AstraZeneca and Daiichi Sankyo said Friday that they will test Datroway, the TROP2-directed antibody drug conjugate they develop together, in combination with ivonescimab, the PD-1/VEGF bispecific antibody that Summit Therapeutics licenses from the Chinese company Akeso. The first study will be a Phase 3 trial in first-line triple-negative breast cancer.

The terms describe something other than a standard drug partnership. Each company contributes its own medicine. The trials will be sponsored by AstraZeneca or Daiichi Sankyo, not by Summit. All three companies share the cost. And each retains development and commercial rights to its own drug, which means Summit keeps ivonescimab and AstraZeneca keeps no piece of it at all.

It is the third arrangement of this kind between the two companies in a week, and it sits on top of a $2 billion equity investment AstraZeneca made in Summit in late September. Read together, the transactions describe a company paying a large sum for access to a molecule it does not own, cannot control, and cannot yet value, and structuring every document so that it never has to.

Two billion dollars bought a place in line, not a share of the drug

The equity investment, disclosed Sept. 28, is for roughly 109,000 shares of convertible preferred stock convertible into common at a ratio of one to a thousand, at an effective common price of $18.36. That gives AstraZeneca rights equivalent to about 12 percent of Summit's outstanding common stock, or about 10.6 percent on a fully diluted basis, at a premium of roughly 10 percent to Summit's five-day volume-weighted average price. Summit's shares rose sharply on the news, and AstraZeneca's barely moved.

What the filing did not contain is the part that matters to anyone modeling the drug. There are no royalties, no milestone payments, and no revenue or profit sharing. The arrangements are explicitly non-exclusive. AstraZeneca did not buy a share of ivonescimab's future sales, which is what a licensing deal would have given it. It bought a share of the company that owns the drug outside China, plus a contractual claim on the data that comes out of trials it helps pay for.

That distinction is not accounting trivia. A royalty rate is a price. To set one, both sides have to agree on what the drug will sell, in which indications, at what net price, against what competition. Ivonescimab's American commercial value depends on a regulatory decision that has not been made, on a data package that has already missed its most important endpoint once, and on a market where the drug has never been prescribed. There is no defensible number to negotiate against, so the parties did not negotiate one. They priced the equity and left the drug unpriced.

AstraZeneca's oncology business is already close to half the company, and it sells its own checkpoint inhibitor, durvalumab, alongside a conjugate portfolio it has said it wants to widen. Susan Galbraith, the company's executive vice president for oncology research and development, described the strategy as "harnessing the power of immunotherapy and antibody drug conjugate combinations" to produce responses that last. Paying for access to a competitor's PD-1/VEGF antibody is not a replacement for either. It is a hedge that costs trial money instead of a licensing premium.

The Phase 3 will be run by the companies that do not own the drug

Sponsorship is the quiet allocation of power in the announcement. A sponsor writes the protocol, picks the sites, owns the statistical analysis plan, controls the data, and holds the regulatory correspondence. Every one of those decisions determines what the trial can prove and what the label can say. In the arrangement announced Friday, those decisions belong to AstraZeneca or Daiichi Sankyo.

Summit supplies the drug and a share of the cost, and receives something it could not easily buy: a first-line breast cancer program run on someone else's infrastructure. Maky Zanganeh, Summit's president and co-chief executive, described the collaboration as evidence of "increasing conviction in ivonescimab's potential as a foundational next-generation immunotherapy," and framed the breast cancer study as an expansion into a tumour type the company had not been developing on its own.

The same disclosure carried a separate collaboration covering sonesitatug vedotin, AstraZeneca's Claudin-18.2-directed conjugate, in gastrointestinal cancers, and a non-binding memorandum of understanding for further combinations with other AstraZeneca cancer medicines. A non-binding memorandum costs nothing to sign and nothing to abandon, which is a useful reminder of how much of what AstraZeneca has promised Summit is optional.

