Proof of Concept Study Evaluating the Efficacy and Safety of MIJ821 in Patients With Treatment-resistant Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Proof of Concept Study Evaluating the Efficacy and Safety of MIJ821 in Patients With Treatment-resistant Depression (NCT03756129), a randomized phase 2 study. The primary outcome is Change From Baseline in the Total Score of the Montgomery Asberg Depression Rating Scale (MADRS) at 24 Hrs.

2. Clinical Question

3. Trial Design

NCT03756129 was a Phase 2, randomized, double-blind, parallel-group proof-of-concept study sponsored by Novartis Pharmaceuticals. Its official title describes a placebo-controlled study with an active comparator and evaluation of efficacy, safety, tolerability and pharmacokinetics. The registry lists 70 enrolled participants and a completed study. The study began on 2019-02-08, and primary completion was recorded as 2020-03-23.

The listed groups were MIJ821 0.16 mg/kg weekly, MIJ821 0.16 mg/kg every other week, MIJ821 0.32 mg/kg weekly, MIJ821 0.32 mg/kg every other week, weekly placebo, and weekly ketamine 0.5 mg/kg. The main posted efficacy summaries pool MIJ821 participants by dose, while the safety record presents the schedule-specific MIJ821 groups separately. The registry summary says treatment was delivered by infusion on a weekly or bi-weekly basis.

The study duration is described as 6 weeks of treatment followed by 1 month of follow-up. For the primary outcome, change in total MADRS score was assessed from baseline to 24 hours. Further listed assessments include MADRS at 48 hours and Week 6, the Young Mania Rating Scale, the Bech-Rafaelsen Melancholia Scale, and MIJ821 plasma pharmacokinetic measures. The registry record does not report the original design assumptions used to choose the planned sample size, nor does it report a detailed allocation ratio or a full protocol-level analysis plan.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent to Treat analysis set

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Only the first 6 of 7 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in the Total Score of the Montgomery Asberg Depression Rating Scale (MADRS) at 24 Hrs, measured over Baseline, and at 24 hours.

The primary endpoint is change from baseline in the total score of the Montgomery-Asberg Depression Rating Scale (MADRS) at 24 hours. The registry describes MADRS as a clinician-rated measure designed to assess depression severity and detect changes associated with antidepressant treatment. It contains 10 items, each scored from 0, meaning an item is not present or normal, to 6, meaning severe or continuous presence of symptoms; the total possible score is 60. Higher scores represent a more severe condition.

Change from baseline expresses how the score at follow-up differs from the participant's baseline score. In this trial's reported summaries, negative values indicate a decrease from baseline. Comparing average change between a treatment group and placebo helps describe whether the groups' changes differed over the stated assessment period. The 24-hour time point is the primary assessment specified in the registry, which also describes the primary efficacy comparison as occurring after single-dose administration.

A scale-based endpoint is useful for quantifying symptom severity on a common measure, but its group average does not show the full distribution of individual responses. Nor does a difference in average change alone describe durability, day-to-day functioning, or every aspect of safety. The registry record identifies the Intent to Treat analysis set for the primary outcome; further detail about the exact endpoint analysis population or handling of missing measurements is not reported in the registry record.

7. Results Posted in the Registry

• Pooled MIJ821 0.16 mg/kg: mean -15.51 versus -7.27 on control; difference -8.24 (95% CI -13.51 to -2.97), standardized effect d = -0.96; z-test p = 0.002.
• Pooled MIJ821 0.32 mg/kg: mean -12.98 versus -7.27 on control; difference -5.71 (95% CI -10.98 to -0.44), standardized effect d = -0.68; z-test p = 0.034.
• Ketamine 0.5 mg/kg Weekly: mean -12.94 versus -7.27 on control; difference -5.67 (95% CI -12.14 to 0.80), standardized effect d = -0.67; z-test p = 0.086.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline in the Total Score of the Montgomery Asberg Depression Rating Scale (MADRS) at 48 Hrs

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Pooled MIJ821 0.16 mg/kg: mean -14.94 versus -7.88 on control; difference -7.06 (95% CI -13.16 to -0.96), standardized effect d = -0.74; z-test p = 0.023.
  • Pooled MIJ821 0.32 mg/kg: mean -15.25 versus -7.88 on control; difference -7.37 (95% CI -13.75 to -0.99), standardized effect d = -0.77; z-test p = 0.024.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Pooled MIJ821 0.16 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Pooled MIJ821 0.32 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

