Akeso's stock has been hammered since early May, tumbling roughly 26% from its 2026 peak as the HARMONi-3 interim PFS miss at Summit Therapeutics spooked the entire ivonescimab complex. Shares trade around HK$116 - down from peaks near HK$158. The narrative is that ivonescimab's late-stage data is cracking.
That narrative got the wrong trial.
HARMONi-3's interim readout was a progression-free survival analysis in a small squamous NSCLC cohort run by Summit in the US. HARMONi-6 is the Phase 3 overall survival trial in first-line squamous NSCLC run by Akeso in China - and its OS data is selected for the plenary session at ASCO 2026. The embargo lifts May 31. That is the most prestigious presentation slot at the world's largest oncology conference. ASCO doesn't hand that to data that might miss.
Here's what matters before the readout:
1. PFS was already dominant - OS confirmation closes the argument. In HARMONi-6, ivonescimab plus chemotherapy delivered a median PFS of 11.14 months versus 6.9 months for the PD-1 control, a hazard ratio of 0.60 - a 40% reduction in the risk of disease progression or death. The PFS curve separation was so wide it landed in The Lancet. In oncology, a PFS HR of 0.60 typically translates into a meaningful OS signal. The question isn't whether there's a benefit; it's how large. A statistically significant OS result here validates the PD-1/VEGF bispecific class head-to-head against the PD-1 standard of care in the first-line squamous setting.
2. The HARMONi-3 miss is a different story in a different population. The interim PFS miss in HARMONi-3's squamous cohort was a single readout in a US-run trial. Interim analyses carry less statistical power than planned final analyses. Akeso still owns the China-run HARMONi-6 data, the commercial rights in China, and the pipeline that's actually generating revenue. Conflating Summit's interim miss with Akeso's plenary-grade OS dataset is how the market creates a disconnect that can be exploited.
3. China revenue growth is accelerating, not stalling. Akeso reported full-year 2025 revenue of RMB 3.06 billion, up 43.9% year over year. That growth is driven by ivonescimab's commercial sales in China following NMPA approval in April 2025, plus cadonilimab (an earlier PD-1/CTLA-4 bispecific). The company is growing its way toward profitability while the stock sells off on a trial miss that doesn't apply to its flagship's largest dataset.
4. The US pathway is on track with a November PDUFA date. Summit filed the BLA for ivonescimab in EGFR-mutated NSCLC (the HARMONi trial indication) in January 2026. The FDA accepted the filing and set a Prescription Drug User Fee Act action date of November 14, 2026. That's an external deadline the market can price. The HARMONi-6 OS data strengthens the entire ivonescimab package by confirming efficacy across indications. A positive plenary readout increases the odds of a smooth FDA review in the EGFRm indication and accelerates discussions for the first-line squamous BLA that follows.
5. Cash is not the problem - and it shouldn't be. Akeso holds more than RMB 4.8 billion in cash and cash equivalents. The full-year 2025 net loss widened to RMB 1.11 billion from RMB 515 million prior year, reflecting R&D spend and the commercial ramp. On a RMB 3 billion revenue base with RMB 4.8 billion in cash, the burn runway is measured in years, not quarters. The company isn't facing a capital overhang. It's facing a sentiment overhang, which is far cheaper to exploit.
The catalyst calendar is loaded. The HARMONi-6 OS data drops May 31. The November PDUFA date for the first US approval follows. After that, the first-line squamous BLA submission becomes the next milestone. Each one is a re-rating event - but only if the market hasn't already repriced the name on data it doesn't deserve.
The risk is clear. If HARMONi-6 OS misses statistical significance, the bispecific thesis takes a real hit and the stock goes lower. The alpha spending in late-stage oncology is real, and plenary selection is not a guarantee. But a PFS HR of 0.60 rarely fades into OS noise. The math works against that outcome.
At HK$116, with an analyst intrinsic value estimate near HK$172-200, Akeso is trading at a discount that reflects a narrative about the wrong trial. The real variable - HARMONi-6 OS - hasn't even been read out yet. If the data confirms what the PFS curve already suggests, the disconnect closes fast. That's the setup.