KRAS G12C: From 'Undruggable' to Druggable with Sotorasib and Adagrasib
核心洞察
KRAS, the most frequently mutated oncogene across cancers, was long considered undruggable due to its smooth surface and high-affinity GTP-binding site.
The G12C mutation's cysteine residue enables covalent inhibition, leading to the approval of sotorasib (2021) and adagrasib (2022) in the US.
In CodeBreaK 200, sotorasib improved progression-free survival versus docetaxel (5.6 vs. 4.5 months, HR 0.66) in previously treated KRAS G12C (搜索) NSCLC.
KRAS, the most frequently mutated oncogene across all cancer types, has transitioned from a target long considered "undruggable" to one now addressable with covalent small-molecule inhibitors. The KRAS protein's smooth surface lacks a drug-binding pocket, and its GTP-binding site has extremely high affinity, making competitive inhibition impossible. However, the G12C mutation introduces a cysteine residue with a free thiol group capable of covalent bonding, opening the door to irreversible inhibitor development. This breakthrough led to the approval of sotorasib in the United States in 2021 and adagrasib in 2022, with Japanese approvals following in 2022 and 2024, respectively.
The Biology of KRAS and the G12C Opportunity
KRAS belongs to the RAS family of small GTPase proteins, alongside NRAS and HRAS. In its normal state, KRAS cycles between a GTP-bound "ON" state and a GDP-bound "OFF" state, regulating the MAPK and PI3K signaling pathways. Oncogenic mutations cause loss of GTP hydrolysis ability, locking KRAS into a constitutively active state that drives sustained proliferative signaling.
KRAS mutation frequency varies by tumor type: pancreatic cancer (搜索) exceeds 90% (mainly G12D), colon cancer ranges from 40–45%, and non-small cell lung cancer (搜索) (NSCLC) adenocarcinoma from 25–30%. Among the specific mutation sites, G12D accounts for approximately 30% of cases overall, G12V approximately 25%, and G12C approximately 13%—with G12C notably concentrated in lung cancer (40%) versus colon cancer (3%).
Sotorasib and the CodeBreaK Program
Sotorasib (Lumakras) is an oral KRAS G12C (搜索) covalent inhibitor that binds to the GDP-bound state, inhibiting conversion to GTP. It is dosed at 960 mg once daily regardless of food intake, with a half-life of approximately five hours. Its indication covers KRAS G12C mutation-positive unresectable advanced or recurrent NSCLC in previously treated patients.
The pivotal CodeBreaK 100 trial (NEJM 2021), a Phase I/II study in previously treated KRAS G12C (搜索) NSCLC, reported a response rate of 37%, median progression-free survival (PFS) of 6.8 months, and overall survival (OS) of 12.5 months, supporting accelerated approval. The confirmatory CodeBreaK 200 trial (NEJM 2023), a Phase III study comparing sotorasib with docetaxel, demonstrated a PFS of 5.6 versus 4.5 months (HR 0.66) and a response rate of 28% versus 13%. No significant difference in OS was observed, attributed to a crossover effect.
Key toxicities include diarrhea (33%, Grade 3+ in 12%), nausea (20–30%), liver dysfunction with elevated AST/ALT (25%, Grade 3+ in 5–10%, with fatal cases reported), fatigue (20%), and rare interstitial lung disease (ILD, <5%). Pharmacist management emphasizes regular liver function monitoring, avoidance of proton pump inhibitor (PPI) co-administration due to reduced absorption, and diarrhea management with loperamide.
Adagrasib and the KRYSTAL Program
Adagrasib (Krazati) is another oral KRAS G12C (搜索) covalent inhibitor, dosed at 600 mg twice daily regardless of food, with a longer half-life of 23 hours. In the KRYSTAL-1 trial (NEJM 2022, Spira et al.) in previously treated KRAS G12C NSCLC, adagrasib achieved a response rate of 43%, a CNS response rate of 33% in brain metastasis cases, PFS of 6.5 months, and OS of 12.6 months. The Phase III KRYSTAL-12 trial (2024) compared adagrasib with docetaxel in second-line KRAS G12C NSCLC, with PFS improvement reported and detailed analysis ongoing.
Adagrasib demonstrates good central nervous system (CNS) penetration, superior to sotorasib, positioning it as a first-line candidate for brain metastasis cases. Its toxicity profile includes diarrhea (63%, Grade 3+ in 6%), nausea (62%), vomiting (45%), fatigue (45%), liver dysfunction (15–25%), and QT prolongation (15–20%), which occurs at a higher frequency than with sotorasib and warrants baseline and periodic ECG monitoring with electrolyte correction.
KRAS G12C in Colorectal Cancer
In colon cancer, KRAS G12C (搜索) accounts for 3–5% of cases (approximately 10% of all KRAS mutations). Unlike NSCLC, KRAS G12C inhibition alone shows limited efficacy in colorectal cancer (搜索), attributed to resistance driven by EGFR signal feedback reactivation. This necessitates combination with anti-EGFR antibodies.
In the KRYSTAL-1 colon cancer cohort, adagrasib monotherapy achieved a 19% response rate, while the combination of adagrasib plus cetuximab reached a 46% response rate with a PFS of 6.9 months. The CodeBreaK 300 trial (NEJM 2023) evaluated sotorasib plus panitumumab versus chemotherapy (regorafenib or trifluridine) in previously treated KRAS G12C (搜索) advanced colon cancer, demonstrating a PFS of 5.6 months versus 2.2 months (HR 0.49), establishing a new standard in this setting.
The Expanding KRAS Pipeline
As of 2026, sotorasib and adagrasib remain the only commercialized KRAS inhibitors, both targeting G12C. Development efforts are advancing across multiple fronts. G12D inhibition—challenging due to the difficulty of covalent binding to the aspartic acid residue—is being pursued with MRTX1133 (Mirati/BMS) in Phase I/II trials, including combination with EGFR antibodies for colorectal cancer (搜索), and HRS-4642. Pan-KRAS inhibition is represented by RMC-6236 (Revolution Medicines), a RAS(ON) inhibitor addressing multiple mutations in Phase I/II testing across pancreatic, colorectal, and NSCLC indications.
Additional strategies target upstream and downstream nodes, including SHP2 inhibitors (TNO155), SOS1 inhibitors (BI 1701963), and MEK plus ERK inhibitor combinations. Resistance mechanisms to G12C inhibitors include KRAS secondary mutations (Y96D, R68S, H95D), NF1 mutation, AKT mutation, MYC amplification, and cluster mutations switching from G12C to G12V or G12D. Given that pancreatic cancer (搜索) carries KRAS mutations in over 90% of cases (mainly G12D, G12V, and G12R), successful G12D inhibition could represent a paradigm shift for a disease that has remained treatment-resistant for many years.