PK Properties of MIJ821 in Plasma - Cmax (ng/mL)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Pooled MIJ821 0.32 mg/kg: mean 149.00 versus 99.50 on control; difference 49.50 (95% CI 4.44 to 94.56), standardized effect d = 0.90; z-test p = 0.031.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Pooled MIJ821 0.32 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

PK Properties of MIJ821 in Plasma - AUClast (h*ng/mL)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Pooled MIJ821 0.32 mg/kg: mean 738.00 versus 496.00 on control; difference 242.00 (95% CI 24.00 to 460.00), standardized effect d = 0.90; z-test p = 0.030.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Pooled MIJ821 0.32 mg/kg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcome record includes MADRS change at 48 hours and Week 6, other symptom scales, and pharmacokinetic measurements. Numerical results posted in the record include MADRS at 48 hours and MIJ821 plasma Cmax and AUClast on Day 1. The 48-hour group summaries show decreases from baseline in the MIJ821 groups and placebo; sponsor ANCOVA results are reported for the two pooled MIJ821 dose comparisons with placebo. These findings concern a secondary time point and do not replace the primary 24-hour result.

The available Day 1 pharmacokinetic summaries show higher reported mean Cmax and AUClast in the 0.32 mg/kg group than in the 0.16 mg/kg group. These are exposure measures, not clinical symptom outcomes, so their pattern addresses a different question from MADRS change. They do not independently confirm a clinical treatment effect. The registry record does not give numerical results for every listed secondary measure in the registry record; unavailable results are not reported in the registry record.

Secondary analyses carry less weight when many outcomes, time points, or comparisons are examined without a reported multiplicity plan. A multiplicity adjustment controls the increased chance of false-positive findings that can arise from testing many hypotheses. The registry record does not report whether the secondary tests were adjusted, and their results should therefore be treated as supportive or exploratory rather than as a collection of independent confirmatory findings. No new conclusion about efficacy should be based on the secondary outcomes alone.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change From Baseline in the Total Score of the Montgomery Asberg Depression Rating Scale (MADRS) at 48 Hrs
• Change From Baseline in the Total Score of the Montgomery Asberg Depression Rating Scale (MADRS) at Week 6
• Change From Baseline in the Young Mania Rating Scale
• Bech-Rafaelsen Melancholia Scale
• PK Properties of MIJ821 in Plasma - Cmax (ng/mL)
• PK Properties of MIJ821 in Plasma - Tmax (ng/mL)
• PK Properties of MIJ821 in Plasma - AUClast (h*ng/mL)
• PK Properties of MIJ821 in Plasma - AUC0-24h (h*ng/mL)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: MADRS is a numerical score, and the reported group summaries are least-squares means with standard errors. A mean summarizes the average value in a group. A least-squares mean is a model-based estimate, typically used to express a group mean under an analysis model; the registry identifies ANCOVA for its reported primary comparisons. The registry does not provide the model covariates or all other model details, so those should not be inferred.

What does an ANCOVA comparison mean?

ANCOVA is a method for comparing group outcomes while accounting for covariates included in the fitted model. In a baseline-and-follow-up setting, a model may account for baseline outcome values, but the specific covariates and implementation are not reported in the registry record. The reported ANCOVA estimates are therefore the sponsor's own model-based results, not estimates reconstructed from the group means alone. They should be distinguished from the simple unadjusted recalculations.

The registry reports different summary measures for its two primary ANCOVA contrasts: a median difference for the 0.16 mg/kg group and a mean difference for the 0.32 mg/kg group. Those values should be described using the measure the registry reports rather than treated as though both were the same estimand. In particular, a median difference and a mean difference summarize distributions differently. The registry record does not explain why the measures differ between these analyses.

What is the recalculation?

The recalculated comparisons use posted group summaries to estimate unadjusted differences between a pooled MIJ821 group and placebo. For the primary comparisons, the posted summary values are least-squares means and standard errors. The recalculation uses those summaries as approximations to derive a pooled standard deviation and a standardized effect size. Because standard errors are not the same as individual-level standard deviations, this conversion is approximate. These results are useful as a transparent descriptive check, but they do not reproduce the sponsor's ANCOVA model or its adjustment.

