Lamei Huang, Zhixing Guo, Fang Wang, Liwu Fu
8 min
Kirsten rat sarcoma viral oncogene homolog (KRAS) is one of the most frequently mutated oncogenes in human cancer, appearing in highly fatal malignancies such as pancreatic, colorectal, and nonsmall-cell lung cancer (NSCLC). For over four decades, KRAS was considered undruggable because its protein structure is small, smooth, and shallow, lacking clear pockets for small-molecule binding, while the intracellular concentration of its preferred nucleotide, GTP, is extremely high. Consequently, early therapeutic efforts focused on indirect approaches, such as inhibiting downstream effectors or blocking membrane localization, but these strategies largely failed due to poor selectivity and limited efficacy.
The landscape shifted with the discovery of a new allosteric regulatory site in the KRAS(G12C) mutant protein, known as the switch-II pocket (S-IIP). Unlike other KRAS mutations, the G12C substitution provides a unique nucleophilic cysteine residue that allows for the development of covalent inhibitors. These small molecules bind to the switch-II pocket and lock the protein in its inactive, GDP-bound state, preventing the signaling cascades (such as RAF-MEK-ERK and PI3K-AKT-mTOR) that drive uncontrolled tumor growth. This structural discovery led to the clinical approval of sotorasib (AMG510), the first direct KRAS(G12C) inhibitor, marking a paradigm shift in precision oncology.
Despite the clinical success of KRAS(G12C) inhibitors, acquired resistance remains a significant hurdle. Cancer cells frequently adapt to inhibitor-induced suppression by reactivating the MAPK pathway through wild-type RAS proteins or bypass mechanisms, such as amplification of receptor tyrosine kinases. Additionally, secondary mutations in the KRAS gene can disrupt the inhibitor's ability to bind to the switch-II pocket. Current research is heavily focused on combination therapies to preempt this resistance. These strategies involve combining KRAS(G12C) inhibitors with upstream signal blockers (like SHP2 or SOS1 inhibitors), downstream pathway inhibitors, cell cycle regulators (CDK4/6 inhibitors), or immune checkpoint inhibitors, which may leverage the shift toward a more immunoreactive tumor microenvironment following KRAS inhibition.
Cancer is the leading cause of death worldwide, and its treatment and outcomes have been dramatically revolutionised by targeted therapies. As the most frequently mutated oncogene, Kirsten rat sarcoma viral oncogene homologue (KRAS) has attracted substantial attention. The understanding of KRAS is constantly being updated by numerous studies on KRAS in the initiation and progression of cancer diseases. However, KRAS has been deemed a challenging therapeutic target, even “undruggable”, after drug-targeting efforts over the past four decades. Recently, there have been surprising advances in directly targeted drugs for KRAS, especially in KRAS (G12C) inhibitors, such as AMG510 (sotorasib) and MRTX849 (adagrasib), which have obtained encouraging results in clinical trials. Excitingly, AMG510 was the first drug-targeting KRAS (G12C) to be approved for clinical use this year. This review summarises the most recent understanding of fundamental aspects of KRAS, the relationship between the KRAS mutations and tumour immune evasion, and new progress in targeting KRAS, particularly KRAS (G12C). Moreover, the possible mechanisms of resistance to KRAS (G12C) inhibitors and possible combination therapies are summarised, with a view to providing the best regimen for individualised treatment with KRAS (G12C) inhibitors and achieving truly precise treatment.
Alex: Yes. The drugs discussed here form a covalent bond, a strong chemical attachment, near a region called switch two. They trap the mutant in its inactive state. Experiments using an early inhibitor also helped establish that this mutant keeps cycling and remains responsive to signals from upstream.
Sam: That gives the review a concrete mechanism. Which example connects that mechanism to patients?
Alex: Sotorasib does. It binds the mutant cysteine in that pocket and locks KRAS inactive. The review records its approval for adults with non-small-cell lung cancer carrying this mutation, after at least one prior systemic therapy.
