Gilead and Merck Scrap Trodelvy-Keytruda Phase 3 in NSCLC After OS Miss

Gilead Sciences (GILD) and Merck (MRK) terminated the Phase 3 KEYNOTE-D46/EVOKE-03 trial of Trodelvy plus Keytruda in first-line PD-L1-positive non-small cell lung cancer on August 6. An external data monitoring committee determined the combination would not significantly improve overall survival versus Keytruda alone. Progression-free survival showed a numerical benefit but did not clear statistical significance. The trial enrolled patients with PD-L1 tumor proportion score of 50% or higher, the population where Keytruda monotherapy already delivers median overall survival exceeding two years.

Gilead closed at $130.86 on August 6, with a market cap of approximately $163 billion. Merck closed at $128.37. Trodelvy generated $457 million in Q2 2026, up 26% year over year, beating the consensus estimate of $448 million. The drug accounted for roughly 6% of Gilead’s $7.8 billion in quarterly product sales. Indication expansion beyond breast cancer was central to the oncology growth narrative.

The NSCLC bar is high. Keytruda monotherapy already delivers durable responses in PD-L1-high patients, and adding a Trop-2 ADC did not clear the survival threshold. The result also clears a lane for Merck’s own sac-TMT (sacituzumab tirumotecan), another Trop-2 ADC from its Kelun partnership. A China-only Phase 3 in PD-L1-negative NSCLC already hit its PFS endpoint, and Merck is evaluating a global registrational strategy. This failure removes a named competitor, making sac-TMT the frontrunner for Merck’s next-generation checkpoint inhibitor backbone in lung cancer.

Trodelvy’s remaining Keytruda combination opportunity is ASCENT-05, an adjuvant triple-negative breast cancer trial evaluating Trodelvy plus Keytruda versus standard of care in patients with residual disease after neoadjuvant therapy. That readout, expected in 2027, is now the event that determines whether the Trodelvy-Keytruda combination has a future beyond the first-line metastatic TNBC indication already approved by the FDA in June 2026 and recommended by the CHMP in July.

The risk: Gilead faces a narrowing oncology path. Trodelvy monotherapy expansion into ES-SCLC and endometrial cancer remains on track, but the NSCLC miss removes a large revenue opportunity. Gilead acquired Trodelvy through its $21 billion Immunomedics acquisition in 2020, and the original investment thesis included lung cancer as a major expansion market. For context on the ADC field and how Trodelvy fits in, see our ADCs explained guide. For a broader look at Gilead’s oncology pipeline, read our coverage of Gilead’s Q2 earnings.

What to watch next: ASCENT-05 adjuvant TNBC data. If that trial misses, the Trodelvy plus Keytruda thesis collapses to the single already-approved metastatic indication.

What It Means for the Trade

The immediate market takeaway is the asymmetry of blame. This was a Merck-branded trial (KEYNOTE-D46) testing Gilead’s drug against Merck’s drug — and the result is arguably worse for Merck’s combo-oncology ambitions than for Gilead’s asset. Keytruda monotherapy in PD-L1-high (TPS ≥50%) NSCLC is the hardest control arm in oncology: median OS north of two years means any add-on must deliver a genuinely large effect to separate. Trodelvy didn’t. But Merck had the same problem with its own MK-2870-family ambitions in this population, and the failure now functions as a data point the whole Trop-2-ADC-plus-checkpoint class has to digest.

For GILD holders, the loss is bounded but strategically clarifying. Trodelvy at $457 million a quarter (+26% YoY) remains a solid breast-cancer-centered franchise, and the first-line metastatic TNBC approval (June 2026, CHMP recommendation July) is still ramping. What died here was the lung expansion — the largest single revenue-escalator in the bull case. Gilead’s $21 billion Immunomedics price tag (2020) included lung as a pillar; that pillar is now formally gone, and the oncology growth narrative narrows toward breast (ASCENT-05 adjuvant), the approved SCLC and endometrial paths, and the next-gen ADCs behind Trodelvy in Gilead’s pipeline.

The subtle winner is Merck’s sac-TMT (sacituzumab tirumotecan, from the Kelun Biotech partnership). The EVOKE-03 failure removed the most advanced Western Trop-2 competitor in NSCLC combination settings, leaving sac-TMT — already PFS-positive in a China-only Phase 3 in PD-L1-negative NSCLC — as the presumptive frontrunner for Trop-2 in lung. Watch Merck’s global registrational strategy announcements here; the lane just cleared.

Mechanism and Class Context

Why Trodelvy-plus-Keytruda failed here is worth internalizing as a class lesson. Trodelvy’s SN-38 payload via a hydrolyzable linker produces moderate bystander killing and a toxicity profile (neutropenia, diarrhea) that constrained dose intensity when combined with an immune checkpoint backbone. In PD-L1-high disease, where single-agent pembrolizumab already achieves deep and durable responses, the marginal benefit of adding a moderately potent ADC has to overcome both the statistical bar set by a strong control arm and the practical bar of added toxicity-driven discontinuations. The PFS numerical-benefit-but-miss pattern is exactly what an underpowered-by-toxicity combination looks like — and it’s a caution for every ADC-plus-PD-1 readout in high-checkpoint-responsive populations (compare the Dato-DXd struggles in non-squamous NSCLC before finding its lane in squamous and other tumors).

The class implication isn’t “Trop-2 is dead” — Trodelvy’s approved TNBC indications and Datopotamab’s emerging wins prove the target works in the right contexts. It’s that Trop-2 ADCs plus checkpoint monotherapy in PD-L1-high NSCLC is now a validated failure mode, and the field’s bets shift to ADC-plus-chemo-plus-checkpoint triplets (where Merck’s sac-TMT and AstraZeneca/Daiichi’s Dato programs are pointed) and to biomarker-selected populations.

Risks and What Could Go Wrong

For Gilead: ASCENT-05 (adjuvant TNBC, 2027) is now the only remaining Trodelvy-Keytruda combination catalyst, and a miss there compresses the franchise to its approved metastatic label while the drug’s patent runway (late-2020s for the base patent, with formulation and manufacturing life beyond) starts to matter more. Gilead’s broader oncology story — Cell Therapy declines, Trodelvy as the growth engine — leans heavily on the breast franchise compounding without new setbacks. For Merck: the sac-TMT global strategy carries China-data credibility questions (the same multiregional-trial bar that ivonescimab faces), and the Kelun partnership’s value depends on clearing it. For the class generally: each ADC-checkpoint failure raises the regulatory and payer bar for the next regimen’s effect size, squeezing the commercial viability of marginal combinations. The trade: watch ASCENT-05 as the swing, treat the EVOKE-03 lesson as a class-wide de-rating input, and don’t extrapolate the failure to settings (post-chemo, biomarker-selected, different payloads) where the biology differs.