The recalculated p-values are unadjusted and depend on the assumptions of that simple comparison. They are not the trial's definitive inferential results when the sponsor's own analysis is available. The record marks a discrepancy between sponsor and recalculated results, so the sponsor's own analysis takes precedence. Differences can arise because the reported analysis uses a model and, for one contrast, a different summary measure, while the recalculation is based on posted aggregate summaries. The registry record does not give enough detail to resolve every source of discrepancy.

How should confidence intervals be read?

A confidence interval gives a range of values compatible with an analysis under its statistical assumptions, and communicates precision as well as direction. A narrower interval indicates greater precision than a wider interval for the same kind of estimate. The registry's reported primary ANCOVA intervals are explicitly 80% confidence intervals. They should not be described as 95% intervals. The recalculated intervals are based on the posted summaries; the registry record does not identify a confidence level for those recalculations. In either case, the interval is about uncertainty in a group contrast, not a range containing most individual outcomes.

Why can over-dispersion matter?

General statistical explanation: over-dispersion occurs when observed variation is greater than a model assumes. If a model understates variability, standard errors can be too small and intervals too narrow; if it accommodates variability appropriately, uncertainty is better represented. The posted primary outcome is a score, and no over-dispersion model or diagnostic is reported in the registry record. This concept is a general consideration, not a claim that over-dispersion occurred in this trial.

The small arm sizes and the pooled summaries limit how much can be learned from aggregate results about distributions, outliers, or subgroup patterns. The registry reports outcome-specific numbers of participants and standard errors, but not individual participant data. A recalculation cannot adjust for baseline characteristics, repeated measurements, missing outcomes, or other model features unless those are encoded in the reported summaries. The registry record does not report enough detail to establish how those features were handled in the sponsor's complete analysis.

Finally, a p-value measures compatibility of the observed data with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor does it measure clinical importance. Interpretation should combine the estimate, interval, design, outcome timing, and limitations of the record rather than relying on a threshold alone.

11. Confidence Intervals and Precision

The sponsor's reported primary ANCOVA estimate for pooled MIJ821 0.16 mg/kg is -8.25, with an 80% confidence interval from -11.67 to -4.83. For pooled MIJ821 0.32 mg/kg, the reported estimate is -5.71, with an 80% confidence interval from -9.22 to -2.20. Both reported intervals lie on the negative side of zero, consistent with the reported direction of a greater score decrease relative to placebo under those analyses.

Intervals communicate uncertainty in the estimated group contrast, not certainty about the true effect and not the range of individual participant changes. The intervals have nontrivial width, so the exact size of the contrast is less certain than its direction in the sponsor's reported analyses. The 80% confidence level is the one stated for these registry analyses; it is important not to silently interpret these as 95% intervals.

The simple recalculated primary intervals are -13.51 to -2.974 for the 0.16 mg/kg comparison and -10.98 to -0.444 for the 0.32 mg/kg comparison. These are based on posted summary statistics and are not the sponsor's ANCOVA intervals. The calculations use approximate standard deviations derived from standard errors, and the registry identifies a discrepancy; the sponsor analysis takes precedence. Precision is also constrained by the modest outcome-group sizes reported in the record and by the absence of individual-level data. The registry record does not report a minimal clinically important difference, so interval limits cannot be translated into a conclusion about clinical importance.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effect, not the trial's original design calculation. The registry record does not report the original sample-size assumptions or justification. The planning assumptions used for the illustration are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. For these continuous outcomes, the entries mean participants per arm, not a total across arms and not total events.

The 80% power column indicates the illustrative number of participants per arm for a design that would have an 80% chance of detecting the specified effect under those assumptions if that effect were the true effect. The 90% power column expresses the corresponding planning target for 90% power. These are design-planning concepts, not probabilities that the observed study result is correct. The entries compare planning for 100%, 75%, and 50% of the observed effect. A smaller target effect is harder to detect, so more participants per arm are needed to reach the same target power.

For example, the primary comparison planning entries are presented for each pooled MIJ821 dose against placebo, at the listed fractions of the observed effect. These illustrations should not be used to infer that the completed trial had those power levels, nor that its results were powered for an effect chosen after observing the data. This is not observed power. The record's own trial design assumptions remain not reported in the registry record.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -8.240; pooled standard deviation σ = 8.605.
2. n = 2 × (1.96 + 0.842)² × 8.605² / -8.240² = 17.1, rounded up to 18 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -8.24                      # observed difference in means
sigma <- 8.60530545949893                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(3756129)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 18 (80% power), 23 (90% power)
 75% of observed effect: n per arm = 31 (80% power), 41 (90% power)
 50% of observed effect: n per arm = 69 (80% power), 92 (90% power)
power.t.test: 18.1 
Simulated power at n = 18 per arm: 0.798 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports ANCOVA results for both pooled MIJ821 dose groups compared with weekly placebo. For 0.16 mg/kg, the sponsor reports a net median difference of -8.25, an 80% confidence interval from -11.67 to -4.83, and p = 0.001. For 0.32 mg/kg, it reports a net mean difference of -5.71, an 80% confidence interval from -9.22 to -2.20, and p = 0.020.

