Mirabegron (YM-178)

Identity and approval status

Mirabegron is a small molecule, not a peptide. PubChem gives it CID 9865528 and the formula C21H24N4O2S. The FDA Global Substance Registration System record under UNII MVR3JL3B2V carries CAS 223673-61-8, INN 8907, USAN code WW-31 and ChEMBL identifier CHEMBL2095212, and describes a single stereocentre in the (2R) configuration. Four molecular weights circulate. Three of them are one number rounded differently: 396.5 on PubChem, 396.5077 on the GSRS structure record, 396.51 on the approved label. ChEMBL returns a fourth, 396.52, for the same structure; that value is not a rounding of the GSRS figure, which rounds to 396.51, and reflects a different atomic-weight table. The development code used by the originating sponsor was YM-178.

Approval came on 28 June 2012, when the FDA cleared extended-release tablets under NDA 202611. The labelled indication is overactive bladder in adults with urge urinary incontinence, urgency and urinary frequency. The labelled maximum dosage is 50 mg once daily; the label's titration schedule is not reproduced here. A second product, extended-release granules for oral suspension, was approved under NDA 213801 on 25 March 2021; generic tablet labelling notes that paediatric information is approved for the innovator's product but withheld from generics under marketing exclusivity. The US register carries no approval for any metabolic indication, and no non-US register was searched for this page.

A ClinicalTrials.gov query for mirabegron as an intervention returned 193 studies when its API was called on 18 August 2026, and the registered volume is almost entirely urological. Narrowing the condition to obesity returned eight. Two of those eight carry a completed status: NCT02919176 at Kentucky with 39 participants and NCT01783470 at Beth Israel Deaconess with 15. The largest protocol in that set, NCT03049462 in the NIDDK intramural programme, has an estimated enrolment of 100 and was still recruiting, with a primary completion date of September 2026; it is registered under a condition of polycystic ovary syndrome, and it is the registration behind the fourteen-woman chronic report published in 2020, so it is an open protocol that has already produced published evidence rather than a future study. Nothing in the metabolic literature approaches the size of a registration trial.

Where the selectivity was measured

Takasu and colleagues, working at the originating sponsor, reported in 2007 that the compound, then YM-178, raised cyclic AMP in Chinese hamster ovary cells expressing the human beta-3 receptor with a half-maximal effective concentration of 22.4 nM, while the corresponding values at human beta-1 and beta-2 receptors were 10,000 nM or higher. Intrinsic activity relative to the maximal isoproterenol response was 0.8 at beta-3 and 0.1 at each of the other two. In isolated human bladder strips precontracted with carbachol the EC50 was 0.78 microM, against 0.28 microM for isoproterenol. Every figure in that paper comes from transfected cells or isolated tissue.

The approved label does not extend that selectivity across the whole dose range. Section 12.1 states that although the compound showed very low intrinsic activity at cloned human beta-1 and beta-2 receptors, results in humans indicate that beta-1 stimulation occurred at a dose of 200 mg. Exposure is not linear either: the label records that a fourfold increase from 50 mg to 200 mg raised peak concentration roughly 8.4-fold and area under the curve roughly 6.5-fold. That 200 mg figure is also the dose most of the metabolic trials used.

Noguchi and colleagues, working in mouse and guinea pig atria in 2026, reported concentration-dependent positive chronotropic effects with maxima of 75.9% in mice and 27.7% in guinea pigs. The selective beta-1 antagonist CGP20172A at 0.03 to 3 nM strongly and non-competitively antagonised that response in both species; the beta-2 antagonist ICI118551 at 30 nM produced a threefold rightward shift; the beta-3 antagonist L748337 antagonised it non-competitively with a Schild plot slope of 0.65. Those authors concluded the chronotropic effect is primarily beta-1 mediated, which is a statement about isolated atria in two rodent species rather than about a person.

