Cabinet I · The Programme

The Trial Programme, Laid Out as Plates

Twelve plates from the published record — mechanism first, then the SURPASS programme, the SURMOUNT programme, and the four newer trials that extend the cabinet beyond glycaemia and weight.

Inscribed · May MMXXVI · Cabinet I

Baroque engraving of an ornate gilded cabinet displaying nine plaques with symbolic ornamental glyphs
Cabinet II — the trial programme, set as a gilded cabinet of nine plaques.

Plate I · The Molecule

Tirzepatide, in atomic detail

Tirzepatide is a thirty-nine-amino-acid linear peptide, formula C225H348N48O68, molecular weight 4810.52 daltons, CAS number 2023788-19-2. It is conjugated to a C20 fatty diacid moiety via a γ-glutamate linker at lysine position twenty, carries α-aminoisobutyric acid substitutions at positions two and thirteen to resist proteolytic cleavage, and is C-terminally amidated. In solution it self-associates in a monomer-trimer-hexamer equilibrium and is predominantly α-helical in secondary structure.

Cryo-electron-microscopy studies of receptor-ligand complexes have shown that tirzepatide engages the GIP receptor in a fully active conformation while adopting a distinct binding mode at the GLP-1 receptor that supports biased signalling — a structural account that aligns the molecule's pharmacological signature with its atomic geometry[13].

Plate II · Dual, Imbalanced, Biased

The pharmacological signature

The phrase used in the seminal pharmacology paper is precise: tirzepatide is an imbalanced and biased dual agonist[4]. Imbalanced means its affinity at the two incretin receptors is not equal — at GIPR the affinity approximates native GIP; at GLP-1R it is approximately five-fold weaker than native GLP-1. Biased means that, at GLP-1R, the molecule preferentially activates the cyclic-AMP pathway over the beta-arrestin pathway, which reduces receptor internalisation and sustains downstream signalling.

In transfected cells and primary islet preparations this biased profile correlated with enhanced glucose-stimulated insulin secretion and reduced beta-arrestin recruitment[4]. The mechanistic story extends downstream — glucose-dependent insulin secretion from pancreatic β-cells, suppression of glucagon at α-cells, delayed gastric emptying, hypothalamic appetite suppression via central GLP-1R and GIPR circuits, and effects on adipose-tissue lipolysis and lipid handling that appear to operate beyond what weight loss alone would predict.

Plate III · SURPASS-1

Monotherapy in type 2 diabetes

SURPASS-1 was a double-blind, randomised, placebo-controlled phase-three trial of tirzepatide monotherapy in adults with type 2 diabetes inadequately controlled by diet and exercise. Across the three approved doses — five, ten, and fifteen milligrams subcutaneous once weekly — HbA1c reductions of 1.87 to 2.07 percent and body-weight reductions of seven to 9.5 kilograms were observed against placebo at forty weeks, without an increased hypoglycaemia signal[1]. The trial established tirzepatide's monotherapy profile and was the first of a sequence that would extend across SURPASS-2 through 6.

Plate IV · SURPASS-2

Head-to-head with semaglutide one milligram

SURPASS-2 placed tirzepatide directly against once-weekly semaglutide one milligram as add-on therapy to metformin. At forty weeks, tirzepatide at all three doses was noninferior to semaglutide, and the ten- and fifteen-milligram doses were superior on both HbA1c and weight outcomes; the fifteen-milligram arm produced nearly twice the weight loss of semaglutide one milligram[2]. The trial was the first direct demonstration that a dual incretin agonist could outperform a market-leading single-receptor GLP-1 agonist on the standard endpoints of the indication.

SURPASS-2 · HbA1c and weight at 40 weeks

Adapted from Frias et al., NEJM 2021 [2]. Approximate dose-arm means at 40 weeks; consult the primary publication for exact estimates and confidence intervals.

Plate V · SURPASS-4

Cardiovascular-risk population

SURPASS-4 enrolled adults with type 2 diabetes and elevated cardiovascular risk and compared tirzepatide against insulin glargine. Tirzepatide produced superior HbA1c and weight reductions and a numerically lower rate of major adverse cardiovascular events; no increased MACE signal was observed against insulin glargine over the trial's median follow-up[12]. A post-hoc kidney-outcomes analysis of the same trial subsequently reported an approximately fifty-percent lower risk of a composite kidney endpoint (eGFR decline ≥ 40 percent, renal death, kidney failure, or new macroalbuminuria), an effect of roughly two millilitres per minute per 1.73 square metres per year against insulin[10].

