Compugen Ltd.: The Deal Whisperer

In December 2025, Compugen did something unusual. It sold a piece of a future royalty stream to AstraZeneca, the very partner already developing the drug. The $65 million upfront payment was non-dilutive, no new shares issued, no warrants attached, no equity stake taken. AstraZeneca simply paid for the right to collect a slice of royalties that Compugen would otherwise have received on rilvegostomig, a bispecific antibody targeting PD-1 and TIGIT that Compugen discovered through its own platform years earlier.

The transaction captured, in a single wire transfer, how Compugen operates. It does not build sales forces. It does not run massive late-stage trials alone. It discovers molecules, proves their biology, then partners with larger players who carry the cost of development and commercialisation. In return, Compugen collects milestones and royalties. And when it needs cash to fund its own pipeline, it monetises those future royalties rather than diluting existing shareholders.

The financial arc tells the story starkly. In Q2 2025, before the AstraZeneca deal, Compugen reported a net loss of $7.3 million and held $93.9 million in cash. By Q4 2025, after the $65 million landed, net income hit $56.8 million and cash surged to $145.6 million. The swing was not the result of operating improvements or pipeline progress. It was the result of a single financial transaction, a reminder that Compugen's income statement is, at any given moment, a reflection of deal timing rather than recurring revenue.

By March 2026, cash had drifted down to $134.9 million as the company funded its clinical programs. Revenue in Q1 2026 fell back to $2.2 million, consisting entirely of milestone payments from Gilead for GS-0321, a program licensed years earlier. No AstraZeneca money arrived in that quarter. The message was clear: between deal events, Compugen looks like a typical pre-commercial biotech, burning cash on R&D and waiting for the next transaction.

Compugen was incorporated in 1993 and is headquartered in Holon, Israel. It has been a public company since August 2000, when it listed on NASDAQ. For most of its existence, it has been a platform company, a computational biology shop that uses proprietary algorithms to identify novel drug targets and antibody candidates. Its longevity in a sector where most small-cap biotechs either get acquired or run out of money is itself a data point. The company has survived because it has learned to generate value from discovery without bearing the full cost of development.

The platform's most advanced output is rilvegostomig, the bispecific antibody Compugen discovered and licensed to AstraZeneca. AstraZeneca is now running the late-stage trials. But the asset that investors watch most closely for Compugen's own account is COM701, an antibody targeting PVRIG, a novel immune checkpoint. COM701 was discovered entirely through Compugen's proprietary platform. It represents the company's best chance to prove that it can do more than discover molecules for others, that it can take an asset through registration itself.

Ophir's comment, made during the FY2025 earnings call, was a response to a question about how Compugen differentiates its assets in a crowded immuno-oncology field. The company's internal discovery approach emphasises specific antibody formats: rilvegostomig uses a reduced-Fc format designed to reduce off-target immune activation, and GS-0321, licensed to Gilead, uses a cytokine-harnessing approach. These are not me-too checkpoint inhibitors. They are molecules designed around specific biological insights about how the tumour microenvironment suppresses immune responses.

The antibody format refrain is more than a technical detail. In a competitive landscape where dozens of companies target PD-1, TIGIT, and other established checkpoints, Compugen argues that the difference between success and failure often comes down to engineering, how a molecule binds, what immune cells it engages, and what side effects it avoids. The claim is difficult to prove preclinically, but it gives the company a narrative framework for why partners like AstraZeneca and Gilead have been willing to pay for access to its discoveries.

In September 2025, Eran Ophir became President and CEO of Compugen, succeeding Anat Cohen-Dayag, who moved into the role of Executive Chair. The succession was planned and orderly, not the kind of emergency replacement that often signals trouble at small biotechs. Cohen-Dayag had led the company through a period of platform validation and partnership-building, including the original rilvegostomig license to AstraZeneca. Ophir inherited a company with a validated discovery engine, a maturing wholly-owned pipeline, and a balance sheet transformed by the December 2025 deal.