Recalculations from posted summary statistics give -8.24 for the 0.16 mg/kg mean difference and -5.71 for the 0.32 mg/kg mean difference, with unadjusted p-values of 0.002 and 0.034, respectively. Although the point estimates are numerically similar, the first sponsor estimate is a median difference whereas the recalculation is a mean difference. For the second comparison, the point estimates match as displayed, but the interval and p-value differ. The recalculation does not reproduce ANCOVA.

The registry flags a discrepancy. Accordingly, the sponsor's own reported analysis takes precedence over the simple recalculations for interpreting the trial's primary comparisons. The recalculations are supplementary checks based on aggregate summary statistics, not a replacement for the sponsor's analysis model.

15. Multiplicity, Missing Data and Analysis Populations

The record lists 19 outcome measures and 8 analyses, alongside multiple treatment groups, time points, and pharmacokinetic measures. Testing many outcomes can increase the chance of observing a statistically notable result by chance, even when each individual test uses a conventional threshold. Statistical methods to address this include prespecifying a hierarchy of tests, adjusting significance thresholds, or controlling an overall error rate. The registry record does not report whether any such multiplicity strategy was planned or used.

The primary outcome and its time point are identified, which helps distinguish the central question from secondary analyses. However, the record does not report a complete statistical analysis plan, the model covariates for ANCOVA, any interim analyses, or details of protocol deviations. These items are not reported in the registry record and should not be inferred from the posted results.

The primary population is described as the Intent to Treat analysis set. That label identifies the stated analysis population, but the registry does not provide enough detail to determine precisely which randomized participants were included in each endpoint analysis or how missing 24-hour measurements were addressed. The outcome summaries show different group sizes across endpoints, including at 48 hours and in pharmacokinetic summaries. The record does not explain all reasons for those differences, the amount or pattern of missing data, or whether imputation was used.

Missing outcomes can affect estimates if participants with unavailable measurements differ systematically from those observed. General statistical practice may address missingness through prespecified methods and sensitivity analyses, but those procedures are not reported in the registry record. Likewise, a listed Intent to Treat population alone does not establish how every missing observation was handled. Interpretation should therefore remain tied to the posted analyses and their stated populations, without assuming a particular missing-data method.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: atrial fibrillation (11.1%), major depression (11.1%), suicide threat (11.1%), asthma (10.0%), suicide attempt (5.0%).

The registry's total safety row reports 5 participants affected by serious adverse events among 70 at risk, 37 affected by other adverse events among 70 at risk, and 0 deaths. The schedule-specific group entries show that serious adverse events were not reported as occurring in every group. The registry also lists serious-event terms with percentages, including atrial fibrillation, major depression, suicide threat, asthma, and suicide attempt. Those terms should be read as registry-listed event categories, not as an assessment of whether an event was caused by treatment.

These counts describe what was recorded during the study, not comparative risks adjusted for follow-up or participant characteristics. The groups are small, and the safety entries are separated by dosing schedule whereas the efficacy summaries pool MIJ821 doses. The registry record does not provide enough detail to assess event severity beyond the serious/non-serious classification, timing, recurrence, or causal attribution. Zero recorded deaths does not mean that every possible safety concern was excluded. Safety counts should be interpreted alongside the limited study size and duration, not as proof of safety or harm.

17. What the Result Does, and Does Not, Mean

The primary result describes a difference in average MADRS change between pooled MIJ821 groups and placebo at 24 hours in this randomized study. The sponsor's reported ANCOVA results indicate more negative change estimates for both MIJ821 dose groups relative to placebo, with the stated 80% intervals and p-values. This is an account of the registry's reported group-level findings at a particular time point, not a prediction for an individual participant.

The trial does not establish from the posted results how durable any difference was beyond the measured period. Although the registry lists Week 6 and other secondary outcomes, the available information does not support treating those as a complete account of sustained response. A change in a rating scale is also not the same as an assessment of every outcome important to patients, and the registry record does not report a threshold for clinical importance.

