KRYSTAL-10 Failure Exposes CRC Resistance Gap; NSCLC Franchise Remains Adagrasib's Only Viable Anchor

BMS' Krazati Fails Confirmatory Colorectal Cancer Trial

Bristol Myers Squibb's Krazati, in combination with cetuximab, failed to meet its primary efficacy endpoints in the Phase 3 KRYSTAL-10 confirmatory study for metastatic colorectal cancer (CRC) patients with the G12C KRAS mutation whose disease progressed after first-line chemotherapy. Patients treated with the Krazati regimen showed a median progression-free survival (PFS) of 7.5 months compared to 8.1 months for those on chemotherapy. Median overall survival (OS) was also similar, at 21.6 months for the Krazati arm versus 21.7 months for controls. These findings, presented at ESMO GI 2026, indicate the investigational combination did not significantly improve survival outcomes, prompting discussions with regulatory authorities regarding next steps for the drug, which had received accelerated approval for this indication.

• The KRYSTAL-10 study was a Phase 3 confirmatory trial enrolling over 460 patients with metastatic colorectal cancer (CRC) harboring the G12C KRAS mutation. These patients had experienced disease progression following first-line chemotherapy, making them a challenging population. The trial investigated Krazati (adagrasib) in combination with Lilly's cetuximab (Erbitux) against a chemotherapy comparator arm, aiming to verify the clinical benefit previously granted under accelerated approval.
• The investigational combination of Krazati and cetuximab did not meet its primary efficacy endpoints of overall survival (OS) and progression-free survival (PFS). Patients in the Krazati arm achieved a median PFS of 7.5 months, numerically shorter than the 8.1 months seen in the chemotherapy comparator arm. Similarly, median OS was 21.6 months for the Krazati group versus 21.7 months for controls, showing no significant improvement. While numerically higher overall response rates were observed with Krazati plus cetuximab, no statistical analysis was presented.
• Krazati had previously received accelerated FDA approval in December 2022 for non-small cell lung cancer and in June 2024 for KRAS G12C-mutated CRC. Confirmatory trials like KRYSTAL-10 are crucial for verifying clinical benefit under the accelerated pathway. The failure to meet primary endpoints in this CRC trial necessitates discussions with regulatory authorities regarding the future of Krazati's CRC indication. This outcome contrasts with Krazati's success in its NSCLC confirmatory trial, KRYSTAL-12, where it significantly improved PFS.

KRYSTAL-10's Outcomes: Krazati's Confirmatory Trial in mCRC

Recent clinical investigation in metastatic colorectal cancer spans a broad range of therapeutic strategies — from targeted HER2 blockade and adenosine receptor antagonism to VEGF/chemotherapy combinations and immunotherapy in molecularly unselected populations. The studies below represent a cross-section of late-line and biomarker-selected trials with notable efficacy and safety readouts.

• STELLAR-303 (2025): Zanzalintinib (100 mg daily) plus atezolizumab (1200 mg q3w) versus regorafenib (160 mg daily, days 1–21 of a 28-day cycle) in 901 patients with previously treated mCRC without MSI-H/dMMR tumors. In the ITT population, the combination demonstrated a statistically significant OS benefit over regorafenib (median 10.9 vs. 9.4 months; HR 0.80, 95% CI 0.69–0.93; p=0.0045), marking the first phase 3 trial to show significant OS improvement with an immunotherapy-based regimen in relapsed/refractory MSS/MSI-L mCRC. Grade ≥3 treatment-related adverse events were substantially higher in the combination arm (60% vs. 37%), with five treatment-related deaths versus one in the regorafenib arm.

• ARC-9 Study — Cohort B (2026): Etrumadenant (A2a/A2b receptor antagonist) plus zimberelimab (anti-PD-1), FOLFOX, and bevacizumab (EZFB) versus regorafenib in 112 third-line mCRC patients randomized 2:1. EZFB demonstrated markedly improved median PFS (6.2 vs. 2.1 months; HR 0.27, 95% CI 0.17–0.43; p<0.0001), OS (19.7 vs. 9.5 months; HR 0.37, 95% CI 0.22–0.63; p=0.0003), and confirmed ORR (17% vs. 3%). Grade ≥3 TEAEs occurred in 82% of EZFB-treated patients versus 49% in the regorafenib arm, though TEAEs leading to discontinuation were lower in the EZFB arm (5% vs. 17%).

