Retatrutide Phase 2: Obesity

Matched-maintenance arms (ID 2 vs ID 4) converge by wk 16–24 regardless of starting step — the trajectories are organized by maintenance dose, not titration speed. Steady-state receptor occupancy tracks current exposure, not the path taken to it.

Body weight curves separate from placebo within the first weeks; waist circumference separates later and continues diverging through wk 48. Early weight loss includes water and glycogen shifts from reduced intake; waist tracks adipose volume reduction, which requires time on dose for lipolysis and hepatic fatty-acid oxidation to mobilize stored fat.

At wk 4, weight reduction outpaces waist reduction in every arm; by wk 24 and wk 48 the two metrics converge. The early gap is intake-driven fluid and glycogen loss; the late convergence reflects adipose volume catching up as lipolysis mobilizes stored fat across sustained exposure.

Dose ordering holds within each BMI band and within each sex stratum. Females show larger proportional weight loss than males at every active dose — lower body weight produces higher per-kg exposure at the same nominal dose, and there is likely estrogen amplification of hypothalamic GLP-1R appetite suppression.

The ≥30% threshold is reached almost only at 8 and 12 mg. Depth-of-loss responders cluster at these doses because GLP-1R climbs into the steep part of its dose-response curve and GCGR engagement scales high enough to layer hepatic and adipose lipid mobilization onto the appetite suppression already established at lower doses.

Any-AE incidence is high across all active arms because most events are GLP-1R/GIPR-mediated GI and autonomic effects every dose produces. Serious-AE rates stay low and dose-flat, and discontinuation clusters at the slow-titration high-dose arms where cumulative step shocks across the climb produce sustained low-grade burden.

Only the 8 mg / ID 4 arm holds nausea prevalence elevated through wk 48; every other arm peaks early and declines. The brainstem houses a GIPR-driven inhibitory circuit that suppresses GLP-1R-triggered nausea, but the connection between the two populations only forms if both fire at low intensity in parallel for several weeks first. Jumping straight to 4 mg drives the nausea circuit hard before that connection establishes; once the window closes, slow titration later cannot recover it.

Arrhythmia and hyperesthesia rates distribute idiosyncratically across active arms, not as a clean dose gradient. Palpitations track the sinoatrial-node chronotropic signal (GLP-1R + GCGR cAMP at the funny channel), but perception threshold is individual — measurable HR rise does not produce reported palpitations in every user. Hyperesthesia mechanism is less characterized — a small-fiber sensory effect without clear dose-scaling.

Peak heart-rate elevation appears at wk 24 (+10 bpm at 12 mg) and falls to +5–6 bpm by wk 48. The attenuation is not pharmacological — primate data at matched chronic exposure shows no receptor-level decline — it reflects downstream cardiovascular adaptation as 15–25 kg of weight loss lowers blood pressure and restores vagal tone. Users who do not lose that much weight (maintenance dosing, low-BMI use) should expect the early HR elevation to persist rather than fade.

At matched maintenance, the ID 4 arm shows pulse elevation earlier — the first dose at 4 mg drives an acute GLP-1R/GCGR Cmax jump the ID 2 arm only reaches after stepping. By wk 24 both arms converge because steady-state receptor occupancy is determined by the dose held, not the path taken. The 4→8 mg step in the 8 mg comparison adds its own transient pulse rise visible in the ID 4 trajectory.

ABPM is ambulatory blood pressure monitoring, measured continuously over 24 hours.

Ambulatory heart rate peaks at wk 24 across active arms; by wk 36 the elevation has begun to attenuate — most clearly in the 12 mg arm where the peak was highest. The receptor-level chronotropic signal does not decline at this exposure; the attenuation reflects ongoing cardiovascular adaptation as blood pressure drops and vagal tone recovers downstream of weight loss.

ABPM is ambulatory blood pressure monitoring, measured continuously over 24 hours.

