Proof-of-Concept Trial of CERC-501 Augmentation of Antidepressant Therapy in Treatment-Resistant Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Proof-of-Concept Trial of CERC-501 Augmentation of Antidepressant Therapy in Treatment-Resistant Depression (NCT01913535), a randomized phase 2 study. The primary outcome is Change in Hamilton Rating Scale for Depression - 6 Items (HAM-D-6).

2. Clinical Question

3. Trial Design

The registry identifies NCT01913535 as a Phase 2 randomized, quadruple-masked, parallel-group proof-of-concept trial sponsored by Massachusetts General Hospital. Its title describes CERC-501 augmentation of antidepressant therapy in treatment-resistant depression. The recorded start date is 2013-09-12, the primary completion date is 2016-01-22, and the trial status is terminated. Enrollment is listed as 8.

The arm descriptions have a sequential structure. The low-dose drug-drug sequence received CERC-501 10.0 mg/day for three days in Phase 1 and for three subsequent days in Phase 2. The high-dose drug-drug sequence used 20.0 mg/day over the same phase durations. Other sequences began with placebo for three days in Phase 1 and then received either low-dose CERC-501, high-dose CERC-501, or placebo for three days in Phase 2. The record also describes the primary endpoint population through the sequential parallel comparison design analyses through 72 hours. It says placebo non-responder data from Phase 2 were pooled with Phase 1 data, and that all Phase 1 placebo participants were non-responders based on HAM-D-6 reduction and MADRS criteria.

The posted HAM-D-6 summaries list CERC-501 with n=7 and placebo with n=5. These are the group counts attached to those endpoint summaries; the registry record does not explain how those counts relate to overall enrollment, the sequences, or the analysis population. The intended sample-size justification, detailed randomization procedure, and further operational details are not reported in the registry record. The record contains no posted statistical analyses for this endpoint, so the design description should not be mistaken for evidence that a planned comparison was completed or reported.

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.

Before assignment: Only 8 participants were randomized because the study terminated early due to slow enrollment.

Phase 1 (Days 0-2)

Phase 2 (Days 3-5)

Analysis population for the primary outcome, as defined in the registry: In the sequential parallel comparison design (SPCD) analyses through 72 hours, placebo non-responder data is included from phase 2 and is pooled with phase 1 data. All placebo participants from phase 1 were non-responders based on HAM-D-6 reduction and MADRS criteria.

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: Only 8 participants were randomized because the study terminated early due to slow enrollment.

6. Primary Endpoint

The registry lists the primary outcome as Change in Hamilton Rating Scale for Depression - 6 Items (HAM-D-6), measured over Baseline and 72 hours after initiating treatment.

The primary endpoint is change in the six-item Hamilton Rating Scale for Depression, or HAM-D-6, from baseline to 72 hours after initiating treatment. A clinician completes the instrument using a structured interview guide and assesses symptoms within the past 24 hours. The registry states that the structured interview has been validated for time frames shorter than one week. Scores range from 0 (normal) to 22 (severe), with higher scores indicating more depression symptoms.

Change is used to describe how symptom scores differ from a starting point over the specified follow-up. For this trial, the registry says baseline was estimated as the average during the screening period rather than taken from a single cross-sectional baseline assessment. That choice is described as an effort to better estimate baseline; the registry does not provide further details of the calculation.

The endpoint is reported as a median with a full range. The CERC-501 group has a median change of -3.0 units, with a full range from -13.5 to 0.0; the placebo group has a median change of 0.0 units, with a full range from -6.0 to 1.5. A negative change is in the direction of a lower HAM-D-6 score, while a positive change is in the direction of a higher score. These summaries describe the recorded group values, not the precision of a treatment comparison. No confidence interval or inferential analysis for the contrast is posted in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies the primary endpoint summary as a median and full range. It describes a randomized, quadruple-masked, parallel-group Phase 2 trial and a sequential parallel comparison design analysis through 72 hours. The listed summaries are CERC-501, n=7, median change -3.0 units, full range -13.5 to 0.0; and placebo, n=5, median change 0.0 units, full range -6.0 to 1.5. The record does not post an inferential analysis for these values.

What does the median describe?

General statistical context: the median is the middle of an ordered set of observations. It describes a central location and is less affected by an unusually high or low observation than the arithmetic mean. With small groups, however, the median still depends on a limited number of observations. A median alone does not show the full shape of the data, and it does not quantify uncertainty about a group contrast.

The reported full range gives the smallest and largest observed changes in each group. It conveys how far apart the observed extremes are, but it does not show how values are distributed between those extremes. It is also not a confidence interval. A confidence interval is a statistical range intended to express uncertainty around an estimate, under specified assumptions and a specified method. No confidence interval for either median or for a treatment contrast is reported in the registry record.

How should change scores be read?

