Extension Study of Pimavanserin for the Adjunctive Treatment of Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Extension Study of Pimavanserin for the Adjunctive Treatment of Schizophrenia (NCT03121586), a na phase 3 study. The primary outcome is Treatment Emergent Adverse Events.

2. Clinical Question

3. Trial Design

This was a phase 3, open-label extension study of pimavanserin as an adjunct to background antipsychotic treatment for schizophrenia. The record identifies the design as single-group, with allocation not applicable and no masking. The listed pimavanserin regimens were 34 mg, 20 mg, or 10 mg, taken once daily by mouth alongside a background antipsychotic. The record does not explain how dose selection or changes were handled.

The study was intended to assess long-term safety and tolerability. Its primary outcome was the count of participants with treatment-emergent adverse events in the Safety Analysis Set. The planned observation period was 52 weeks of treatment followed by 30 days of follow-up, described as a planned total of 56 weeks. The study was terminated, and the registry notes that this shortened the reporting period. Adverse events were assessed to the end of treatment, at a mean duration of 319 days (or 46 weeks). The reason for termination and the timing of the decision are not reported in the registry record.

Participants were grouped according to rollover from ACP-103-034, ACP-103-038, or ACP-103-064. The adverse-event summaries give separate at-risk and affected counts for these groups. Because these are rollover categories rather than randomized treatment and control arms, differences between their counts cannot be interpreted as comparative treatment effects. The record does not report a concurrent comparator group or an allocation procedure.

Enrollment is listed as 995. That registry figure alone does not indicate how many participants entered the Safety Analysis Set, how many were followed for the planned duration, or how many contributed to a particular outcome. Participant flow and group-specific outcome denominators beyond the adverse-event summaries are not reported in the registry record.

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: Safety 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.

Population: Safety Analysis Set

6. Primary Endpoint

The registry lists the primary outcome as Treatment Emergent Adverse Events, measured over Treatment period (52 weeks) and follow-up period (30 days): planned total of 56 weeks. Since the study was prematurely terminated, the reporting period was shortened; AEs were assessed to the end of treatment, at a mean duration of 319 days (or 46 weeks)..

The primary endpoint is the number of participants with treatment-emergent adverse events. It is a count of people who experienced an event, rather than a count of all event episodes. The registry identifies the measurement unit as participants and names the Safety Analysis Set as the population for this outcome. It does not provide an overall count of participants with treatment-emergent adverse events, nor a corresponding proportion or uncertainty interval.

Counting participants with events is a way to describe how often an adverse event occurred among those included in a safety analysis. It can provide a broad account of safety experience during the observed period, but a participant count does not by itself describe the severity, duration, recurrence, or timing of every event. Nor does it establish that treatment caused an event. Those questions require details about event definitions, assessment, attribution, and analysis, which are not reported in the registry record for this primary outcome.

The intended observation period included 52 weeks of treatment and 30 days of follow-up, planned as a total of 56 weeks. The study was terminated early, shortening the reporting period. The registry says adverse events were assessed through the end of treatment at a mean duration of 319 days (or 46 weeks). This shorter, variable observation experience matters when interpreting counts: participants may not all have been observed for the same duration, and the registry does not give individual follow-up times or a time-adjusted estimate for the primary outcome.

Adverse-event monitoring is relevant to a long-term extension because the study aimed to characterize safety and tolerability during continued adjunctive treatment. However, the endpoint as reported cannot establish comparative safety without a suitable control group, and the overall primary result is not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods help determine what a study’s results can and cannot establish. In this record, the design information is limited: the study is identified as single-group, open-label, and phase 3, with allocation not applicable and masking absent. The primary outcome is a participant count. The registry record does not report an analysis plan, inferential methods, confidence intervals, hypothesis tests, or a statistical analysis of the primary outcome.

What does a participant count describe?

A count of participants with an event records how many people in a defined analysis population experienced at least one event meeting the outcome definition. It differs from an event count, which can include multiple episodes in the same person. A participant count can be useful for a straightforward description of safety experience, but its interpretation depends on who was included, how events were defined, and how long participants were observed. For the primary endpoint, the registry names the Safety Analysis Set but does not report the overall count or further details about the population’s construction.

