A Biomarker-Guided, Randomized, Placebo-Controlled Efficacy and Safety Study of Liafensine in Patients With TRD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Biomarker-Guided, Randomized, Placebo-Controlled Efficacy and Safety Study of Liafensine in Patients With TRD (NCT05113771), a randomized phase 2 study. The primary outcome is Change in Montgomery Åsberg Depression Rating Scale (MADRS) Total Score From Baseline to Day 42, in DGM4-positive Patients.

2. Clinical Question

The primary outcome is a change score on a depression-severity scale, rather than a count of events or a time-to-event outcome. The registry describes the primary objective as testing whether liafensine is superior to placebo in the DGM4-positive subgroup. For interpretation, a more negative change represents a larger reduction in MADRS score because higher scores indicate more severe depression. The registry record does not report the biomarker test’s operating characteristics, the full eligibility criteria, or the rationale for selecting DGM4-positive patients as the primary analysis population.

3. Trial Design

The registry identifies NCT05113771 as a completed, randomized, double-blind, placebo-controlled, parallel-group Phase 2 study sponsored by Denovo Biopharma LLC. The condition is treatment-resistant depression. The study began on 2022-06-29, and primary completion is recorded as 2024-02-06. The registry lists enrollment as 197. Its brief summary describes an approximate enrollment of 180 eligible patients; the registry record does not explain the difference between that planning description and the recorded enrollment.

The treatment arms were liafensine 1 mg once daily, liafensine 2 mg once daily, and placebo once daily, each described for 6 weeks. For the primary outcome, results are reported separately for the 1 mg arm and the 2 mg arm, as well as for a combined liafensine 1 mg or 2 mg group. The posted primary results list 62.0 participants for the 1 mg group, 61.0 for the 2 mg group, 123.0 for the combined group, and 63.0 for placebo. These registry outcome counts do not by themselves describe how participants were assigned or included in every analysis.

The primary endpoint was change in MADRS total score from baseline to Day 42 among DGM4-positive patients. The registry also lists changes in Clinical Global Impression-Severity (CGI-S) and the Clinical Global Impression-Improvement (CGI-I) assessment. The registry record does not report the randomization sequence method, allocation concealment details, study sites, full protocol, or the original sample-size justification. It identifies double masking but does not specify which roles were masked.

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

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.

Population: Intent to treat (ITT) population is equal to196 (65 participants in liafensine 1mg arm, 64 patients in liafensine 2mg arm, and 67 patients in placebo arm), which is including all patients who took at least 1 dose of study drug. Total of 197 patients were initially randomized and one randomized patient was not dosed.

6. Primary Endpoint

The registry lists the primary outcome as Change in Montgomery Åsberg Depression Rating Scale (MADRS) Total Score From Baseline to Day 42, in DGM4-positive Patients, measured over Baseline to Day 42.

The primary endpoint is change in MADRS total score from baseline to Day 42 in DGM4-positive patients. MADRS is a continuous measure of depression severity; the registry describes its range as 0 60 and states that higher scores indicate more severe depression. A change score expresses how the measured score at follow-up differs from the baseline score. Negative values in the posted results indicate reductions in score over the assessment period.

Using a continuous scale preserves information across the range of measured severity rather than reducing the outcome to a binary event. This can make changes in the degree of symptom severity visible in the group average. However, the mean alone does not show how individual participants’ scores were distributed or how many experienced changes of particular sizes. The registry record does not report a threshold for interpreting a change as clinically important, so the posted mean differences should not be treated as a direct measure of the proportion of patients who experienced meaningful improvement.

The endpoint is focused on DGM4-positive patients rather than all enrolled patients. That focus matches the stated biomarker-guided objective, but it also means the primary result does not establish how the same outcome compares in DGM4-negative patients. The registry record does not report the DGM4-negative subgroup’s primary outcome results.

7. Results Posted in the Registry

• Liafensine 1mg: mean -15.40 versus -11.00 on control; difference -4.40 (95% CI -7.95 to -0.85), standardized effect d = -0.43; z-test p = 0.015.
• Liafensine 2mg: mean -15.50 versus -11.00 on control; difference -4.50 (95% CI -8.12 to -0.88), standardized effect d = -0.44; z-test p = 0.015.
• Liafensine 1mg or 2 mg: mean -15.40 versus -11.00 on control; difference -4.40 (95% CI -7.52 to -1.28), standardized effect d = -0.43; z-test p = 0.006.

