NSI-189

NSI-189 came from a phenotypic screen: researchers looked for small molecules that made human hippocampus-derived neural stem cells proliferate, then advanced the hit without first identifying a conventional receptor or enzyme target [1][5]. That origin explains both the interest and the uncertainty. It is not a monoamine reuptake inhibitor, but "neurogenic" describes an observed laboratory phenotype rather than a complete mechanism.

Phase 1b randomised 24 adults with recurrent major depressive disorder to placebo or 40 mg once, twice or three times daily for 28 days. The active groups improved on a self-rated depression questionnaire and the Cognitive and Physical Functioning Questionnaire, while differences on MADRS and the clinician global rating were not statistically significant. The active groups contained only six people each and all were pooled for the efficacy analysis [1].

The 220-person phase 2 trial is the result that should govern the verdict. Neither 40 mg nor 80 mg daily beat placebo on the primary MADRS endpoint (pooled differences −1.8 and −1.4 points; p=0.22 and p=0.34). Forty milligrams did favour NSI-189 on two self-rated scales and selected CogScreen tasks, but not on the Cogstate battery [2]. A later post-hoc paper found an 80 mg MADRS-6 signal after dividing patients by baseline severity, but that unplanned subgroup cannot reverse the negative primary analysis [6].

The binding target is unknown. The initial human paper says explicitly that the preclinical mouse neurogenesis, hippocampal-volume and behavioural results were company data on file [1]. A review reports activity in human hippocampal stem cells and mouse hippocampus, but still does not identify a receptor [5].

Peer-reviewed injury models give the most concrete downstream picture. In rats given fractionated cranial irradiation, four weeks of oral NSI-189 restored performance on four spontaneous-exploration tasks, increased the number of newly born granule neurons by 55% and reduced activated microglia by 56% versus irradiated controls. It did not change hippocampal volume, and the fear-conditioning result was not significant [3].

In stroke rats, 30 mg/kg daily for 12 weeks improved motor and neurological scores from day 3 through six months. It increased MAP2 staining in hippocampus and peri-infarct cortex but did not shrink the primary infarct and did not increase Ki67 in the canonical subgranular or subventricular neurogenic niches [4]. In oxygen-glucose deprived hippocampal cultures, the compound increased BDNF and stem cell factor; neutralising either signal reduced the survival effect [4]. That supports a remodelling and trophic account, but not a single direct target.

NSI-189 phosphate is the development form used in the human trials. The free base has molecular weight 366.5; the phosphate material in the stroke paper is listed at 464.5, and the authors report animal doses as free-base equivalent [4].

In phase 1b, dose escalation to 40 mg three times daily produced no serious adverse event. Headache occurred in half of active and placebo participants; dizziness occurred in 28% of active participants and 17% of placebo, and nausea in 11% and 0%, respectively [1]. The sample was far too small to define uncommon risks.

Phase 2 reported no serious adverse event among NSI-189 recipients and no intolerance discontinuations in either active arm during its first six-week stage [2]. That supports short-term tolerability, not long-term safety: only 220 patients were randomised, treatment lasted 12 weeks, and the compound has no established clinical use.