A 220-patient Phase 3 trial of oral ketamine for depression was posted to ClinicalTrials.gov on September 24, with an estimated October 2026 start at five Indian sites. Nothing has been enrolled, so KOMET has produced no efficacy evidence. The registration is worth reading for what it will be tested against: a small set of blinded oral-ketamine trials that mostly reported an advantage for the drug, but that differ so much in dose, schedule, comparator, and endpoint that they cannot be read as one answer.
What KOMET will test
KOMET is registered by the Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER) in Puducherry, with Vikas Menon as principal investigator. It is a randomized, triple-masked, active-placebo-controlled trial of adjunctive oral racemic ketamine against oral midazolam in adults with a major depressive episode. The registry classifies the sponsor as a government institution and lists the Indian Council of Medical Research as a collaborator without describing what that role involves, so the trial’s funding is not characterized here.
Participants receive nine sessions. Ketamine starts at a fixed 150 mg, sipped in about 200 mL of water over 20 to 30 minutes. The registry allows the dose to rise to 175 mg in the second session and 200 mg in the third, and to 250 mg in the final week if response is suboptimal, or to fall to 100 to 125 mg for people who do not tolerate 150 mg. Midazolam is dosed at 5 to 7.5 mg, adjusted to physiological response to preserve blinding. The primary endpoint is the HAM-D score at the end of week 3. Secondary measures run through week 4 and include MADRS, PHQ-9, suicidal ideation, response and remission, functioning, and adverse effects.
Entry requires a HAM-D of at least 20 in an episode of major depressive disorder or bipolar disorder. The registry lists no treatment-resistance requirement. It excludes obsessive-compulsive disorder, psychotic symptoms, substance dependence other than nicotine, concurrent neuromodulation, and outpatients with active suicidal ideation with a plan or intent in the past three days. Estimated completion is October 2029. The registry marks the trial as not an FDA-regulated drug study and posts neither a sample-size rationale nor a statistical analysis plan.
What the earlier controlled trials found
Oral racemic ketamine is a different intervention from intravenous ketamine, intranasal esketamine, R-ketamine, proprietary prolonged-release ketamine, and ketamine-derived compounds such as ACADIA’s ACP-211. Only trials of standard oral racemic ketamine are counted here. Among the five randomized, blinded trials of it in depression identified for this analysis, none enrolled more than 81 patients.
Three used controls other than midazolam. A 2016 trial compared 150 mg of daily oral ketamine with the analgesic diclofenac in patients with chronic pain and mild-to-moderate depression, with 20 completers per arm, and reported lower depression scores with ketamine at week 6. A 2018 trial of 81 patients added ketamine or placebo to sertraline and reported greater early improvement with ketamine, 85.4 percent against 42.5 percent. A 2019 proof-of-concept trial gave 41 outpatients with treatment-resistant depression 1 mg/kg or placebo three times a week for 21 days. MADRS scores fell 12.75 points with ketamine and 2.49 with placebo at day 21, and 27.3 percent of the ketamine group reached remission against none of the controls.
Two used midazolam, the comparator KOMET has chosen, and they reported different patterns. A 45-patient Vienna trial, published in the Journal of Affective Disorders in 2026, gave 1 mg/kg of oral ketamine or 0.03 mg/kg of midazolam six times over two weeks. Its primary endpoint, MADRS reduction at one week, did not reach statistical significance. Response rates favored ketamine at that point, with a number needed to treat of 4.6, but the confidence interval ran from 2.4 to 62.6.
An 80-patient trial in adults with major depressive disorder and suicidal ideation, published in the Asian Journal of Psychiatry in 2025, gave a single session of 3 mg/kg of ketamine, a mean of 180 mg, or 0.3 mg/kg of midazolam. Suicidal-ideation scores were lower with ketamine at 4 hours, day 3, and day 7. Depression scores were significantly lower at 4 hours and day 3; the abstract does not report a significant difference at day 7, when study-defined response was 25 percent against 0 percent and remission 5 percent against 0 percent. Because the trial enrolled suicidal patients and KOMET excludes outpatients with active suicidal ideation with a plan or intent, its results do not transfer automatically to KOMET’s population.