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 to Day 42 in Clinical Global Impression-Severity Scale (CGI-S) Score in DGM4 Positive Patients

• 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:
  • Liafensine 1mg: mean -1.50 versus -1.10 on control; difference -0.40 (95% CI -0.82 to 0.02), standardized effect d = -0.34; z-test p = 0.059.
  • Liafensine 2mg: mean -1.50 versus -1.10 on control; difference -0.40 (95% CI -0.83 to 0.03), standardized effect d = -0.33; z-test p = 0.068.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Liafensine 1mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Liafensine 2mg and the control. The interval gives the range of effects that remain compatible with the data.

The Clinical Global Impression-Improvement Scale (CGI-I) (Range = 1-7, With Higher Score Indicating Worsening) Was Assessed in DGM4 Positive Patients

• 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:
  • Liafensine 1mg: mean 2.30 versus 2.90 on control; difference -0.60 (95% CI -0.98 to -0.22), standardized effect d = -0.55; z-test p = 0.002.
  • Liafensine 2mg: mean 2.40 versus 2.90 on control; difference -0.50 (95% CI -0.92 to -0.08), standardized effect d = -0.42; z-test p = 0.019.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Liafensine 1mg 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 Liafensine 2mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Across the two posted secondary endpoints, the group summaries generally point in a direction consistent with the primary outcome: the liafensine groups have more negative mean changes for CGI-S than placebo, and lower CGI-I mean scores than placebo. On CGI-I, the registry states that higher scores indicate worsening. The combined-group analyses reported by the sponsor are described as superiority analyses. The posted secondary findings therefore provide supportive context for the primary result, but they do not independently establish the primary outcome or resolve its limitations.

The statistical evidence for a secondary measure must be interpreted in light of the full set of outcomes and comparisons. Multiple endpoints, dose groups, subgroup definitions, and analyses create more opportunities for a low p-value to arise by chance when results are assessed separately. The recalculated secondary p-values are unadjusted. The registry record does not report a multiplicity procedure or a hierarchy that would establish how secondary tests were intended to support the primary claim.

The registry reports MMRM for the combined CGI-S analysis and Cochran-Mantel-Haenszel for the combined CGI-I analysis. It provides a p-value for each reported combined analysis but does not provide model-based effect estimates and confidence intervals for those combined results. Consequently, the registry record offers less information for judging the magnitude and precision of the sponsor’s secondary contrasts than it does for the primary outcome. Secondary outcomes can add coherence to an overall pattern, but without reported multiplicity handling and fuller effect estimates, they should carry less inferential weight than the prespecified primary question. The registry record does not identify the full prespecified endpoint hierarchy or explain whether all analyses planned for the trial are posted.

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

• Change From Baseline to Day 42 in Clinical Global Impression-Severity Scale (CGI-S) Score in DGM4 Positive Patients
• The Clinical Global Impression-Improvement Scale (CGI-I) (Range = 1-7, With Higher Score Indicating Worsening) Was Assessed in DGM4 Positive Patients

9. Statistical Interpretation

10. Statistical Methods Explained

The registry reports a continuous primary outcome summarized as a mean change in MADRS score. A mean is the arithmetic average of measured values within a group. It describes the group’s central level of change, but does not show the range of individual experiences or the shape of the score distribution. The listed spread for the primary arm summaries is a standard error, not a standard deviation. A standard error describes the uncertainty in estimating a group mean; it should not be read as the typical spread of individual scores.

What does the reported MMRM estimate represent?

The sponsor’s primary analysis is described as an MMRM, or mixed model for repeated measures. In general, this class of model can analyze measurements collected repeatedly over time while accounting for their relationship within a participant. It can estimate a treatment contrast using the available repeated measurements and model assumptions. The reported parameter is a least-squares mean difference, which is a model-based contrast rather than simply the difference between the two posted group means. The registry record does not report the model’s covariates, repeated-measures covariance structure, visit terms, missing-data assumptions, or other implementation details. The exact model therefore cannot be reconstructed from the registry record.

The primary reported estimate is -4.4 for the combined liafensine groups versus placebo, with a 95% confidence interval from -7.6 to -1.3. The interval gives a range of values compatible with the data under the model and its assumptions, using the stated confidence level. It is not a range containing 95% of individual participant outcomes, nor does it assign a 95% probability to the fixed treatment effect. A confidence interval that excludes zero is consistent with a difference between the groups under the analysis, but it does not measure the clinical importance of that difference.

How do the recalculated comparisons differ?

