In a recent study published in Science Translational Medicine, Huang et al. reveal that KRAS G12C and KRAS G12D mutations are not biologically equivalent, but instead encode distinct oncogenic, immunologic, and therapeutic states, with important implications for precision oncology and combination immunotherapy.1
Oncogenic KRAS has long been considered one of the most challenging targets in cancer therapy due to its high affinity for guanosine 5′-triphosphate (GTP) and the absence of suitable drug-binding pockets, earning its reputation as “undruggable.” However, this paradigm has shifted dramatically with the development of allele-specific inhibitors, most notably those targeting KRAS G12C, which covalently lock the mutant protein in an inactive guanosine 5′-diphosphate (GDP)-bound state and have demonstrated clinical efficacy in pancreatic ductal adenocarcinoma (PDAC), colorectal cancer (CRC), and non-small cell lung cancer (NSCLC). More recently, the therapeutic landscape has expanded to include KRAS G12D-targeted therapies, such as zoldonrasib and setidegrasib, marking a new phase in KRAS-directed precision oncology. As these agents undergo early clinical development, a critical question has emerged: do distinct KRAS mutations represent interchangeable oncogenic drivers, or do they encode fundamentally different tumor states that require mutation-specific therapeutic strategies? Addressing this question is paramount for understanding the biological effects induced by the selective targeting of common oncogenic KRAS mutations and essential for guiding the rational deployment of KRAS-targeted therapies and their integration with other drug modalities. In this context, Huang and colleagues provide compelling evidence suggesting that individual KRAS alleles define unique signaling outputs and therapeutic liabilities in lung adenocarcinoma.1
Using genetically engineered mouse models (GEMMs) that faithfully recapitulate autochthonous lung tumor development, the authors demonstrate that KRAS G12D is a more potent oncogenic driver than KRAS G12C. KRAS G12D tumors exhibit accelerated initiation, enhanced proliferation, and stronger mitogen-activated protein kinase (MAPK) pathway activation, consistent with higher levels of GTP-bound KRAS and downstream signaling. In contrast, KRAS G12C tumors display delayed onset and slower growth kinetics driven by MAPK signaling (Fig. 1). These findings challenge the prevailing paradigm of KRAS as a monolithic oncogene and instead position allele-specific signaling as a determinant of tumor fitness and evolutionary trajectory. Such heterogeneity aligns with recent conceptual advances emphasizing context-dependent oncogene addiction and signaling plasticity in cancer progression.2,3
Fig. 1The alternative text for this image may have been generated using AI.
Allele-specific KRAS signaling shapes tumor immunogenicity and therapeutic response in lung cancer. Left panel (oncogenic strength): KRAS G12D functions as a more potent oncogenic driver than KRAS G12C, promoting earlier tumor initiation, accelerated tumor growth, and enhanced downstream MAPK pathway signaling. In contrast, KRAS G12C-driven tumors exhibit delayed onset and comparatively lower proliferative signaling, consistent with reduced oncogenic intensity. Middle panel (immune microenvironment): KRAS G12C tumors display a more inflamed (“hot”) microenvironment characterized by increased CD8 + T cell infiltration, elevated IFN-γ signaling, and higher expression of immune checkpoint molecules such as PD-L1, reflecting greater immune engagement. In contrast, KRAS G12D tumors exhibit a less immunogenic (“cold”) microenvironment with reduced CD8 + T cell infiltration, weaker antigen presentation, and a more immune-suppressive milieu. Right panel (therapeutic response and immune modulation): Inhibition of KRAS G12C (e.g., adagrasib) induces tumor regression with relatively durable responses but eventual resistance. Inhibition of KRAS G12D (e.g., MRTX1133) also drives tumor regression but is associated with earlier relapse, consistent with stronger oncogenic drive. Notably, KRAS G12D inhibition enhances tumor antigen presentation, promotes CD8+ T cell infiltration, induces IFN-γ and granzyme B expression, and upregulates PD-L1 and MHC I expression, suggesting an immunomodulatory effect that may sensitize tumors to immune checkpoint blockade. GzmB granzyme B, ICB immune checkpoint blockade, IFN-γ interferon-γ, MAPK mitogen-activated protein kinase, MHC I major histocompatibility complex class I, PD-L1 programmed death-ligand 1, PFS progression free survival, OS overall survival. Created with BioRender.com
