Five Gastrointestinal Cancer Abstracts You May Have Missed at ESMO GI 2026

The 2026 European Society for Medical Oncology World Congress on Gastrointestinal (GI) Cancers delivered a mix of practice-informing and practice-questioning data across the GI cancer landscape, from a phase 3 trial evaluating a KRAS G12C inhibitor combination that failed to beat chemotherapy in colorectal cancer to early phase 1/2 signals for a new RAS(ON) inhibitor in pancreatic cancer. Investigators also reported mixed results for chemoembolization-based combinations in hepatocellular carcinoma (HCC) and landmark long-term survival data for a checkpoint inhibitor doublet in refractory microsatellite-stable (MSS) colorectal cancer.

We rounded up 5 abstracts of interest from the meeting:

How did adagrasib plus cetuximab perform in the KRYSTAL-10 trial?

The chemotherapy-free combination of adagrasib (Krazati), a KRAS G12C inhibitor, and cetuximab (Erbitux), an anti-EGFR monoclonal antibody, did not outperform chemotherapy in patients with previously treated KRAS G12C-mutated metastatic colorectal cancer (CRC), according to final data from the phase 3 KRYSTAL-10 trial (NCT04793958) presented at the meeting.¹

KRYSTAL-10 randomly assigned 461 patients with metastatic CRC who had progressed on first-line fluoropyrimidine-based doublet chemotherapy 1:1 to adagrasib plus cetuximab (n = 231) or chemotherapy with or without a VEGF/VEGF receptor inhibitor (n = 230). Although the adagrasib combination produced a numerically higher objective response rate (ORR) than chemotherapy (47% vs 16%), the trial did not meet its dual primary end points. The median progression-free survival (PFS) by blinded independent central review was 7.5 months (95% CI, 6.3-9.2) with adagrasib plus cetuximab vs 8.1 months (95% CI, 7.3-9.2) with chemotherapy (HR, 0.89; 95% CI, 0.71-1.13; P = .3241), and the median overall survival (OS) was 21.6 months (95% CI, 18.4-25.5) vs 21.7 months (95% CI, 18.0-24.8), respectively (HR, 0.83; 95% CI, 0.67-1.03; P = .0938).

Regarding toxicity, no new safety signals were reported with the combination.

Per Pfeiffer, MD, of the University of Southern Denmark and Odense University Hospital in Denmark, stated, “It seems that the experimental arm performed as expected based on the phase 1/2 results, but the chemotherapy arm performed much better than would be predicted based on prior randomized trials. The median survival of 21.7 months with chemotherapy with or without VEGF(R) inhibitor is very difficult to understand, especially as patients with the KRAS G12C mutation tend to have a worse prognosis than those with other KRAS-mutated tumors and wild-type tumors. These patients would be expected to have a median OS of 12 to 14 months.”¹ He also pointed to a higher rate of subsequent KRAS G12C inhibitor use in the chemotherapy arm (30% vs 4%) as a possible confounder for the OS findings, though not the PFS findings.

CancerNetwork® previously covered the 2024 accelerated approval of adagrasib plus cetuximab in this population, which was based on response and duration-of-response data from the CRC cohort of the phase 1/2 KRYSTAL-1 trial (NCT03785249) rather than a randomized comparison.² KRYSTAL-10 was designed as the confirmatory trial for that accelerated approval, and these final results raise questions about whether the combination's regulatory status could be affected going forward.

What activity did zoldonrasib plus chemotherapy show in first-line RAS G12D mPDAC?

An investigational RAS(ON) G12D-selective inhibitor, zoldonrasib, combined with standard first-line chemotherapy produced high response rates and reduced circulating tumor DNA (ctDNA) in patients with previously untreated RAS G12D-mutated metastatic pancreatic ductal adenocarcinoma (PDAC), according to phase 1/2 data from the RMC-GI-102 trial (NCT06445062).^3,4

The multicenter study enrolled 81 patients with previously untreated metastatic PDAC harboring a RAS G12D mutation; 41 patients received zoldonrasib plus modified FOLFIRINOX (mFOLFIRINOX; oxaliplatin, irinotecan, leucovorin, and fluorouracil) and 40 received zoldonrasib plus gemcitabine and nab-paclitaxel (GnP). Among evaluable patients, the ORR was 82% (95% CI, 60%-95%) with the mFOLFIRINOX regimen and 61% (95% CI, 42%-78%) with GnP, and the disease control rate (DCR) was 96% (95% CI, 77%-100%) and 90% (95% CI, 74%-98%), respectively. All evaluable patients experienced at least a 50% reduction in RAS G12D-mutant ctDNA, with complete clearance achieved in 47% of patients receiving mFOLFIRINOX and 71% of those receiving GnP.