KOMET’s registry cites three background papers. Two were unblinded: an uncontrolled series of 30 inpatients, and a 61-patient randomized comparison of oral and intravenous ketamine that found lower dropout with oral treatment, 26.7 percent against 54.8 percent, but could not settle relative efficacy. The third is the Vienna trial. The midazolam-controlled suicidality trial is not among them.
How KOMET differs
KOMET departs from the earlier trials on several dimensions at once. It is larger, at 220 planned participants against 81 or fewer, and is registered across five sites, in Bengaluru, Udupi, Mumbai, Bhubaneswar, and Puducherry. It doses in absolute milligrams, 150 to 250 mg with flexible titration, where the Vienna and 2019 trials dosed by weight and the suicidality trial averaged 180 mg from a weight-based dose. It runs nine sessions, more than the one and six of the two midazolam-controlled trials and about the count of the 2019 trial’s thrice-weekly, 21-day schedule, though the registry gives no session frequency. It places its primary endpoint at week 3, between the one-week Vienna endpoint and the day-21 endpoint of the 2019 trial. It includes bipolar depression, as the Vienna trial did, and sets no treatment-resistance requirement, unlike the 2019 trial. And it adjusts the midazolam dose flexibly where the two midazolam-controlled trials fixed it by weight.
The registry does not say why any of these choices were made, and none should be read as a response to a particular earlier result. The practical consequence is different: because so many things change together, a KOMET result in either direction will not isolate which one mattered. Different scales and time points also mean its numbers cannot be lined up against the earlier trials as a head-to-head comparison.
Why the active comparator matters
Midazolam produces sedation and other perceptible effects, which can make it harder for participants to conclude they received an inert pill than a placebo would. That is the logic behind active-placebo designs in ketamine research. KOMET also adjusts the midazolam dose to physiological response. An active comparator does not solve functional unblinding. Ketamine’s dissociative effects differ from sedation, and the registry’s outcome list includes no measure of whether participants or raters guessed their assignment. The ketamine-arm description also permits benzodiazepines such as clonazepam for anxiety or agitation, and the registry does not say how rescue use will be handled in the analysis, a point that matters when the comparator is itself a benzodiazepine.
What it could settle, and what it could not
If it enrolls and reports, KOMET could show whether adjunctive oral ketamine beats a midazolam control on HAM-D at week 3 in the trial’s enrolled population of adults with major depressive disorder or bipolar depression, and whether early separation at days 2 and 7 holds. It would add tolerability and discontinuation data across nine sessions. It could not speak to durability past week 4, the last registered time point, or establish effects in outpatients with active suicidal ideation with plan or intent, given that exclusion, or to how oral ketamine compares with intravenous ketamine or esketamine, since neither is an arm. A Phase 3 label does not establish efficacy, and here it does not indicate an FDA registrational program either.
The commercial picture differs from proprietary programs. ACP-211 is a patented deuterated compound. Racemic ketamine is generic, and the registry describes the goal as an affordable, scalable option for routine psychiatric services that could inform treatment guidelines. The registry does not say what regulatory or commercialization path, if any, would follow a positive result.
What to watch
The first signals are procedural: whether enrollment begins in October as estimated, whether all five sites activate, and whether a protocol paper or an Indian trial-registry record appears, since the ClinicalTrials.gov entry lists neither. After that come recruitment pace against the 220 target and any change to dosing or endpoints before enrollment closes. The result itself, the week-3 HAM-D comparison, is estimated for 2029. Alongside it, watch discontinuation across nine sessions, whether early effects persist through week 4, how blinding held, and how rescue benzodiazepines were counted.