Beyond intrinsic oncogenic strength, Huang et al. uncover striking differences in the tumor immune microenvironment. KRAS G12C-driven tumors are characterized by increased infiltration of CD8+ T cells, elevated expression of interferon-γ (IFN-γ)-related genes, enhanced MCH I expression, and higher programmed death-ligand 1 (PD-L1) levels, reflecting an inflamed (“hot”) phenotype. Conversely, KRAS G12D tumors exhibit reduced immune cell engagement and a comparatively immunologically “cold” microenvironment (Fig. 1). These observations are corroborated in human lung adenocarcinoma datasets, underscoring their clinical relevance. Notably, these findings extend recent insights into the bidirectional crosstalk between oncogenic signaling and immune surveillance, whereby oncogene-specific signaling outputs actively shape tumor immunogenicity rather than merely reflecting it.4
The therapeutic implications of these differences are profound. Both KRAS G12C and KRAS G12D inhibitors (adagrasib and MRTX1133, respectively) induce robust tumor regression in vivo. However, KRAS G12D-driven tumors relapse more rapidly, exhibiting shorter progression-free survival (PFS) despite comparable initial responses. Mechanistically, resistance in both contexts is associated with MAPK pathway reactivation, yet KRAS G12D tumors retain higher proliferative capacity at relapse, implying a deeper oncogenic “set point” that facilitates rapid tumor regrowth (Fig. 1). These findings highlight that therapeutic durability is not solely determined by drug potency but is intrinsically linked to allele-specific tumor biology.
Importantly, Huang et al. observed that KRAS G12D inhibition exerts pronounced immunomodulatory effects in lung adenocarcinoma. Short-term treatment with MRTX1133 enhances antigen presentation through upregulation of major histocompatibility complex class I (MHC I), increases CD8+ T cell infiltration, and promotes effector cytokine production, including IFNγ. Functionally, these changes translate into enhanced antigen-specific cytotoxicity and improved responses to immune checkpoint blockade (ICB) in immunogenic models. In contrast, KRAS G12C inhibition yields comparatively modest immune remodeling, likely reflecting the already inflamed baseline state of these tumors (Fig. 1). Together, the above data position KRAS G12D inhibition as a dual-function therapeutic strategy that simultaneously suppresses oncogenic signaling and reprograms the tumor immune microenvironment.
These observations resonate with a broader paradigm shift in oncology, wherein targeted therapies are increasingly recognized as immunomodulatory agents capable of reshaping tumor–immune interactions.5 In this context, KRAS G12D inhibitors may be particularly well-suited for rational combination with immune checkpoint inhibitors (ICIs), especially in tumors with low baseline immunogenicity. However, the durability of such responses appears contingent on tumor context, as immune memory was not universally established in non-immunogenic settings, highlighting the need for additional strategies to sustain antitumor immunity.
Despite the conceptual advances of this work, several questions remain. First, the molecular mechanisms underlying the differential oncogenic potency and immunogenicity of KRAS alleles remain incompletely defined. Whether these differences are driven by the distinct biochemical properties of each KRAS allele, downstream signaling networks (e.g., differential engagement of rapidly accelerated fibrosarcoma (RAF)–mitogen-activated protein kinase kinase (MEK)–extracellular signal-regulated kinase (ERK) versus phosphoinositide 3-kinase (PI3K) pathways) or by non-cell-autonomous effects on stromal and myeloid compartments warrants further investigation. Second, the mechanisms of acquired resistance to KRAS G12D inhibition, particularly in the context of immune engagement, remain to be fully elucidated. Third, translating these findings into the clinic will require careful patient stratification, as tumor immunogenicity and co-mutation status are likely to modulate therapeutic outcomes.
In summary, Huang et al. provide a compelling framework for understanding how allele-specific KRAS signaling governs tumor biology, immune contexture, and therapeutic response in lung adenocarcinoma. By demonstrating that KRAS mutations encode distinct oncogenic and immunological states, this study challenges the notion of KRAS as a uniform therapeutic target and underscores the need for mutation-specific treatment strategies. As KRAS-targeted therapies continue to expand beyond KRAS G12C, including KRAS G12D small-molecule inhibitors and degraders, as well as pan-KRAS inhibitors and degraders, these findings offer critical insight for the rational design of combination regimens and the development of precision oncology approaches tailored to patients with KRAS-mutant lung cancers.