The recalculated comparisons use posted summary statistics to make simple, unadjusted comparisons of group means. They are not replacements for the sponsor’s MMRM. For the combined liafensine group, the recalculated mean difference is -4.4, with a 95% confidence interval from -7.515 to -1.285 and p = 0.006. The recalculated comparisons for the 1 mg and 2 mg groups are also unadjusted. They do not account for any covariates or repeated-measures structure that may be included in the sponsor’s analysis. The registry record does not report a discrepancy flag for these results, but the methods remain different and should not be treated as interchangeable.

A p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the analysis assumptions held. It is not the probability that the null hypothesis is true, and it is not a measure of the size or importance of an effect. The recalculated p-values are unadjusted; the registry record does not state whether the sponsor adjusted the primary or secondary testing strategy for multiplicity.

What is the role of standardized effects?

A standardized effect expresses a mean difference relative to a measure of variability, allowing outcomes measured on different scales to be placed on a common scale. The recalculated primary comparisons include standardized effect estimates, but those estimates are not the sponsor’s primary model result and are not needed to interpret the original MADRS units. Standardization does not resolve questions about clinical meaning, analysis population, missing data, or the model assumptions. It can also make an estimate less intuitive than the original scale, so the MADRS-unit difference and its interval remain central to interpreting the primary endpoint.

Why can over-dispersion matter?

For count outcomes, over-dispersion means that variability is greater than a simple model assumes; ignoring it can make standard errors too small and uncertainty appear overstated. That general issue is not the central feature of this primary endpoint, which is continuous rather than a count. The registry record does not report a count-based primary outcome. For continuous measures, the relevant questions instead include how the outcome is distributed, how repeated values are modeled, and how missing measurements are handled. Those details are not reported in the registry record for the primary MMRM.

The other posted outcomes use different summaries and analysis methods. CGI-S is reported as a least-squares mean with standard error, while CGI-I is reported as a mean with standard deviation. The registry reports MMRM for the combined CGI-S analysis and Cochran-Mantel-Haenszel for the combined CGI-I analysis. In general, the Cochran-Mantel-Haenszel method compares groups across strata, but the registry record does not report the strata or the exact coding used for CGI-I. The methods and summaries should therefore be read as reported rather than used to infer unreported details of the analysis plan.

11. Confidence Intervals and Precision

The sponsor’s primary analysis reports a least-squares mean difference of -4.4 MADRS units for the combined liafensine groups versus placebo, with a 95% confidence interval from -7.6 to -1.3. The interval is entirely below zero and spans a range of possible differences that is meaningful for judging precision: it is compatible with a more modest contrast as well as a larger contrast than the point estimate. It does not imply that the true effect has a 95% probability of lying inside those limits.

The recalculated unadjusted comparison for the combined group also estimates a difference of -4.4, with a 95% confidence interval from -7.515 to -1.285. This interval is based on posted group summaries and a simple unadjusted comparison, not the sponsor’s MMRM. The two intervals are close in location and width, but they should not be regarded as estimates from one identical analysis. The registry record does not report additional model details needed to explain their exact relationship.

For the separate dose comparisons, the recalculated confidence intervals also point in the same direction as the combined comparison. The posted data do not establish whether the apparent similarity between dose estimates reflects a genuine dose pattern; the record does not report a formal dose-response analysis. Confidence intervals communicate uncertainty around estimated group contrasts, not variation among individual patients. The registry record also does not report a clinically meaningful threshold for MADRS, so precision relative to statistical zero should not be confused with precision 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 sample-size figures are planning illustrations based on the observed effect, not a description of the trial’s original design calculation. They give participants per arm for 80% power and participants per arm for 90% power. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. These figures are not observed power.

For the combined primary comparison, planning for 100% of the observed effect gives 84 participants per arm for 80% power and 112 participants per arm for 90% power. Planning for 75% of the observed effect gives 148 participants per arm for 80% power and 198 participants per arm for 90% power. Planning for 50% of the observed effect gives 333 participants per arm for 80% power and 445 participants per arm for 90% power. The separate dose comparisons have their own planning figures because their estimated effects differ slightly.

In general, the 80% column describes a planning target for the chance of detecting the specified effect under the stated assumptions, and the 90% column uses a higher target. Smaller effects are harder to distinguish from random variation and therefore require more participants per arm under these assumptions. The figures are conditional on the planning model and the chosen effect fraction; they do not show the probability that the observed trial result is correct. The original trial’s sample-size assumptions and justification are not reported in the registry record. For time-to-event outcomes the described convention is total events, but that convention does not apply to this continuous primary endpoint.

