeed, a recent study has reported that the co-inhibition of SHP2 and KRAS in xenograft models is effectively able to suppress feedback reactivation of KRAS pathway, producing a significant tumor shrinkage (35). With regard to immunotherapy, KRAS mutations have been associated with immune checkpoint inhibitors (ICIs) benefit in NSCLC patients. In particular, a subgroup analysis of Checkmate 057 has shown greater outcomes wit ICIs in KRAS-mutant patients (36). Kim et al. (37) compared in a meta-analysis ICIs treatment and docetaxel in pretreated NSCLC patients, stratifying results by KRAS status. They have found that, in KRAS-mutant NSCLC patients, overall survival (OS) was improved with ICI versus chemotherapy; on the other hand, this benefit was not reported in KRAS wild-type patients. Moreover, through the ImmunoTarget registry, it has been evaluated the sensitivity of ICIs in NSCLC patients with a variety of driver mutations. Accordingly, KRAS mutations resulted in increased immunotherapy benefit compared to other driver mutations (38). Recently, new data are available regarding (chemo)immunotherapy efficacy according to KRAS status. Sun et al. have demonstrated that among patients with PD-L1>50% treated with ICI, KRAS mutant tumors are associated with a superior overall survival compared with KRAS wild type (mOS 21.1 vs 13.6 months, p=0.03). However, among patients treated with chemoimmunotherapy, there was no significant survival difference between patients with KRAS mutant and KRAS wild type status (mOS, 20.0 vs 19.3 months; p=0.93). Moreover, among patients with KRAS mutant status, OS did not differ between those treated with ICI monotherapy and chemoimmunotherapy (mOS, 21.1 vs 20.0 months; p= 0.78). The authors conclude that among patients with PD-L1 expression of 50% or greater NSCLC treated with ICI monotherapy, KRAS wild type is associated with worse survival compared with KRAS mutant patients; in contrast, survival did not differ appreciably between patients with KRAS mutant and KRAS wild type status who were treated with chemoimmunotherapy. Definitely patients with PD-L1>50% and KRAS-mutant NSCLC had favorable survival (mOS ≥20 months) with either ICI monotherapy or chemoimmunotherapy, while patients with KRAS wild type who were treated with ICI monotherapy had worse survival compared with those treated with chemoimmunotherapy (39). With regard to KRAS mutation subtype, another study has shown that treatment outcomes with chemoimmunotherapy are similar in patients with G12C and non-G12C subtypes (40). All these results could have a potential biological rationale because of the correlation between a higher incidence of KRAS mutations in smokers and, as already known, the elevated rate of tumoral DNA mutations and the higher tumor mutational burden (TMB) in smoker patients compared to never-smokers (41, 42). Another important issue is the potential clinical relevance of KRAS mutations in combination with other genetic alterations. In particular, variations in KRAS mutation subtypes have been associated with distinct biological behaviors that may led to different clinical outcomes (43, 44). For example, tumors with KRAS G12C mutations exhibited higher ERK1/2 phosphorylation than those with KRAS G12D (45); to support this data, a recent study using mouse models with KRAS-mutant tumors, demonstrated higher efficacy of the MEK inhibitor selumetinib in KRAS G12C cancers compared with KRAS G12D tumors. Concerning additional alterations, EGFR/ALK/BRAF co-mutations are rare in KRAS-mutant lung adenocarcinoma; concomitant mutations in onco-suppressor genes, especially TP53, STK11 and KEAP1, are more often found (11, 14, 46, 47). Furthermore, additional alterations may affect the immunogenicity of KRAS-mutant tumors. For example, it has been recently demonstrated that tumors carrying KRAS/TP53 mutations have an enhanced tumor inflammation, increased PD-L1 expression and tumor-infiltrating lymphocytes (TILs) and, consequently, there is a remarkable clinical benefit wit ICIs (48). In contrast, additional STK11 mutations in KRAS-mutant NSCLC could decrease immune surveillance through NF-kB pathway modulation. Indeed, these tumors are associated with low TILs infiltration that further reduces immune surveillance (49, 50). Several trials have shown that STK11/LKB1 mutations in KRAS-mutant NSCLC might be predictive of primary resistance to ICIs or chemotherapy as well (51–53). Taken together, all this evidence suggests that KRAS mutations cannot be evaluated as a single entity, but it needs to be placed in the context of the specific tumor features and seen as part of a complex process in cancer development and progression. The deeper understanding of underlying KRAS biology and function has allowed, after many years of failures, to develop agents able to stop the winding path towards KRAS unstoppable tumors ( Tables 1 , 2 ).