The presence of a ketamine group does not establish a direct comparison between ketamine and MIJ821. The primary reported comparisons are with placebo, and the record does not report a sponsor analysis comparing the active comparator directly with either MIJ821 dose. Nor does the result establish that one MIJ821 dose is superior to the other: the reported comparisons each use placebo as the reference.

The findings should not be generalized automatically beyond the study population and settings described. The registry record does not provide a complete account of participant characteristics, missing-data handling, or all analysis details needed to judge every source of uncertainty. Statistical evidence against a no-difference null is not proof of a specific mechanism, long-term benefit, or favorable overall balance of benefits and harms. Clinical decisions require evidence and context beyond these trial statistics.

18. Why This Trial Matters Statistically

This study is statistically instructive because the registry presents both model-based sponsor analyses and aggregate summaries from which simple unadjusted comparisons can be recalculated. The estimates appear similar in some respects, but the methods are not interchangeable. The first sponsor comparison is reported as a median difference, while the simple recalculation is a mean difference; the second comparison also has a difference between sponsor and recalculated uncertainty and p-value. Recognizing these distinctions prevents a summary-statistic check from being mistaken for the trial's own inferential analysis.

The study also illustrates how an endpoint's timing defines the question. MADRS change at 24 hours is the primary outcome, while later MADRS assessments and other measures are secondary. Findings at different times or on different scales need not answer the same question, and results from many analyses should be weighed in light of multiplicity. The registry does not report a multiplicity plan, so secondary results warrant caution.

Finally, precision and design scale matter. The reported group sizes are modest, and planning illustrations based on observed effects are not evidence of the study's original planning rationale or its observed power. Safety counts likewise describe recorded events in this study but cannot establish the frequency of uncommon events with precision. Together, these points show why statistical interpretation requires attention to estimand, model, uncertainty, outcome hierarchy, and limits of the registry record.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary endpoint is a clinician-rated symptom scale assessed at 24 hours. The registry does not report a clinically important difference threshold, so the size of the numerical contrast cannot be classified from this record alone as clinically meaningful. A group mean also conceals variation in individual changes and the proportion of participants with different response patterns.

The sponsor's primary results use ANCOVA, but the registry record does not report the full model specification, including covariates and other analytic details. It reports a median difference for one primary dose comparison and a mean difference for the other without explaining the choice. The simple recalculations are unadjusted approximations and are explicitly discrepant from aspects of sponsor results; they should not be used as replacements.

The listed treatment groups differ by dose and schedule, while the main efficacy summaries pool schedules by dose. That pooling means the posted primary estimates do not separately show the effect estimates for weekly and every-other-week schedules. The record does not provide a sponsor comparison of MIJ821 with ketamine or a direct comparison between MIJ821 doses. The active-comparator group is also smaller than the placebo and pooled MIJ821 efficacy groups in the primary summaries.

The record does not report the original sample-size rationale, a multiplicity plan, interim analyses, or detailed missing-data methods. Outcome counts vary across measures, but reasons for the variation are not fully reported. The study lasted 6 weeks of treatment plus 1 month of follow-up, limiting what these findings alone can establish about longer-term outcomes. The safety summaries are descriptive, and the study size and duration limit assessment of uncommon or delayed events.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Proof of Concept Study Evaluating the Efficacy and Safety of MIJ821 in Patients With Treatment-resistant Depression — NCT03756129.
• Registry record last updated 2021-10-08; status last verified 2021-10.
• Results first posted 2021-05-19.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT03756129 is a completed Phase 2 randomized, double-blind, parallel-group proof-of-concept study of MIJ821 in treatment-resistant depression. The record lists 70 enrolled participants, placebo and ketamine comparator groups, and MIJ821 regimens using two doses and two schedules. The primary endpoint was change from baseline in total MADRS score at 24 hours.

The sponsor's ANCOVA analyses report a net difference favoring greater MADRS decrease for each pooled MIJ821 dose versus placebo, with 80% confidence intervals. Simple recalculations from posted summaries are supplementary and discrepant; the sponsor's analyses take precedence. Secondary clinical and pharmacokinetic outcomes provide additional, but not independently confirmatory, information. The registry reports serious and other adverse-event counts and no deaths. Important details, including the original sample-size rationale, multiplicity strategy, and missing-data methods, are not reported in the registry record.