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 = -4.400; pooled standard deviation σ = 10.126.
2. n = 2 × (1.96 + 0.842)² × 10.126² / -4.400² = 83.1, rounded up to 84 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 <- -4.4                      # observed difference in means
sigma <- 10.1262203319591                     # 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(5113771)
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 = 84 (80% power), 112 (90% power)
 75% of observed effect: n per arm = 148 (80% power), 198 (90% power)
 50% of observed effect: n per arm = 333 (80% power), 446 (90% power)
power.t.test: 84.1 
Simulated power at n = 84 per arm: 0.804 (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 sponsor’s reported primary comparison is an MMRM analysis of the combined liafensine groups versus placebo. It gives a least-squares mean difference of -4.4, a 95% confidence interval from -7.6 to -1.3, and identifies the test as a superiority analysis. The registry reports the sponsor’s p-value as 0.006; that value is reported as part of the sponsor analysis, not the recalculation.

The recalculated comparison from posted summary statistics is an unadjusted mean difference of -4.4, with a 95% confidence interval from -7.515 to -1.285 and p = 0.006. The numerical estimates are similar, but the calculations are not the same: the sponsor’s result is model-based, whereas the recalculation uses summary statistics and does not reproduce the MMRM. A p-value from one method should not be substituted for the other.

The registry record does not flag a discrepancy between the results. The sponsor’s analysis is therefore best read as the trial’s reported inferential analysis, while the recalculation is a transparent summary-statistic comparison. The registry record does not provide enough information to assess all model specifications or independently reproduce the sponsor’s analysis.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several hypotheses are tested, such as multiple treatment doses, outcomes, time points, or subgroups. If each test is interpreted independently without an appropriate strategy, the chance of at least one apparently notable result can increase. This general concern applies to interpreting the separate dose comparisons and the additional CGI-S and CGI-I analyses alongside the primary comparison. It does not invalidate any individual estimate, but it affects how strongly a set of separate p-values can support a broad conclusion.

The registry record does not report a multiplicity adjustment, a testing hierarchy, or whether the primary outcome was the only confirmatory test. It also does not report whether the separate 1 mg and 2 mg comparisons were prespecified as inferential tests or whether the combined comparison was the sole primary contrast. The reported p-values should therefore be understood as the values attached to their stated analyses, without assuming control of error across all outcomes and comparisons.

Missing data are another important consideration for a repeated-measures endpoint. The sponsor reports an MMRM, but the registry record does not describe the amount or pattern of missing MADRS measurements, how participants with incomplete follow-up were handled, or the model’s assumptions about missingness. It does not report sensitivity analyses for alternative missing-data assumptions. These details matter because an analysis can depend on assumptions about how unobserved outcomes relate to observed outcomes.

The registry record identifies DGM4-positive patients as the primary endpoint population, but does not report the precise analysis-population definition, such as whether the posted analysis included all randomized participants meeting a particular rule. It also does not report the number screened for the biomarker, the number found positive, or results for the DGM4-negative subgroup. These omissions restrict assessment of selection, exclusions, and generalizability from the registry record alone.

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: depression (3.0%), ankle fracture (1.6%), hip fracture (1.6%), rib fracture (1.6%), abortion spontaneous (1.6%).

The registry lists serious adverse events and deaths by treatment group, with an at-risk count for each group. For serious adverse events, the posted figures are 0 among 65 at risk in the liafensine 1 mg group, 3 among 64 at risk in the liafensine 2 mg group, and 2 among 67 at risk in the placebo group. The death counts are 0, 1, and 0, respectively. These are descriptive registry counts; they do not establish that treatment caused or prevented an event.

The registry also lists other adverse events affecting 36 among 65 at risk in the 1 mg group, 37 among 64 at risk in the 2 mg group, and 38 among 67 at risk in the placebo group. The record lists Depression at 3.0% among the top serious-event terms and Ankle Fracture, Hip Fracture, Rib Fracture, Abortion Spontaneous, and Intestinal Haematoma at 1.6% each. The record does not provide enough detail in these summaries to assess event timing, duration, severity beyond the serious classification, or causal attribution. Counts across groups are not a substitute for a full safety analysis, particularly when denominators and exposure periods may differ. The registry record does not report additional safety-analysis methods.

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

The posted primary results show a larger average reduction in MADRS score for the combined liafensine groups than for placebo among DGM4-positive patients over the reported baseline-to-Day-42 period. The sponsor’s MMRM estimate and the simple recalculated comparison are both -4.4 units, and their reported confidence intervals lie below zero. This describes the observed group contrast and its statistical uncertainty; it is not a statement about what any individual participant experienced.