Grade 3 or higher treatment-related adverse events (TRAEs) occurred in 61% of patients receiving zoldonrasib plus mFOLFIRINOX and 80% of those receiving zoldonrasib plus GnP, with no treatment-related deaths reported. Teresa Macarulla, MD, head of the medical oncology department at Hospital Clínic Barcelona in Spain and congress co-chair, stated, “These findings are encouraging because they suggest that targeting KRAS G12D can be combined with standard chemotherapy without introducing unexpected safety concerns.”⁴ These data support the ongoing phase 3 RASolute 305 trial (NCT07621718), which is comparing zoldonrasib plus chemotherapy with placebo plus chemotherapy in this population.

How effective were zoldonrasib plus daraxonrasib in previously treated RAS G12D pancreatic cancer?

The phase 1 RMC-9805-001 trial (NCT06040541) evaluated zoldonrasib plus daraxonrasib, a RAS(ON) multi-selective inhibitor, in 60 patients with RAS G12D-mutated metastatic PDAC who had received 1 or more prior lines of therapy.³ This analysis was conducted with a data cutoff date of February 9, 2026.

In the second-line cohort (n = 30), the ORR was 50% (95% CI, 31%-69%) and the DCR was 97% (95% CI, 83%-100%); the median PFS was 9.6 months (95% CI, 7.1-not evaluable [NE]), with a 6-month PFS rate of 71%. The median OS was not yet estimable, with a 6-month OS rate of 89%. In the third-line-or-later cohort (n = 30), ORR was 47% (95% CI, 28%-66%) and disease control rate was 90% (95% CI, 74%-98%). The median PFS was 7.6 months (95% CI, 4.6-10.5), with a 6-month PFS rate of 59%, and median OS was 10.5 months (95% CI, 6.7-NE), with a 6-month OS rate of 82%.

Grade 3 or higher TRAEs occurred in 35% of patients, with rash (12%), anemia (10%), and stomatitis/mucositis (7%) among the most common events. TRAEs led to discontinuation of zoldonrasib in 2% of patients and daraxonrasib in 5%. The mean dose intensity was 88% and 76% for zoldonrasib and daraxonrasib, respectively. These data support the planned phase 3 RASolute 309 trial, which will evaluate zoldonrasib plus daraxonrasib vs GnP in patients with previously untreated RAS G12D-mutated metastatic PDAC.

CancerNetwork previously covered daraxonrasib monotherapy in the previously treated metastatic PDAC setting, where the phase 3 RASolute 302 trial (NCT06625320) showed a median OS of 13.2 months vs 6.7 months with chemotherapy (HR, 0.40; P <.0001) across patients with previously treated metastatic PDAC.⁵

What do mixed results for TACE combinations signify in hepatocellular carcinoma?

Findings from the phase 3 EMERALD-1 (NCT03778957) and EMERALD-3 (NCT05301842) trials highlighted uncertainty over the role of immunotherapy combinations with transarterial chemoembolization (TACE) in patients with unresectable, embolization-eligible HCC.⁶

In the final analysis of EMERALD-1, durvalumab (Imfinzi) with or without bevacizumab (Avastin) plus TACE did not improve OS compared with TACE alone among 616 patients with HCC. The median OS was 29.9 months (95% CI, 23.5-34.8) with durvalumab plus bevacizumab plus TACE vs 33.3 months (95% CI, 27.7-40.0) with TACE alone (HR, 1.10; 95% CI, 0.87-1.39; P = .470), and 33.6 months (95% CI, 24.9-40.0) with durvalumab plus TACE vs 33.3 months with TACE alone (HR, 0.93; 95% CI, 0.74-1.19; P = .666). These findings were consistent across subgroups.

Jens Ricke, MD, of University Hospital, LMU Munich in Germany, stated, “The OS data from EMERALD-1 are disappointing since the previously reported benefit in PFS with the TACE combination…did not translate into prolonged OS.”⁶ These findings raise questions about the regimen's role in this setting given the accompanying toxicity; grade 3 or 4 adverse events occurred in 47% of patients receiving durvalumab plus bevacizumab plus TACE compared with 31% receiving durvalumab plus TACE and 25% receiving TACE alone. Serious adverse events occurred in 52% of the durvalumab plus bevacizumab plus TACE arm vs 38% in the durvalumab plus TACE arm and 34% in the TACE alone arm.

More encouraging results came from EMERALD-3, which evaluated the STRIDE regimen (tremelimumab [Imjudo] plus durvalumab) with or without lenvatinib plus TACE in 760 patients with embolization-eligible HCC. At the second data cutoff of February 2026, the PFS benefit with STRIDE plus TACE vs TACE alone continued to be observed (HR, 0.70; 95% CI, 0.57-0.86), and ORR improved with STRIDE plus lenvatinib plus TACE (38.9%) and STRIDE plus TACE (40.8%) compared with TACE alone (27.0%).