• MOUNTAINEER Trial — Final Analysis (2026): Tucatinib plus trastuzumab (dual HER2-targeted, chemotherapy-free regimen) in 84 patients with chemotherapy-refractory, HER2-positive, RAS wild-type unresectable or metastatic colorectal cancer, with a median follow-up of 32.4 months. Confirmed ORR was 39.3%, median duration of response 15.2 months, median PFS 8.1 months, and median OS 23.9 months. Efficacy was consistent across central HER2-testing methods with no clear association identified between treatment response and co-occurring biomarker alterations. The safety profile was favorable — few patients discontinued due to adverse events and no treatment-emergent deaths occurred.

• TRIPLETE Study — 5-Year Analysis (2026): Modified FOLFOXIRI plus panitumumab versus modified FOLFOX plus panitumumab in 435 patients with RAS/BRAF wild-type mCRC (first-line). After a median follow-up of 60.2 months, the triplet chemotherapy arm demonstrated a statistically significant OS benefit (41.1 vs. 33.3 months; HR 0.79, 95% CI 0.63–0.99; p=0.049), with no significant differences in ORR (75% vs. 78%), early tumor shrinkage, depth of response, R0 resection rate, or PFS (HR 0.95; p=0.606). Median post-progression survival trended in favor of the experimental arm (24.6 vs. 17.7 months; HR 0.79; p=0.062).

• NSABP FC-11 — Arm 2 (2025): Neratinib plus cetuximab in 21 patients with quadruple wild-type (KRAS, NRAS, BRAF, PIK3CA) mCRC with prior oxaliplatin and irinotecan exposure. The ITT ORR/PFS-at-cycle-6 rate was 28% (6/21), comparing favorably to historical single-agent cetuximab or panitumumab monotherapy response rates of 13–17%. No grade 5 or otherwise unexpected adverse events were reported.

• Network Meta-Analysis of Refractory mCRC Treatments (2025 — 28 RCTs): FTD/TPI plus bevacizumab demonstrated statistically superior OS and PFS relative to placebo/BSC (OS HR 0.41, 95% CI 0.28–0.58; PFS HR 0.21, 95% CI 0.14–0.31), FTD/TPI monotherapy (OS HR 0.59; PFS HR 0.46), regorafenib (OS HR 0.60; PFS HR 0.49), fruquintinib (OS HR 0.62), cetuximab (OS HR 0.47; PFS HR 0.32), and panitumumab (OS HR 0.46; PFS HR 0.35) across 16 connected trials, positioning FTD/TPI plus bevacizumab as the most favorable regimen in the refractory setting based on current network evidence.

The Role of KRAS G12C in Metastatic Colorectal Cancer

Metastatic colorectal cancer (mCRC) is driven by the convergence of genetic mutations, dysregulated signaling networks, and epigenetic reprogramming. Among the most commonly altered pathways are the MAPK and PI3K/AKT/mTOR axes, with mutations in APC, KRAS, BRAF, and p53 representing foundational oncogenic events. Constitutive phosphorylation of MAPK correlates specifically with B-Raf mutations, and proliferation of B-Raf-mutant cells is exquisitely sensitive to MAPK pathway inhibition. Concurrently, PI3K pathway mutations influence treatment responsiveness, and combinatorial inhibition of mTOR, AKT, and MEK has demonstrated profound tumor cell destruction in three-dimensional culture models. Aberrant Wnt/β-catenin signaling is present in the majority of CRCs; cells harboring APC mutations exhibit low-level autocrine Wnt activation, while β-catenin-mutant cells are entirely unresponsive to extracellular Wnt inhibition. Additional transcriptional regulators, including FOXK1, further amplify the malignant phenotype — FOXK1 overexpression promotes cellular proliferation and metastatic dissemination by directly binding the GRB2 promoter, thereby activating a pro-tumorigenic signaling axis. Alteration of miRNA profiles has also been correlated with malignant transformation and metastasis, implicating PTEN/PI3K, EGFR, TGF-β, and p53 signaling pathways in disease progression.

