For the first time ever, a drug has beaten Keytruda head-to-head in first-line lung cancer, on both progression-free survival and overall survival. The challenger? A Chinese bispecific antibody that attacks tumors from two directions at once.
Keytruda has been untouchable for years. The world's best-selling cancer drug. A $31.7 billion juggernaut. The cornerstone of first-line lung cancer treatment across the globe.
Nobody had beaten it head-to-head in a phase 3 trial. Until now.
At the World Conference on Lung Cancer (WCLC) 2026, China's Akeso dropped the kind of data that makes oncologists sit up straight: its bispecific antibody ivonescimab didn't just match Keytruda (pembrolizumab) in advanced lung cancer patients. It beat it. On progression-free survival and overall survival. In a randomized, double-blind, head-to-head trial.
This is the first time any drug has pulled that off in first-line PD-L1-positive non-small cell lung cancer (NSCLC), the most common type of lung cancer. And if the data hold up globally, the immunotherapy landscape may never look the same.
One Drug, Two Jobs
To understand why ivonescimab worked so well, you need to understand what makes it different from Keytruda.
Keytruda does one thing brilliantly. It blocks a protein called PD-1 on T cells, which is like removing the brakes from your immune system's attack squad. Tumors exploit PD-1 to hide from immune detection. Block it, and the T cells can find and kill cancer cells again. Simple, elegant, and wildly effective for millions of patients.
Ivonescimab does that and something extra. It's a bispecific antibody, meaning it's engineered to grab two different targets at once. One arm blocks PD-1 (just like Keytruda). The other arm grabs VEGF, a protein that tumors use to grow new blood vessels and feed themselves.
Think of it like this: Keytruda sends the cops after the tumor. Ivonescimab sends the cops and cuts the power lines to the hideout. You're attacking the cancer from two directions with a single molecule.
What's especially clever is something called cooperative binding. When ivonescimab latches onto VEGF, it actually gets better at grabbing PD-1, and vice versa. The two arms make each other stronger. It's a rare case where 1 + 1 genuinely equals more than 2.
Get tomorrow's biotech intelligence before your competitors.
Join thousands of biotech professionals who start their day with our free, daily briefing.
The Numbers That Shook the Room
The HARMONi-2 trial enrolled patients with locally advanced or metastatic NSCLC whose tumors expressed PD-L1 (a marker that predicts response to checkpoint drugs like Keytruda). Patients were randomized to receive either ivonescimab or pembrolizumab, with neither doctors nor patients knowing which drug they got.
The primary endpoint was progression-free survival (PFS): how long patients lived without their cancer getting worse. Ivonescimab delivered a median PFS of 11.14 months compared to just 5.82 months for pembrolizumab. The hazard ratio was 0.51, which translates to a 49% reduction in the risk of disease progression. The p-value was less than 0.0001, which in statistical terms means "this almost certainly wasn't a fluke."
But PFS has always been the appetizer. Overall survival (OS) is the main course, and it's what regulators, doctors, and patients care about most. Does the drug actually help people live longer?
The answer: yes. Median overall survival was 30.8 months with ivonescimab versus 22.6 months with pembrolizumab. That's an extra eight months of life, on average. In oncology, an eight-month OS improvement in a head-to-head trial against the standard of care is a seismic result.
The Catch (Because There's Always a Catch)
Ivonescimab's side effect profile deserves an honest look. The trial showed more grade 3 or higher treatment-related adverse events with ivonescimab compared to pembrolizumab. That makes intuitive sense: you're hitting two targets instead of one, so there's more biological disruption.
The rates of severe immune-related side effects were similar between the two drugs, which is reassuring. But the overall toxicity burden was higher with ivonescimab. For doctors making treatment decisions, this creates a real tradeoff: more efficacy, but more side effects. That calculus will play out differently depending on the patient.
There's also a geographic asterisk. HARMONi-2 was conducted entirely in China. That's not disqualifying by any means; Chinese clinical trials are rigorously designed and regularly inform global practice. But regulators, payers, and guideline committees in the U.S. and Europe will want to see confirmatory data from a more geographically diverse population before overhauling treatment protocols.
