🇪🇸 Leer este artículo en español

Title: KRAS G12C responds better to immunotherapy: the KRAS subtype matters in NSCLC (real-world) Type: B (clinical / real-world) Tags: KRAS, NSCLC, immunotherapy, biomarkers, real-world PMID: 42607379 DOI: 10.1016/j.lungcan.2026.109560 Journal: Lung Cancer (2026) Category: Oncology

---

⚖️ Transparency notice: this article was written with AI assistance and reviewed by the author, a medical oncologist.

198 | Patients with advanced KRAS-mutated NSCLC 49.5% | Carriers of the G12C variant 15 vs 9 months | Median OS with ICI (G12C vs non-G12C) HR 0.71 | p = 0.031 · in favor of G12C

The finding

In the tumor board, when a patient with advanced NSCLC and a KRAS mutation comes, the classic question is: do we give immunotherapy or not? We know that KRAS is the most frequent oncogene in NSCLC and that response to checkpoint inhibitors is not uniform. But almost always we treat “KRAS” as a single block. This multicenter real-world study says that the specific allele matters: patients with G12C mutation had significantly longer overall survival with immunotherapy than the rest of KRAS mutations [PMID 42607379].

198 patients with advanced KRAS-mutated NSCLC treated with checkpoint inhibitors (ICI) were analyzed, of whom 49.5% carried the G12C variant. 81% received immunotherapy in first line, alone or combined with chemotherapy [PMID 42607379].

The numbers

Variable | G12C | Non-G12C Patients | 98 (49.5%) | 100 (50.5%) Median OS with ICI | 15 months | 9 months Hazard ratio (OS) | HR 0.71 (p = 0.031) | — Median follow-up | 48 months | 48 months

In multivariate analysis, the G12C mutation was independently associated with better overall survival, along with good performance status (PS) and absence of CNS metastases. 81% of the cohort received ICI in first line [PMID 42607379].

1. Retrospective design, always with caution. Real-world data are valuable but carry selection biases: G12C and non-G12C patients may differ in unmeasured characteristics (tumor burden, comorbidity, subsequent lines).

2. Confounding subsequent lines. Today there are KRAS G12C inhibitors (sotorasib, adagrasib) approved in advanced NSCLC. If some G12C patients received them after immunotherapy, part of the OS advantage could be due to the sequence and not the ICI effect. The abstract does not detail this point — the full text would need to be checked.

3. Magnitude and robustness of the effect. An HR of 0.71 with p = 0.031 is a moderate effect and a p-value not very robust, without correction for multiple comparisons. It is a hypothesis-generating finding, as the authors themselves acknowledge.

4. Co-mutations, the missing piece. STK11 and KEAP1 strongly condition response to ICI in KRAS-mutated. The study collected them, but distribution by allele is not detailed in the abstract — without that data, we don’t know if the G12C effect is due to the allele or the co-mutations that accompany it.

5. PFS and response rates? The abstract reports OS; without progression or response data, part of the mechanism remains unclear.

Does not change practice today. The decision to treat with ICI remains guided by PD-L1, comorbidities, and co-mutations (STK11/KEAP1), not by the KRAS allele. But this study reinforces an idea that is gaining ground: KRAS-mutated NSCLC is not a homogeneous group, and the specific allele should be included in routine molecular reports — not only because G12C has targeted therapy, but because it also conditions response to immunotherapy.

Reference: Masfarré L, et al. Lung Cancer. 2026;220:109560. doi:10.1016/j.lungcan.2026.109560. PMID: 42607379.

— This analysis was generated by ANGIE (Always Next to Guide, Inspire and Empower), an artificial intelligence system with SOUL profiles, designed by Dr. Javier Pumares Pérez.

Connect on LinkedIn · ppj.es