Summit Therapeutics Was Supposed to Dethrone Keytruda. Then the Data Arrived.

But lung cancer isn't one disease. The squamous subtype (named for the flat cells where the cancer originates) tends to be trickier. Squamous tumors are more commonly found in smokers, respond differently to treatment, and have historically been harder to crack with VEGF-targeting drugs. The interim analysis in this particular cohort is where the wheels came off.

Full trial data from HARMONi-3 is still expected in the second half of 2026. So it's possible the non-squamous cohort tells a different story. But "just wait for the rest of the data" is a tough sell when your stock just lost a quarter of its value.

The Earnings Didn't Help Either

The clinical miss would have been painful on its own. Pair it with Summit's first-quarter financials, and you've got a one-two punch that rattled even the bulls.

Summit reported a net loss of $189.4 million for Q1 2026, more than triple the $62.9 million loss from the same quarter a year ago. Operating expenses ballooned to $195.2 million, up from $66.8 million. The GAAP loss per share came in at $0.24, missing the consensus estimate of $0.27. Running five global Phase III trials simultaneously is expensive, and when the data disappoints, that burn rate goes from "aggressive" to "alarming" in a hurry.

Piper Sandler responded by trimming its price target to $16 from $17 while keeping a Neutral rating. At least one analyst maintained a Buy with a $40 target, arguing the November FDA decision could still unlock major value. That's a wide spread, which tells you how divided the Street is right now.

A Crowded Race Gets More Interesting

Summit isn't the only company chasing the PD-1/VEGF dream. This drug class has attracted significant licensing deal activity since 2024, as Big Pharma scrambles to defend against (or join) the bispecific revolution.

Pfizer licensed SSGJ-707 from 3SBio. BioNTech grabbed PM8002 from Biotheus in a deal worth $55 million upfront plus over $1 billion in milestones. Merck KGaA struck a deal with Hengrui worth €1.4 billion covering HRS-1167 and SHR-A1904. And Merck (the U.S. one, Keytruda's maker) is running its own candidate, MK-2010, which showed a 55% response rate in early NSCLC data.

The core question now is whether all PD-1/VEGF bispecifics are created equal, or whether ivonescimab's squamous miss reveals something specific about this drug. If it's a class-wide issue with squamous tumors, the entire sector takes a hit. If it's ivonescimab-specific, competitors might actually benefit from Summit's stumble.

The Bigger Picture for I/O Combinations

Summit's bad week fits into a broader pattern. The immuno-oncology combination space has been humbling investors lately.

IO Biotech's Keynote-D18 trial failed in melanoma in August 2025. The TIGIT pathway, once considered a promising checkpoint target, has racked up serial failures from Roche, Merck, BMS, and iTeos/GSK. The common thread? Early-stage excitement that didn't survive the rigor of large, randomized trials. iTeos/GSK's belrestotug showed promising response rates in mid-stage testing, then delivered no meaningful benefit in the pivotal study. It's like acing every practice exam and bombing the final.

Investors are getting pickier as a result. "Rational combinations" (checkpoint inhibitors paired with chemo or antibody-drug conjugates) still get funding. But novel dual-target experiments face a higher bar, both from regulators and from the market.

What Comes Next

Summit's story isn't over. The FDA decision on ivonescimab for EGFR-mutant lung cancer arrives in November. And full HARMONi-3 data later this year could redeem (or further damage) the narrative.

But May 1st was a reminder that in biotech, positioning and promise only take you so far. At some point, the data has to deliver. Summit bet billions that one molecule could do the work of two. The molecule just blinked.