Genentech agreed to pay Hanmi Pharmaceutical up to $2.3 billion for an obesity drug that has not finished its first human trial.
The deal, announced Monday, gives Genentech (part of the Roche Group) global rights, minus South Korea, to a peptide called HM17321. Hanmi gets $190 million up front and stands to collect the rest through development, regulatory, and sales milestones, plus royalties. Hanmi will run the Phase 1 trial it was cleared to start; Genentech takes over from Phase 2 on. Details are in Hanmi's licensing announcement ↗ and were reported by Pharmaceutical Technology ↗.
Why this one is not another GLP-1 drug
Almost every obesity drug on the market or in late testing works the same way. Semaglutide (Wegovy), tirzepatide (Zepbound), and the triple-agonist retatrutide all copy incretins, the gut hormones that blunt appetite after a meal. HM17321 does not. It is an engineered version of urocortin-2, a natural peptide the body already makes, and it switches on a different receptor entirely, the corticotropin-releasing factor 2 receptor (CRFR2). That receptor sits on muscle and heart tissue, not just the appetite circuits in the brain.
The pitch built on that difference is body composition. When people lose weight on a GLP-1 drug, a large share of what they shed is not fat. In Eli Lilly's SURMOUNT-1 trial of tirzepatide, about a quarter of the total weight lost was lean tissue, mostly muscle, and across the GLP-1 class the figure runs as high as 40 percent, a pattern reviewed in Diabetes, Obesity and Metabolism ↗. Hanmi is positioning HM17321 as a drug that drops weight while sparing muscle. If that holds up, it would matter most for older patients and anyone already at risk of frailty.
The part the money is ahead of
None of the muscle-sparing story has been shown in people yet. Hanmi received a green light from the US Food and Drug Administration to begin a Phase 1 trial only in November 2025, and Phase 1 measures safety and dosing, not weight loss and not muscle. The lean-mass claim rests on animal and lab work. CRFR2 is also tied to stress-hormone signaling and cardiovascular effects, which is exactly the kind of biology that can produce off-target surprises once a drug reaches a real dosing study.
So this is a bet on a mechanism, not a readout. What Genentech bought is a differentiated shot at the one weakness the incretin drugs are still trying to patch, backed by preclinical data and a first-in-class receptor rather than human numbers.
That mechanism is the reason it is worth watching. peptidemodel hosts a card for the parent peptide, urocortin-2 ↗, alongside the incretin drugs it aims to differentiate from: semaglutide ↗, tirzepatide ↗, and retatrutide ↗. The gap between them is not appetite. It is what is left standing after the weight comes off.