MAP4343

MAP4343 is an oral neurosteroid Mapreg is testing for treatment-resistant major depression, meaning patients who have failed prior antidepressants. It's a chemical cousin of pregnenolone, a steroid the brain makes naturally from cholesterol, modified at the 3β position so the body doesn't convert it into other active steroids. Pregnenolone itself has been studied as a depression treatment and didn't produce a clean signal; MAP4343 attempts to fix that metabolic instability problem with a single ether modification at the 3β hydroxyl. The Phase 2 RESIST trial (NCT03870776) enrolled 126 antidepressant non-responders against a 110-patient target and is listed as completed [1]; topline results have not been publicly disclosed in a form I can cite. Mapreg is a small French biotech built around work by Etienne-Emile Baulieu, the chemist behind mifepristone (RU-486). Because the company and the trial are both French-based, the natural first regulatory path is EMA, not FDA. If MAP4343 produces an antidepressant signal through microtubule stabilization rather than the GABA-A route taken by zuranolone and brexanolone, it would be a genuinely new approach in a category that hasn't seen one in decades.

MAP4343 is a novel chemical entity, never approved for anything. Mapreg has been developing it since the mid-2000s, with most preclinical work coming out of Baulieu's lab [2]. Phase 2 RESIST (NCT03870776) completed enrollment of 126 patients against an original target of 110 (115% of target - a positive enrollment signal in a category notorious for slow recruitment) and is listed as completed on ClinicalTrials.gov [1]. The primary endpoint was Hamilton Depression Rating Scale (HAM-D) change from baseline to day 43, against placebo. HAM-D is a 17-item clinician-rated scale ranging from 0 to 52, where higher scores mean more severe depression; FDA generally treats a placebo-adjusted improvement of 3 or more points as the minimum bar for clinical meaningfulness, and most approved antidepressants land in the 2-to-4-point range. No FDA breakthrough therapy, fast track, or orphan drug designation could be verified - and given this is a French sponsor running French sites, the obvious regulatory path is an EMA CTD submission, with FDA engagement contingent on either a separate Mapreg IND or a US licensee taking on that risk. Mapreg has not published full RESIST results in a peer-reviewed journal to date, and no published Phase 1 PK/safety dataset in healthy volunteers could be found - odd for a program of this vintage and worth flagging. The company's next step depends on what the data showed: a clear signal would justify Phase 3 planning and almost certainly require a partnership, since Mapreg is a small private French biotech with opaque funding and lacks the capital to run a registrational program alone. A weak or mixed signal would likely end the program or push it toward a smaller indication. Timeline for any public readout is uncertain; small biotechs often delay disclosure until they have a deal or financing in place.

Pregnenolone is a steroid the brain manufactures from cholesterol; it sits at the top of the neurosteroid pathway, with most other brain steroids derived from it. Some studies show lower brain pregnenolone in depression, which is part of the rationale here. The problem with giving pregnenolone itself as a drug is that the body rapidly converts it into other active steroids (allopregnanolone, DHEA, progesterone), so you cannot isolate any single effect. MAP4343 is 3β-methoxy-pregnenolone - a methoxy group (OCH3) is installed at the 3β hydroxyl position, converting the hydroxyl to an ether. That ether resists the sulfotransferases and 3β-hydroxysteroid dehydrogenase enzymes that normally chew up pregnenolone, so the compound circulates and acts as itself rather than as the metabolic cascade [2]. As a lipophilic steroid analog, MAP4343 should cross the blood-brain barrier readily - pregnenolone and related neurosteroids do so passively - and rodent work from Baulieu's group demonstrates CNS exposure adequate to drive behavioral effects [2]. Human CNS exposure has not been confirmed via CSF sampling in any public RESIST disclosure, which is a real gap for a mechanistically novel CNS compound. The proposed target is microtubule-associated protein 2 (MAP2), a protein that helps stabilize the internal scaffolding inside neurons, especially in dendrites - the branches that receive incoming signals. Depression is associated with shrinkage of these dendrites in the hippocampus and prefrontal cortex; chronic stress reproduces the same shrinkage in rodents. If MAP4343 stabilizes microtubules via MAP2, the idea is that dendrites regrow and synaptic plasticity is restored. How strong is the case? Weak. MAP2 has never been a validated drug target. The preclinical work (mostly from Baulieu's group) shows MAP4343 reverses stress-induced behavioral changes in rodents [2], but no other drug works through this mechanism, so there is no clinical precedent for whether MAP2 binding translates to patient benefit. The biology is plausible. The bar from plausible to approvable in CNS is enormous.