Datroway is already approved in this cancer, for the patients immunotherapy cannot reach

Triple-negative breast cancer has no hormone receptor and no HER2 amplification to target, which is why it has been the hardest of the three breast cancer subtypes to treat and why it has attracted the most antibody drug conjugates. Datroway is approved in more than 30 countries for unresectable or metastatic triple-negative disease in adults who are not candidates for PD-1 or PD-L1 inhibitor therapy, on the strength of the TROPION-Breast02 trial.

That population is defined by exclusion. It is the group immunotherapy cannot help, which is precisely why a TROP2 conjugate was tested there and why the approval is narrower than the headline indication suggests. A first-line trial moves the drug to the opposite end of the treatment sequence, where the patients are immunotherapy candidates and where pembrolizumab plus chemotherapy has been the standard for the PD-L1-positive subset since 2020.

Datroway carries two other approvals. It is cleared in more than 45 countries for HR-positive, HER2-negative breast cancer after endocrine-based therapy and chemotherapy, based on TROPION-Breast01, and in the United States and several other countries for EGFR-mutated non-small cell lung cancer after prior EGFR-directed therapy and platinum chemotherapy, based on the TROPION-Lung05 and TROPION-Lung01 trials. That last approval is contingent on a confirmatory trial in the United States, which means one of the three indications the combination strategy leans on is not yet fully earned.

The program behind the drug runs to more than 20 trials, including eight Phase 3 studies in lung cancer and five in breast cancer. Adding a sixth breast cancer study on a partner's balance sheet is cheaper than running it internally, which explains why Daiichi Sankyo's research chief, John Tsai, described the collaboration in terms of maximising an existing portfolio rather than opening a new one.

Ivonescimab's American approval is six weeks away and the survival question is open

The FDA accepted Summit's biologics license application for ivonescimab in combination with chemotherapy in EGFR-mutated non-small cell lung cancer after third-generation TKI therapy on Jan. 29, and set a goal action date of Nov. 14, 2026. It is the first application the agency has accepted for a PD-1/VEGF bispecific antibody.

The submission rests on HARMONi, a global Phase 3 trial that randomised 438 patients to ivonescimab plus pemetrexed and carboplatin or to placebo plus the same chemotherapy. Median progression-free survival was 6.8 months against 4.4 months, a hazard ratio of 0.52 with a p value below 0.0001, and the benefit held across subgroups including patients with brain metastases. Overall survival did not reach statistical significance. Reported hazard ratios were 0.79 at the primary analysis and 0.78 in September 2025, and the FDA had asked for a statistically significant survival result in this setting.

Summit later shared an updated analysis, with a June 2026 data cutoff, showing an overall survival hazard ratio of 0.76 in the Western subgroup of 165 patients alongside a result in the full trial population that matched it. Subgroup analyses are not the primary endpoint, and a major amendment submitted to a pending application can extend the review by three months, which would push a decision past the November date. The agency has approved drugs in this setting on progression-free survival before, including amivantamab in 2024 and datopotamab deruxtecan itself in 2025, so a negative survival result is not automatically disqualifying.

Roughly 14,000 patients a year in the United States are estimated to be eligible in the post-TKI setting the application covers. That figure describes the market the application addresses, not the market the collaboration announced Friday is aimed at. The breast cancer program depends on an approval that has not happened, for an indication the drug has not been tested in. Every document signed this week is a bet placed before that first answer arrives.

The molecule belongs to someone else and the deal locks nobody in

Ivonescimab was engineered by Akeso, which sells it in China under a different name. Summit's rights cover the rest of the world, which is what gives the company its value and also what fixes its ceiling. A United States approval is not one asset among several for Summit. It is the asset, and the company's own disclosures describe the drug as investigational everywhere outside China.

