Study Evaluating Nelotanserin for Treatment of REM Sleep Behavior Disorder in Subjects With Dementia (DLB or PDD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study Evaluating Nelotanserin for Treatment of REM Sleep Behavior Disorder in Subjects With Dementia (DLB or PDD) (NCT02708186), a randomized phase 2 study. The primary outcome is Change in the Frequency of REM Sleep Behavior Disorders (RBD) From Baseline to the End of the Treatment Period (28 Days) Based on a Clinical Evaluator..

2. Clinical Question

The clinical question is specific to short-term change in an event count recorded under sleep-laboratory assessment. It is not a question about all symptoms of DLB or PDD, longer-term outcomes, or outcomes in people without the listed dementias. Nor does the endpoint by itself establish whether a change is noticeable or important to participants. The registry record does not report a threshold for a clinically meaningful change.

3. Trial Design

This was a Phase 2, randomized, double-blind, placebo-controlled study with a parallel-group model. The registry record lists quadruple masking. The intervention was nelotanserin 80 mg, and the comparator was placebo. Enrollment was 34. The listed conditions were dementia with Lewy bodies, REM sleep behavior disorder, and Parkinson’s disease dementia.

The registry record lists the study start as 2016-03, primary completion as 2018-05, and status as completed. Those dates describe the study record and do not, by themselves, provide details about when each participant began or completed treatment. The treatment period relevant to the primary endpoint was 28 days.

The primary endpoint was change in RBD frequency from baseline to the end of treatment, measured as the number of events per 10 minutes. The assessment used video and audio recordings conducted at a sleep laboratory. The reported arm estimates are least-squares means, with standard errors as the dispersion measure. There were 16 participants in the nelotanserin group and 18 in the placebo group for these reported results.

The record identifies Axovant Sciences Ltd. as sponsor. It does not report the original sample-size justification, the intended allocation ratio, or detailed procedures for randomization and masking. The record also does not report the analysis population, handling of missing observations, or a detailed statistical analysis plan. These omissions limit how far the study design can be reconstructed from the registry results alone. The posted results include one outcome measure and no posted secondary outcomes.

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: The Safety population included subjects who took at least 1 dose of double-blind study drug

6. Primary Endpoint

The registry lists the primary outcome as Change in the Frequency of REM Sleep Behavior Disorders (RBD) From Baseline to the End of the Treatment Period (28 Days) Based on a Clinical Evaluator., measured over 28 days.

The primary endpoint is change from baseline in the frequency of REM sleep behavior disorder events at the end of the 28-day treatment period. Frequency is expressed as the number of RBD events per 10 minutes. It is based on video and audio assessment in a sleep laboratory and combines simple or major events with complex events. The description characterizes simple or major behaviors as jerks with higher movement amplitude or intensity, such as whole-body or single-limb jerks. Complex behaviors involve more muscle groups or more complex or violent movements and may enact dream content.

Using an event frequency gives a quantitative way to compare recorded behavior over the assessment period. In general, a baseline-to-endpoint change focuses on how the measure moved during treatment rather than only on its value at the end. The registry reports least-squares mean changes for each arm, accompanied by standard errors. Those are model-based arm summaries as labeled in the record; the detailed model specification is not reported in the registry record.

The endpoint captures a defined, observed behavior under sleep-laboratory conditions. It does not, on its own, describe every aspect of sleep, dementia symptoms, daily functioning, or participant experience. The registry record does not report a threshold that would distinguish a clinically important change from a smaller change, nor does it report additional primary outcomes.

7. Results Posted in the Registry

• Nelotanserin: mean -1.47 versus -0.26 on control; difference -1.21 (95% CI -4.03 to 1.61), standardized effect d = -0.29; z-test p = 0.400.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous measure: a change in the number of RBD events per 10 minutes. The registry labels the arm summaries as least-squares means and reports standard errors. A least-squares mean is a model-based estimate of an average, usually adjusted for terms included in a statistical model. The registry record does not report the model terms or enough detail to determine what adjustments were used, so no particular adjustment should be assumed.

The posted arm estimates are -1.47 for nelotanserin and -0.26 for placebo. The comparison is a mean difference of -1.21. Its direction is consistent with subtracting the placebo estimate from the nelotanserin estimate. This comparison is identified as a recalculated, simple unadjusted calculation from the posted summary statistics, rather than a reported adjusted treatment-effect model. The discrepancy flag is false, so there is no stated conflict requiring a sponsor analysis to take precedence. This unadjusted comparison should not be treated as a substitute for a fully specified trial analysis.