Although KRAS was one of the earliest oncogenic drivers discovered, for decades a therapeutic breakthrough in its targeting has been hampered by several biochemical obstacles (such as the high affinity of GTP for KRAS, the difficulty of finding pockets on KRAS large enough for small-molecule drugs), prompting the increasingly widespread opinion that KRAS was definitely undruggable (54, 55). Therefore, tackling KRAS alterations has been focused on inhibition of its membrane binding (or subcellular localization), identification of lethality partners and on inhibition of its downstream effectors. Concerning direct inhibition, the peculiar crystal structure of the KRAS G12C subtype and the unique reactivity of the cysteine thiol result in the ability to covalently bind the cysteine residue within the region adjacent to the nucleotide-binding pocket, inducing a significant perturbation of protein functional domains and a negative allosteric regulation of RAS signaling in tumor cells (56). This feature can be exploited to create covalent small-molecule inhibitors, binding to KRAS and locking cells in the inactive GDP-bound state (57–59). Thus, various pharmaceutical companies have begun to develop KRAS G12C-specific inhibitors, many of which are being explored in ongoing clinical trials. AMG 510/Sotorasib was the first compound to enter clinical trials with promising results since early trials phases, especially in NSCLC patients. Canon et al. (60) have identified this highly selective molecule inducing tumor regression in preclinical models. In particular, AMG 510 potently inhibits cellular viability in KRAS G12X cell lines and induces tumor regression in xenograft models (60). Furthermore, AMG 510 has shown synergistic growth inhibitory effects when combined with other inhibitors of protein that regulate KRAS activation (such as MEK, AKT, PI3K and members of EGFR family) or with immunotherapy (60). Preliminary results of AMG 510 clinical activity have been presented at the 2019 American Society of Clinical Oncology meeting (61), showing a very promising activity in a cohort of 34 NSCLC patient harboring KRAS G12C mutation achieving a surprising objective response rate (ORR) of 48% and a disease control rate (DCR) of 96%. Regarding its safety profile, the adverse events (AEs) have been manageable with only 9% of grade 3-4 AEs (mostly anaemia and diarrhoea) (62). The final results from the phase I trial have shown a median progression-free survival (PFS) of 6.3 months and a median duration of response of 10.9 months in NSCLC subgroup (59 patients), with a confirmed ORR in 32.2% and DCR in 88.1% of patients (63). Thus, the trial has confirmed the antitumor activity of AMG 510 in pretreated NSCLC patients carrying KRAS G12C mutation, consistent with preliminary results (64). The phase II CodeBreak 100 trial has validated the power of KRAS inhibition with sotorasib in patients with advanced NSCLC. In the phase II trial sotorasib has now shown a durable ORR of 37.1%, a DCR of 80.6%, a median PFS of 6.8 months and a median OS of 12.5 months, in the face of grade 3-4 toxicities rate, mostly diarrhea and hepatic transaminase elevation (11%) (65, 66). These exciting data have led to the fast-track designation of AMG 510 by Food and Drug Administration (FDA). Another cysteine 12-directed covalent inhibitor is MRYX849/adagrasib. Similar to sotorasib, adagrasib is a KRAS G12C inhibitor that has demonstrated a great activity on tumor regression in cell lines and patient-derived xenograft models, as well as in vivo, showing tumor regression in 65% of KRAS G12C mutant preclinical models including different tumor histotypes (67). Indeed, this agent is under investigation in the KRYSTAL trial, a phase I/II multiple expansion cohort trial of MRTX849 in patients with advanced solid tumors carrying KRAS G12C mutation. In this trial, 79 patients were treated with adagrasib 600 mg BID. Overall, the treatment has been well tolerated and most patients has experienced grade 1-2 AEs, such as nausea, diarrhea, vomiting, fatigue and increased ALT. The only commonly reported (>2%) grade 3-4 treatment-related serious AE has been hyponatremia. Among the 51 patients evaluable for clinical activity, the trial has reported an ORR of 45% and DCR of 96% (68). Interestingly, other studies have highlighted that MRTX849 has synergistic effects when combined with EGFR, Src homology 2 domain-containing tyrosinephosphatase-2 (SHP2), mammalian target of rapamycin (mTOR) or cyclin-dependent kinase 4 (CDK4) and CDK6 inhibitors and its efficacy has resulted increased in case of SHP2, MYC or mTOR gene loss (67). A third KRAS G12C covalent inhibitor is JNJ-74699157 (ARS-3248), currently evaluated in a phase I clinical trial enrolling different KRAS G12C mutated tumor types (NCT04006301). A fourth KRAS G12C covalent inhibitor is LY3499446, evaluated in a phase I/II clinical trial (NCT04165031): the study was early terminated due to an unexpected toxicity finding. Concerning pan-RAS inhibitors, many compounds have been tested but the major issue reported is the difficult discrimination between active and inactive KRAS state, with a consequent high off-target activity and a marked toxicity (69). Regarding preclinical evidence, two previous compounds (ARS-853 and ARS-1620) have demonstrated to reduce cell growth and inhibit downstream signaling to MAPK exclusively in tumor cell lines with KRAS G12C mutations (57, 58, 70). Notably, ARS-853 is a potent G12C allele-specific inhibitor that reduces cancer cell proliferation in vitro; its mechanism exploits the need of KRAS G12C nucleotide exchange to switch in the activated status, trapping it in the GDP-RAS state (58, 70). However, ARS-853 has not entered clinical evaluations in KRAS mutant lung cancer yet. Instead, ARS-1620, a covalent compound with high potency and selectivity for KRAS G12C, have shown rapid and sustained tumor regression in vivo (57). The interesting feature of tumor cells harboring KRAS mutations is that they develop, over time, escape mechanisms to break free from the KRAS activation status and undergo adaptative changes to became insusceptible to the drug (71). Instead of relying on the drug-inhibited MEK pathway to proliferate, these adapted cells are capable of using EGFR and aurora kinase downstream signaling cascades to remain in the active form (35). These data may help elucidating why most patients enrolled in clinical trials tend to exhibit partial responses to KRAS inhibitors. For this reason, overcome treatment resistance by developing combination therapies might be a promising approach for the future to create a more durable response to KRAS G12C inhibitors ( Table 3 ).

