Four Hormones, One Failure Mode

A pattern worth naming

Between 1994 and 2017 the field identified leptin, characterised PYY(3-36)'s anorectic action, established ghrelin as the one circulating orexigenic hormone, and found GFRAL as the receptor for GDF15. Each was reported as a landmark, each produced striking data in animals, and none has produced a licensed weight-management medicine. Meanwhile a hormone known since the 1980s, GLP-1, produced two. The gap between those outcomes is not luck, and understanding it is the most useful thing a reader can take from this part of the literature.

Failure mode one: the signal is already loud

Leptin is the clearest case. It reports stored energy, and in obesity there is a great deal of stored energy, so circulating leptin is high. Administering more of a signal that is already elevated and being disregarded achieves little, and the founding 1995 paper already contained the demonstration - db/db mice had elevated levels and did not respond. GDF15 may have a version of the same problem arriving from a different direction: the evidence that prior exposure blunts the acute response suggests a system that adapts to sustained elevation rather than continuing to respond to it. A hormone that habituates is a poor foundation for chronic therapy.

Failure mode two: redundancy

The systems defending against energy deficit are old, layered and duplicated, for the obvious reason that failing at them was historically fatal. Removing one orexigenic signal, as ghrelin antagonism attempts, leaves the rest of that architecture intact and free to compensate. Adding one anorectic signal, as PYY administration attempts, presents a system built to counteract exactly that. Redundancy is why single-target approaches to appetite have repeatedly produced an effect that shrinks over weeks, and it is a substantial part of the argument for multi-receptor agonists.

Failure mode three: aversion arrives first

The circuits that read circulating appetite signals sit in the hindbrain, close to or overlapping the machinery that produces nausea. That is not a coincidence of anatomy - it reflects that reducing food intake and rejecting food are related functions. The practical consequence is that increasing the dose of an appetite-suppressing signal tends to produce nausea before it produces the full available reduction in intake, which caps the achievable exposure below the level the pharmacology could otherwise reach. This has been dose-limiting for PYY, it defines the GDF15 problem, and it is the reason incretin dosing is titrated slowly rather than started at target.

What the incretins actually did differently

The incretin class did not succeed by replacing a deficient hormone. It succeeded by delivering an exposure far above anything physiology generates, held there continuously by half-life engineering, at a receptor that does not appear to develop the resistance leptin encounters. Slow titration manages the aversive component rather than eliminating it - the gastrointestinal adverse event rates in every trial in this class make that plain. And the more recent compounds attack redundancy directly by engaging two or three receptors at once. Those are engineering answers to each of the three failure modes, not a discovery that GLP-1 was the right hormone all along.

How to read the next announcement

When a new appetite hormone or receptor is announced, four questions separate the interesting cases from the rest. Is the human population deficient in this signal, or already replete? Does the response persist on sustained exposure, or adapt? What compensates when this pathway is engaged? And how much separation is there between the dose that reduces intake and the dose that produces nausea? These have determined the outcome for every candidate so far, and they are answerable long before Phase 3. None of the compounds discussed here is supplied on this site, and licensed medicines in this area are prescribed and monitored by clinicians.

Quick reference