The comparison interval runs from -4.028 to 1.608, and the unadjusted p-value is 0.4. In general, a confidence interval describes a range of effect estimates compatible with the observed data under the method used. Its width communicates precision: a wider interval indicates less precision than a narrower one. Because this interval crosses zero, the data do not rule out a zero difference at the interval’s stated confidence level. The registry record does not identify a confidence level for this interval, so no level should be inferred.

The posted effect size, d, is -0.288. It is a standardized description of the difference relative to a pooled variability estimate, rather than a result expressed in the endpoint’s original units. The pooled standard deviation is 4.204 and is explicitly approximate. The arm-specific standard errors, 1.006 and 1.027, are used in the posted recalculation, and the result should be read as approximate because the pooled variability is approximate. Standardized effects can help compare scales in general, but they do not supply a clinical threshold for meaningful change.

The p-value is an unadjusted value. In general, a p-value measures how unusual data at least as incompatible with a specified null model would be if that model and its assumptions were true. It does not give the probability that the null hypothesis is true, and it does not measure the magnitude or practical importance of an effect. The registry record does not report whether any adjustment for multiple testing was planned or used.

Random allocation is intended, in general, to support a fair comparison by reducing systematic differences between groups at assignment. A parallel-group design keeps participants in their assigned group rather than comparing different treatment periods within the same participant. Quadruple masking indicates that the record describes masking at multiple levels, but the registry record does not specify which individuals were masked or how masking was maintained. These design labels do not resolve the limitations of a small enrollment or missing reporting details.

Over-dispersion is a general concern when count outcomes vary more than a model’s assumed distribution allows. If present and not accounted for, uncertainty estimates can be too small. The endpoint is an event count, but the registry record does not report the model’s distributional assumptions or whether over-dispersion was assessed. It is therefore not possible to say whether this issue affected the posted analysis. The record also does not report transformations, covariates, or sensitivity analyses.

General statistical context: an analysis of change scores can be affected by baseline measurement variability, and repeated event counts may vary across assessment periods. The registry provides the endpoint definition and summary estimates but does not report the detailed analysis structure or participant-level data. This prevents independent evaluation of model fit, baseline balance, or the impact of individual observations. The posted summaries support a descriptive reading of the outcome and its uncertainty, not a reconstruction of every analytic decision.

10. Confidence Intervals and Precision

The reported between-arm mean difference is -1.21 RBD events per 10 min, and its interval extends from -4.028 to 1.608. The interval spans zero. In general, an interval gives a range of effect estimates that remain compatible with the data and the method used to construct it. A narrow interval would indicate greater precision; a broad interval indicates that materially different values remain plausible. The registry record does not state the confidence level attached to this interval.

The interval includes estimates in the negative and positive directions as well as zero. Consequently, the posted results do not pin down the direction of the underlying between-group difference with precision. The estimate itself is centered on a negative difference, but an estimate is not the same as certainty about the true effect. The unadjusted p-value of 0.4 is consistent with the interval’s inclusion of zero, but the p-value and interval answer related, not identical, questions.

For the individual arms, the least-squares mean changes are -1.47 and -0.26, with standard errors of 1.006 and 1.027, respectively. Standard errors describe uncertainty in the estimated arm means; they do not show the full spread of individual participant outcomes. The registry record does not provide arm-specific confidence intervals or participant-level distributions.

The enrollment of 34 and the reported outcome group sizes of 16 and 18 are relevant context for precision. Small studies often leave uncertainty substantial, particularly where outcomes vary. That is a general statistical consideration, not a claim that enrollment alone explains the interval width. The posted pooled standard deviation is approximate, and the record lacks the detailed model specification needed to assess all sources of uncertainty. No clinical-importance boundary is reported, so the interval cannot be classified as excluding or including a prespecified meaningful benefit or harm.

11. 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 participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s original design assumptions. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is set as a stated fraction of the observed effect. The registry record does not report the trial’s original sample-size justification or its original planning assumptions.

For planning at 100% of the observed effect, the illustration gives 190 participants per arm for 80% power and 254 participants per arm for 90% power. At 75% of the observed effect, the figures are 337 participants per arm for 80% power and 452 participants per arm for 90% power. At 50% of the observed effect, they are 759 participants per arm for 80% power and 1,015 participants per arm for 90% power.