The endpoint compares the follow-up HAM-D-6 score with an estimated baseline. The instrument score ranges from 0 (normal) to 22 (severe); higher scores indicate more depression symptoms. Accordingly, a negative change represents a decrease in score, and a positive change represents an increase. The registry states that baseline was estimated using the average during the screening period rather than a single assessment. This baseline definition is relevant when interpreting a change score because the reference value determines what counts as change. The record does not give further calculation details.

What is the role of the sequential design note?

The registry describes analyses through 72 hours using a sequential parallel comparison design. Its population note says Phase 2 placebo non-responder data were pooled with Phase 1 data, and that all Phase 1 placebo participants were non-responders according to HAM-D-6 reduction and MADRS criteria. In general, a sequential comparison design can use information from a later phase according to response status in an earlier phase. The exact analysis model, pooling weights, handling of repeated observations, and inferential procedure are not reported in the registry record. The posted group summaries therefore cannot be used to recreate the design-specific estimate.

Why are inferential results important?

General statistical context: a treatment contrast estimates how outcomes differ between groups, while a confidence interval describes uncertainty around that contrast under the chosen model. A p-value, when reported, assesses how compatible the data are with a specified null hypothesis under the test assumptions; it is not a measure of clinical importance. None of these inferential quantities is posted for the primary result. The record lists zero analyses, so no reported test or adjusted model can be discussed. The summaries should not be converted into a claim of statistical significance or a precise effect.

What do planning illustrations mean?

General statistical context: sample-size planning estimates how many participants might be needed to detect a specified effect under assumed design conditions. Such estimates depend on assumptions, including the significance level, allocation, statistical approximation, and effect targeted. The registry’s planning description specifies a two-sided significance level of 0.05, equal allocation, a normal approximation, and an effect defined as a stated fraction of the observed effect. It also distinguishes participants per arm for non-time-to-event outcomes from total events for time-to-event outcomes. These are planning assumptions, not results from the trial and not observed power. The record does not report the trial’s own sample-size justification, and the planning illustration does not explain the enrollment of 8.

Over-dispersion is a general issue in count outcomes: it occurs when variability exceeds what a simpler count model assumes, and can make standard errors too small if ignored. It is not a reported feature of this continuous scale endpoint, and the registry record does not identify a model for the HAM-D-6 analysis. Similarly, general statistical methods such as covariate adjustment or robust standard errors should not be attributed to this trial unless the record reports them. It does not.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 13 outcome measures and 12 posted secondary outcomes, in addition to the primary HAM-D-6 endpoint. The secondary outcome list includes HAM-D-6 at Day 20, HAM-D-6 response, changes in MADRS, CGI-S, SDQ, PSS, and PAS, and CGI-I. The registry record does not report a multiplicity adjustment strategy or identify which outcomes, time points, or analyses were considered confirmatory. General statistical context: examining multiple outcomes can increase the chance of at least one result appearing notable by chance; methods for addressing this depend on the analysis plan. No adjustment should be presumed.

Although the primary result is posted as medians and full ranges, there are zero analyses listed. A statistical analysis plan, interim analysis plan, and rules for choosing analysis populations are not reported in the registry record. The record does describe pooling of Phase 2 placebo non-responder data with Phase 1 data for the sequential parallel comparison design analysis through 72 hours. It does not provide the full analysis algorithm or the number of participants contributing at each stage beyond the posted endpoint group counts.

Missing-data amounts, reasons for missingness, imputation procedures, and sensitivity analyses are not reported in the registry record. The registry’s terminated status and enrollment of 8 are not enough to establish whether data were missing, why the trial terminated, or whether the posted endpoint summaries include every participant. The posted analysis population and the relationship between the endpoint counts and overall enrollment are also not explained. These gaps limit assessment of both representativeness and robustness.

11. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent non-serious adverse event terms

The adverse-event record lists no serious adverse events among participants at risk in the low-dose drug group, high-dose drug group, or placebo group. The respective serious-event counts are 0 among 3, 0 among 4, and 0 among 4. Death counts are not reported in the registry record. No top serious adverse-event terms are listed.

The record also lists other adverse events affecting 2 participants in each group: the low-dose drug group had 2 affected among 3 at risk, the high-dose drug group had 2 among 4, and placebo had 2 among 4. These are counts, not evidence that event types or severity were the same across groups. The registry record does not give event-level details in the information.

General statistical context: a zero count in a small group does not establish that an event cannot occur, and counts from small groups provide limited precision for comparing risks. These figures describe what was recorded for the listed groups; they do not establish a comparative safety profile. The registry record does not report the reasons for termination, and termination should not be attributed to safety on the basis of these counts.

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

The result establishes that the registry posted descriptive HAM-D-6 change summaries for two groups. The CERC-501 summary is a median change of -3.0 units, with a full range from -13.5 to 0.0, among 7 participants. The placebo summary is a median change of 0.0 units, with a full range from -6.0 to 1.5, among 5 participants. Because lower scores indicate fewer symptoms, the negative median indicates a decrease in the scale score within that group. These summaries do not, by themselves, establish a causal or statistically supported difference between groups.