A trend toward improved OS was also reported for STRIDE plus lenvatinib plus TACE (HR, 0.84; 95% CI, 0.65-1.09) and STRIDE plus TACE (HR, 0.70; 95% CI, 0.51-0.95) vs TACE alone. Ricke added, “This combination may be worth the price patients are willing to pay in treatment-related toxicity, although we need to see mature data. An additional benefit for the patient of omitting lenvatinib at this stage is that it will be available for subsequent lines of treatment.”⁶

CancerNetwork previously covered the initial EMERALD-3 readout, in which the STRIDE regimen plus lenvatinib and TACE met its primary end point of PFS compared with TACE alone, with a trend toward improved OS reported at that earlier interim analysis.⁷

What was the 3-year survival of botensilimab plus balstilimab in refractory MSS metastatic CRC?

Extended follow-up from a phase 1b study (NCT03860272) evaluating botensilimab, an Fc-enhanced anti-CTLA-4 antibody, plus balstilimab, an anti-PD-1 antibody, showed a 3-year OS rate of 33% in patients with refractory microsatellite-stable (MSS) metastatic CRC without active liver metastases, a population in which durable long-term survival is rarely reported.⁸

The fully enrolled C-800-01 cohort included 123 patients who had received a median of 3 prior lines of therapy; 67% had received at least 3 prior lines and 15% had received prior anti–PD-(L)1 with or without anti–CLTA-4 therapy.

With extended follow-up, the median OS was 21.2 months, and 24-month and 36-month OS rates were 41% and 33%, respectively, with the Kaplan-Meier curve showing a plateau beyond 2 years. The confirmed ORR was 21%, including 3 complete responses and 23 partial responses, and the median duration of response was not reached (range, 1.6-37.4) Notably, 17% of patients (n = 21/123) were alive and off all systemic anticancer therapy at last follow-up, including 13 responders. No new safety signals were observed and no treatment-related deaths occurred.

Benjamin L. Schlechter, MD, of Dana-Farber Cancer Institute and presenting author of the study, said, “Seeing a subset of patients remain alive and off systemic anticancer therapy after treatment speaks to the clinical relevance of these results and the potential for botensilimab plus balstilimab to change expectations for what immunotherapy may achieve in this setting.”⁸

No new safety signals or treatment-related deaths were observed with extended follow-up, and 98% of cases of immune-mediated diarrhea/colitis resolved. These data support the ongoing randomized phase 3 BATTMAN trial (NCT07152821) evaluating botensilimab plus balstilimab in refractory MSS/proficient mismatch repair metastatic CRC.

CancerNetwork previously covered earlier findings from this same C-800-01 cohort, when a 2-year OS rate of 42% was reported alongside an FDA clinical outcomes meeting that supported omitting a botensilimab monotherapy arm from the phase 3 BATTMAN trial design.⁹ The updated 3-year data reflect a longer follow-up on that same patient population.

References

1. Adagrasib plus cetuximab fails to outperform chemotherapy in KRAS G12C-mutated metastatic colorectal cancer. News release. ESMO Daily Reporter. July 3, 2026. Accessed July 7, 2026. https://tinyurl.com/5n75p754
2. FDA grants accelerated approval to adagrasib with cetuximab for KRAS G12C-mutated colorectal cancer. News release. FDA. June 21, 2024. Accessed July 7, 2026. https://tinyurl.com/2app6ahp
3. Revolution Medicines presents phase 1/2 clinical data for zoldonrasib combination regimens in patients with RAS G12D metastatic pancreatic cancer at ESMO Gastrointestinal Cancers Congress 2026. News release. Revolution Medicines, Inc. July 2, 2026. Accessed July 7, 2026. https://tinyurl.com/yyvf9sjf
4. New targeted therapy shows promising early results in advanced pancreatic cancer. News release. ESMO. July 2, 2026. Accessed July 7, 2026. https://tinyurl.com/324ubk2x
5. Daraxonrasib demonstrates unprecedented overall survival benefit in pivotal phase 3 RASolute 302 clinical trial in patients with metastatic pancreatic cancer. News release. Revolution Medicines, Inc. April 13, 2026. Accessed July 7, 2026. https://tinyurl.com/44t5vh5d
6. The place for TACE combinations in hepatocellular carcinoma is still uncertain. News release. ESMO Daily Reporter. July 2, 2026. Accessed July 7, 2026. https://tinyurl.com/2sb6ca5r
7. Imfinzi plus imjudo combined with lenvatinib and TACE demonstrated a statistically significant and clinically meaningful improvement in progression-free survival in embolisation-eligible unresectable liver cancer in EMERALD-3 phase III trial. News release. AstraZeneca. April 2, 2026. Accessed July 7, 2026. https://tinyurl.com/yesrt38e
8. Agenus reports landmark BOT+BAL data showing 33% three-year overall survival in refractory MSS metastatic colorectal cancer without active liver metastases at ESMO GI 2026. News release. Agenus Inc. July 6, 2026. Accessed July 7, 2026. https://tinyurl.com/9949ebna
9. Agenus' BOT/BAL achieves 42% two-year survival in refractory MSS CRC, advances toward registration with FDA alignment on phase 3. News release. Agenus Inc. July 7, 2025. Accessed July 7, 2026. https://tinyurl.com/yxkc3atd