In plain language, the 80% power column describes an assumed probability of detecting an effect of the specified size under the stated planning assumptions, if that effect were the true one. The 90% column uses a higher target probability under those same assumptions. Smaller effects need more participants to distinguish them from random variation, which is why the illustrative requirements increase as the target effect becomes a smaller fraction of the observed effect. These figures are not observed power.

They are planning illustrations only. They do not establish what the trial was designed to detect, whether the original design had adequate power, or how a future study should be sized. The registry record does not report the trial’s own sample-size calculation.

12. 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 = -1.210; pooled standard deviation σ = 4.204.
2. n = 2 × (1.96 + 0.842)² × 4.204² / -1.210² = 189.5, rounded up to 190 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 <- -1.21                      # observed difference in means
sigma <- 4.20429733278939                     # 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(2708186)
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 = 190 (80% power), 254 (90% power)
 75% of observed effect: n per arm = 337 (80% power), 452 (90% power)
 50% of observed effect: n per arm = 759 (80% power), 1,015 (90% power)
power.t.test: 190.5 
Simulated power at n = 190 per arm: 0.799 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and no posted secondary outcomes. It also lists zero posted analyses. This does not establish that no additional outcomes or analyses were considered during the study; it means they are not reported in the registry record. The record does not report a prespecified hierarchy of outcomes, a multiplicity adjustment, or a plan for handling multiple statistical tests.

In general, multiplicity arises when several outcomes, time points, subgroups, or analyses are examined. Repeated testing can increase the chance of a false-positive finding if no adjustment or interpretation plan is used. Because only one outcome measure is posted and the detailed analysis plan is not reported, it is not possible to assess the scope of testing or determine whether the unadjusted p-value reflects any broader testing strategy.

The registry record also does not report the analysis population, the number of participants with missing endpoint data, or the method used to handle missing data. It is therefore not possible to determine whether the posted summaries reflect all randomized participants, participants who completed treatment, or another analysis set. Nor can the impact of missing measurements be assessed from these summaries alone.

General statistical context: excluding participants with missing outcomes can change the composition of the groups being compared, while imputing missing values requires assumptions about why data are missing and what the unobserved values might have been. Those assumptions can affect estimates and uncertainty. No particular approach should be attributed to this study because the registry record does not report one.

The detailed statistical analysis plan, model specification, and sensitivity analyses are not reported in the registry record. These gaps limit evaluation of whether the posted estimates are robust to alternative analysis choices. They do not alter the numeric summaries themselves, but they constrain how confidently those summaries can be generalized beyond the outcome and analysis as posted.

14. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: gastroenteritis viral (6.2%), migraine (6.2%), mental status changes (6.2%).

The serious adverse event figures are reported for 16 participants at risk in the nelotanserin group and 18 in the placebo group. Serious adverse events affected 2 participants in the nelotanserin group and 0 in the placebo group. The record reports 0 deaths in each group. Other adverse events affected 11 participants in the nelotanserin group and 8 in the placebo group.

The serious-event terms listed in the record are viral gastroenteritis, migraine, and mental status changes, each with a reported frequency of 6.2%. The registry excerpt does not give an arm assignment for these listed terms, so they should not be attributed to a particular group. The event-term percentages also should not be treated as group-level rates beyond what the record states.

These counts describe events recorded during the trial; they do not establish that an event was caused by the assigned treatment. The groups are small, and the record does not provide detailed event timing, severity breakdowns, exposure-adjusted rates, or a fuller account of how events were collected. The counts therefore provide a limited descriptive safety summary, not a stand-alone comparative safety conclusion.

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

The primary results describe a short-term comparison of changes in a sleep-laboratory measure of RBD event frequency. The reported mean change was -1.47 events per 10 min in the nelotanserin group and -0.26 events per 10 min in the placebo group. The unadjusted between-group estimate was -1.21, with an interval from -4.028 to 1.608 and an unadjusted p-value of 0.4. The interval includes zero, and the estimate is imprecise.

These figures do not establish that the groups differ, nor do they demonstrate that their effects are equivalent. The registry record does not report a clinical-importance threshold that would allow the observed estimate or interval to be judged against a prespecified meaningful change. It also does not report enough detail about the model, missing data, or analysis population to evaluate all assumptions behind the comparison.

The findings apply to the people enrolled in this Phase 2 study and to the measured endpoint over the 28-day treatment period. They do not directly address longer-term outcomes, other measures of sleep or dementia, or people outside the stated population. A sleep-laboratory event count is a specific outcome and should not be treated as a complete measure of daily experience or overall health.

