Repeated intranasal ketamine for treatment-resistant depression - the way to go? Results from a pilot randomised controlled trial

Earlier randomised controlled trials have shown that a single sub-anaesthetic dose of ketamine produces rapid and large antidepressant effects in both unipolar and bipolar depression, but these benefits are often short-lived, typically lasting 1–2 weeks. Open-label reports and case series suggest repeated ketamine dosing can prolong response, yet there have been no prior published randomised trials that systematically compare repeated ketamine administration with an active control. Most controlled work has used intravenous (IV), weight-based dosing, which requires infusion equipment and anaesthetic monitoring and is therefore challenging to translate into routine clinical settings. This pilot randomised, double-blind, placebo-controlled study examined the feasibility and safety of a repeated intranasal (IN) ketamine regimen for treatment-resistant depression (TRD). Gálvez and colleagues chose a fixed 100 mg IN dose given as ten 10 μL sprays (self-administered) across eight sessions over 4 weeks to approximate the antidepressant exposure of the common 0.5 mg/kg IV protocol, and they measured plasma ketamine and norketamine concentrations after the first treatment to assess bioavailability and inter-individual variability. The primary focus was feasibility and acute plus cumulative safety; mood and cognitive outcomes were also collected but the trial was explicitly a pilot study to inform future approaches.

This was a small, double-blind, parallel randomised pilot trial comparing eight intranasal treatments of 100 mg ketamine to an active control (4.5 mg midazolam) over 4 weeks (three times weekly for 2 weeks, then weekly for 2 weeks). Permuted block randomisation was used and allocation was concealed from treating staff and raters; only the trial pharmacist and statistician were unblinded. The original sample target was 10 participants, but the study was suspended after five participants were enrolled because of tolerability problems. Participants met DSM-5 criteria for a unipolar major depressive episode of at least 3 months’ duration, had MADRS ≥20 and had failed at least two adequate antidepressant trials in the current episode. Key exclusions included primary diagnoses other than major depressive disorder, unstable medical/neurological conditions, significant suicidal risk, recent substance dependence, lifetime ketamine/phencyclidine misuse, contraindications to ketamine, pregnancy risk, smoking, and anatomical issues likely to compromise nasal administration. Medication doses were required to be stable for 4 weeks pre-study and during the trial. Study drug was prepared in 1 mL atomisation devices delivering 100 μL per spray; each treatment comprised ten sprays (total 100 mg ketamine hydrochloride or 4.5 mg midazolam). Participants were trained on an insufflation protocol (nose-blowing, nozzle angle, alternate nostril dosing, head tilt, sniffing) and supervised during every session. The initial administration schedule (two sprays every 10 min) was modified early in the study to a single spray every 5 min (alternate nostrils) because of acute adverse effects; further delays or longer intervals between sprays were permitted per an operationalised protocol to manage elevations in blood pressure, psychotomimetic effects or motor incoordination. Participants were monitored for 4 h post-treatment and had to meet specified discharge criteria before leaving. Acute effects were assessed at pre-treatment, ~20 min after the last spray (capturing effects over the dosing period) and at 4 h using CADSS (dissociation), selected BPRS positive symptom items, YMRS item 1 (elevated mood), orientation testing, and vital signs (BP, HR). Cumulative safety measures included the Patient-Rated Inventory of Side Effects (PRISE), liver function tests (LFTs) and the Bladder Pain—Interstitial Cystitis Symptoms Scale (BPIC-SS) at baseline and after the treatment course. Neuropsychological functioning was measured with a CogState battery at baseline and 48–72 h after the course, selecting single primary outcomes per task. Blood for plasma ketamine and norketamine assays was collected 20 min after the last spray in the first treatment session and analysed by liquid chromatography–mass spectrometry. Mood was rated with the MADRS at baseline, pre-treatment and 48–72 h after the course; secondary clinician and self-report measures included HAM-A, CGI scales and SF-12. The extracted text does not clearly report statistical analysis methods or formal power calculations given the pilot nature and early termination.

