Ivonescimab trumps tislelizumab for squamous NSCLC in HARMONi-6

In the phase III HARMONi-6 trial, ivonescimab plus chemotherapy significantly improved progression-free survival (PFS) compared with tislelizumab plus chemo as first-line treatment for individuals with advanced squamous non-small cell lung cancer (NSCLC).

After a median follow-up of 10.28 months, the median PFS with ivonescimab + chemo was 11.1 months, whereas with tislelizumab + chemo, it was only 6.9 months. A comparison between arms yielded stratified hazard ratios (HRs) of 0.60 (95 percent confidence interval [CI], 0.46–0.78; p<0.0001) by IRRC* and 0.64 (95 percent CI, 0.50–0.84) by investigator assessment.

The PFS benefit with ivonescimab was sustained across key subgroups, with the most pronounced treatment effect observed among participants aged <65 years (HR, 0.40) and those with ≥3 metastatic sites (HR, 0.46). [Lu, S, et al, ESMO Asia 2025]

PFS was also better with the ivonescimab vs the tislelizumab regimen, irrespective of PD-L1 expression, with HRs ranging from 0.55 (PD-L1 TPS** <1 percent) to 0.71 (PD-L1 TPS ≥50 percent).

Compared with the tislelizumab arm, the ivonescimab arm had better tumour responses (objective response rates, 75.9 percent vs 66.5 percent; p=0.008 [any PD-L1], 69.5 percent vs 61 percent [PD-L1 TPS <1 percent], and 80.1 percent vs 70.2 percent [PD-L1 TPS ≥1 percent]) and a longer median duration of response (11.2 vs 8.38 months; p=0.0219).

Moreover, 75.6 percent of participants on the experimental regimen had a partial response, and only 2.3 percent had progressive disease. With the comparator regimen, the corresponding rates were 66.5 percent and 5.6 percent, respectively.

The proportion of participants reporting a grade ≥3 treatment-related adverse event (TRAE) was greater in the investigational vs the comparator arm (63.9 percent vs 54.3 percent), but there were similar incidences of serious TRAEs (32.3 percent vs 30.2 percent), TRAEs leading to ivonescimab or tislelizumab discontinuation (3.4 percent vs 4.2 percent), and TRAEs leading to death (3 percent vs 3.8 percent).

The most common TRAEs with ivonescimab were reductions in neutrophil (32 percent) and white blood cell (10.9 percent) counts and leukopenia (5.6 percent).

“Ivonescimab also exhibited similar immune-related AEs to tislelizumab,” said study investigator Dr Shun Lu from the Shanghai Chest Hospital, Shanghai, China, who presented the results at ESMO Asia 2025.

Ivonescimab + chemo was tied to higher incidences of possibly VEGF-related AEs than tislelizumab + chemo, but these were mostly grade 1/2. The most common any-grade VEGF-related AEs with ivonescimab were proteinuria (27.1 percent), haemorrhage (21.4 percent), and hypertension (10.2 percent).

The median time to deterioration of global health status/quality of life was not reached in the ivonescimab arm and 12.4 months in the tislelizumab arm; the between-group comparison yielded a stratified HR of 0.94.

Poorer prognosis than non-squamous NSCLC

“Squamous NSCLC is typically centrally located, with rarely actionable driver gene mutations and poorer prognosis than non-squamous NSCLC,” said Lu.

“[Moreover,] squamous NSCLC has few driver mutations … and median PFS remains around 6–9 months, suggesting that we need to work harder in this area,” added discussant Dr Kumar Prabhash from the Tata Memorial Hospital, Mumbai, India, at ESMO Asia 2025.

PD-(L)1 inhibitors, including tislelizumab + chemo, are the standard first-line treatments for advanced squamous NSCLC, noted Lu.

The results of HARMONi-A and HARMONi-2 have led to the approval of ivonescimab, an anti-PD-1/VEGF bispecific antibody, for the treatment of patients with EGFR-sensitive mutation non-squamous NSCLC who have progressed on EGFR-TKI therapy, and those with PD-L1 TPS ≥1 percent advanced NSCLC as first-line treatment in China. [JAMA 2024;332:561-570; Lancet 2025;405:839-849]

HARMONi-6 comprised patients with pathologically confirmed, stage IIIB–IV squamous NSCLC who had no prior systemic therapy and no EGFR mutations or ALK rearrangements. A total of 532 participants were randomized 1:1 to receive up to four cycles of chemotherapy with either ivonescimab 20 mg/kg Q3W or tislelizumab 200 mg Q3W. The chemo regimen consisted of carboplatin AUC 5 Q3W and paclitaxel 175 mg/m² Q3W. Following which, ivonescimab or tislelizumab was continued for up to 24 months or until unacceptable toxicity.

Overall, approximately 93 percent of the participants were men, and about 52 percent were <65 years. Ninety-two percent had stage IV disease, approximately two-thirds had central type tumours, 60.5 percent had PD-L1 TPS 1 percent, and about 15 percent had ≥3 metastatic sites.

Taken together, treatment with ivonescimab plus chemo led to a significant PFS improvement, with consistent benefits across key subgroups, a more durable tumour response, and a manageable safety profile in squamous NSCLC.

“Hence, ivonescimab plus chemo may be a new standard of care for advanced squamous NSCLC,” Lu said.

A promising new frontline Tx option

According to Prabhash, HARMONi-6 demonstrated a robust PFS benefit with a novel bispecific strategy. While vigilance is warranted for VEGF-related events, he noted that the toxicities were acceptable.

“Ivonescimab + chemo is a promising new frontline treatment alternative that validates PD-1 plus VEGF synergy and addresses the unmet need in squamous NSCLC, which has very limited treatment options,” Prabhash said.

Overall survival (OS) was immature at the time of reporting. “[The results have] high relevance for Asian practice. The OS results are eagerly awaited for full confirmation,” Prabhash said.