The record gives group-level counts for serious adverse events, other adverse events, and deaths. Those summaries are descriptive. A count alone does not express statistical uncertainty, and a difference in counts across rollover groups is not automatically evidence of a meaningful difference in risk. In general, statistical uncertainty reflects how much an estimate might vary across repeated samples under a specified model. A confidence interval is one way to communicate that uncertainty; no confidence interval for the primary outcome is reported in the registry record.

Why does the study design matter?

A randomized comparison can help distinguish treatment-related differences from differences in the people assigned to each group, subject to the design and conduct of the study. This study has no randomized control group. The rollover groups are defined by the earlier study from which participants entered, not by a randomized assignment described in this record. Their adverse-event counts therefore do not provide a like-for-like control comparison. Background antipsychotic treatment was also part of the listed regimen, so the record concerns adjunctive treatment rather than pimavanserin in isolation.

Open-label status means that masking was absent. In general, knowledge of treatment can influence reporting, assessment, or management of outcomes, particularly when outcomes depend on observation or judgment. This does not establish that such influence occurred in this study; it identifies a design consideration. The registry record does not describe procedures used to reduce potential assessment or reporting effects.

How does follow-up affect interpretation?

The planned period was 52 weeks of treatment and 30 days of follow-up, with a planned total of 56 weeks. The record says premature termination shortened the reporting period and that adverse events were assessed to the end of treatment at a mean duration of 319 days (or 46 weeks). In general, a participant observed for less time has less opportunity for an event to be recorded than someone observed longer. A simple count does not adjust for differences in time under observation. Time-to-event or person-time approaches can address aspects of unequal follow-up, but the registry record does not report such an analysis for the primary outcome.

Variation in follow-up also means that an event count should be read in the context of the observation period, rather than as a rate over a common fixed duration. The registry does not report individual follow-up distributions, the number completing planned treatment, or a time-adjusted estimate. The mean duration is informative about the shortened reporting period but does not establish the experience of each participant.

What can be inferred from the adverse-event summaries?

The serious-event summaries list participants at risk and participants affected within each rollover group. They permit a descriptive reading of the reported counts within those groups. They do not establish causality, because an event occurring during treatment is not necessarily caused by treatment. General causal assessment would require information about timing, clinical review, alternative explanations, and the study’s attribution procedures. Those details are not reported in the registry record.

Likewise, a percentage attached to a serious-event term should not be treated as a treatment-versus-control estimate. The record lists schizophrenia at 0.9%, and chest pain, drowning, cellulitis, pneumonia, and sepsis at 0.3% each. Without a randomized comparator and full details of the denominator and event coding for those terms, these values are descriptive entries, not evidence of comparative risk. The registry record also does not report confidence intervals or tests for these terms.

Why should absent analyses be noted?

The registry lists no analyses. This means inferential results, adjusted estimates, and analysis procedures are not reported in the registry record. It does not justify assuming that an analysis was or was not performed outside the record. Readers should distinguish between information that is reported and information that is unavailable in the registry entry. In particular, no statistical significance claim, treatment-effect estimate, or precision statement can be made for the primary endpoint from the reported material.

General sample-size planning methods estimate how much information a future study might need under specified assumptions. Such planning is not an observed result and does not establish the power of a completed study. The registry record does not report a trial-specific statistical plan or a sample-size justification. It also does not report an inferential primary analysis. Any planning illustration should therefore be interpreted as a hypothetical calculation under its stated assumptions, not as evidence that this extension study had a particular probability of detecting an effect.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Rollover From ACP-103-034", "Rollover From ACP-103-038", "Rollover From ACP-103-064")
n      <- c(323, 325, 347)
events <- c(128, 111, 110)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Rollover From ACP-103-034  323  128  39.6  34.3  45.2  34.4  45.1  34.4  45.1
ARM  Rollover From ACP-103-038  325  111  34.2  29.0  39.6  29.2  39.5  29.2  39.5
ARM  Rollover From ACP-103-064  347  110  31.7  26.8  36.9  27.0  36.8  27.0  36.8

11. Multiplicity, Missing Data and Analysis Populations

The registry record identifies Treatment Emergent Adverse Events as the primary outcome and lists one outcome measure, but it does not report an analysis plan or multiplicity procedure. Multiplicity concerns the increased chance of false-positive findings when many outcomes, comparisons, or repeated tests are assessed. This is a general statistical issue; the available record does not establish which analyses were performed or whether any adjustment was used. No significance tests for the primary endpoint are reported.