Sam: Approval is a milestone, but it doesn’t tell us how often treatment helped. What clinical evidence should we remember?
Alex: The phase two trial evaluated 126 patients with this mutation-defined lung cancer. They received sotorasib. The objective response rate, meaning complete or partial tumour responses, was about 37 percent.
Sam: That’s evidence of activity, but most patients did not have that kind of response. What does the review say about how long control lasted?
Alex: Median progression-free survival, the time before progression or death, was 6.8 months. That describes the treated group, not an improvement over another treatment. The excerpt gives no randomized comparison for these results.
Sam: And among those who did respond, was the response sustained?
Alex: Median response duration was ten months. That is a different measure from progression-free survival because it concerns the responders. Complete responses were rare, so tumour response should not be confused with eradication.
Sam: What about harm? A targeted drug can still carry substantial toxicity.
Alex: The review reports grade three adverse events in roughly one in five patients in that phase two trial. No treatment-related deaths were reported. My reading is that this supports clinical activity with meaningful safety concerns, not a claim that treatment is harmless.
Sam: Before we get to resistance, how far can we generalize beyond that lung cancer population?
Alex: Not automatically to other mutations, or even other cancers carrying this mutation. The review emphasizes biochemical differences between KRAS mutants. They differ in signalling behaviour, treatment sensitivity, and dependence on KRAS itself.
Sam: So the same mutation in a different tissue can also sit inside a different signalling network.
Alex: The review describes stronger receptor signalling in colorectal cancer models than in lung cancer models. After KRAS inhibition, the growth-signalling pathway rebounded more strongly in those colorectal models. That provides a possible explanation for the generally poorer clinical responses reported in colorectal cancer.
Sam: Let’s turn to resistance. Is the main problem that the drug stops binding its target?
Alex: That is one route, but not the only one. Cells can reactivate signalling through other RAS proteins that these selective drugs do not inhibit. They can also engage alternative survival pathways, or produce new mutant KRAS that becomes activated.
Sam: And changes to the target itself can happen alongside those escape routes?
Alex: Yes. The review describes secondary mutations that alter drug sensitivity, including changes affecting the binding pocket. Clinical evidence also shows multiple resistance mechanisms within a single patient. That makes resistance a moving collection of problems, rather than one defect with one obvious replacement drug.
Sam: Then combining treatments sounds reasonable. What separates a plausible combination from something ready for routine use?
Alex: The evidence level. Much of the combination evidence here comes from cells and animal models, with some early clinical testing. One strategy blocks upstream signals that help reactivate KRAS. Another blocks survival pathways that remain available after KRAS inhibition.
Sam: Can you give one example without taking us through a catalogue of experimental drugs?
Alex: SHP2 is a signalling protein involved in reactivation downstream of several growth-factor receptors. Inhibiting it alongside mutant KRAS reduced tumour growth in preclinical models. The attraction is blocking a shared signalling node, rather than guessing which single receptor matters in each tumour.
Sam: But a shared node could also mean broader unwanted effects.
Alex: The authors explicitly flag tolerability as a concern for combinations. They also discuss immune combinations because KRAS affects the tumour microenvironment: the surrounding cells and signals. In mouse experiments, sotorasib altered immune-cell infiltration, and sustained regression depended on an intact immune system.
Sam: That supports a biological rationale, rather than proving that adding immunotherapy improves patient outcomes.
Alex: That is the right boundary. This is a historical review mapping mechanisms and candidate treatments, not a definitive comparison of combination regimens. Its central limitation is the distance between selective inhibition, initial patient response, and durable disease control.
Sam: Who should read the full review, and where should they begin?
Alex: Researchers in targeted cancer therapy should start with “Clinical Exploration” to anchor the mechanism in patient outcomes. Then read the acquired-resistance section, followed by the combination strategy relevant to their work. Treat it as a map of the field at publication, not a current treatment guide.
Sam: And what should everyone else carry away?
Alex: Making a cancer target druggable does not make every tumour dependent on it—or stop tumours finding another route.
Sam: Thanks for listening.