British pharmaceutical giant GSK and Chinese innovative drugmaker Hutchmed have reached a licensing agreement worth up to $1.295 billion for a drug candidate that has not yet entered human clinical trials. The core significance of this deal: GSK has, for the first time, given top-tier multinational pharmaceutical endorsement to a novel drug technology paradigm defined by a Chinese company—ATTC (Antibody-Targeted Conjugate).

On September 3, the two parties announced that a GSK subsidiary will acquire global rights to HMPL-A830 (excluding Greater China), while Hutchmed retains development and commercialization rights in mainland China, Hong Kong, Macau, and Taiwan. The deal structure includes a $110 million upfront payment plus up to $1.185 billion in development, regulatory, and commercial milestone payments.

HMPL-A830 is a KRAS-EGFR antibody-targeted conjugate designed to deliver a KRAS small-molecule inhibitor as its payload, carried on an anti-EGFR antibody. Hutchmed plans to initiate a global Phase I clinical development program (NCT07718581) in the second half of 2026. The study will include both dose escalation and dose optimization components, enrolling patients with confirmed, unresectable, advanced or metastatic solid tumors who are refractory or unresponsive to standard therapy.

Hesham Abdullah, GSK’s Global Head of Oncology R&D, stated: “The dual KRAS-EGFR mechanism of HMPL-A830 has the potential to significantly improve upon current standard of care.”

From “Poisoned Arrows” to “Drug-Loaded Darts”

The fundamental difference between ATTC and ADC (Antibody-Drug Conjugate) lies in the philosophy of the payload. Traditional ADCs use highly toxic cytotoxins as “warheads”—while they precisely hit tumors, the blast fragments still damage surrounding normal tissue, causing severe side effects such as bone marrow suppression, neurotoxicity, and interstitial lung disease. This has forced most ADCs to retreat to second- and third-line treatment settings.

Hutchmed’s ATTC platform retains the classic antibody-linker-payload ternary structure but replaces the payload from “cytotoxin” to “signaling pathway inhibitor”—for example, PI3K/PIKK inhibitors or KRAS inhibitors. These small-molecule targeted drugs, which are limited by systemic toxicity in traditional oral administration, theoretically release only within tumors once equipped with the antibody’s precision navigation system, maintaining low exposure in normal tissues.

Mechanistically, the design of HMPL-A830 has inherent logical coherence. KRAS mutations are common driver mutations in colorectal cancer, lung adenocarcinoma, and pancreatic cancer, but the core obstacle limiting KRAS inhibitor monotherapy efficacy lies in EGFR-mediated feedback activation: when KRAS is inhibited, EGFR compensatorily upregulates, reactivating the MAPK pathway. Clinically, KRAS G12C inhibitors combined with EGFR inhibitors have improved objective response rates in colorectal cancer from 19% to 46%, but at the cost of systemic combination toxicity.

HMPL-A830’s approach is to have the anti-EGFR antibody simultaneously serve as the delivery vehicle for the KRAS inhibitor: after the antibody binds to EGFR-positive tumor cells, the KRAS inhibitor is released in lysosomes while EGFR signaling is blocked. This reduces systemic KRAS inhibition exposure on one hand, while EGFR blockade reduces KRAS GTP loading on the other, theoretically enhancing KRAS inhibitor binding.

KRAS is the most prominent subtype of the RAS family, mutated in approximately 44% of colorectal cancers, 34% of lung adenocarcinomas, and up to 89% of pancreatic ductal adenocarcinomas. Mutations across the entire RAS family drive approximately 15% to 20% of human cancers. Hutchmed stated that clinical development of HMPL-A830 will initially focus on these three cancer types.

Hutchmed’s ATTC Pipeline

HMPL-A830 is the third drug candidate from Hutchmed’s ATTC platform and the first molecule licensed to a global partner. The previous two are both in clinical stages:

CandidatePayload TypeAntibody TargetClinical StageHMPL-A251PI3K/PIKK inhibitorAnti-HER2Phase I/IIa (initiated simultaneously in China and US, December 2025)HMPL-A580PI3K/PIKK inhibitorAnti-EGFRPhase I/IIa (global initiation, March 2026)HMPL-A830KRAS inhibitorAnti-EGFRPreclinical (Phase I to start H2 2026)

Note: Preclinical data for HMPL-A251 showed activity in Enhertu-resistant cell lines; HMPL-A580’s design differentiation lies in simultaneously blocking EGFR signaling and acting on the PAM pathway via the PI3K/PIKK payload. The ATTC format mitigates the hyperglycemic toxicity of PI3K inhibitors, but the metabolic risk of this payload has not been eliminated—it has merely been redistributed to the delivery stage.