NSCLC is a heterogeneous disease, in which the only histological classification is largely inadequate to completely define its intrinsic complexity. Molecular markers have dramatically reshaped NSCLC treatment, but KRAS mutations still represent a challenge for physicians and patients. Recently, promising compounds have opened the doors to a new era in which KRAS targeting is possible. The most promising therapeutic approaches are represented by KRAS G12C direct inhibitors, AMG 510 and MRTX849, that have shown encouraging preliminary results in ongoing clinical trials. Actually, the impressive preclinical and clinical activity showed by AMG 510, has led to the fast-track designation by the FDA, representing one of the major breakthrough in the lung cancer research of the last years (142). Other treatment strategies including the inhibition of specific pathways, such as RAF/MEK/ERK, are under investigation as well (NCT03745989; NCT04418167; NCT03415126). The role of ICIs in KRAS mutant NSCLC treatment is also under consideration, alone or in combination with targeted therapy approaches (NCT04613596). Also epigenetic deregulation has been increasingly recognized as one of the major mechanisms of gene expression modulation in lung cancer. Furthermore, some data support the hypothesis that there could be an interplay between KRAS epigenetic regulation and lung cancer pathogenesis (141), suggesting a sophisticated balance between genomic and epigenomic features and lung cancer development. However, although much progress has been made in KRAS-mutant NSCLC treatment, clinical trials have shown that outcomes cannot be given for granted even when the same subtype mutation occurs. This limited efficacy underlies a relevant molecular heterogeneity among KRAS-mutant lung cancer, with such an innate or acquired resistance to targeted therapy, requiring further investigations. Overcoming mechanisms of resistance, implementing gene expression profiling in routine practice, tailoring treatment strategies and extending pharmacological approaches to KRAS mutations other than G12C currently represent the biggest challenge to be addressed for the near future. On the whole, the historical undruggable KRAS alteration is likely to be targeted and a significant turning point has been reached, allowing to offer a range of opportunities in the growing individualized treatment paradigm for KRAS-mutant NSCLC patients ( Figure 1 ).

MF, AS, and EB, conceived the original idea of the article, drafting, and writing the paper. CC, SP, LB, MM, and GP revised the scientific content of specific sections of the manuscript and participated in drafting specific section of the paper. EV, MD, IS, FM, AV, GV, and ED, participated in the critical revision of the paper. MM, GT, and EB participated in the critical revision and editing of the manuscript. GT and EB conceived the original idea and provided critical revision of the manuscript as well as the final approval of the version to publish. All authors contributed to the article and approved the submitted version.

EB is supported by Institutional funds of Università Cattolica del Sacro Cuore (UCSC-project D1-2019-2021). EB is currently supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC) under Investigator Grant (IG) No. IG20583. GT is supported by AIRC under IG No. IG18599. CC is supported by AIRC Grant “Luigi Bonatti and Anna Maria Bonatti Rocca”.

EB received speakers’ and travels’ fee from MSD, Astra-Zeneca, Pfizer, Eli-Lilly, BMS, Novartis and Roche. EB received institutional research grants from Astra-Zeneca, Roche.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.