The class is crowded. Several companies are developing bispecifics that pair checkpoint blockade with VEGF inhibition, on the theory that blocking the vessel-forming pathway changes the immune microenvironment in a way checkpoint blockade alone does not. Ivonescimab's advantage is timing: it is approved and selling in China, and its American application is ahead of any competitor's. Timing is not a moat, and the non-exclusive structure leaves Summit free to sign the same molecule to another partner.

AstraZeneca has spent a decade building an antibody drug conjugate platform and has said publicly that combinations, not single agents, are the next step. Buying 12 percent of the company that owns the most advanced PD-1/VEGF antibody is a way to be in those combinations without acquiring the company, without booking its losses, and without taking on the obligation that a licensing deal would create. It is a position, sized in dollars, that can be expanded or left alone.

What the combination has to prove is that two mechanisms beat one

The scientific premise is straightforward. Datroway delivers a topoisomerase I inhibitor payload into cells that express TROP2, a protein present on most triple-negative tumours. Ivonescimab blocks PD-1 to release the brakes on T cells and blocks VEGF to normalise tumour vasculature. The theory is that the conjugate kills cells in a way that provokes an immune response, and the bispecific makes that response larger and more durable than either mechanism manages alone.

First-line triple-negative breast cancer is a demanding place to test that theory. The control arm will be an active regimen, not a placebo, and the trial will have to show the combination beats what patients already receive. Datroway's existing approval in this cancer sits in the patients who are not immunotherapy candidates, which means the drug has not yet been shown to add to a checkpoint inhibitor. Ivonescimab has not been approved anywhere outside China. The trial is therefore being asked to validate two unproven propositions at once, and it will be run by the two companies whose revenue depends on the conjugate rather than the one whose valuation depends on the antibody.

None of that makes the combination a bad bet. Deeper responses in a disease with no targeted option would be worth a great deal, and the partner that owns a lung cancer franchise and a breast cancer franchise is better placed to run the study than the partner that owns one antibody. What the structure reveals is how the parties priced the uncertainty. AstraZeneca paid for the company and not for the drug. Summit accepted trial sponsorship by its partners in exchange for a tumour type it could not fund alone. Both sides left the royalty question for a day when there is a number to argue about, and the November decision is the first thing that will produce one.

Primary sources

  1. AstraZeneca, AZ and DS collaboration with Summit for Datroway, regulatory news service announcement, Oct. 2, 2026, for the collaboration terms, the Phase 3 first-line triple-negative breast cancer plan, the sponsorship and cost-sharing provisions, the executive statements and the Datroway and ivonescimab approval and development detail.
  2. Summit Therapeutics, Summit Therapeutics Announces Clinical Trial Collaboration with AstraZeneca and Daiichi Sankyo to Evaluate Ivonescimab in Combination with TROP2-Directed ADC Datroway, Oct. 2, 2026, for Summit's account of the agreement and its development program.
  3. AstraZeneca, AstraZeneca Makes $2 Billion Strategic Equity Investment in Summit Therapeutics, Sept. 28, 2026, for the equity terms, the share conversion ratio and effective price, the ownership percentage, and the statement that the arrangements carry no royalties, milestones or revenue sharing.
  4. BioPharma Dive, AstraZeneca invests $2B in Summit, gaining a seat at the table for PD-1/VEGF drug push, for the market reaction, the analyst assessment of the deal structure and the competitive context for the PD-1/VEGF class.
  5. Summit Therapeutics, Summit Therapeutics Announces U.S. FDA Acceptance of Biologics License Application Seeking Approval for Ivonescimab in Combination with Chemotherapy in Treatment of Patients with EGFRm NSCLC Post-TKI Therapy, Jan. 29, 2026, for the application's acceptance, the November 2026 goal action date and the eligible patient estimate.
  6. CancerNetwork, FDA Accepts BLA for Ivonescimab Regimen in Pretreated EGFR-Mutated NSCLC, for the HARMONi progression-free and overall survival results, the subgroup analyses and the regulatory history of the application.