What the acute trials measured

Cypess and colleagues screened fifteen healthy men with cold exposure and carried forward the twelve who had detectable brown adipose tissue, then gave each a single 200 mg oral dose and a placebo. Against placebo, brown adipose tissue glucose uptake on 18F-FDG PET/CT was higher in all twelve, median 132 with an interquartile range of 70 to 253 mL·SUVmean·g/mL (p = 0.001), and resting metabolic rate rose 203 ± 40 kcal per day, or 13% (p = 0.001). Heart rate rose 14 ± 3 bpm and systolic pressure 11 ± 2 mmHg, both at p = 0.002. That 2015 report is the founding trial of the metabolic literature, and it is that small.

Baskin and colleagues separated the doses three years later, giving twelve healthy men one-time randomised doses of placebo, the approved 50 mg and 200 mg, and reported a more-than-dose-proportional increase in brown adipose tissue metabolic activity, with medians of 0.0, 18.2 and 305.6 mL·SUVmean·g/mL. Only the 200 mg dose elevated both non-esterified fatty acids, by 68%, and resting energy expenditure, by 5.8%. The same paper described previously unreported effects elsewhere: a 35% increase in gallbladder size and reductions in conjugated bile acids, following receptor profiling that found beta-3 messenger RNA highly expressed in gallbladder as well as brown fat.

Nahon and colleagues ran a randomised, double-blinded crossover in ten lean Dutch South Asian men and ten age- and BMI-matched Europid men, twenty participants in total, comparing about two hours of cold exposure, a single 200 mg oral dose and placebo. Cold raised several serum lipid species in both groups while the drug raised only free fatty acids, and lipid oxidation rose with the drug in the Europid group alone. The two stimuli were separate arms of the crossover and were never administered together.

Gorini Pereira and colleagues placed eleven healthy adults, five of them women, in a whole-room calorimeter at 20 °C for six hours after 100, 150 or 200 mg. Cumulative energy expenditure was greater than placebo at every dose — 494 ± 75 kcal at 100 mg (p < 0.001), 481 ± 67 at 150 mg (p = 0.017) and 492 ± 69 at 200 mg (p = 0.001) against 456 ± 67 on placebo — with no separation between the three doses. Their brown fat measure, the area under the curve for supraclavicular skin temperature relative to sternal, was higher than placebo at 100 mg (12.54 ± 7.51 °C × min, p = 0.011) and at 150 mg (10.01 ± 6.05, p = 0.021) but not at 200 mg (9.17 ± 5.95, p = 0.067), against 4.74 ± 3.86 on placebo. That ordering runs against the pattern in the PET-based trials, where 200 mg is the dose that produced the signal.

The chronic trials, and what did not move

Chronic exposure appears across five published reports and none of them carried a placebo group: Finlin 2018 at ten weeks and 50 mg daily, Finlin 2020 at twelve weeks and 50 mg daily, the combination arm of Finlin 2021, O'Mara 2020 at four weeks and 100 mg daily, and Walker 2024 at 28 days and 100 mg daily. Two of those come from one Kentucky cohort family and two from one NIDDK protocol. Finlin 2018 and Finlin 2020 state ten and twelve weeks respectively for overlapping registrations, NCT02596776 and NCT02919176, and neither paper reconciles the difference; the duration of chronic exposure in that cohort is therefore recorded two ways in the published record.

O'Mara and colleagues treated fourteen healthy women of diverse ethnicities, mean age 27.5 ± 1.1 years and mean BMI 25.4 ± 1.2 kg/m2, with 100 mg daily for four weeks in an open-label study registered as NCT03049462 — the NIDDK protocol registered under a polycystic ovary syndrome condition, still recruiting toward an estimated 100. Brown adipose tissue metabolic activity, the primary endpoint, was higher after treatment, as was whole-body resting energy expenditure. Body weight and body composition did not change. Plasma HDL, ApoA1 and total bile acids rose, adiponectin was 35% higher at completion, and a frequently sampled intravenous glucose tolerance test showed higher insulin sensitivity, glucose effectiveness and insulin secretion.

Thirteen obese, insulin-resistant participants, the majority female, received 50 mg daily for twelve weeks in the Finlin trial, and all thirteen completed. Oral glucose tolerance improved (P < 0.01), haemoglobin A1c fell (P = 0.01), insulin sensitivity on euglycaemic clamp improved (P = 0.03) and beta-cell function improved (P = 0.01). Subcutaneous white adipose tissue showed stimulated lipolysis and more alternatively activated macrophages; skeletal muscle showed reduced triglycerides, higher PGC1A expression (P < 0.05) and more type I fibres (P < 0.01). The paper's own limitations section states the study was not placebo controlled.

