Frequently Asked Questions

No. In this 24-week randomized, double-blind, placebo-controlled phase 2b study, adjunctive basmisanil 240 mg twice daily did not improve the primary cognitive outcome versus placebo. Mean change from baseline on the MCCB neurocognitive composite T-score was 1.08b1 5.78 with placebo and 1.36b1 4.80 with basmisanil, with an effect size of 0.05 and P=.793.

The primary outcome was the MCCB neurocognitive composite T-score. After 24 weeks, this measure showed no meaningful difference between adjunctive basmisanil 240 mg twice daily and placebo, with mean changes from baseline of 1.08b1 5.78 for placebo and 1.36b1 4.80 for basmisanil (ES=0.05; P=.793).

Overall, no consistent cognitive benefit was shown. The study found no significant treatment effects on MCCB verbal learning, working memory, or any other individual MCCB cognitive domains, but one secondary measurethe delayed recall condition VPA total raw score IIshowed a positive signal at week 24 with ES=0.45 and P<.1.

The authors did not consider that isolated finding clinically meaningful because it was small overall and was not supported by improvement on other verbal memory tasks, global cognition, functional measures, or symptom ratings.

No. The prespecified subgroup analyses did not confirm the hypothesis that younger patients or those with deteriorated cognition would benefit more from basmisanil. Numerically, older patients showed a treatment effect of ES=0.39, whereas younger patients showed a numerically larger benefit in the placebo group (ES=-0.32).

By cognitive subtype, the largest numerical treatment effect was in the preserved group (ES=0.27), with no effect in the deteriorated group (ES=0) and a slight negative effect in the compromised group (ES=-0.04). None of these subgroup findings reached significance on primary, secondary, or exploratory outcomes.

The 80 mg arm was stopped after a preplanned interim futility analysis. Once about 30 patients per arm had completed 12 weeks of treatment, the 80 mg group had a less than 20% conditional probability of reaching a final effect size of at least 0.35 on the primary endpoint, so recruitment into that dose group was discontinued.

The study administered the MCCB three times before randomizationat screening, baseline 1, and baseline 2to reduce training and practice effects during the treatment phase. According to the authors, no practice effect was observed at the 12-week assessment, and only small treatment and placebo effects emerged at 24 weeks.

The discussion notes that the placebo-group change at the second assessment was 1.08 points, which was below the 1.9-point mean improvement cited from prior controlled CIAS trials, suggesting the repeated pre-baseline testing approach helped contain placebo response.

Patients were stratified into deteriorated, compromised, and preserved cognitive subtypes using the difference between premorbid IQ and current full-scale IQ. Deteriorated patients had a premorbid IQ minus current FSIQ difference of 10 points or more.

Among patients with a difference of less than 10 points, compromised patients had a premorbid IQ of 90 or lower, and preserved patients had a premorbid IQ of greater than 90.

Basmisanil was generally well tolerated in this study, although adverse events were somewhat more common with active treatment than with placebo. Adverse events were reported in 46% 49% of patients treated with basmisanil versus 40% with placebo, and the most frequent events were headache, diarrhea, fatigue, and psychotic disorder.

Headache and somnolence were more frequent in the treatment groups. Serious adverse events occurred at similar rates in the 240 mg basmisanil group and placebo group, with 4 events and 5 events, respectively.

No clear signal of worsened psychosis or major CNS safety liability was seen in this trial. The adverse event of special interest, worsening of psychosis, was reported in 6.2% of placebo-treated patients versus 3.7% in the 80 mg group and 3.8% in the 240 mg group.

The study also reported no CNS-related safety findings such as proconvulsive liabilities, insomnia, anxiety, depression, sedation, or suicidality.

The study enrolled 214 patients across 37 US sites, and 213 received treatment. A total of 81 patients (37.9%) withdrew prematurely.

Withdrawal rates were 46.3% in the 80 mg group, 41.0% in the 240 mg group, and 29.3% in the placebo group. Safety-related withdrawals were lower, ranging from 3.7% to 5.6%.

The authors identified three main limitations: a relatively modest sample size, a relatively high dropout rate, and the absence of biomarkers associated with NMDA receptor functioning such as resting-state and evoked gamma oscillations. They noted that the modest sample size limited power to detect more subtle effects in subgroups.