From Selling Molecules to Selling Platforms: Hansoh's USD 15.5B+ ADC Deal with Roche as China's Out-Licensing Tops USD 130B in Nine Months

Hansoh Pharma announced a platform-level ADC partnership with Roche on 19 October: USD 400M upfront, USD 110M on the one-year anniversary and more than USD 15B in potential milestones, a total exceeding USD 15.5B and the second-largest China out-licensing deal on record. The subject is no longer a single molecule but two preclinical programmes plus multiple Roche-proposed targets developed on Hansoh's own ADC platform. In the same window, METiS partnered with Lilly on an AI-driven extrahepatic LNP platform and Hutchmed closed a USD 110M upfront, while the NDRC reported roughly USD 130B of innovation-drug technology licensing in the first nine months.

On the evening of 19 October, Hansoh Pharma (03692.HK) announced a strategic collaboration and licence agreement with Roche. The structure: USD 400M upfront, USD 110M payable on the one-year anniversary, more than USD 15B in potential development, regulatory and commercial milestones, plus tiered royalties on future sales — a potential total above USD 15.5B, the second-largest out-licensing deal by a Chinese innovative drug company. The next day the stock rose as much as 9% intraday and closed up 8.32% at HK$34.38, taking market capitalisation back above HK$200B to HK$208.5B.

The subject changed from a molecule to a platform

Unlike a typical single-asset licence, this collaboration covers a portfolio of ADC programmes: two of Hansoh's preclinical ADC projects plus multiple projects against targets proposed by Roche, developed using Hansoh's proprietary ADC technology platform. Under the split, Hansoh leads early development for each programme up to completion of global Phase Ia at the latest, Roche takes over development and commercialisation outside Greater China, and Hansoh retains all Greater China rights. Hansoh may also designate up to two programmes to participate in overseas development, earning additional US commercial economics on those two in return for its investment.

Two features of the terms carry the real information. First, the payment structure prices a platform option: USD 400M upfront for molecules that have not yet entered the clinic and for platform capability rather than clinical data. Second, Hansoh keeps all Greater China rights and can opt into overseas development economics, meaning it has not exited downstream investment in these programmes — which preserves its negotiating position but also commits it to domestic clinical supply and downstream CMC for those assets, demand that ultimately lands with CDMOs holding ADC conjugation and high-potency capacity.

A second deal within a year, ten times the size

This is the second collaboration between the two companies in twelve months. In October 2025, Hansoh licensed ex-Greater China rights to HS-20110, a CDH17-targeting ADC, for USD 80M upfront, up to USD 1.45B in milestones and sales royalties, a total of about USD 1.53B. The jump from USD 1.53B to USD 15.5B is not simply a better asset: the counterparty now pays for a platform's ability to keep producing molecules, not for one candidate. Multinational interest in Chinese ADC assets is shifting from buying a molecule with attractive clinical data to locking in the design and process capability that generates them.

Three deals in three days, one direction

Between 19 and 21 October, platform-level and early-stage licensing landed in clusters. METiS announced a strategic collaboration with Lilly to develop next-generation extrahepatic LNP delivery using its AI-driven NanoForge / AiLNP platform, with an upfront above USD 10M and up to USD 1.1B in milestones — among the largest AI-driven extrahepatic LNP platform deals globally. On 21 October, Hutchmed closed its licence deal with GSK for HMPL-A830, receiving a USD 110M upfront plus up to USD 1.185B in development, regulatory and commercial milestones — for an asset still in Phase Ia. HMPL-A830 is an antibody-targeted therapeutic conjugate (ATTC) pairing an EGFR antibody with a KRAS small-molecule inhibitor payload. A Phase Ia molecule attracting a USD 110M upfront means the buyer is paying for molecular design and conjugation-delivery logic, not clinical data.

At the aggregate level, the National Development and Reform Commission said at its 20 October press conference that China's innovation-drug technology out-licensing reached roughly USD 130B in the first nine months of 2026, close to the whole of last year. The half-year data is more granular: according to Industrial Securities, China recorded 100 out-licensing deals in H1 2026 with USD 5B in upforwards and USD 99.7B in total deal value, all records for the period; for January-September the count reached 158 deals (+60% YoY) with USD 129.3B in disclosed total value (+33%).

Hard validation on the demand side: 29 Chinese late-breakers at ESMO

Licensing valuations ultimately rest on data. At the 2026 ESMO Congress on 23-27 October, 29 Chinese innovative-drug studies were selected as late-breaking abstracts — roughly one third of the total, a record — with five earning places in the highest-format presidential sessions. Guangda Securities reads this as a key window on the industry inflection, with companies shifting from a 'R&D spending narrative' to a 'product revenue plus milestone delivery' logic. For CROs and CDMOs, every link in that conversion chain requires outsourcing: registration batches, clinical supply, comparators and biomarker testing all have to be completed before the readout.

Three layers of transmission to CRO and CDMO demand

Platform-level licensing transmits to outsourcing demand in three thicker layers than single-asset deals. The first is funnel expansion: platform collaborations mean multiple programmes advancing in parallel, concentrating demand for early process development, analytical methods, stability and toxicology batches within a single year — and these orders are far stickier than single-asset work, because once a process is locked early, subsequent clinical phases and commercial supply usually stay with the same provider. The second is where Greater China capacity sits: Hansoh retains all Greater China rights, so domestic clinical supply and eventual commercial manufacturing for these programmes must run through Chinese capacity, further sharpening scarcity in ADC conjugation and high-potency suites. The third is that milestone realisation depends on CMC execution: whether USD 15B of milestones is ever paid depends on programmes reaching IND, pivotal trials and launch on schedule, and the biggest variable in any schedule is process scale-up and batch consistency — which is why buyers now weight schedule certainty alongside price when selecting a CDMO.

The sober side deserves equal space. First, milestone payments are conditional; the USD 15.5B headline and actual cash flows may differ by an order of magnitude, and only USD 510M of upfront and anniversary payments is currently recognisable, under 4% of the nominal total. Second, ADC target crowding is already high: whether Roche-proposed targets can be differentiated depends on how fast Hansoh's linker and payload platform iterates, not on the number of molecules. Third, the repeat-business character of platform deals cuts both ways — the deeper the client lock-in, the more directly a single client's pipeline volatility hits a CDMO's utilisation.

Taken together, the shape of China's innovation-drug out-licensing has changed in 2026: from selling molecules to selling platforms, from selling late-stage assets to selling early design capability, from one-off transactions to repeat business with the same counterparty inside a year. For CROs and CDMOs the implication is order front-loading — licensing now happens before the clinic, so process development moves earlier and revenue recognition shows up sooner in early-stage outsourcing. If you want a leading indicator for 2027 CXO orders, the repeat rate of multinationals licensing Chinese early-stage platforms may be more reliable than any single deal value.