Plate VI · SURMOUNT-1

Obesity at seventy-two weeks

SURMOUNT-1 enrolled adults with obesity or overweight with at least one weight-related comorbidity, excluding type 2 diabetes. At seventy-two weeks, mean body-weight reductions were sixteen, 21.4, and 22.5 percent at five, ten, and fifteen milligrams, against 2.4 percent on placebo. Eighty-nine to ninety-six percent of treated participants achieved at least a five-percent reduction, against twenty-eight percent on placebo[3]. A 176-week post-hoc analysis subsequently estimated a 4.6 to 9.2 percent reduction in ten-year predicted atherosclerotic-cardiovascular-disease risk, dose-dependently, against an increase in the placebo arm[15].

Plate VII · SURMOUNT-5

Head-to-head with semaglutide 2.4 milligrams

Published in 2025, SURMOUNT-5 was the head-to-head trial that the field had been waiting for: tirzepatide versus semaglutide 2.4 milligrams in adults with obesity without diabetes. At seventy-two weeks, mean weight reduction was 20.2 percent with tirzepatide and 13.7 percent with semaglutide — a 47 percent relative difference. Thirty-one-and-a-half percent of tirzepatide-treated participants achieved at least twenty-five percent weight loss, against sixteen percent on semaglutide, and gastrointestinal-related discontinuations were numerically lower with tirzepatide[9].

Baroque Vanitas allegory of a gilded balance-scale weighing two engraved laurel wreaths
Plate VII allegory — comparative weight as a baroque Vanitas, the larger laurel hanging lower on the gilded balance.

Plate VIII · SUMMIT

Heart failure with preserved ejection fraction

SUMMIT enrolled adults with heart failure with preserved ejection fraction (HFpEF) and obesity. Over a median of two years on up to fifteen milligrams once weekly, tirzepatide lowered the composite risk of cardiovascular death or worsening heart failure against placebo, improved Kansas-City-Cardiomyopathy-Questionnaire scores, reduced left-ventricular mass by approximately eleven grams, and reduced paracardiac adipose-tissue volume by approximately forty-five millilitres[6]. The trial is the first to demonstrate a cardiovascular-event benefit from incretin therapy in the HFpEF population.

Plate IX · SURMOUNT-OSA

Obstructive sleep apnoea

Two parallel phase-three trials, reported jointly as SURMOUNT-OSA, enrolled adults with moderate-to-severe obstructive sleep apnoea and obesity. At fifty-two weeks, tirzepatide at ten or fifteen milligrams reduced apnoea-hypopnoea index by approximately twenty-five to twenty-nine events per hour, against roughly five events per hour with placebo, across both PAP-using and non-PAP cohorts. Reductions in body weight, hypoxic burden, high-sensitivity CRP, and systolic blood pressure tracked alongside the AHI improvement[7].

SURMOUNT-OSA · AHI reduction at 52 weeks

Adapted from Malhotra et al., NEJM 2024 [7]. AHI = apnoea-hypopnoea index; PAP = positive airway pressure.

Plate X · SYNERGY-NASH

Metabolic dysfunction-associated steatohepatitis

SYNERGY-NASH enrolled adults with biopsy-proven metabolic dysfunction-associated steatohepatitis (MASH) and stage two or three fibrosis. At fifty-two weeks, 51.8, 62.8, and 73.3 percent of participants on tirzepatide five, ten, and fifteen milligrams achieved resolution of steatohepatitis without worsening fibrosis, against 13.2 percent on placebo; gastrointestinal adverse events were the most common findings, consistent with the broader programme[8]. Preclinical mouse work in MASLD models has separately reported reductions in hepatic CD36 and OBP2A expression and decreased hepatic lipid accumulation, suggesting hepatic lipid-handling effects beyond those attributable to weight loss alone[14].

SYNERGY-NASH · MASH resolution without worsening fibrosis at 52 weeks

Adapted from Loomba et al., NEJM 2024 [8]. Percent of participants meeting the primary endpoint.

Plate XI · Caveat

Recurring safety signals

Across the programme, the most common adverse events have been gastrointestinal — nausea, diarrhoea, vomiting, and constipation — dose-dependent, predominantly during titration, and largely mild to moderate. GI-related discontinuations are the leading reason participants leave trials[17]. Acute pancreatitis has been reported as a rare event without a clear dose-dependence. A boxed warning addresses the risk of thyroid C-cell tumours, derived from rodent two-year carcinogenicity studies; the human relevance is unconfirmed but the contraindication is unambiguous in subjects with a personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia type 2.

Plate XII · Animal and Mechanistic

Beyond the human programme

Beyond the human programme, tirzepatide has been examined in preclinical models that probe mechanism rather than indication. In a rat MPTP model of Parkinson's disease, tirzepatide preserved dopaminergic neurons and improved motor performance at ten nanomoles per kilogram intraperitoneal — a finding consistent with the broader neuroprotective literature on incretin receptor agonism[16]. The mechanistic neuroprotection story is preliminary and well short of any clinical claim, but it illustrates the breadth of the receptor systems engaged.

“An imbalanced and biased dual incretin agonist.” — Willard et al., JCI Insight, MMXX