Heart rate rises while both systolic and diastolic pressure fall — a divergence that distinguishes this cardiac signal from sympathomimetic stimulation, which raises both. The chronotropy is direct cAMP signaling at the sinoatrial-node funny channel; norepinephrine and HRV are unchanged. The BP drop reflects weight-loss-driven vasodilation and reduced vascular tone, independent of the SA-node pathway.

MASLD (metabolic dysfunction-associated steatotic liver disease) indicates excess liver fat tied to metabolic dysfunction. MRI-PDFF is the scan used to measure liver-fat fraction.

In this subcohort, 12 mg measured 86% liver-fat reduction at week 48; 93% met the steatosis-resolution threshold. The driver is sustained low-fractional GCGR engagement: native glucagon engages this receptor in brief fasting/stress pulses, while retatrutide holds the same engagement amplitude continuously, producing hepatic fatty-acid oxidation drive that pulsed glucagon does not. Participants without baseline liver-fat elevation show no change at any dose — there is no substrate to mobilize.

Liver volume drops dose-dependently at wk 24 and continues falling through wk 48. The volume reduction is the structural readout of the fat reduction — hepatocytes shrink as lipid droplet content clears under sustained GCGR-driven fatty-acid oxidation.

VAT and ASAT both fall with dose at wk 24 and continue through wk 48. VAT drops proportionally more than ASAT because visceral adipocytes are more lipolytically responsive — they release stored fat at higher rate per unit caloric deficit and respond more strongly to the GCGR-driven hepatic pull on circulating substrate.

ALT and AST fall in parallel with liver-fat reduction across active arms. Both enzymes leak from hepatocytes under lipotoxic stress — when triglyceride accumulation overwhelms cellular oxidative capacity, membrane integrity declines and cytosolic enzymes spill into circulation. As GCGR-driven fatty-acid oxidation clears the substrate load, the spill normalizes.

Body weight, waist, VAT, and ASAT all correlate with liver-fat reduction at wk 48, but the slopes differ. Body weight and waist track LF most tightly because they integrate the full calorie-deficit and adipose-mobilization response; VAT correlates particularly because it drains directly into the portal vein, so visceral loss cuts the lipid flux delivered to the liver and reduces the substrate driving steatosis.

FGF21 (fibroblast growth factor 21) is a hormone the liver produces during metabolic stress.

FGF21 falls dose-dependently at wk 24 and wk 48, with the deepest reductions at 8 and 12 mg. FGF21 is a hepatokine the liver releases under metabolic stress — predominantly oxidative and ER stress from lipotoxic burden; as fatty-acid oxidation clears the substrate load and cellular stress drops, production declines in parallel with liver fat and the transaminases.

eGFR (estimated glomerular filtration rate) comes from the obesity-cohort supplementary arm-level week 48 table. UACR (urine albumin-to-creatinine ratio) is omitted because the available extract collapses arms differently, so it would not match this split-titration chart.

eGFR rises in the 8 and 12 mg arms while placebo and 1 mg stay flat — an unusual direction for a metabolic drug, where weight loss typically preserves rather than improves filtration. The rise is GCGR-mediated: glucagon receptor engagement at the renal vasculature alters glomerular hemodynamics directly, separate from the weight-loss-driven UACR drop. Cystatin-C confirmation rules out muscle-mass artifact.

Both dimers are inhibitor signals on lipid clearance. ANGPTL3/8 (hepatic origin) suppresses lipoprotein lipase in muscle and heart, restricting fat oxidation; ANGPTL4/8 (adipose origin) regulates fatty-acid uptake into storage. Retatrutide’s GCGR engagement reduces both dose-dependently, restoring lipid flux from circulation into oxidation — the upstream driver of the triglyceride and LDL responses shown elsewhere in this section.

Triglycerides fall most steeply — the direct downstream readout of ANGPTL3/8 suppression and restored lipoprotein-lipase activity in muscle and heart. LDL-C and PCSK9 fall less steeply but in parallel: PCSK9 normally targets hepatic LDL receptors for degradation, so dropping PCSK9 raises receptor density on hepatocytes and increases LDL clearance from circulation.