Hanmi Pharmaceutical has signed an exclusive global licensing agreement with Genentech, a member of the Roche Group, for HM17321, turning an early-stage non-incretin obesity programme into a transaction worth up to approximately $2.3 billion before the candidate has produced human efficacy data. Genentech receives exclusive development, manufacturing and commercialization rights outside South Korea, while Hanmi will receive $190 million upfront and can earn additional development, regulatory and commercial milestones that take the potential contract value to about $2.3 billion, plus tiered royalties on future sales. Hanmi will complete the ongoing Phase 1 programme before Genentech assumes development beginning with Phase 2.
The economics reveal how Roche is valuing differentiation rather than another conventional GLP-1 asset. Only $190 million is guaranteed upfront, leaving more than $2 billion of the headline consideration dependent on HM17321 progressing successfully through human development and commercialization. The candidate is a long-acting urocortin-2 analog designed to activate corticotropin-releasing factor receptor 2, or CRFR2, rather than the GLP-1, GIP or glucagon receptors that dominate the current obesity-development landscape. Hanmi’s proposition is that the mechanism can reduce fat mass while preserving or potentially increasing lean tissue, but those body-composition claims are still derived from animals and remain one of the central questions the human programme must resolve.
Why is Genentech paying heavily for an obesity drug that does not target appetite through GLP-1?
The extraordinary efficacy of modern incretin medicines has changed the competitive question in obesity. Developers no longer need merely to demonstrate that a drug can reduce weight; they increasingly need to explain how it could improve the quality, durability or tolerability of that weight loss. One area receiving particular attention is body composition, because clinically important weight reduction generally includes some loss of lean tissue alongside adipose tissue. The amount and consequences vary between patients and studies, and lean-mass reduction should not automatically be equated with harmful muscle wasting, but the issue has created room for mechanisms that might shift more of the total weight reduction toward fat.
HM17321 approaches obesity through CRFR2-selective UCN2 signaling. Hanmi’s preclinical rationale links UCN2 biology with energy metabolism, skeletal-muscle effects and adipose-tissue catabolism rather than primarily suppressing food intake. In diet-induced obese mice, HM17321 produced dose-dependent body-weight reduction and greater fat-mass loss than semaglutide in the reported experiments, while lean mass increased rather than declined. Similar body-composition patterns were observed in obese rats, and monkey experiments showed weight reduction driven predominantly by fat-mass loss. These are intriguing cross-species findings, but comparison with semaglutide in animals does not establish equivalent or superior human weight loss, nor does an observed lean-mass increase prove a clinically meaningful improvement in strength or physical function.
That distinction helps explain why Roche’s stated interest extends beyond weight alone. The company is building a broader cardiometabolic portfolio and has described HM17321 as a potential approach to reducing fat while improving muscle mass and function. Genentech is effectively acquiring the option to test whether a mechanism developed around body recomposition can complement a market already populated with very powerful appetite-regulating therapies. The scientific opportunity is substantial precisely because the clinical proof is still absent.
What does the current Phase 1 study actually test?
HM17321 is already in a randomized, double-blind and placebo-controlled Phase 1 study expected to enroll approximately 90 participants. Part A plans about 40 healthy volunteers across five single-ascending-dose cohorts, while Part B plans around 50 people with obesity across five multiple-ascending-dose cohorts. Participants in the obesity portion receive subcutaneous HM17321 or placebo once weekly for 12 weeks, allowing investigators to evaluate safety, tolerability, pharmacokinetics and pharmacodynamic responses across a treatment interval long enough to begin observing metabolic and body-composition signals.
The trial started in November 2025 after FDA IND clearance and is expected to continue into 2027 under the current ClinicalTrials.gov schedule. Its primary role is dose finding rather than proving an obesity indication. Investigators need to establish how human exposure changes with dose, whether the candidate’s long-acting design supports predictable once-weekly administration and whether CRFR2 activation produces cardiovascular, endocrine, gastrointestinal or other adverse effects before development expands into much larger efficacy studies.
The multiple-dose cohort will nevertheless provide the first opportunity to see whether the central body-composition proposition translates at all. If patients lose fat while maintaining lean mass, Genentech would have an early clinical reason to pursue the differentiated mechanism aggressively. If weight reduction proves modest or muscle effects disappear in humans, a large portion of the scientific argument behind the transaction would need to be reconsidered.
Could HM17321 be more valuable in combination with incretin drugs than as a stand-alone obesity medicine?
Hanmi has deliberately explored combination strategies rather than presenting HM17321 solely as a GLP-1 competitor. In obese mice, combining HM17321 with Hanmi’s experimental GLP-1/GIP/glucagon triple agonist produced greater reductions in body weight and fat mass than either compound alone while increasing skeletal-muscle weight and improving measured muscle function. Switching animals from the incretin programme to HM17321 also maintained fat loss while preserving lean mass in the reported experiments.
The strategic implication is potentially more important than direct monotherapy competition. Incretins can deliver strong appetite suppression and major total weight reduction, while a UCN2-based therapy might theoretically alter the composition or maintenance of that loss through different peripheral metabolic mechanisms. A combination capable of retaining incretin-level efficacy while reducing unwanted lean-tissue loss could become commercially attractive even if HM17321 itself never produces the largest stand-alone weight reduction in the market.
Combination therapy raises a much higher evidentiary bar, however. Adding another injectable medicine increases cost, treatment burden and the opportunity for toxicity. Human trials would need to show that the additional body-composition or metabolic benefit is sufficiently large to justify a second mechanism rather than simply demonstrate that two drugs can be administered together.
Why does the $190 million upfront payment matter more than the $2.3 billion headline?
The $2.3 billion number reflects potential value rather than money already transferred to Hanmi. Genentech is committing $190 million upfront, while the overwhelming majority of the transaction depends on HM17321 reaching development, regulatory and commercial milestones. That structure allows Roche to secure worldwide rights outside South Korea before human efficacy data potentially increase the asset’s value, while limiting the amount of capital exposed if the Phase 1 or subsequent trials fail.
For Hanmi, the transaction validates an asset that remains early enough for Genentech to shape the later development programme while allowing the Korean company to complete the first human study it has already initiated. Tiered royalties preserve additional upside if the drug eventually reaches commercialization.
The deal also illustrates how competitive obesity business development has become. A company no longer needs a Phase 3 candidate to attract billion-dollar potential economics if it can offer a credible biological answer to a problem the first generation of blockbuster obesity medicines has exposed.
What would turn HM17321 from an expensive experiment into a differentiated obesity franchise?
Human body-composition evidence comes first. Genentech and Hanmi will need to show not merely that participants lose weight but how much comes from adipose tissue, what happens to lean mass and whether any apparent muscle preservation corresponds to function rather than imaging measurements alone. Larger studies will then need to compare efficacy and tolerability with contemporary obesity therapies rather than historical placebo performance.
Cardiometabolic outcomes could broaden the opportunity further. Hanmi is positioning HM17321 for obesity-associated conditions including Type 2 diabetes and cardiovascular disease, while preclinical experiments have shown effects on glucose and lipid-related measures. Those observations provide hypotheses for human development, not established clinical benefits.
Genentech has therefore purchased an unusually large option on a fundamentally different way of treating obesity. The transaction recognizes that the next phase of the market may reward more than the largest percentage shown on a weight-loss graph. HM17321’s $2.3 billion potential value depends on whether its animal-model advantage survives the transition into people, where fat loss, muscle preservation and meaningful physical function all have to coexist in the same patient.
