Tesofensine (NS-2330)

Identity, and a development history spanning three sponsors

Tesofensine's identity is not in dispute; its chemical class is another matter. PubChem holds one entry under the name — CID 11370864, CAS 195875-84-4, molecular formula C17H23Cl2NO, molecular weight 328.28 g/mol, InChIKey VCVWXKKWDOJNIT-ZOMKSWQUSA-N — and the FDA Global Substance Registration System carries the same structure under UNII BLH9UKX9V1 with INN number 8343. The molecule is a phenyltropane: an 8-azabicyclo[3.2.1]octane cage bearing a 3,4-dichlorophenyl group at position 3 and an ethoxymethyl group at position 2. That scaffold runs through the cocaine-analogue series of transporter ligands, which is why the compound appears in structural work on the dopamine transporter rather than in peptide chemistry. Material sold as a peptide is mislabelled at the level of chemical class.

Development began at NeuroSearch in Denmark under the code NS-2330. Boehringer Ingelheim ran the neurology programme, registering three phase 2 trials in Parkinson's disease and Alzheimer's disease between 2005 and 2007. NeuroSearch then took the compound into obesity, and the asset later passed to Saniona, which licensed Mexican and Argentine rights to Medix in 2016 and developed a fixed-dose combination with metoprolol. The FDA Office of Orphan Products Development records two designations for that combination, both to Saniona: Prader-Willi syndrome on 2 March 2021, and hypothalamic obesity on 21 July 2021. Both carry the status Not FDA Approved for Orphan Indication.

Approval status is straightforward to state and frequently misstated. Tesofensine holds no marketing authorisation from the FDA or the EMA. In Mexico, the COFEPRIS technical committee on new molecules issued a favourable opinion in February 2023, a non-binding step in the review of a new molecule rather than an authorisation. Saniona disclosed in November 2024 that Medix had not received approval, and in February 2025 that a revised dossier was being resubmitted. Krug and colleagues, writing in Drug Testing and Analysis in 2026, described the compound as still under regulatory review while being marketed online as a dietary supplement.

The trial the reputation rests on, and the notice attached to it

TIPO-1 is the trial that every secondary summary of this compound traces back to. Astrup and colleagues randomised 203 adults with a body-mass index of 30 to 40 across five Danish obesity management centres to tesofensine 0.25 mg, 0.5 mg or 1.0 mg, or placebo, once daily for 24 weeks alongside an energy-restricted diet. It was published in The Lancet in November 2008. Weight loss after 24 weeks was reported as 4.5%, 9.2% and 10.6% on the three ascending doses, against 2.0% on diet and placebo, the three figures described in the abstract as "greater than diet and placebo" (Astrup 2008, PMID 18950853). That phrasing does not settle whether they are total mean losses or amounts in excess of placebo. The same author group's Danish secondary publication of the trial renders them as reductions higher than placebo (PMID 19824222), and the absolute kilogram figures reported from the trial, 6.7, 11.3 and 12.8 kg against 2.2 kg on placebo, imply consistent baseline weights across the randomised arms only on that second reading. This page records the ambiguity rather than resolving it, publishes no derived placebo-subtracted value, and notes only that the largest figure belongs to the 1.0 mg arm.

In April 2013 The Lancet printed an expression of concern about that paper. It followed an inspection by the Danish Health and Medicines Authority of two of the five trial sites and of the sponsor, conducted in November 2011 under a random inspection programme. The inspectors recorded that informed consent had been delegated to non-medical personnel at one site, questioned whether blinding had held, and found the recording and assessment of adverse events by the contract research organisation to be incomplete. The authority's stated conclusion was that the published side-effect profile did not accord with the actual course of the trial.

Astrup and colleagues replied in The Lancet in July 2013 under the title Under-reporting of adverse effects of tesofensine. Their account was that monitors had wrongly instructed investigators not to register headache, migraine, stress and depression as adverse events where a participant had reported the condition before randomisation, and that this under-reporting extended across all centres. The efficacy data were not withdrawn and the paper has not been retracted. PubMed still types it as carrying an expression of concern, and that notice remains unresolved as of August 2026. Any weight-loss figure quoted from this trial carries the notice with it.

Weight loss was first recorded as an adverse event

Before obesity, the compound was a neurology candidate, and the neurology results were unpersuasive. Hauser and colleagues randomised 261 people with early Parkinson's disease of under five years' duration to 0.25, 0.5 or 1.0 mg or placebo for 14 weeks; adjusted mean differences in total UPDRS score were −0.7, −1.3 and −1.7, none of them significant. Rascol and colleagues ran the ADVANS study in advanced Parkinson's disease with motor fluctuations and reported a −4.7 point difference on UPDRS subscales II plus III at 0.5 mg and a 7.1% reduction in off time at 0.25 mg, while stating that no dose-response relationship for efficacy could be established.

