The failed melanoma vaccine showed a 16.6-month benefit in one subgroup but missed its primary endpoint. Now a bankruptcy trustee is selling the IP — and the price will tell us if the field sees salvage value in a statistical miss.
image from grok
IO Biotech filed Chapter 7 bankruptcy in March 2026 after the FDA rejected Cylembio, their off-the-shelf IDO1/PD-L1-targeting therapeutic cancer vaccine, despite a Phase 3 trial showing a clinically meaningful PFS improvement in advanced melanoma patients (19.4 vs 11.0 months). The primary endpoint narrowly missed statistical significance (HR 0.77, p=0.056 vs required p<0.045), though a prespecified subgroup analysis revealed a robust signal in PD-L1 negative patients (HR 0.54, p=0.006). The company's assets of $1-10M against liabilities of $10-50M with fewer than 50 creditors reflect the rapid collapse following the regulatory setback.
IO Biotech filed for Chapter 7 bankruptcy on March 31, 2026, six months after the FDA rejected its cancer vaccine Cylembio, and five months after the company had roughly $31 million in the bank. The bankruptcy petition, filed in U.S. Bankruptcy Court for the District of Delaware, lists assets of $1 million to $10 million against liabilities of $10 million to $50 million, with fewer than 50 creditors, according to a petition summary from Bankruptcy Observer. All six board members resigned. CEO Mai-Britt Zocca and CFO Amy Sullivan were terminated without cause. The company's workforce, once roughly 80 people, had already been halved in September 2025 after an FDA meeting and cut again in January, BioSpace reported.
The immediate question is what happens to Cylembio.
Cylembio — imsapepimut and etimupepimut, adjuvanted — is an investigational immune-modulatory off-the-shelf therapeutic cancer vaccine, as BioPharma Dive explained. The two adjuvanted peptide sequences are designed to provoke an immune response to IDO1 and PD-L1, checkpoint proteins that tumors use to suppress T-cell attacks. Unlike the personalized mRNA cancer vaccines in development at Moderna and BioNTech, Cylembio was designed as an off-the-shelf product: a patient walks in, gets an injection, and the vaccine is ready. No sequencing, no custom manufacturing. That simplicity was the pitch.
The Phase 3 trial enrolled 407 patients with previously untreated advanced melanoma and gave half Cylembio plus Keytruda, the other half Keytruda alone. The combination extended median progression-free survival to 19.4 months from 11.0 months, a meaningful gain, but the hazard ratio of 0.77 produced a p-value of 0.056, just shy of the 0.045 threshold required for statistical significance, IO Biotech reported. The FDA told IO Biotech not to file for approval, FierceBiotech reported.
The overall result missed. But the subgroup did not.
In patients whose tumors were PD-L1 negative, typically a poorer prognosis population, Cylembio plus Keytruda delivered a median PFS of 16.6 months versus 3.0 months for Keytruda alone. The hazard ratio was 0.54 with a nominal p-value of 0.006. That is not a marginal signal. In BRAF-mutated tumors and in patients with elevated LDH, the same pattern appeared: consistent favor toward the combination. "This is not the outcome we had hoped for," Zocca said in September. She was right. But the subgroup data suggested the vaccine was doing something real in the patients who needed it most.
The financial collapse was not subtle. IO Biotech had drawn 22.5 million euros from the European Investment Bank, and the Chapter 7 filing triggered an immediate event of default, making that sum immediately due, the SEC filing states. With assets measured in the low millions and liabilities measured in the tens of millions, there is no restructuring path. The trustee will sell what remains.
What remains is primarily the IP.
Whether Cylembio draws a buyer depends entirely on who is doing the math. A company with an existing immunotherapy franchise and the manufacturing capacity to scale an off-the-shelf vaccine might see value in a subgroup-split result that Moderna's and BioNTech's personalized approaches cannot easily replicate. Personalized vaccines require individual manufacturing; Cylembio does not. In a PD-L1 negative melanoma patient, the difference between 3.0 months and 16.6 months is the difference between progression and something approaching disease control.
The cancer vaccine field has bet heavily on personalization. The logistics are formidable and the costs are likely formidable too. An off-the-shelf vaccine that works in a defined subgroup occupies a genuinely different commercial and clinical niche.
Whether that niche is worth anything is an open question. IO Biotech's Chapter 7 filing suggests the market was not assigning it much value before the trustee's sale. But trustee sales are information-forcing events. The eventual buyer, if there is one, and the price they pay will say something concrete about whether the oncology field thinks a failed Phase 3 with a striking subgroup is a discarded asset or an unfinished story.
The trial data, at minimum, deserves a closer look than a bankruptcy court will give it.
Story entered the newsroom
Research completed — 0 sources registered. IO Biotech filed Chapter 7 March 31 2026, assets $1M-$10M vs liabilities $10M-$50M. Phase 3 IOB-013 missed primary endpoint p=0.056 vs threshold 0.045
Draft (663 words)
Reporter revised draft (671 words)
Reporter revised draft (644 words)
Reporter revised draft (675 words)
Approved for publication
Published (662 words)
Get the best frontier systems analysis delivered weekly. No spam, no fluff.
Biotech & Life Sciences · 18h 28m ago · 4 min read
Biotech & Life Sciences · 1d ago · 4 min read
