Two Survival Wins in NSCLC With Bispecific Seeking Worldwide Status

The investigational bispecific antibody ivonescimab significantly improved overall survival (OS) in PD-L1-positive advanced non-small cell lung cancer (NSCLC) compared with single-agent pembrolizumab (Keytruda), an updated analysis of the randomized HARMONi-2 trial showed.

Median OS — the key secondary endpoint of the trial — improved from 22.6 months with pembrolizumab to 30.8 months with ivonescimab. The survival curves began to separate at 12 months and steadily swung in favor of ivonescimab, with no narrowing of the gap. A consistent benefit was observed across subgroups, irrespective of PD-L1 expression cutoff, age, or tumor histology.

Previously reported results from the study showed superior progression-free survival (PFS) with ivonescimab, bolstering a case for a first-line indication, said Caicun Zhou, MD, PhD, of Shanghai East Hospital and Tongji University School of Medicine in Shanghai, at the World Conference on Lung Cancer (WCLC) in Seoul, South Korea.

“With longer follow-up, ivonescimab maintained a favorable and manageable safety profile with no new safety signals,” said Zhou. “These findings support ivonescimab as a first-line treatment for PD-L1-positive non-small cell lung cancer. A global phase III trial of ivonescimab versus pembrolizumab in PD-L1-high metastatic NSCLC is ongoing to provide a potential chemo-free regimen for patients worldwide.”

Not Ready for Prime Time … Yet

HARMONi-2 was conducted entirely in China, and PD-L1 expression was ascertained by means of an assay currently available only in China, noted invited discussant Mariana Brandão, MD, PhD, of Institut Jules Bordet in Brussels. Moreover, the pembrolizumab monotherapy control arm is not a global standard for the population.

“What we saw was, indeed, an improvement in overall survival among the PD-L1-positive population, which is not only statistically significant but also clinically relevant,” said Brandão. “Of course the control arm … is not what most of us are using in clinical practice. In these patients, we prefer to use anti-PD-1 plus chemotherapy.”

Closer inspection of the subgroup analysis showed that the results were driven primarily by the PD-L1-high subgroup (tumor proportion score [TPS] ≥50%) and, surprisingly, patients with squamous histology, who usually have a worse prognosis, she pointed out. Data on concomitant genomic alterations would have been informative, and the trial included relatively few non-smokers and patients with liver metastases, though both groups benefited from ivonescimab.

“Am I changing my clinical practice tomorrow for this PD-L1-high non-small cell lung cancer subset, replacing anti-PD-1 with ivonescimab?” asked Brandão. “The answer is no. This was an underpowered subgroup analysis of a secondary endpoint, and we know that ivonescimab, although well tolerated, has an increased toxicity compared to pembrolizumab. Also, we don’t have any other predictive biomarkers.”

“Am I excited by these data?” she continued. “Absolutely. I think this is really interesting, and of course, now we have to wait for the phase III global HARMONi-7 trial that is being conducted specifically in the PD-L1-high population. We also have to compare this to the other results that we already have in this space.”

In a separate presentation, ivonescimab picked up another OS win in an updated analysis of the global phase III HARMONi trial, involving patients with EGFR-mutated NSCLC and resistance to EGFR inhibitors.

At last year’s WCLC, the HARMONi trial showed a statistically significant improvement in PFS with ivonescimab plus chemotherapy versus pembrolizumab and chemotherapy. At that time, the second primary endpoint of OS trended in favor of the ivonescimab arm but did not achieve statistical significance.

The updated analysis showed a median OS of 16.8 months with ivonescimab-chemotherapy and 14 months with pembrolizumab-chemotherapy, representing a 24% reduction in the survival hazard (95% CI 0.61-0.95, P=0.0151), reported Antonio Passaro, MD, PhD, of the European Institute of Oncology in Milan.

A first-in-class therapy, ivonescimab pairs a PD-1 inhibitor and a VEGF inhibitor. The drug is approved in China but nowhere else in the world, and the bispecific has been studied most extensively in Chinese patients. Studies such as HARMONi-7 and HARMONi are evaluating ivonescimab in different populations to demonstrate broader applicability for treating PD-L1-positive NSCLC.

HARMONi-2 Details

HARMONi-2 included 398 patients with locally advanced/metastatic NSCLC, a PD-L1 TPS ≥1%, and no prior systemic therapy. Patients were randomized to ivonescimab or pembrolizumab and treated for as long as 2 years in the absence of progression or unacceptable toxicity.

The OS analysis showed an 8-month difference in favor of ivonescimab, translating into a 27% reduction in the survival hazard (95% CI 0.57-0.95, P=0.009). Landmark OS analyses favored ivonescimab at 24 months (57.9% vs 48%) and 36 months (45% vs 33.1%). For context, Zhou referred to the China-based KEYNOTE-042 study, which showed 2- and 3-year OS rates of 43.8% and 28.1%, respectively, in a population treated with pembrolizumab alone.

As Brandão pointed out, the overall benefit was driven by statistically significant advantages for ivonescimab in two subgroups: patients with PD-L1 TPS ≥50% (median OS not reached vs 23.2 months, HR 0.58, 95% CI 0.38-0.89) and squamous histology (30.5 vs 19.3 months, HR 0.65, 95% CI 0.45-0.95).

OS also favored ivonescimab in the PD-L1 TPS 1-49% population, but the difference did not achieve statistical significance (28.5 vs 22.1 months, HR 0.85, 95% CI 0.61-1.18). The same was true of patients with non-squamous tumors (33.6 vs 25.6 months, HR 0.79, 95% CI 0.55-1.14).

Twice as many patients in the ivonescimab arm developed grade ≥3 treatment-related adverse events (TRAEs, 41.6% vs 21.6%), but TRAE-related discontinuation rates were similar (4.1% vs 5%).