At the cellular level, cancer stem cells (CSCs) are recognized as central mediators of metastasis, therapeutic resistance, and disease relapse. Defined by surface markers including CD44, CD133, CD24, LGR5, epithelial cell adhesion molecule, and aldehyde dehydrogenase, colorectal CSCs are highly tumorigenic, chemoresistant, and capable of self-renewal and differentiation into heterogeneous tumor cell populations. Their generation is attributed to dysregulation of signaling pathways — including Wnt, Notch, TGF-β, and Hedgehog — that govern self-renewal and pluripotency in normal intestinal stem cells, altered through cumulative genetic and epigenetic changes. Hyper-activation of Wnt/β-catenin signaling through concurrent APC and KRAS mutation expands the CSC pool, while β-catenin additionally upregulates STT3A/B expression, promoting PD-L1 glycosylation and immune evasion. The homeobox transcription factor PROX1, itself a target of oncogenic Wnt signaling, regulates autophagy-dependent metabolic adaptation in metastatic CSC subpopulations by directly repressing the pro-apoptotic BCL2L15; PROX1 inactivation following metastatic establishment abrogates further lesion growth, and autophagy inhibition effectively targets these PROX1-positive cells.

Epithelial-mesenchymal transition (EMT) and the tumor microenvironment (TME) constitute additional critical drivers of CRC dissemination and distant organ colonization. EMT induction enriches CD24+CD44+ CSC subpopulations, imparting mesenchymal morphology, reduced proliferation, and significantly enhanced colony-forming and tumorigenic capacity. The TME orchestrates disease progression through both cell-autonomous and non-cell-autonomous mechanisms: CRC-associated CSCs secrete exosomes and cytokines that shape an immunosuppressive milieu involving regulatory T cells, myeloid-derived suppressor cells, tumor-associated macrophages, and cancer-associated fibroblasts. Intestinal enteric serotonergic neurons further modulate CSC self-renewal by producing 5-hydroxytryptamine (5-HT), which engages HTR1B/1D/1F receptors on CSCs to activate Wnt/β-catenin signaling — a process potentiated by the CRC-enriched microbiota metabolite isovalerate, which suppresses NuRD complex enrichment on the Tph2 promoter to drive 5-HT biosynthesis. Intratumoral heterogeneity, arising from genetic mutations, microenvironmental interactions, and CSC plasticity, collectively underlies primary and acquired resistance to both chemotherapeutic and biologic agents, rendering metastatic disease particularly refractory to eradication.

Krazati's mCRC Setback: A Test for KRAS G12C Strategy

The recent announcement regarding Krazati (adagrasib) in combination with cetuximab for metastatic colorectal cancer (mCRC) patients with the KRAS G12C mutation marks a pivotal moment for targeted oncology. Despite earlier promising data that led to its accelerated approval, the Phase 3 KRYSTAL-10 confirmatory study failed to demonstrate a significant improvement in progression-free survival or overall survival compared to standard chemotherapy. This outcome is particularly impactful given the historical context of KRAS mutations being considered 'undruggable' and the excitement surrounding the development of direct KRAS G12C inhibitors.

This result underscores the inherent challenges of treating KRAS G12C-mutated mCRC, an indication where these inhibitors have consistently shown lower response rates compared to non-small cell lung cancer. The combination with an anti-EGFR antibody like cetuximab was intended to overcome resistance mechanisms, and while early trials suggested benefit, the confirmatory data did not bear this out. This raises critical questions for regulatory bodies regarding the accelerated approval pathway and the evidence required for full approval, potentially leading to a withdrawal of Krazati's approval in this setting.

For the broader pharmaceutical landscape, this event highlights the urgent need to deepen our understanding of intrinsic and acquired resistance mechanisms in mCRC. Future strategies must focus on:

• Identifying more precise biomarkers to select patients most likely to respond.

• Developing novel combination therapies that target bypass pathways or secondary KRAS mutations.

• Exploring next-generation KRAS inhibitors with improved potency or broader activity.

The path forward for KRAS G12C-mutated mCRC remains complex, demanding innovative research and development to deliver truly transformative outcomes for patients. This setback, while disappointing, serves as a stark reminder of the complexities of cancer biology and the high bar for demonstrating meaningful clinical benefit in confirmatory trials.