Astrup and colleagues then pooled four of those randomised neurology trials, covering 740 participants on tesofensine and 228 on placebo, all dosed orally once daily for 14 weeks with no weight-loss programme attached. Weight change across the cohort was +0.5% on placebo and −0.5%, −0.9%, −1.8% and −2.8% at 0.125, 0.25, 0.5 and 1.0 mg. In the obese subgroup the changes were −0.2%, −1.7%, −1.6%, −1.5% and −3.7%, and the proportion reaching a 5% reduction rose from 2.1% on placebo to 32.1% at 1.0 mg. Heart rate rose by 2.1 to 6.8 beats per minute across the dose range.

The order of events is worth holding onto when reading claims about this compound. Its obesity programme exists because an unwanted effect in a failed neurology programme was large enough to redirect the asset. The dataset that established the weight signal was not designed to measure weight, ran for 14 weeks, and involved populations with neurodegenerative disease rather than otherwise healthy adults with obesity.

Transporter occupancy, and where the potency numbers came from

Appel and colleagues measured striatal dopamine transporter occupancy directly, using positron emission tomography with the radioligand [11C]βCIT-FE after multiple oral doses of 0.125 to 1 mg given over 8 to 12 days. Mean striatal occupancy varied with dose between 18% and 77%. A sigmoid Emax model fitted the relationship, with maximum achievable occupancy estimated at about 80% and half of that effect reached at roughly 0.25 mg and a plasma concentration near 4 ng/mL. The report does not state the number of subjects scanned in its indexed abstract.

Structural work arrived much later. Li and colleagues published cryo-electron microscopy structures of the human dopamine transporter bound to five triple reuptake inhibitors in Nature Communications in 2025. The tesofensine complex was resolved at 2.8 Å and captured the transporter in an outward-facing conformation, with the dichlorophenyl group in a T-shaped π-stacking interaction with Y156 on transmembrane helix 3. Alanine substitutions at F326 and S422 retained uptake activity while substantially reducing inhibitory potency. Coordinates were deposited as PDB 9J6S, with noradrenaline and serotonin transporter complexes deposited as 9VWR and 9VWS.

The reuptake potency figures that circulate for this compound do not reconcile, and their provenance varies sharply. Three sets of transporter IC50 values appear in the secondary literature, one of them the product of a decimal-point error in a 2025 paper reproducing a 2008 review table, and the review table itself cites a rat neurogenesis study that reports no such assay. Two figures do trace to a retrievable primary assay: that same 2025 structural paper reports its own measured IC50 of 9.12 nM for inhibition of [3H]dopamine uptake at the wild-type human dopamine transporter across three independent experiments, and the NeuroSearch tropane patent US 6,288,079, filed in 1997 and issued in 2001, tabulates rat striatal synaptosome uptake values for the same molecule. The conflicting-figures entry below records what each search returned.

The cardiovascular signal and the beta-blocker combination

Heart rate is the consistent finding across the human record. In TIPO-1, heart rate rose by 7.4 beats per minute in the 0.5 mg group against placebo, while systolic and diastolic blood pressure showed no significant increase at 0.25 mg or 0.5 mg. The pooled neurology analysis recorded dose-related heart-rate increases up to 6.8 beats per minute with no blood-pressure effect. Bentzen and colleagues examined the mechanism in conscious telemetrised rats, recording food intake and cardiovascular variables simultaneously, and reported that oral metoprolol at 10 to 20 mg/kg prevented the sympathetic cardiovascular effects while leaving the hypophagic response intact.

That rat experiment is the stated rationale for the fixed-dose combination. In a phase 2 trial in type 2 diabetes, registered as NCT02737891, 60 adults were randomised evenly to tesofensine 0.5 mg with metoprolol 50 mg or to placebo for 90 days across two centres. Results posted to ClinicalTrials.gov for NCT02737891 give a net difference in 24-hour mean heart rate of −3.8 beats per minute (95% CI −6.36 to −1.29, p=0.004) and a net body-weight difference of −3.5 kg (95% CI −4.65 to −2.30, p<0.0001). The HbA1c difference was 0.1 percentage points (p=0.724). No journal publication of that trial was located.

Huynh and colleagues published the combination trial in hypothalamic obesity in the European Journal of Endocrinology in 2022. Twenty-one adults, 16 of them female, were randomised to the combination or placebo for 24 weeks with diet and lifestyle counselling; 18 completed. Additional mean weight change against placebo was −6.3% (95% CI −11.3 to −1.3, p=0.017), and 8 of 13 on active treatment against 1 of 8 on placebo reached a 5% reduction. One participant had a serious adverse event attributed to the drug, an exacerbation of pre-existing anxiety leading to discontinuation. Sleep disturbance occurred in 50% against 13%, dry mouth in 43% against 0%.

Registered trials that reported nothing, and two that were withdrawn

A ClinicalTrials.gov intervention search returned 13 registered studies naming tesofensine, sponsored by Boehringer Ingelheim, NeuroSearch and Saniona. Five carry posted results. The largest obesity dataset in existence is not among them: the phase 3 programme run by Medix in Mexico, described by the sponsor as covering 372 patients with an average weight reduction near ten percent at 24 weeks, has no entry in ClinicalTrials.gov, no PubMed record and no Europe PMC record. Searches of both databases for tesofensine combined with phase 3 and with Mexico returned only reviews and unrelated trials.