Five participants were enrolled before suspension: three randomised to ketamine and two to midazolam. The sample was treatment resistant, with a mean of 3 (SD 1.7) failed antidepressants in the current episode and a mean of 5 (SD 2.6) failed lifetime. Feasibility: All three ketamine recipients experienced motor incoordination that interfered with self-administration from as early as the second spray (after ~20 mg delivered); none were able to self-administer all ten sprays at any session. By contrast, both midazolam participants self-administered the full ten sprays at every session. Several ketamine sessions required extension of treatment time under the operational protocol; midazolam sessions did not. Acute side effects and tolerability: Ketamine recipients generally had higher CADSS (dissociation) scores than midazolam recipients, with one ketamine participant showing markedly elevated CADSS after each session; CADSS scores returned to baseline by 4 h in all participants. Peak blood pressure values were higher in the ketamine group, with three ketamine participants and one midazolam participant recording peak BP ≥150/80 at various sessions; these episodes required only longer intervals between sprays in the ketamine group and no pharmacological intervention. No increases in selected BPRS positive symptoms or orientation impairments at 4 h were observed. Other acute physical side effects were more common with ketamine than midazolam. Cumulative safety: There were no clinically important changes in liver function tests, PRISE scores, or BPIC-SS urinary symptom scores over the treatment course in either group. Neuropsychological testing showed only isolated decrements: one ketamine participant had slower choice reaction time (Identification Task) and one midazolam participant showed poorer working memory on the n-back task; no systematic cognitive decline was evident. Pharmacokinetics and variability: Ketamine and norketamine plasma concentrations sampled 20 min after the last spray in the first session varied by approximately two-fold between participants. The participant with the highest plasma concentrations (k1) also experienced the poorest acute tolerability. Mood outcomes: Two participants met response (and remission) criteria at the end of the treatment course—one in the ketamine group (k1) and one in the midazolam group (m2). At one month follow-up only the ketamine participant k1 remained a responder. Inspection of MADRS trajectories suggested an antidepressant effect after the first and second ketamine sessions but little maintenance of benefit thereafter across sessions for most ketamine recipients. Comparison with subcutaneous (SC) ketamine within-subjects: Two participants who received both IN and subsequent SC ketamine under the same study design reported and objectively experienced fewer and less severe acute side effects with SC administration. Peak BP and peak CADSS scores were lower with SC than with IN dosing for these participants, and several acute symptoms (palpitations, sweating, blurred vision, oral numbness/tingling, dry mouth) were reported only with IN ketamine. Participants spontaneously commented that SC administration felt better than the sprays.

Gálvez and colleagues interpret these pilot data to show that, with the specific drug formulation and atomiser device used here, a fixed 100 mg intranasal ketamine regimen was not a practical approach for TRD. Acute tolerability problems—particularly motor incoordination, dissociation and transient blood pressure elevations—occurred at low cumulative exposure (as early as after 20 mg in some participants), prevented self-administration of the intended dose and sometimes prolonged administration beyond the planned 45 min. The investigators contrast their findings with a prior IN trial that used 50 mg and reported fewer adverse effects, noting that the prior study achieved lower plasma concentrations; thus they attribute differences to dosing and to the effective speed and extent of nasal absorption rather than to the intranasal route per se. The authors note substantial inter-individual variability in plasma ketamine and norketamine concentrations following a fixed IN dose, with up to two-fold differences between participants; the highest concentrations correlated with the greatest side-effect burden but also with the single ketamine remitter at one month. They therefore argue that weight-based or individually titrated dosing may be preferable to a one-size fixed dose, and they highlight that IN administration lends itself to titration because dose can be adjusted by number of sprays during a session. The study team also reports that, for two participants who later received subcutaneous ketamine in the same design, SC administration was better tolerated and led to fewer and milder acute effects. Longer-term safety signals were not observed in this small sample: no hepatic, cognitive or urinary adverse changes were detected after eight treatments, and no participants spontaneously reported subsequent drug craving or misuse when followed at 12 months (follow-up response 4/5). The authors acknowledge the primary limitation is the very small sample size due to early termination, which prevents reliable efficacy conclusions. They further emphasise that tolerability and efficacy with IN ketamine are likely to depend on multiple factors including drug formulation, delivery device, insufflation technique and individual nasal anatomy or vascularity. The investigators conclude that while the IN route remains of interest, careful optimisation of formulation, device and dosing strategy (for example titration or weight-based dosing) will be needed before repeated IN ketamine can be recommended for routine clinical use.