The primary outcome result itself is not reported in the registry record. The record also does not state how missing outcome information was handled, whether participants with incomplete follow-up were excluded or retained, or whether any imputation method was used. Those details matter because missingness can affect estimates when the availability of outcome information is related to participants’ experiences or follow-up. No such relationship can be inferred from the record.

The Safety Analysis Set is named as the analysis population for the primary outcome, but its membership and construction are not described. Enrollment is listed as 995; this does not establish the number included in the safety analysis or the number with evaluable follow-up. The registry record does not report participant flow, exclusions, treatment discontinuations, or completeness of event collection by group.

The study’s premature termination shortened the reporting period. The reason for termination, any interim analyses, and any effect of the termination on the planned statistical analysis are not reported in the registry record. These omissions prevent assessment of whether the observed reporting period or available outcome data departed from a prespecified analysis plan. No protocol-level details should be inferred from the study description alone.

12. 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: schizophrenia (0.9%), chest pain (0.3%), drowning (0.3%), cellulitis (0.3%), pneumonia (0.3%).

The serious adverse-event summaries are presented by rollover group. For Rollover From ACP-103-034, 13 participants were affected among 323 at risk. For Rollover From ACP-103-038, 4 were affected among 325 at risk. For Rollover From ACP-103-064, 4 were affected among 347 at risk. These counts describe the entries for each group; they are not results from randomized treatment and control arms.

In general, a serious adverse event is a category distinguished by seriousness criteria, not simply by how often an event occurs. The registry record does not give the criteria or additional event-level details in the information reported. The listed serious-event terms include schizophrenia at 0.9%, and chest pain, drowning, cellulitis, pneumonia, and sepsis at 0.3% each. The record does not provide enough detail to interpret these percentages as comparative rates or to establish treatment causation.

Deaths are listed as 2 in the ACP-103-034 rollover group, 0 in the ACP-103-038 rollover group, and 0 in the ACP-103-064 rollover group. These are descriptive counts, not estimates of treatment-attributable mortality. The registry does not report a controlled comparison or details on attribution. The counts should therefore be read as reported safety events during the observed period, with the study’s shortened follow-up in mind.

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

The record documents an open-label extension study intended to evaluate long-term safety and tolerability of pimavanserin as adjunctive treatment for schizophrenia. It reports serious adverse-event counts, other adverse-event counts, deaths, and selected serious-event terms for rollover groups. These entries describe events recorded in the registry; they do not establish that pimavanserin caused those events or that the observed counts represent a comparative treatment effect.

The primary outcome is the number of participants with treatment-emergent adverse events, but its overall result is not reported in the registry record. The record therefore does not give a complete numerical answer to its stated primary question. Nor does it provide an uncertainty interval or inferential analysis for that outcome. A lack of a reported result should not be interpreted as a result of zero events or as evidence of no effect.

The study had no randomized control group, and the rollover categories are not like-for-like control arms. Their counts cannot establish whether event occurrence differed because of treatment. The open-label design also means masking was absent. These features do not invalidate the descriptive safety information, but they limit what can be inferred from it about causal effects or comparative risk.

The intended treatment and follow-up period was 52 weeks plus 30 days, planned as 56 weeks in total. Premature termination shortened reporting; events were assessed to the end of treatment at a mean duration of 319 days (or 46 weeks). The reported experience therefore does not represent complete observation through the full planned period for every participant. The record does not report individual follow-up times or a time-adjusted primary estimate.

