ALTO-300

ALTO-300 is Alto Neuroscience's (ANRO) lead depression candidate, a 25 mg adjunctive (added on top of a patient's existing antidepressant) Phase 2 therapy for major depressive disorder (MDD) where enrollment is restricted to patients whose resting EEG shows a specific neural-irregularity signature derived by Alto's machine-learning pipeline [10]. The molecule has been publicly described as repurposed agomelatine, the same compound Servier sells across Europe as Valdoxan [7] and that never reached US approval after Novartis returned its license following Phase 3 failure [6]. Alto's thesis is that prior failures in unselected populations buried a real treatment effect in a biomarker-defined subgroup whose brain electrical activity predicts response. NCT05922878, the ongoing Phase 2b readout the equity story rides on, was expanded after a pre-specified interim from 150 to roughly 200 biomarker-positive patients within a total enrollment target of 321 [1][9]. Two smaller biomarker-discovery studies (NCT05118750 [2] and NCT05157945 [3]) have already completed and form the basis of the selection algorithm being tested prospectively here. If positive, ALTO-300 becomes a regulatory test case for whether EEG-guided enrichment can rescue a molecule with a known mechanism but a checkered clinical history. If negative, it tells investors how much to discount the rest of Alto's precision-psychiatry platform.

ALTO-300 is in Phase 2b. The active trial, NCT05922878, is recruiting toward a total of 321 patients (primary efficacy population: roughly 200 biomarker-positive after a pre-specified interim sample-size increase) with Alto Neuroscience listed as sponsor [1][10]. No breakthrough therapy, fast track, orphan, or RMAT designation has been disclosed, and none would be expected for an MDD adjunctive program at this stage. The compound is not novel chemistry; agomelatine has been on European pharmacy shelves since 2009 as Servier's Valdoxan [7], but the US development program ended after Phase 3 failure and Novartis returned its US license to Servier in 2011 [6]. Alto positions ALTO-300 as a development asset only in the sense that no one has previously sought US approval with EEG-stratified enrollment at the 25 mg dose. Company guidance issued on the Q2 2025 earnings call and reiterated since calls for topline data in mid-2026 [9]. The 2026-03-16 8-K disclosed a $120 million PIPE financing led by Commodore Capital that brought cash to approximately $264 million as of Q1 2026, with management guiding to runway through 2029 covering the ALTO-300 readout and a planned ALTO-207 Phase 3 start [4][9]. Cash is not a near-term constraint; the binary on this name is the mid-2026 readout, not the balance sheet.

Agomelatine hits two pieces of brain chemistry at once. The first is melatonin signaling: it activates MT1 and MT2, the two receptors melatonin uses to tell the brain it is night. Disrupted sleep-wake cycles and flattened circadian rhythms are a common feature of depression, and re-anchoring the night signal is the proposed mechanism by which the drug normalizes mood over weeks. The MT1/MT2 target itself is not first-in-class in the US: ramelteon (Rozerem), an MT1/MT2 agonist, has been FDA-approved for insomnia since 2005 [11], so the receptor pharmacology has a precedent regulators are familiar with. The second arm is serotonergic: agomelatine blocks the 5-HT2C receptor, a subtype that normally restrains release of dopamine and noradrenaline in frontal cortex. Blocking it disinhibits those circuits, which is the same general lever many established antidepressants pull on, but from a different angle than reuptake inhibition. The case for the mechanism is mixed. Agomelatine is approved in Europe as Valdoxan [7], where regulators accepted a small effect size on the MADRS depression scale. The FDA never granted approval; US Phase 3 trials missed their primary endpoints [6], and published meta-analyses have placed the effect somewhere between trivial and modest. The pharmacology is real but the clinical magnitude is the question. Alto's bet is that the molecule works well in a subset of patients defined by a resting EEG pattern (see trial_design), and that earlier diluted trials washed the signal out. ALTO-300 is dosed at 25 mg, the lower of the two clinical strengths used historically; Alto has argued the 25 mg dose retains antidepressant activity while reducing the liver-function-test elevations that drove most of the safety scrutiny on the 50 mg label [10].

NCT05922878 enrolls adults with MDD who have an inadequate response to a current antidepressant, randomizes them between ALTO-300 25 mg and placebo added to that existing antidepressant, and reads out the change in MADRS (Montgomery-Asberg Depression Rating Scale, the standard clinician-rated depression score where higher is worse) at six weeks [1]. The primary endpoint is specified in a pre-defined EEG biomarker-positive subgroup rather than the whole intention-to-treat population. The biomarker is not a single peak frequency. Per Alto's 2025 ASCP presentation, it is a measure of EEG irregularity (sample entropy) derived by a machine-learning model trained on Alto's own Phase 2 data; the biological rationale offered is that elevated 5-HT2C tone and reduced dopaminergic drive produce more disordered (higher-entropy) cortical activity at rest, and that those are exactly the patients in whom an MT1/MT2 + 5-HT2C drug would be expected to act [10]. That choice is the entire scientific bet: if the subgroup analysis is positive but the overall population is negative, the FDA will need to be persuaded that the biomarker is genuinely prognostic rather than a post-hoc rescue, and a companion-diagnostic (CDx) regulatory pathway will need to be aligned. The design is competent and tracks how Alto has run its other programs. The 6-week MADRS window is industry standard, the adjunctive use makes ethical recruitment easier, and a pre-specified interim led Alto to increase the biomarker-positive population from 150 to roughly 200 [10]. Concerns: MDD trials have notoriously high placebo response, often 30 to 40% on MADRS, which can swamp drugs with real but modest pharmacology. The two completed precursor trials (NCT05118750 n=91 [2], NCT05157945 n=148 [3]) were biomarker-discovery rather than confirmation, which means the EEG signature being tested here was derived on internal data, not pre-registered from an external source. That is the standard playbook for precision-psychiatry startups but it is also the standard place these programs go wrong.

