Ivonescimab plus chemo extended survival in EGFR lung cancer after TKI failure

Adding ivonescimab, a bispecific antibody, to chemotherapy lengthened survival for patients whose EGFR-mutated lung cancer had progressed after targeted therapy, according to the final analysis of the phase 3 HARMONi-A trial published in JAMA [s1]. Median overall survival was 16.8 months with the antibody plus chemotherapy versus 14.1 months with chemotherapy alone, an absolute gain of 2.7 months (hazard ratio 0.74; 95% confidence interval 0.58 to 0.95; P=0.02) [s1].

The result matters because this is a hard place to treat. Patients with EGFR-variant non-small-cell lung cancer (NSCLC) whose disease grows despite an EGFR tyrosine kinase inhibitor (TKI) — the mainstay targeted pills such as osimertinib — have few good options, and chemotherapy alone buys limited time [s1]. Ivonescimab is a first-of-kind molecule that blocks two targets at once: PD-1, the immune checkpoint that drugs like pembrolizumab release, and VEGF, the signal tumours use to grow blood vessels [s1].

What the trial did

HARMONi-A was a randomised, double-blind, placebo-controlled phase 3 trial conducted at 55 sites, all in China [s1]. Between 25 January and 2 November 2022 it enrolled 322 adults with locally advanced or metastatic EGFR-variant non-squamous NSCLC who had already been treated with an EGFR-TKI [s1]. They were assigned 1:1 to ivonescimab at 20 mg/kg (161 patients) or placebo (161 patients), each added to pemetrexed and carboplatin every three weeks for four cycles, followed by maintenance therapy [s1].

The trial's primary endpoint was progression-free survival, assessed by an independent radiology committee, and that result was reported earlier: in the 2024 analysis, ivonescimab plus chemotherapy extended median progression-free survival to 7.1 months from 4.8 months, a hazard ratio of 0.46 (95% CI 0.34 to 0.62; P<0.001) [s2]. Overall survival was a key secondary endpoint, tested only after the primary succeeded, and the 2026 paper reports it with the data matured — a median follow-up of 32.5 months against a cutoff of 12 April 2025 [s1].

What it found

Beyond the 2.7-month median gain, the survival curves separated durably: estimated 30-month survival was 29.1% (95% CI 22.1 to 36.4) with ivonescimab and 18.4% (95% CI 12.8 to 24.8) with placebo [s1]. Roughly three in ten patients on the antibody were alive at two and a half years, against fewer than one in five on chemotherapy alone [s1].

The cost was added toxicity. Grade 3 or higher treatment-emergent adverse events occurred in 67.1% of the ivonescimab group versus 54.7% of the chemotherapy-alone group [s1]. That is a real increase, though the earlier analysis noted most severe events were chemotherapy-related rather than specific to the antibody [s2].

How to read it

Three cautions belong next to the headline. First, the survival benefit is statistically significant but modest in absolute terms — 2.7 months at the median, with a hazard-ratio confidence interval (0.58 to 0.95) whose upper bound sits close to 1.0, meaning the effect is real but not large [s1]. Second, overall survival was a secondary endpoint; the trial was designed and powered around progression-free survival, so the survival result, while prespecified and hierarchically tested, carries the weight of a supporting finding rather than the primary one [s1].

Third, and most consequential for readers outside China: every one of the 55 sites was in China [s1]. Trial populations, prior treatment patterns and access to later therapies differ across health systems, and a China-only phase 3 does not automatically transfer to patients elsewhere — a limitation regulators in the United States and Europe weigh heavily when a single-country trial supports a marketing application. The drug's developer, Akeso, has claimed broad activity for ivonescimab across lung-cancer settings, but those are company positions; the independent, peer-reviewed trial result is what is reported here [s1].

For the wider immunotherapy question — whether a PD-1/VEGF bispecific can beat single-checkpoint blockade — the more direct comparison lies in ivonescimab's separate head-to-head trials against pembrolizumab, not in HARMONi-A, which used chemotherapy as its backbone [s1]. Readers can set this against another targeted approach to driver-mutation lung cancer covered here, the early-phase RAS inhibitor daraxonrasib in RAS-mutant NSCLC, and the broader case for catching these cancers earlier through low-dose CT screening.

Why it matters

For EGFR-mutant lung cancer after TKI failure, incremental gains are the currency, and a validated survival signal in a randomised trial is uncommon in this setting. HARMONi-A establishes that adding a dual PD-1/VEGF antibody to chemotherapy helps — in the population and health system it studied. Whether that translates to a global standard of care depends on confirmatory trials enrolling more diverse populations, several of which are under way.

What to watch

The near-term questions are regulatory and geographic: whether agencies outside China accept a single-country phase 3 or require multiregional confirmation, and how the survival benefit holds up against the newer combinations entering the same post-TKI space [s1][s2]. The toxicity trade-off — a meaningfully higher rate of severe adverse events — will also shape where oncologists place the antibody for patients who have already been through targeted therapy and are weighing quality against quantity of remaining time [s1].

This article describes clinical-trial research and is not medical advice. Treatment decisions are for patients and their treating clinicians.

Sources

• Bispecific Antibody Ivonescimab Added to Chemotherapy in EGFR-Variant NSCLC: The HARMONi-A Randomized Clinical Trial — JAMA, 28 July 2026 (final overall-survival analysis)
• Ivonescimab Plus Chemotherapy in NSCLC With EGFR Variant: The HARMONi-A Randomized Clinical Trial — JAMA, 20 August 2024 (primary progression-free-survival analysis)