In short, the registry offers selected descriptive safety information from an extension study and identifies the primary safety question. It does not provide the overall primary endpoint result, a controlled efficacy comparison, or enough statistical detail to support broader causal or precision claims. The absence of those details in the registry record is not itself evidence for or against a treatment effect.

14. Why This Trial Matters Statistically

This study raises statistical issues that commonly arise in long-term extension research. Its single-group, open-label design means the main contribution is descriptive safety information rather than a randomized estimate comparing treatment with a control. Understanding that distinction is essential: counts observed during treatment can characterize reported events, but cannot by themselves identify what would have happened under another treatment or no study treatment.

The rollover structure also matters. Participants are grouped by the earlier study from which they entered, and the adverse-event summaries are organized accordingly. These groups may be useful for presenting the recorded safety experience, but they are not randomized arms. A numerical contrast between the groups would not remove confounding or establish a treatment-related difference. The registry record does not describe an adjusted comparative analysis.

Follow-up duration is another key issue. The study planned 52 weeks of treatment and 30 days of follow-up, but premature termination shortened the reporting period. The mean duration to the end of treatment was 319 days (or 46 weeks). For safety outcomes, the length of observation affects the opportunity to identify events. A count without a time-adjusted analysis should be read alongside the duration actually described, rather than assumed to represent a complete planned period for every participant.

Finally, the primary endpoint is named but its overall result is not reported in the registry record. This illustrates why readers should distinguish a registered outcome from a reported analysis. The adverse-event summaries provide some specific counts, but not a complete primary-outcome result or measures of statistical precision. Recognizing what is and is not reported is central to responsible interpretation of clinical trial records.

15. 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 limitations specific to this study constrain interpretation. It was a single-group, open-label extension with no randomized control group. The rollover groups from ACP-103-034, ACP-103-038, and ACP-103-064 do not provide a like-for-like treatment comparison. Differences in their reported counts cannot be attributed to pimavanserin or interpreted as evidence of comparative safety.

The study was terminated, and the registry states that the reporting period was shortened. The planned period was 52 weeks of treatment plus 30 days of follow-up, while adverse events were assessed to the end of treatment at a mean duration of 319 days (or 46 weeks). The reason for termination is not reported in the registry record. Nor are individual follow-up durations, the number who completed planned treatment, or the effect of early termination on outcome ascertainment reported.

The overall result for the primary outcome, treatment-emergent adverse events, is not reported in the registry record. The record also does not report confidence intervals, hypothesis tests, analysis methods, or the construction of the Safety Analysis Set. Although enrollment is listed as 995, the number contributing to the primary endpoint cannot be established from that figure alone.

The adverse-event summaries are descriptive and do not establish causation. Details about event attribution, event timing, recurrence, and missing safety information are not reported. The record also does not describe a statistical analysis plan, interim analyses, sample-size justification, or procedures for handling missing data and multiplicity. These limits mean that the available information supports a cautious account of recorded events, not a definitive causal or comparative conclusion.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Extension Study of Pimavanserin for the Adjunctive Treatment of Schizophrenia — NCT03121586.
• Registry record last updated 2025-08-24; status last verified 2025-07.
• Results first posted 2025-08-24.
• 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.

19. Record Summary

NCT03121586 is a phase 3, open-label, single-group extension study of pimavanserin for adjunctive treatment of schizophrenia. The listed regimen was once-daily oral pimavanserin at 34 mg, 20 mg, or 10 mg with a background antipsychotic. Enrollment is listed as 995, and the study status is terminated.

The primary endpoint was the number of participants with treatment-emergent adverse events in the Safety Analysis Set. The planned observation period was 52 weeks of treatment plus 30 days of follow-up, for a planned total of 56 weeks. Premature termination shortened the reporting period; adverse events were assessed to the end of treatment at a mean duration of 319 days (or 46 weeks). The overall primary-outcome result and statistical analyses are not reported in the registry record.

Serious adverse events, other adverse events, and deaths are reported descriptively by rollover group. These groups do not form a randomized comparison, and the reported counts do not establish causation or comparative treatment effects.