Our model estimates this drug has a 5% chance of eventually being approved. That figure starts from the historical approval rate for drugs at this phase (about 24%), then adjusts based on ten specific facts about the trial and its sponsor. Larger-than-typical enrollment pushes the number up, while heavier-than-usual blinding, a thin or weak sponsor approval record, and weak earlier-phase results pull it down. The remaining factors are near average for this stage, so they leave the estimate close to where those main adjustments landed it.

Efficacy risk is the dominant failure mode. Agomelatine ran a US registrational program in the past and missed [6]; the molecule simply may not produce a large enough effect to register over placebo, even in an enriched subgroup. The MADRS placebo response in adjunctive MDD trials regularly hits 30 to 40%, and a 2-point separation, while statistically significant, would be commercially marginal against generic SSRIs and SNRIs. The biomarker hypothesis adds its own failure mode: the EEG signature was derived from Alto's internal Phase 2 data [2][3][10], and a prospectively defined subgroup analysis that fails to replicate would tank the program even if the overall trend looked directional. Safety risk is concrete but mitigated by dose. Agomelatine carries a hepatotoxicity warning in Europe, with cases of severe liver injury that led to mandatory liver function monitoring on the Valdoxan label [8]. Alto's 25 mg dose was chosen specifically because retrospective and ongoing data suggest lower LFT-elevation rates than at 50 mg [10], but the FDA will scrutinize US liver-event rates closely regardless. Execution risk is low on the financial side after the March 2026 PIPE: cash of approximately $264 million funds operations through 2029 per management, well past the ALTO-300 readout [4][9]. Commercial risk is the kicker. Even with approval, ALTO-300 enters a market with two recent branded adjunctive entrants: Auvelity (dextromethorphan/bupropion, AXS-05, approved August 2022 for MDD) and Spravato (esketamine, approved 2019 for TRD and as adjunctive MDD), both of which have struggled with payer coverage and prescriber inertia despite stronger mechanistic novelty than agomelatine. ALTO-300 would also compete with atypical antipsychotics (brexpiprazole, aripiprazole, quetiapine) that have decades of prescriber familiarity. IP risk is structural: agomelatine itself is off-patent globally, so Alto's only durable moat is the EEG biomarker selection method and any associated companion-diagnostic IP. That changes the deal economics versus a typical novel-compound asset and raises the question of generic entry the moment any composition-of-matter or formulation IP lapses.

Worth watching, with a specific date attached. The NCT05922878 readout (mid-2026 per Alto guidance [9]) is the binary event for ANRO [1] and likely the cleanest test in the public market of whether EEG-based patient selection can rescue an otherwise modest antidepressant. A clear positive on both the biomarker-positive subgroup and the overall population would be a genuine signal not just for ALTO-300 but for the precision-psychiatry thesis more broadly. A biomarker-only positive with a small effect size is the most likely outcome and the hardest to read; expect the stock to move but the regulatory path to remain murky because an EEG-based companion diagnostic would also need FDA alignment to support a label. Cash is no longer the swing factor: the March 2026 PIPE took runway to 2029 [4][9], so the company has time to follow a partial signal with a confirmatory study if needed. Commercial framing for the BioCosm reader: the right comparators are Auvelity and Spravato, not generic SSRIs; both branded adjuncts have shown that getting paid for new MDD mechanisms is hard even when the science is cleaner than agomelatine's, and ALTO-300's commercial case ultimately rests on whether the EEG test creates a defensible payer story rather than on the molecule itself. Because agomelatine is off-patent, the durable asset here is the biomarker IP and the CDx workflow; that is what an acquirer or partner would actually be buying. Cross-read from the rest of Alto's pipeline matters: the company's broader precision-psychiatry platform is the real product, and ALTO-300 is the first chance to validate it prospectively. If the platform fails here, the read-across to other Alto assets (notably ALTO-207 in TRD) is negative even though those programs have separate mechanisms. Watch for: an enrollment-completion 8-K, interim safety reports, and any commentary on LFT event rates at 25 mg. The only number that ultimately matters at readout is MADRS change from baseline at week 6, in the biomarker-positive arm, versus placebo, plus the overall-ITT comparison.