Weight That Won't Respond to Willpower: Two Real Drugs for a Hunger the Surgery Left Behind
A single patient, two years past craniopharyngioma resection, with severe hyperphagia-driven obesity that hasn't responded to any lifestyle intervention. The disagreement is which of two real pharmacologic options — one repurposed, one built specifically for this problem — fits her mechanism better.
O.J., a 24-year-old woman, was two semesters from finishing her nursing degree when a craniopharyngioma, diagnosed after months of headaches and progressive vision changes, required surgical resection that necessarily involved the hypothalamus itself, not just the pituitary — the tumor's own growth pattern along the hypothalamic-pituitary axis left her surgeons no clean plane to spare it. She has since gained 62 pounds over two years despite what her food diary, kept meticulously for the past six months at her endocrinologist's request, shows as genuinely calorie-restricted intake; her own account, consistent with the hyperphagia hypothalamic injury classically produces, is a hunger that doesn't resolve after eating and returns within an hour regardless of meal size, a drive she describes as "not the same feeling as being hungry before, it doesn't turn off."
That distinction — hunger that doesn't turn off, rather than ordinary appetite — maps onto real, different pharmacology than typical obesity, and onto a genuine current choice between two mechanistically distinct drugs. GLP-1 receptor agonists work through the gut-brain axis, amplifying satiety signals that arrive at the hypothalamus after eating; real-world cohorts and small prospective studies in craniopharyngioma-related hypothalamic obesity, most substantially with semaglutide, have shown clinically meaningful weight loss, though generally more modest than semaglutide's effect in ordinary obesity, plausibly because the signal it amplifies still has to be received by hypothalamic circuitry the surgery itself damaged. Setmelanotide, an MC4R agonist approved by the FDA for acquired hypothalamic obesity in March 2026 on the strength of the phase 3 TRANSCEND trial (Miller et al., New England Journal of Medicine, 2026), works downstream of that damaged relay, directly activating the melanocortin-4 receptor pathway that GLP-1 signaling would otherwise have to reach through an already-injured hypothalamus — a mechanistic case for a larger, more reliable effect in patients like her, with trial results showing substantially greater BMI reduction than GLP-1 cohorts have reported, at the cost of a drug that is new, expensive, requires daily injection, and carries two warnings added to its label at that same approval which describe her specifically rather than the average trial participant — acute adrenal insufficiency in acquired-HO patients who already have secondary adrenal insufficiency, reported in 5% of treated patients against none on placebo, and sodium imbalance in those with concomitant AVP deficiency. Her panhypopituitarism and her desmopressin put her inside both of those groups, not outside them.
Two drugs, two different broken signals, one patient
I'd start with semaglutide rather than setmelanotide. It's the more established option in this population at this point, real-world and small prospective cohorts have shown clinically meaningful weight loss specifically in craniopharyngioma-related hypothalamic obesity, and starting with the more familiar drug lets us establish a response before committing to a newer agent with a thinner track record.
I'd make the case for setmelanotide first, precisely because of her mechanism. Her hunger doesn't turn off — that's a description of a broken melanocortin pathway, and setmelanotide was built to activate that exact pathway directly, downstream of the hypothalamic damage a GLP-1 signal still has to pass through on its way to a receptor field the surgery injured. The phase 3 trial's own results, a placebo-adjusted BMI reduction of roughly twenty percentage points, are a materially larger effect than GLP-1 cohorts in this same population have reported.
I don't think "more established" should outweigh "mechanistically matched" here — semaglutide's modest results in hypothalamic obesity, compared to its much larger effect in ordinary obesity, is itself evidence that the damaged relay it depends on is limiting how well it can work in patients like her.
That's a real point about the mechanism, and given her early metabolic complications — glucose and LDL already trending the wrong way — I'd rather start with the drug more likely to produce a larger effect than default to the more familiar one for its own sake. I'd start setmelanotide directly — but not on the adverse-event profile you'd quote to an average patient. The hyperpigmentation and injection-site reactions are the common ones and she should hear about them, but they are not what makes this drug risky in her. The March 2026 label carries two warnings written for her exact situation: acute adrenal insufficiency in acquired-HO patients who already carry secondary adrenal insufficiency — 5% of treated patients in the trial, against none on placebo — and sodium imbalance in patients with concomitant AVP deficiency. She has both deficits. So her adrenal axis gets formally re-evaluated before the first dose rather than assumed adequate because she is already replaced and stable, she goes home with explicit sick-day glucocorticoid rules, and her sodium gets followed as her intake falls — which is the whole point of the drug.
Agreed: start setmelanotide given the mechanistic match to her specific hyperphagia phenotype and the substantially larger effect size shown in its pivotal trial for this exact population, with adrenal function formally re-evaluated before the first dose, sick-day glucocorticoid rules given explicitly, and serum sodium followed as her intake falls — the two label warnings that name her own deficits — alongside counseling on the common injection-site reactions and skin hyperpigmentation. Semaglutide held in reserve if setmelanotide is not tolerated or access becomes a genuine barrier.
Both physicians converged on setmelanotide as the better first choice once the mechanistic reasoning was laid out fully; the endocrinologist's initial preference for the more established drug was revised specifically in light of her early metabolic complications, which argued for prioritizing effect size over familiarity.