In 2018, FDA published draft guidance on drug development for major depressive disorder, which included advice for evaluating novel antidepressants (FDA, 2018), said Tiffany Farchione, director of the Division of Psychiatry in the Office of Neuroscience at FDA. Although the draft guidance was published before the approvals of esketamine and brexanolone, it was informed by the agency’s regulatory considerations from these drugs and other products under development at the time, she said.

Traditional antidepressants include all antidepressants approved prior to 2019—SSRIs, serotonin and norepinephrine reuptake inhibitors (SNRIs), MAO inhibitors, and other drugs that are taken orally on a daily basis, said Farchione. By contrast, novel antidepressants could include drugs given as a single dose with long-lasting effects, drugs given at fixed intervals, drugs given on an as-needed basis, and drugs given to jumpstart treatment for an initial period before the effects of traditional antidepressants are noticeable, she said. To approve such drugs, Farchione said, FDA needs to see data that enable the agency to write a label that tells prescribers how to use the product. She added that the study design will vary depending on the intended use of the drug. For example, to assess the efficacy of a rapidly acting antidepressant, an early endpoint is useful. In addition to that, later endpoints are also needed to characterize the durability of the effect. The population for whom the drug is intended also must be clearly defined, she said. For example, a drug may be developed for the treatment of treatment-resistant depression, PPD, or other types of depression.

Farchione noted that for traditional antidepressants, FDA has allowed maintenance studies to be conducted after market approval through randomized withdrawal studies because experience has demonstrated that these treatments continue to work. However, for novel antidepressants, data are needed to demonstrate appropriate dosing regimens over the long term for this chronic cyclical illness.

at the day 28 endpoint. Interestingly and unexpectedly, participants who received the intranasal placebo also improved when they switched their oral antidepressants; however, those in the esketamine group improved more, which led to the product approval, she said.

THE APPROVAL OF BREXANOLONE

Prior to the introduction of brexanolone, PPD was treated similarly to an episode of major depressive disorder, said Farchione. However, as noted by Samantha Meltzer-Brody, a more rapidly acting drug was needed to prevent the risk of lasting negative effects on both the mother and child.

Administered as a one-time, 60-hour infusion in a targeted population, the trials for brexanolone were complex, said Farchione. Participants in these studies were no more than 6 months postpartum. To assess the durability of the antidepressant effect while limiting the number of participants who might otherwise drop out of the study, the primary endpoint was the

Hamilton Depression Rating Scale (HDRS), assessed at hour 60 (the end of the infusion) and the secondary endpoint at day 30. Farchione said the rationale for making the second assessment at day 30 was that most participants did not start the trial immediately after giving birth, and if depression recurred more than 30 days later, it was more likely to reflect a traditional depressive episode that could be treated with a traditional antidepressant.

Administration of the drug required continuous monitoring, said Farchione, because in the development program some patients experienced excessive sedation or even loss of consciousness. When this occurred, patients were aroused without incident when the drug infusion was suspended, she said. Because investigators were unable to identify predictors of loss of consciousness, a more stringent monitoring plan was necessary. In addition, the company agreed to a post-marketing commitment to evaluate alternative dosing regimens, such as in an outpatient, non-24-hour facility. As with esketamine, the time to onset efficacy curves were included in the FDA label (see Figure 6-2). Farchione noted that a second study (not shown here) showed no difference from placebo at day 30 because the placebo group continued to improve and eventually caught up with the brexanolone-treated group.

Farchione noted that neither the Montgomery-Asberg Depression Rating Scale (MADRS), which was used in the esketamine trials, nor the HDRS are particularly good for detecting early signals of efficacy because many of the items rated by these measures would not be expected to change quickly. She said FDA accepted letters of intent for two patient-reported clinical outcome measures in development that focus on recall of depression symptoms over hours rather than days. She added that if either esketamine or brexanolone is evaluated for other chronic conditions such as PTSD, maintenance studies will most likely be required.