The Keytruda Problem: What AstraZeneca's Volrustomig Failure Reveals About the Bispecific Antibody Race in Lung Cancer

On August 17, 2026, AstraZeneca announced it was discontinuing the Phase 3 eVOLVE-Lung02 trial of volrustomig, its experimental PD-1/CTLA-4 bispecific antibody, in metastatic non-small cell lung cancer. The Independent Data Monitoring Committee had reviewed interim data and concluded that the drug, combined with chemotherapy, was unlikely to beat pembrolizumab plus chemotherapy on either of the trial's dual primary endpoints: progression-free survival or overall survival. The trial enrolled 895 patients across 25 countries. AstraZeneca did not release specific data.

The headline is a clinical failure. The more interesting story is what it reveals about the structural difficulty of displacing Keytruda in lung cancer, and about the diverging fortunes of different bispecific antibody strategies in the same disease.

The Logic Behind Volrustomig

Volrustomig was designed around a compelling mechanistic premise. PD-1 inhibitors like pembrolizumab have transformed lung cancer treatment, but their benefit is uneven. Patients whose tumors express low levels of PD-L1, the protein that PD-1 binds to, tend to respond less durably to single-agent checkpoint blockade. The hypothesis behind volrustomig was that adding CTLA-4 inhibition to PD-1 blockade on the same molecule, rather than as a separate drug, could provide deeper immune activation in precisely this harder-to-treat population without the toxicity that comes from combining two separate checkpoint inhibitors at full doses.

The eVOLVE-Lung02 trial targeted patients with PD-L1 expression below 50%, with the primary analysis focused on the PD-L1 negative subgroup, those with less than 1% expression. This is the population where pembrolizumab alone performs least well, and where the theoretical advantage of dual checkpoint blockade should be most apparent. The trial was designed to find the signal where it should be strongest. It did not find it.

What the Failure Means for the PD-1/CTLA-4 Approach

The combination of PD-1 and CTLA-4 blockade is not a new idea. Nivolumab plus ipilimumab, two separate antibodies targeting those same pathways, is an approved regimen in lung cancer. The question volrustomig was trying to answer was whether a bispecific format, which preferentially delivers CTLA-4 blockade to PD-1-positive activated T cells rather than to resting peripheral T cells, could achieve comparable or superior efficacy with a more favorable safety profile than the combination of two separate drugs.

The eVOLVE-Lung02 result does not answer that question cleanly, because the comparator was pembrolizumab plus chemotherapy, not nivolumab plus ipilimumab. What it does establish is that volrustomig plus chemotherapy could not outperform the current standard of care in the population where it was most expected to. That is a meaningful negative signal, even without granular data on the magnitude of the shortfall.

AstraZeneca's other Phase 3 volrustomig trials, in cervical cancer, head and neck squamous cell carcinoma, and mesothelioma, continue as planned. The company has been careful to frame the lung cancer discontinuation as disease-specific rather than platform-specific. That framing is defensible, but it will face scrutiny as the remaining trials read out.

The Contrast With Ivonescimab

The volrustomig failure arrives at a moment when a different bispecific strategy in lung cancer is generating considerable excitement. Ivonescimab, developed by Akeso and licensed to Summit Therapeutics, targets PD-1 and VEGF rather than PD-1 and CTLA-4. At ASCO 2026, ivonescimab became the first bispecific antibody to demonstrate an overall survival benefit in a Phase 3 trial in any solid tumor, doing so in squamous non-small cell lung cancer. Summit has a PDUFA date in November 2026 for an EGFR-mutated NSCLC indication, and the drug's broader lung cancer program is advancing on multiple fronts.

The divergence between these two bispecific strategies is instructive. Both are designed to improve on single-agent PD-1 blockade. Both target lung cancer. But the mechanisms they pair with PD-1 inhibition are fundamentally different, and the clinical results so far suggest that the PD-1/VEGF combination may be finding a more durable path to differentiation than PD-1/CTLA-4 in this disease. VEGF inhibition addresses the tumor microenvironment in ways that complement immune checkpoint blockade, potentially normalizing the vasculature that tumors use to exclude immune cells. CTLA-4 inhibition works earlier in the immune activation cascade, and its benefit in lung cancer has historically been most pronounced in combination with PD-1 blockade at doses that carry significant toxicity.

AstraZeneca's Broader Clinical Moment

The volrustomig discontinuation is the latest in a series of late-stage setbacks for AstraZeneca. In July, the company disclosed that Wainua, its antisense therapy developed with Ionis Pharmaceuticals, failed to show significant cardiovascular benefit in a Phase 3 study of transthyretin amyloidosis-cardiomyopathy. That same month, Ultomiris failed to achieve statistically significant event-free survival in a Phase 3 trial in transplant-associated thrombotic microangiopathy. An antibody-drug conjugate, sonesitatug vedotin, also missed its primary endpoint in gastric cancer.

AstraZeneca's CEO Pascal Soriot has maintained that the company remains on track for its $80 billion revenue target by 2030, pointing to the continued strength of Tagrisso, Enhertu, and Imfinzi. That confidence is not unfounded. The company's approved oncology portfolio is genuinely strong, and it reported separate Phase 3 wins for Tagrisso and Enhertu in lung cancer on the same day the volrustomig discontinuation was announced. But the accumulation of late-stage failures raises legitimate questions about the productivity of its experimental pipeline, and about whether the company's ambition to replace established checkpoint inhibitors with next-generation bispecifics is running ahead of the biology.

What Keytruda's Dominance Actually Means

The deeper lesson from eVOLVE-Lung02 may be about the difficulty of the benchmark itself. Pembrolizumab is not a weak comparator. It is one of the most successful drugs in the history of oncology, with more than 40 approved indications and annual revenues exceeding $25 billion. In lung cancer specifically, it has been refined through years of biomarker-driven patient selection, combination strategies, and dosing optimization. Beating it in a head-to-head trial, in a population where it already performs reasonably well, requires a drug that offers something genuinely different, not just a mechanistic variation on the same theme.

The bispecific antibody field is still early, and the volrustomig result in lung cancer does not foreclose the approach in other tumor types or other disease contexts. But it does reinforce a pattern that has emerged across immuno-oncology: the drugs that succeed against Keytruda tend to be those that address a distinct biological vulnerability, not those that attempt to do what Keytruda does, only slightly better. The eVOLVE-Lung02 discontinuation is a reminder that mechanistic elegance and clinical differentiation are not the same thing, and that the distance between a compelling hypothesis and a Phase 3 win in lung cancer remains one of the most demanding gaps in medicine to close.