The Prader-Willi syndrome trial illustrates what unpublished data can look like. Saniona registered a two-centre double-blind study of the combination in 18 participants, with results posted in February 2024 and no journal publication located. Among adults, six on active treatment showed a mean body-weight change of −4.09% against −0.38% in three on placebo, an ANCOVA least-squares mean difference of −5.4% with a 95% confidence interval from −12.3 to 1.5 and p=0.1045. Among adolescents at the lower 0.125/25 mg strength, both arms gained weight, and the difference was 0.7 percentage points with p=0.7422.

Two confirmatory phase 2 trials, one in Prader-Willi syndrome and one in hypothalamic obesity, were registered and then withdrawn on 9 December 2022 with zero participants randomised. Both registry records give the phase as PHASE2; the 2b label attached to them in sponsor material does not appear in the registry. The reason recorded in both entries is financial: the sponsor stated it was unable to complete the trials and that no safety concern was involved. The confirmatory human evidence for the two indications carrying FDA orphan designations was therefore never generated.

Anti-doping status and detection window

Tesofensine is named on the WADA Prohibited List. Extracting the text of the 2025 list directly from the published PDF places the entry in section S6.B, specified stimulants, between tenamfetamine and tuaminoheptane, which makes it prohibited in-competition only and a Specified Substance. The 2026 list could not be downloaded during preparation of this page; Krug and colleagues, writing in 2026, describe the compound as classified under S6 stimulants and prohibited in-competition only, which is consistent with the 2025 document.

That same 2026 paper is the only published human excretion study. Six volunteers ingested 483 µg of tesofensine sold as a dietary supplement, and urine was collected for up to 600 hours and analysed by liquid chromatography with high-resolution mass spectrometry. Four principal metabolites were identified: three dealkylated species and one hydroxylated and glucuronidated species. Peak urinary concentrations of 1 to 4 ng/mL were reached between 4 and 46 hours, with marked variation between individuals, and detection windows extended to 500 hours at a limit of detection of 0.01 ng/mL. The authors reported that recommended dosing was unlikely to produce urinary concentrations reaching an Adverse Analytical Finding under the stimulant minimum reporting levels applicable at the time of writing (PMID 42320973). The paper does not address therapeutic equivalence.

The long detection window follows from the pharmacokinetics. Lehr and colleagues fitted a population model to 1,969 tesofensine and 1,714 metabolite concentrations from 320 people with Alzheimer's disease on multiple oral dosing, and reported apparent half-lives of 234 hours for the parent compound and 374 hours for the M1 metabolite. Female participants with a creatinine clearance of 35.6 mL/min showed 62% greater exposure than males without renal impairment.

What is not known

The published record has a hole where its confirmatory evidence should be. No phase 3 trial of tesofensine has been published anywhere: the Medix programme in Mexico, described by the licensor as covering 372 patients, has no ClinicalTrials.gov registration, no PubMed record and no Europe PMC record, so its methods, endpoints and adverse-event tables cannot be examined. Of 13 registered studies, five carry posted results and only a minority have appeared in journals; the type 2 diabetes combination trial (NCT02737891) and the Prader-Willi trial (NCT03149445) exist solely as registry postings, as does the 48-week hypothalamic obesity study with its open-label extension (NCT03845075), for which no journal publication of the extension was located. The longest exposure that has actually been published is the two-part phase 2 obesity programme reported by Gilbert and colleagues, which ran 24 weeks randomised, a drug-free period of 12 plus or minus 3 weeks, then 24 weeks open-label, with assessments to week 60 (PMID 21720440). The largest published randomised obesity dataset remains 203 people over 24 weeks in a trial carrying an unresolved expression of concern about the completeness of its adverse-event recording. Two confirmatory phase 2 trials in the indications holding FDA orphan designations were withdrawn in December 2022 with nobody randomised. What is recorded on discontinuation in humans is narrow: in that same programme, after drug withdrawal the composite satiety score returned to baseline values (50 plus or minus 17 mm) while the reduced-weight state was maintained at minus 7.2 plus or minus 6.7 kg (P<0.0001), and no human data on weight regain after discontinuation were located. In diet-induced obese rats, four groups of 15 given 2.0 mg/kg for 28 days with one arm followed by a 28-day treatment-free period, route not stated in the indexed abstract, caloric intake and weight gain gradually increased again after treatment stopped (van de Giessen 2012, PMID 21889317). Body composition, cardiovascular outcomes, use in pregnancy or lactation, use in adolescents outside nine Prader-Willi participants, and interaction with GLP-1 receptor agonists are all uncharacterised in the published literature. Tesofensine holds no marketing authorisation in any jurisdiction, and material sold outside a registered trial has not been subject to the identity, purity or sterility controls that apply to investigational supply.

Questions