The timing of the leadership change matters because it coincided with the most consequential financial transaction in the company's recent history. The AstraZeneca royalty monetisation was signed two months after Ophir took over. It is possible the deal was teed up under Cohen-Dayag and executed under Ophir. Either way, it gave the new CEO an immediate credibility boost with investors. On earnings calls, Ophir projects a tone of measured confidence, collaborative, partnership-focused, but clear-eyed about the risks.

This is not trivial. In small-cap biotech, CEO transitions often trigger strategic pivots or cultural disruption. Compugen's continuity, Cohen-Dayag remaining as Executive Chair, Ophir having been inside the company before taking the top role, suggests a deliberate attempt to maintain strategic course while refreshing leadership. The message to investors is that the platform model does not depend on any single executive. The company, not the CEO, is the discovery engine.

The MAIA-ovarian adaptive platform trial, initiated in the first quarter of 2026, is the most important clinical event in Compugen's history as an independent company. The trial randomises COM701 as maintenance monotherapy versus placebo in patients with relapsed platinum-sensitive ovarian cancer. It is designed as a potentially registration-enabling study, meaning that if the data are positive, Compugen may be able to file for approval based on this single trial.

The study builds on a pooled Phase I analysis presented at ESMO in 2025, which showed durable responses with COM701 in the more difficult platinum-resistant ovarian cancer setting. The decision to test COM701 first in platinum-sensitive disease, where patients have a better prognosis, reflects a strategic choice to pursue a cleaner registration path. If COM701 can prolong progression-free survival in patients whose cancer has already responded to platinum chemotherapy, the argument for using it as a maintenance treatment becomes straightforward.

The risk is evident. R&D expenses rose to $6.9 million in Q1 2026 from $5.8 million a year earlier, driven primarily by MAIA-ovarian clinical costs and drug supply. Those costs will increase as the trial enrols. The company has enough cash to fund the study into 2029, but the trial's outcome is binary. If COM701 demonstrates a meaningful PFS benefit, Compugen will have a registrable asset and a potential commercial product. If the trial misses, the wholly-owned pipeline loses its most mature candidate, and the company reverts to being purely a discovery platform dependent on partner-funded assets.

Ovarian cancer is a logical indication for PVRIG targeting. The biology suggests that PVRIG is upregulated in ovarian tumours, and COM701's mechanism of action, blocking an immune checkpoint that suppresses T-cell activity, has shown signals of activity in early testing. But the history of checkpoint inhibitors in ovarian cancer is littered with negative trials. PD-1 inhibitors have shown only modest single-agent activity in the disease. COM701 targets a different pathway, but the burden of proof remains high for a small company running its own pivotal trial.

The financial statements tell a story that the narrative around platform innovation and partnership strategy cannot fully mask. Compugen is dependent on deal timing. The numbers from fiscal 2025 illustrate the point dramatically.

Revenue jumped from $27.9 million in FY2024 to $72.8 million in FY2025, but almost all of the increase came from the AstraZeneca upfront payment. Strip out that transaction, and the underlying revenue trajectory would have been modestly higher due to Gilead milestones for GS-0321, but nothing approaching the reported figure. The gross margin swing from negative 32.5 percent in Q2 2025 to 94.7 percent in Q4 2025 and back to 16.2 percent in Q1 2026 is not a story of operational leverage. It is a story of deal accounting.

The cash runway is the single most important financial metric for investors. As of the Q1 2026 earnings call, Ophir stated that the company's cash position supports operations into 2029, with one crucial caveat: assuming no further cash inflows. That caveat matters because the company's business model depends on further cash inflows. Without new milestone payments from Gilead for GS-0321, without potential downstream payments from AstraZeneca on rilvegostomig development or commercial milestones, without any new partnership or royalty monetisation, the cash balance will decline steadily toward zero.

GS-0321, the program licensed to Gilead, dosed its first patient in January 2025. It is a novel cytokine-harnessing immunotherapy that represents another example of Compugen's platform output being developed by a larger partner. Under the terms of the Gilead license, Compugen is eligible for development and commercial milestones plus tiered royalties. Those milestones, if they materialise, represent the most probable source of the further cash inflows that the runway assumption excludes. But GS-0321 is early stage. Phase I data are not yet available. The timeline to meaningful milestones is uncertain.