Three further papers from that cohort narrow what the result covers. An earlier report established that ten weeks at 50 mg induced UCP1, TMEM26, CIDEA and phosphorylated hormone-sensitive lipase in the subcutaneous white adipose tissue of obese subjects. A 2021 arm added pioglitazone and found that the combination produced less beiging than either drug alone, and that neither induced brown adipose tissue in these participants; that trial randomised 39 participants across mirabegron, pioglitazone and combination groups, so a third arm received chronic mirabegron in combination and no arm received placebo. A 2025 transcriptomic follow-up from the same group opens by restating the 2020 result as an improvement in insulin sensitivity, beta-cell function and glucose tolerance obtained without weight loss and without a change in brown adipose tissue.

Which receptor is doing the work

Blondin and colleagues reported in 2020 that human brown adipose tissue thermogenesis is not mediated by beta-3 stimulation: oral mirabegron elicited increases only when ingested at the maximal allowable dose, which produced off-target binding at beta-1 and beta-2 receptors, raising cardiovascular responses and white adipose tissue lipolysis respectively. ADRB2 was co-expressed with UCP1 in human brown adipocytes, and pharmacological stimulation, pharmacological inhibition and knockdown of ADRB1, ADRB2 or ADRB3 in those cells all pointed to beta-2 signalling. The human arm was registered as NCT02811289 at Sherbrooke with an actual enrolment of 22, under a type 2 diabetes condition rather than an obesity one, which is why it does not appear in an obesity-condition registry query; per-experiment group sizes are not stated in the abstract.

A follow-up tested whether the cardiovascular response could be separated from the metabolic one. Dumont and colleagues ran a randomised crossover in eight lean men, giving 200 mg alone or together with 10 mg of the beta-1 antagonist bisoprolol. Against room temperature, mirabegron alone raised brown adipose tissue oxidative metabolism from 0.84 ± 0.46 to 1.79 ± 0.91 per minute (p = 0.0433), and the metabolic rate of glucose in that tissue was higher with mirabegron alone than with the combination, 24 ± 10 against 16 ± 8 nmol/g/min (p = 0.0284). Bisoprolol suppressed the rise in systolic pressure and heart rate and blunted the thermogenic response at the same time.

Ma and colleagues published a systematic review and meta-analysis in 2024, registered as CRD42023413446, that reviewed ten papers and pooled six. Brown adipose tissue volume showed no significant change (p = 0.72). Brown adipose tissue activity, non-esterified fatty acids, body temperature, resting energy expenditure, heart rate, diastolic pressure and blood insulin each increased significantly (p < 0.01). Systolic pressure did not reach significance in that same pooled analysis (p = 0.25), nor did blood glucose (p = 0.52); the paper's own abstract nonetheless lists systolic pressure among its significant increases while printing that p value beside it, so the source is internally inconsistent on this point. Volume and activity are separate quantities derived from the same scan, and the pooled record supports an increase in measured activity while not supporting an increase in measured volume.

The cardiovascular record at the doses the metabolic trials used

In a randomised, placebo- and moxifloxacin-controlled thorough QT study in 352 healthy subjects dosed for ten days, the label records maximum mean increases in heart rate against placebo of 6.7 bpm at 50 mg, 11 bpm at 100 mg and 17 bpm at 200 mg. Placebo-adjusted QTcI at four to five hours post-dose was 3.7, 6.1 and 8.1 milliseconds across the same doses; in women at 200 mg the mean effect was 10.4 milliseconds, upper 95% confidence bound 13.4. Twenty-four-hour average systolic pressure rose 3.0, 5.5 and 9.7 mmHg respectively. The only pooled human analysis this page cites disagrees with the label on that last measure: Ma 2024 found no significant pooled systolic effect (p = 0.25) against the label's dose-graded rise, and the disagreement is left standing here rather than averaged.