That confirmatory data is coming. The HARMONi-3 study, a global phase 3 trial in first-line NSCLC, has interim readouts expected as early as 2026, with additional data through 2027. If it replicates these results in a broader population, the conversation shifts from "interesting China data" to "new global standard of care."
The Road to American Pharmacies
Ivonescimab is already approved in China. Getting it to American patients is Summit Therapeutics' job.
Summit licensed exclusive rights from Akeso to develop and sell ivonescimab in the Americas, Europe, Japan, and other international markets. They filed a Biologics License Application (BLA) with the FDA, which was accepted on January 29, 2026, with a decision deadline of November 14, 2026.
One important detail: the U.S. filing isn't for the same indication that just produced these headline results. Summit's initial BLA is for EGFR-mutated NSCLC after prior targeted therapy, a different (and smaller) patient population. Think of it as getting a foot in the door. The blockbuster first-line PD-L1-positive indication will likely require the global HARMONi-3 data before it can be filed in the U.S.
So the timeline looks something like this: potential first U.S. approval by late 2026 in a niche lung cancer setting, with the broader (and far more lucrative) first-line opportunity potentially following in 2028 or 2029.
Why Merck Should Be Nervous
Keytruda isn't just a drug. It's an economic engine. Merck reported $31.7 billion in global Keytruda sales in 2025, with analysts projecting north of $32 billion for 2026. Lung cancer is its single biggest indication, accounting for roughly 38 to 40% of total Keytruda revenue.
For years, Merck has operated from a position of strength. Competitors tried to go around Keytruda (different combinations, different patient populations) rather than through it. Nobody dared run a head-to-head superiority trial in Keytruda's strongest setting.
Akeso did. And won.
Now, this doesn't mean Keytruda revenue collapses tomorrow. The drug is approved in dozens of cancer types, deeply embedded in treatment guidelines worldwide, and backed by an enormous commercial infrastructure. But the narrative has shifted. Keytruda is no longer unbeatable. And in the pharmaceutical world, narrative matters almost as much as data.
China's Bispecific Antibody Factory
Ivonescimab isn't an isolated success story. It's the tip of a Chinese biotech iceberg.
China now accounts for 8 of 13 global PD-1/VEGF bispecific antibodies. A 2026 review counted 121 interventional trials for PD-1-containing bispecifics, underscoring just how crowded (and ambitious) this space has become.
Akeso itself has another major bispecific, cadonilimab, which targets PD-1 and CTLA-4 instead of VEGF. It's already approved in China for cervical and gastric cancer. Other Chinese companies like 3SBio (whose PD-1/VEGF drug SSGJ-707 was licensed by Pfizer), Junshi Biosciences, and Kelun-Biotech are all pushing their own bispecific candidates through development.
The pattern is clear: China is emerging as the world's bispecific antibody powerhouse. The question is whether these programs can clear the higher regulatory and commercial bars required for U.S. and European markets. Ivonescimab's HARMONi-2 data suggest the science is real. Converting it into global market access is the next challenge.
The Bottom Line
For the first time, a drug has beaten Keytruda head-to-head on both PFS and overall survival in first-line lung cancer. That sentence alone would have seemed impossible five years ago.
Ivonescimab's dual-target approach (blocking both the immune checkpoint and the tumor's blood supply in one molecule) appears to offer a genuine step forward over single-target PD-1 inhibitors. The survival benefit is clinically meaningful; the mechanism is scientifically elegant.
But this isn't a coronation yet. The toxicity tradeoff needs careful management. The geographic limitation of HARMONi-2 demands global confirmation. And the U.S. regulatory path, while progressing, still has milestones to clear.
What is undeniable: the era of Keytruda's unchallenged dominance in lung cancer just ended. The next chapter belongs to bispecifics, and a Chinese company wrote the opening page.
Clinical & Regulatory5 min read
Sarepta's Worst Week: A Black Box, 500 Pink Slips, and a Gene Therapy in Crisis
Sarepta hit its Duchenne gene therapy Elevidys with the FDA's most serious safety warning after fatal liver failures, then cut 36% of its workforce in the same week. It's the biggest crisis yet for one of gene therapy's most prominent companies.