Two Readings of the $110 Million Upfront Payment

For an asset that has not yet entered human clinical trials, a $110 million upfront payment is notably high. This can be interpreted in two ways: first, GSK is buying priority positioning in the KRAS space for the ATTC platform, along with an option on this technology concept; second, the $1.185 billion in milestone payments is entirely tied to future clinical milestones, with actual payment contingent on development success. In other words, the “total value” of this deal is more like an installment-based option menu than immediate value.

There is another easily overlooked detail in the announcement: GSK holds a right of first negotiation on another of Hutchmed’s early-stage ATTC candidates. This is a carefully designed option clause—if HMPL-A830 clinical data is positive, GSK has incentive to deepen collaboration; Hutchmed, in turn, locks in an out-licensing channel for subsequent ATTC assets.

From a financial perspective, as of the end of 2025, Hutchmed had cash reserves (cash plus short-term investments) of approximately $1.4 billion. The $110 million upfront payment will further extend its cash runway, supporting the global clinical advancement of HMPL-A251 and HMPL-A580.

A Shift in How Chinese Innovative Drugs Are Valued

Within Hutchmed’s own out-licensing history, this deal carries landmark significance. In 2023, Hutchmed licensed fruquintinib—already approved for marketing in China—to Takeda Pharmaceutical for a total of $1.13 billion, with a $400 million upfront payment. That was a model of “mature product going global”—a clinically validated, approved molecule exchanged for global market rights.

The 2026 GSK deal follows a completely different logic. HMPL-A830 has not yet started Phase I clinical trials; Hutchmed is selling a “pre-proof-of-concept” platform asset, and the $1.295 billion total deal value has already exceeded fruquintinib’s $1.13 billion. This signals that the valuation logic for Chinese innovative drugs is shifting tracks: from “looking at clinical data” to “looking at platform potential,” from “selling the fruit” to “selling the tree.”

In the broader industry context, this deal is not an isolated case. In the first half of 2026, total out-licensing deal value for Chinese innovative drugs reached approximately $99.7 billion, with China occupying 8 of the top 10 global BD deals; first-quarter upfront payments grew 124% year-over-year, and multinational pharmaceutical companies’ upfront payments for China BD deals accounted for 52% of their total global BD upfront payments. CSPC Pharmaceutical Group’s $18.5 billion platform collaboration with AstraZeneca, Hengrui Pharmaceuticals’ $15.2 billion multi-pipeline deal with BMS, and Innovent Biologics’ $10.5 billion collaboration with Pfizer together paint a clear picture: Chinese innovative drugs are leveraging platform power to transform from followers into rule-makers.

For GSK, in-licensing HMPL-A830 also aligns with the R&D strategy announced by new CEO Luke Miels in July 2026. That strategy plans to save $2.5 billion by 2029, with most of it reallocated to pipeline building, including bringing in external assets through deals, regular portfolio reviews, and accelerating late-stage development programs.

The existing competitive landscape of the KRAS inhibitor market also provides context for this deal. Amgen’s Lumakras/Lumykras (sotorasib) is the highest-selling KRAS inhibitor, reaching $363 million in 2025; BMS’s Krazati (adagrasib) grew sales 62% to $205 million. But the limitations of monotherapy efficacy and the toxicity of combination therapy leave clear clinical unmet need for ATTC’s “one molecule, dual mechanism” design.

The potential of ATTC extends far beyond these three candidates. PI3K/PIKK inhibitors can be swapped for other signaling pathway inhibitors, KRAS inhibitors can be expanded to other RAS family inhibitors, and antibody targets can extend from HER2 and EGFR to more tumor-associated antigens. The versatility of the platform means ATTC is a forest that can keep growing, not a few isolated trees.