Those figures diverge sharply from what the same label reports in the approved population. Across three twelve-week placebo-controlled trials in overactive bladder at 25, 50 and 100 mg, mean systolic and diastolic increases against placebo were approximately 0.5 to 1 mmHg, and mean pulse rate change at 50 mg was about 1 bpm. Morning systolic pressure rose by at least 15 mmHg from baseline in 5.3% of placebo patients, 5.1% at 25 mg and 6.7% at 50 mg. The label carries a warning for increases in blood pressure and states the product is not recommended in severe uncontrolled hypertension, defined as 180 mmHg systolic or 110 mmHg diastolic and above.

Sui and colleagues reported in 2019 that oral administration of clinically relevant doses to apolipoprotein E knockout and LDL receptor knockout mice markedly accelerated atherosclerotic plaque growth and instability, by a mechanism they attributed to raised LDL cholesterol and very-low-density remnants. Deleting uncoupling protein 1 completely abrogated the effect, tying it to thermogenesis-driven lipolysis rather than to an off-target action. Group sizes are not stated in the abstract. No human study has measured an atherosclerotic endpoint after mirabegron, and both strains are engineered hyperlipidaemic models.

Pharmacokinetics and interactions

Absolute bioavailability on the label rises from 29% at 25 mg to 35% at 50 mg, peak concentration is reached at about 3.5 hours, and both peak concentration and exposure increase more than dose-proportionally above 50 mg. Terminal elimination half-life is approximately 50 hours in patients, steady state is reached within seven days of once-daily administration, and steady-state exposure is roughly double single-dose exposure. Volume of distribution at steady state is approximately 1670 L, plasma protein binding approximately 71%, and erythrocyte concentrations of radiolabelled drug were about twice those in plasma in vitro. Handling in the body is therefore unusual for an oral tablet.

Metabolism runs through several routes at once — dealkylation, oxidation, direct glucuronidation and amide hydrolysis — with two phase 2 glucuronides accounting for 16% and 11% of total exposure and neither active at the beta-3 receptor. In vitro work implicates CYP2D6 and CYP3A4, but in genotypically poor CYP2D6 metabolisers peak concentration and exposure were only about 16% and 17% higher than in extensive metabolisers, so those isozymes play a limited role in overall elimination. The compound is itself a moderate CYP2D6 inhibitor. Coadministration of 100 mg raised mean digoxin peak concentration by 29% and exposure by 27%.

What is not known

Chronic dosing has been reported from two cohorts, neither placebo controlled, with thirteen and fourteen participants and maximum exposures of twelve and four weeks — and one of those cohorts is described as ten weeks in one paper and twelve in another without the discrepancy being reconciled. Every other human dataset is a single-dose experiment in eight to twenty-two people. No trial has run long enough, or in enough people, to establish whether the metabolic changes recorded are durable, and none has measured a clinical outcome of any kind. Body weight and body composition were measured and did not change. Which receptor mediates the thermogenic effect is unresolved in the literature: the compound's class name says beta-3, the label states beta-1 stimulation occurs at 200 mg, and Blondin's 2020 cell and human work attributes brown adipose thermogenesis to beta-2. That dispute matters because the doses producing measurable brown fat activity on PET are two to four times the approved maximum, at which the label's own thorough QT study recorded a 17 bpm mean heart rate rise and a 9.7 mmHg twenty-four-hour systolic increase — though the one thermography-based trial found its brown fat measure separated from placebo at 100 and 150 mg and not at 200 mg, and the one pooled human analysis found no significant systolic effect at all. Long-term cardiovascular safety at those doses has not been studied; the label's safety database covers 25 to 100 mg. Populations across the metabolic trials skew narrow — lean young men in the acute studies, exclusively women in the chronic 100 mg study, whose protocol is registered under a polycystic ovary syndrome condition, mostly women in the 50 mg study — and the one trial that compared ethnic groups found the lipid oxidation response in Europid participants only. No published human study has measured atherosclerotic endpoints, and the only animal work addressing them found accelerated plaque growth. Mirabegron is an approved medicine for overactive bladder and nothing more; the US register carries no approval for any metabolic indication, and material obtained outside a pharmacy or a registered trial has not passed the identity, purity or content controls that apply to an approved product.

Questions