Open in interactive viewer → charts, metric popovers & call review

Compugen Moves the Goalposts Forward — Then Rewrites What the Goal Is

AstraZeneca's Phase III biliary readout jumps from 2028 to H2 2027, while Compugen quietly replaces 'ovarian PFS' with 'extending the platinum-free interval' — and unveils UnigenX with no candidates attached.
CGEN · Earnings Call · 2026-11-09

The readout that arrived early

The single most consequential sentence on Compugen's Q3 call was not about COM701. It was a scheduling change buried in the second paragraph of CEO Eran Ophir's prepared remarks: the partnered AstraZeneca program ARTEMIDE-Biliary01 is now expected to produce results sooner than the company last told investors.

“the Phase III readout from AstraZeneca now anticipated in the second half of 2027 from the ARTEMIDE-Biliary01 trial” — Eran Ophir, President and CEO · 2026-11-09

Three months ago, on the August 3 call, the same question from H.C. Wainwright's RK produced a materially different answer:

the Phase III results according to their guidance is after '27, meaning '28, and no, it doesn't mean that it couldn't be earlier interim analysis and other options, but the actual formal guidelines are after '27 for the Phase III studies.

Eran Ophir, President and Chief Executive Officer · 2026-08-03

So the formal guide has been pulled from "2028" to "second half of 2027" — a six-to-twelve-month acceleration on a 760-patient adjuvant biliary tract cancer study that Compugen does not run and cannot control. Management deserves credit for surfacing it. But note the phrasing they chose: high level results, not data. That is the language of a company managing expectations on a trial owned by someone else, where the actual disclosure decision — content, format, timing — sits in AstraZeneca's hands, exactly as it does with Gilead on GS-0321.

And the payment mechanics are narrower than the headline. Asked directly whether the Biliary01 readout triggers cash, Ophir was precise: “the next milestone is BLA acceptance” — Eran Ophir, President and CEO · 2026-11-09 — the $25 million added when Compugen monetized a sliver of its royalty stream. A positive Phase III is necessary but not sufficient; a regulatory submission is the trigger. The remaining $195 million in regulatory and commercial milestones, plus mid-single-digit royalties, stay back-loaded.

The quiet reframing of MAIA-Ovarian

MAIA-Ovarian is fully enrolled — 62 patients, randomized 2:1, maintenance monotherapy in second- and third-line relapsed platinum-sensitive disease — with data still guided to the end of Q1 2027, a timeline that has held for three consecutive calls despite earlier slippage from 2026. No news there.

The news is in the vocabulary. The company's keyword profile for this quarter is topped by free interval — platinum-free interval — which did not exist as a top-30 theme in any prior quarter in the dataset. Simultaneously, Ovarian trial registers as one of the sharpest conceptual decliners of the quarter. Compugen has not stopped running an ovarian trial; it has stopped leading with one. The pitch now runs through “the proportion of patients who remain progression-free at 6 months will be an important measure because it reflects the opportunity to extend the platinum-free interval” — Eran Ophir, President and CEO · 2026-11-09

That is a meaningful, and defensible, shift in how success gets defined. A six-month landmark is a lower bar than a hazard ratio, and it lets management lean on the placebo arm assumption they have been walking down all year. In May the benchmark was 5.5 months with a 3-month improvement hypothesis; by August it was “we've adjusted it to approximately 4 months” — Michelle Mahler, Chief Medical Officer · 2026-08-03; today two recent European studies in similar-but-heavier populations reportedly showed control-arm median PFS under three months. Every revision makes the separation easier to demonstrate. Ophir is candid that this is assumption, not knowledge: “The actual outcome eventually will depend on the characteristics of the enrolled and treated populations on the study.” — Eran Ophir, President and CEO · 2026-11-09 Investors should treat the placebo number as an input to the framing, not to the trial.

Also gone from the top keywords: PD L1, which ranked second as recently as Q2 2026. That tracks the disclosure that MAIA is not stratified by PD-L1 and that responses have been seen in both PD-L1 positive and negative patients — a differentiated-biology claim, but also a confession that there is no enrichment biomarker to fall back on if the data disappoints.

UnigenX: a new modality, zero candidates

The genuinely new corporate asset this quarter is UnigenX, an extension of the AI powered discovery engine into novel tumor associated antigens for T-cell engagers and other immune activation modalities. It is a real strategic broadening — from checkpoint biology into targeted immune activation — and it lands on a market that has re-rated on exactly this theme. Note the global tape: anti PD 1 and immune checkpoint inhibitor names are among the strongest twelve-month and ninety-day clusters, and small cell lung cancer has broad participation from the biotech complex. Compugen is riding a live wave, not inventing one.

But the substance lagged the slideware. When Leerink's Bill Ling asked directly for timing on new clinical candidates, the answer was mechanistic, not calendrical: “the challenge we are taking ourselves is to exploit all our experience in the platform we built along years, Unigen, now using UnigenX to identify more such targets” — Eran Ophir, President and CEO · 2026-11-09 — illustration via DLL3, no named asset, no IND date. That is consistent with the company's stated preference to disclose early assets only when de-risked, but it means UnigenX is currently narrative optionality rather than a catalyst.

The balance sheet does the patient work

Financially this was a quiet quarter, and quiet is the point. Revenue of roughly $3.8 million versus $1.9 million a year ago is Gilead upfront and IND-milestone amortization, not a commercial business. R&D held flat at about $5.8 million and G&A at $2.3 million, producing a $5.8 million net loss, or $0.06 per share. Against roughly $117.5 million in cash, deposits and marketable securities, CFO David Silberman reiterated “Cash runway, assuming no further cash inflows is expected to fund our operating plans into 2029” — David Silberman, Chief Financial Officer · 2026-11-09

For a roughly $200 million market-cap name, that runway buys the entire catalyst set without dilution risk: MAIA data, the Biliary01 readout, and a Gilead-partnered Phase I that remains entirely opaque — “due to the nature of the collaboration with Gilead, we are unfortunately unable to give guidance on the specifics of the stage we are” — Eran Ophir, President and CEO · 2026-11-09. That opacity is now a two-year-old feature, not a bug; the same non-answer was given in August.

The story, then, is a micro-cap that has bought itself time and is using it to move the goalposts in two directions at once — forward on AstraZeneca's calendar, sideways on its own endpoint language. The Q1 2027 MAIA readout will not be judged against a hazard ratio; it will be judged against whether six-month progression-free numbers clear a placebo bar that management has spent two quarters ratcheting lower. That is a cleaner, more binary test than investors had three months ago — and a more forgiving one.