The registry does not post a confidence interval, p-value, or model-based estimate for the between-group contrast. It also does not report the full statistical procedure used for the sequential parallel comparison design. The group summaries therefore cannot show how precisely a comparative effect was estimated or how sensitive it might be to modeling choices, participant inclusion, or missing data. The ranges describe observed extremes rather than the uncertainty around an effect.

The record describes a randomized, quadruple-masked Phase 2 trial, but the reported design label does not fill in the missing analysis details. Overall enrollment is 8, whereas the primary endpoint summaries list group counts of 7 and 5. The registry record does not explain this relationship. It is not appropriate to infer the reasons from the counts alone.

The safety record reports zero serious adverse events in the listed groups, while also recording other adverse events in each group. The small at-risk counts and limited posted details do not support broad conclusions about comparative safety. In the same way, trial termination does not explain itself: its reason is not reported in the registry record.

In short, the record supports a description of the posted scores, ranges, design, and adverse-event counts. It does not supply the inferential evidence needed to determine the precision of a treatment contrast, the influence of multiplicity, or the robustness of the result. These limits matter especially in a small proof-of-concept trial with no posted analysis results.

13. Why This Trial Matters Statistically

This trial is statistically notable because its primary endpoint summaries appear alongside a design that the registry describes as sequential and parallel. The endpoint population note says Phase 2 placebo non-responder data were pooled with Phase 1 data through 72 hours. That feature means the analysis context is not fully represented by a simple reading of two group medians. Without the analysis details, a reader cannot reconstruct the design-specific comparison from the posted summaries.

The listed endpoint counts are 7 for CERC-501 and 5 for placebo, while overall enrollment is 8. The registry record does not explain how the group counts, phase sequences, and enrollment relate. This makes transparent reporting of analysis populations especially important: readers need to know which participants contributed to each estimate and how repeated or pooled observations were handled. Those details are not reported in the registry record.

The record also illustrates the distinction between descriptive and inferential statistics. Medians and full ranges provide a compact account of observed outcomes, but no analysis result or uncertainty measure is posted. The sample-size description concerns planning illustrations based on an observed effect under stated assumptions; it is not evidence of observed power or a substitute for an estimate with uncertainty. The trial’s own sample-size justification is not reported.

Finally, multiple outcome measures are listed, but a multiplicity strategy is not reported. This is a relevant reporting issue because interpretation can depend on whether outcomes were prespecified and how multiple comparisons were treated. Taken together, the design, small posted group counts, missing analysis details, and absence of inferential results make clear why statistical interpretation depends on transparent reporting—not merely on the direction of a pair of medians.

14. 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 most important trial-specific limitation is the lack of posted inferential analysis for the primary endpoint. The registry gives a median and full range for each listed group, but no comparison estimate, confidence interval, or p-value. As a result, precision of the comparative effect cannot be assessed. The group counts are small, and the registry record does not explain why the endpoint summaries list 7 and 5 participants when total enrollment is 8.

Interpretation is further limited by the sequential parallel comparison design description. The record notes pooling Phase 2 placebo non-responder data with Phase 1 data, but does not report the analysis model, weighting, or complete population flow. It is therefore not possible to determine from the posted summary how the displayed groups map to the design’s phases and sequences.

The trial is listed as terminated, but the reason for termination is not reported in the registry record. Missing-data amounts and handling are also not reported, as are a statistical analysis plan, interim analyses, and the trial’s own sample-size justification. The registry does not report whether all participants were included in the endpoint summaries or how the endpoint group counts relate to the enrolled population.

Lastly, the registry lists multiple outcomes but does not report a multiplicity strategy. Safety counts are based on small at-risk groups, and death counts are not reported in the registry record. These limitations do not change the descriptive values that are posted; they restrict the conclusions that can be drawn from them.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Proof-of-Concept Trial of CERC-501 Augmentation of Antidepressant Therapy in Treatment-Resistant Depression — NCT01913535.
• Registry record last updated 2017-07-02; status last verified 2017-05.
• Results first posted 2017-07-02.
• 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.

18. Record Summary

NCT01913535 is a randomized, quadruple-masked, parallel-group Phase 2 proof-of-concept trial of CERC-501 augmentation of antidepressant therapy for treatment-resistant depression. Massachusetts General Hospital is listed as sponsor. The record gives a start date of 2013-09-12, primary completion date of 2016-01-22, terminated status, and enrollment of 8.

The primary endpoint is change in HAM-D-6 from baseline to 72 hours after initiating treatment. The posted summaries are median change -3.0, full range -13.5 to 0.0, n=7 for CERC-501, and median change 0.0, full range -6.0 to 1.5, n=5 for placebo. The registry lists zero analyses, so no inferential comparison is posted.

The endpoint note describes sequential parallel comparison design pooling of Phase 2 placebo non-responder data with Phase 1 data. The registry also lists secondary outcomes and adverse-event counts. It does not report the reason for termination, the trial’s own sample-size justification, missing-data handling, or the complete statistical analysis details.