The safety counts are likewise descriptive. Serious adverse events were reported in 2 participants receiving nelotanserin and 0 receiving placebo; no deaths were reported in either group. These counts, in a study with limited enrollment, cannot by themselves establish a causal relationship or characterize uncommon events reliably.

The registry results are best read as an uncertain estimate from a small, short-term study rather than as a definitive answer. They identify the observed direction and scale of the reported changes, while leaving substantial uncertainty about the between-group effect and its practical importance.

16. Why This Trial Matters Statistically

This trial illustrates why the estimated difference and its precision need to be considered together. The nelotanserin arm’s mean change was more negative than the placebo arm’s, but the between-arm interval ranges from -4.028 to 1.608 and includes zero. The point estimate gives a concise summary of the observed contrast; the interval indicates that the registry data are compatible with a wide range of contrasts.

It also illustrates the distinction between statistical evidence and clinical importance. The unadjusted p-value is 0.4, but a p-value does not tell readers whether an effect matters to participants. The registry record reports no threshold for clinically meaningful change. Without such a threshold, the practical significance of the estimated change cannot be judged from the endpoint summaries alone.

The endpoint combines observable RBD events into a rate per 10 minutes, assessed using video and audio in a sleep laboratory. This makes the question concrete and measurable, while also defining the scope of the conclusion: the result concerns that specific recorded behavior, in that assessment setting, over 28 days. It does not represent every outcome relevant to people with DLB or PDD.

The enrollment of 34 and the approximate pooled standard deviation of 4.204 place the uncertainty in context. In general, limited sample sizes and variable outcomes can make treatment-effect estimates imprecise. The planning illustrations further show how larger studies may be needed to target smaller effects, but those illustrations are not a record of the original design.

Finally, the registry record’s omissions matter for statistical interpretation. Missing-data procedures, analysis populations, model details, and multiplicity plans are not reported. Recognizing these limits helps distinguish what the posted estimates directly show from what cannot be concluded from them.

17. 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 measured over a 28-day treatment period, so the posted results address short-term change only. The registry record does not report longer follow-up results. The endpoint is an event frequency assessed in a sleep laboratory, and the record does not report how closely those assessments represent behavior outside that setting.

The enrolled population was limited to participants with DLB or PDD and RBD. Enrollment was 34, with 16 participants in the nelotanserin group and 18 in the placebo group in the primary outcome results. The registry record does not provide further demographic or clinical details here, so the applicability of the results to more specific participant groups cannot be assessed from this record.

The comparison interval is broad and includes zero, and the pooled standard deviation used in the recalculation is approximate. The registry record does not report the confidence level for the interval, the statistical model specification, or arm-specific confidence intervals. It also does not report a threshold for clinically important change.

Important analysis details are not reported in the registry record, including the analysis population, missing-data handling, randomization procedures, masking procedures, and detailed sample-size justification. No secondary outcomes or analyses are posted. These omissions restrict assessment of robustness, risk of bias, and the extent to which all planned analyses are represented.

Safety results are also limited by the small groups and lack of detailed event information. Serious adverse events, other adverse events, and deaths are reported as counts, but the record does not include event timing, exposure-adjusted summaries, or causal assessments. The serious-event terms are listed with percentages but without group attribution. The results should therefore be interpreted as a limited registry summary rather than a complete account of efficacy, safety, or study conduct.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Study Evaluating Nelotanserin for Treatment of REM Sleep Behavior Disorder in Subjects With Dementia (DLB or PDD) — NCT02708186.
• Registry record last updated 2020-05-20; status last verified 2020-04.
• Results first posted 2020-05-20.
• 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.

21. Record Summary

NCT02708186 is a completed Phase 2 randomized, quadruple-masked, parallel-group study sponsored by Axovant Sciences Ltd. It compared nelotanserin 80 mg with placebo in participants with DLB or PDD who had RBD. Enrollment was 34, and the treatment period for the primary endpoint was 28 days.

The primary outcome was change in RBD event frequency per 10 minutes, assessed by video and audio in a sleep laboratory. The reported least-squares mean changes were -1.47 for nelotanserin and -0.26 for placebo. The unadjusted mean difference was -1.21, with an interval from -4.028 to 1.608 and an unadjusted p-value of 0.4.

Serious adverse events affected 2 participants in the nelotanserin group and 0 in the placebo group; no deaths were reported. The registry record does not report secondary outcomes, the detailed analysis plan, missing-data methods, or the original sample-size justification.