The result does not establish the clinical importance of the mean difference. The registry record does not report a threshold for a meaningful MADRS change, and group means cannot show the proportion of patients who improved by a particular amount. Nor do the posted results establish how the findings apply to DGM4-negative patients, other populations, or outcomes beyond the reported assessment period.

The result also does not establish a causal explanation beyond the randomized comparison as reported. Interpretation depends on the trial’s conduct, analysis population, measurement, and missing-data assumptions. Several of these details are not reported in the registry record. Although the registry identifies the study as randomized and double-blind, that description alone does not provide all information needed to judge allocation, masking success, adherence, or departures from the intended treatment.

The secondary findings and safety counts add context but should be interpreted separately from the primary estimate. The registry does not report a multiplicity strategy for the secondary analyses, and the safety summaries do not establish event attribution. Overall, the posted results support a statistical difference in the reported primary comparison under the sponsor’s analysis, while leaving clinical meaning and several aspects of robustness and generalizability unresolved.

18. Why This Trial Matters Statistically

Statistically, this trial is notable for making its primary question biomarker-specific. The primary outcome is assessed in DGM4-positive patients, rather than being described only as an average across everyone with treatment-resistant depression. That focus makes the target population explicit, while also making the unreported details about biomarker screening, subgroup definition, and outcomes outside the positive subgroup important to interpretation.

The study reports both separate dose-group summaries and a combined liafensine comparison. That distinction matters because the combined contrast addresses the registry’s stated primary objective, while the separate dose estimates are not independent confirmations: they share the placebo group and contribute to the combined group. Keeping the comparisons distinct helps avoid treating related analyses as a collection of separate trials.

The registry also makes it possible to compare a sponsor-reported MMRM estimate with a simple unadjusted recalculation from group summaries. The estimates are close in value, but their methods differ. This illustrates why a numerical match does not mean that the underlying analyses are interchangeable, and why model specifications and analysis populations are relevant to independent assessment.

Finally, the results highlight the difference between statistical evidence and clinical interpretation. The primary confidence interval excludes zero, but the registry record does not report a clinical-importance threshold or enough detail about missing data and multiplicity to settle all interpretive questions. Reading the estimates together with their precision and reporting limitations is more informative than relying on a p-value alone.

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.

Several trial-specific details needed for a full appraisal are not reported in the registry record. These include the original sample-size justification, the randomization sequence and concealment methods, the roles covered by double masking, the full eligibility criteria, and the biomarker assessment process. The brief summary describes an approximate enrollment of 180, while the registry lists enrollment as 197; the record does not explain the difference.

The primary population is DGM4 positive, but the registry does not report the number screened, the number meeting the biomarker definition, or the primary outcome among DGM4-negative patients. Without those details, the registry results do not show how biomarker status related to treatment contrast across the broader study population. The record also does not report the exact analysis-population rule or participant flow through the outcome assessments.

The MMRM is identified by name, but its model specification and missing-data assumptions are not reported. There are no posted sensitivity analyses in the registry record described for alternative missing-data assumptions. The registry record also does not state a multiplicity adjustment or testing hierarchy for the separate dose comparisons and secondary endpoints. These omissions limit assessment of the robustness and confirmatory status of the reported p-values.

Finally, the primary endpoint is measured through Day 42, and the registry summaries do not report a clinical-importance threshold or individual-level response distribution. The safety summaries provide group counts but not detailed event narratives, timing, exposure-adjusted comparisons, or causal assessment. These limitations mean the registry record supports a focused reading of the reported estimates, not a complete evaluation of long-term outcomes, patient-level benefit, or safety attribution.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Biomarker-Guided, Randomized, Placebo-Controlled Efficacy and Safety Study of Liafensine in Patients With TRD — NCT05113771.
• Registry record last updated 2025-05-15; status last verified 2025-03.
• Results first posted 2025-05-15.
• 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

NCT05113771 is a completed, randomized, double-blind, placebo-controlled, parallel-group Phase 2 study of liafensine in treatment-resistant depression, sponsored by Denovo Biopharma LLC. The primary question concerns change in MADRS score from baseline to Day 42 among DGM4-positive patients. The sponsor’s reported MMRM comparison of combined liafensine doses with placebo gives a least-squares mean difference of -4.4, with a 95% confidence interval from -7.6 to -1.3.

The posted group summaries and recalculated unadjusted comparison are consistent in direction with the sponsor estimate. Two secondary outcomes and descriptive safety counts are also reported. Important details—including the original sample-size justification, analysis-population definition, missing-data assumptions, and multiplicity strategy—are not reported in the registry record. The results should therefore be read as estimates for the specified population, outcome, and assessment period, with those reporting limitations in view.