
Dr Leighl on HARMONi Survival Data and the FDA's Decision on Ivonescimab in Previously Treated EGFR-Mutated NSCLC
Natasha B. Leighl, MD, MMSc, FRCPC, FASCO, discusses the FDA's decision on ivonescimab plus chemotherapy in EGFR-mutated NSCLC.
"From my perspective, we believe that this [overall survival] difference is real…. I think that this post hoc analysis actually represents more mature follow-up and stronger data."
Natasha B. Leighl, MD, MMSc, FRCPC, FASCO, division head of Medical Oncology and Hematology at Princess Margaret Cancer Centre and a professor of medicine at the University of Toronto, discussed the FDA's decision not to approve ivonescimab plus chemotherapy and updated overall survival (OS) data from the phase 3 HARMONi trial (NCT06396065) evaluating ivonescimab plus carboplatin and pemetrexed in patients with EGFR-mutated non–small cell lung cancer (NSCLC) whose disease progressed on a third-generation EGFR TKI, as well as what ongoing frontline trials of the agent would need to show to change the regulatory picture.
In November 2026, the FDA issued a
HARMONi randomly assigned 438 patients 1:1 to ivonescimab or placebo, each with carboplatin and pemetrexed, and
PFS has historically been an acceptable end point in the United States, Leighl noted, and the PFS benefit in HARMONi was sustained. From a Canadian perspective, however, OS is the key end point. She suggested the FDA's stance may partly reflect the VEGF-inhibitor component of ivonescimab, because VEGF inhibitors have previously improved PFS without an OS benefit in lung and other advanced cancers.
An updated analysis extended follow-up in Western patients to a data cutoff of June 30, 2026, raising their median follow-up from 9.2 months at the primary analysis to 23.2 months. In this
Short follow-up may miss a survival benefit that emerges later with biologic agents, Leighl explained. Hazard ratios rarely improve as data mature, but in HARMONi the curves continued to separate. With novel biologics, response rate, PFS, and OS gains do not always track together, she added. Leighl, who acknowledged her bias as a trial investigator, considers the survival difference real.
With respect to safety, grade 3 or higher treatment-related adverse effects occurred in 51.4% of patients who received ivonescimab plus chemotherapy (n = 218) vs 43.6% of those who received placebo plus chemotherapy (n = 218) at the June 30, 2026, data cutoff.
Looking ahead, the frontline phase 3 HARMONi-3 trial (NCT05899608) is comparing ivonescimab plus chemotherapy vs pembrolizumab (Keytruda) plus chemotherapy, and the phase 3 HARMONi-7 trial (NCT06767514) is comparing ivonescimab vs pembrolizumab in patients with high PD-L1 expression. These trials will show whether the VEGF component and bispecific construct add benefit beyond PD-1 inhibition, Leighl said. An OS benefit would change practice, whereas a PFS gain alone would leave open whether the effect reflects VEGF inhibition. The trials must also show that ivonescimab is not more toxic, she noted.
Related to this article