RESIST (NCT03870776) was a Phase 2, double-blind, randomized, placebo-controlled, multicenter study in adults with major depressive disorder who had not responded adequately to prior antidepressant treatment in the current episode [1]. Original target was 110 patients; final enrollment landed at 126. Primary endpoint: change in HAM-D from baseline to day 43 versus placebo. Trial was sponsored by Mapreg and conducted at French sites. Mapreg has not publicly disclosed the MAP4343 dose(s) tested in RESIST, which is itself a small flag - lack of dose transparency complicates any Phase 3 design inference and any preclinical-to-clinical exposure bridging. Several design concerns. Day 43 (six weeks) is a reasonable window for an antidepressant effect, but HAM-D as a primary endpoint is sensitive to placebo response, which routinely runs 30 to 40 percent in TRD studies. Detecting a drug effect on top of that requires either a large sample or a strong effect size; 126 patients is small. Second, the entry definition of 'non-responder' matters a lot. Failing one antidepressant is a much weaker bar than the standard TRD definition of two or more failures, and a softer bar pulls in patients who may still respond to placebo or to a second-line standard treatment, blunting the signal. Third, no biomarker stratification. Everyone with the diagnosis goes in. If MAP2-related biology is relevant only in a subset of depressed patients, the trial cannot find that subset, and the average effect dilutes.

Our model gives this drug a 7% chance of eventually being approved. That number starts from the historical approval rate for Phase 2 drugs in this area, which is about 24%, then adjusts based on ten facts about the trial and the sponsor. The estimate goes up because the trial has more secondary endpoints than usual, but comes down because of heavier-than-usual blinding, the sponsor's thin or weak approval record, and limited earlier-phase results. The remaining factors were close to average and didn't push the number much in either direction.

Efficacy risk is the dominant concern. Placebo response in MDD trials swallows weak signals, and 126 patients is not much statistical power. A 2-point HAM-D advantage is borderline clinically meaningful and will not move FDA or EMA. No separation means the program ends. Because RESIST was not stratified by biomarker, even a positive average signal will not tell anyone which patients are most likely to benefit, and that is an expensive problem to carry into Phase 3. Safety risk is moderate. Pregnenolone-derived compounds have generally been tolerable in short-term human exposure; they do not cause the abuse liability of ketamine or the heavy sedation seen with GABA-A neurosteroids. The unknowns are chronic dosing effects on the HPA axis, possible downstream interference with sex steroid pathways via metabolites, and any cumulative endocrine signal beyond six weeks. None of that has been ruled out at scale, and the absence of a public Phase 1 human PK/safety paper means tolerability claims cannot be grounded in a clean reference. Execution risk is severe. Mapreg cannot fund a 1,000-patient Phase 3 alone, so a partnership or licensing deal has to materialize. Big-pharma appetite for unprecedented CNS mechanisms has been low after a decade of expensive failures across glutamatergic, anti-inflammatory, and serotonergic programs. Commercial risk: even if it works and gets approved, MAP4343 enters a TRD market that already includes esketamine (Spravato, J&J) [5] and Auvelity (dextromethorphan-bupropion, Axsome), both with established payer positions. Payers will demand differentiation on speed of onset, durability, or a clean tolerability story.

Worth watching, low priority. The interesting question is not whether MAP4343 specifically works - the prior is against it - but whether a non-GABAergic neurosteroid can produce any antidepressant signal at all. That would be a real mechanistic data point regardless of which company captures the value. The data point that flips this from noise to signal: a published or presented RESIST readout showing greater than a 4-point HAM-D placebo-adjusted reduction with a clean p-value (versus the 3-point FDA-meaningful threshold), plus a subgroup analysis hinting at who responded. Catalyst checklist: (1) EMA CTD submission or scientific advice announcement - the primary regulatory path given Mapreg's French base - (2) a Mapreg partnership announcement with a large CNS pharma, which would be a stronger commercial signal than a sponsor press release because partner diligence is more demanding than a sponsor's own framing, (3) an FDA pre-IND or IND filing, which would signal real US ambition rather than EU-only positioning, (4) any peer-reviewed publication of RESIST topline. Realistic timeline is 2026 given the trial completed and the company has been quiet. If 12 months pass with no announcement, assume the data were not compelling. Mapreg's credibility rests on Baulieu, who is a serious scientist (Lasker laureate). That gets a fair hearing from partners. It does not, on its own, get you a drug.