A Study to Investigate the Efficacy and Safety of Balovaptan (RO5285119) in Participants With Autism Spectrum Disorder (ASD): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Investigate the Efficacy and Safety of Balovaptan (RO5285119) in Participants With Autism Spectrum Disorder (ASD) (NCT02901431), a randomized phase 2 study. The primary outcome is Change From Baseline in Vineland™-II Adaptive Behavior Scale Two Domain Composite (2DC) Score at Week 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo.

2. Clinical Question

3. Trial Design

This was a Phase II, multicenter, randomized, double-blind, parallel-group, placebo-controlled study of balovaptan in children and adolescents with autism spectrum disorder. The official study description specifies a 24-week treatment comparison and an age range of 5–17 years. The registry records enrollment of 339 and a terminated status. The reason for termination is not reported in the registry record.

The protocol description changed over time. For participants enrolled before Version 6, the main study was described as a three-arm study, including balovaptan doses equivalent to 4 mg/d and 10 mg/d and placebo. The 4 mg equivalent arm was open only to participants enrolled before Version 6. For participants enrolled under Version 6, the description specifies a two-arm comparison of 10 mg equivalent balovaptan with placebo. The primary posted comparison is between the 10 mg equivalent group and its concurrently randomized placebo group.

The registry describes three study parts: a pharmacokinetic part with up to 8 weeks of treatment, a main treatment part with 24 weeks of treatment, and an optional open-label extension from Week 24 to Week 76. Participants completing the 24-week period were eligible for the extension, during which they received balovaptan. The posted primary efficacy result is for the randomized comparison, not a comparison within the open-label extension.

The study began on 2016-11-21, and the recorded primary completion date is 2020-04-15. Further details about the original sample-size justification, specific randomization procedures, or reasons for the protocol-version change 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.

Before assignment: PK review revealed that age-adjusted doses of 4 and 10 mg in 5-17 year olds were not equivalent to target exposure. For the Main Study Part, data was summarised by exposure ranges (tertiles) based on individual participants PK exposure at Week 12, estimated as the average plasma concentration since treatment start. To allow clear analysis by exposure tertiles, participants with dose adjustment were excluded from analysis by tertiles.

Only the first 6 of 13 groups are shown.

PK Part

Reasons for not completing this period:

Main Treatment Part

Reasons for not completing this period:

Open Label Part

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy Inferential Population included participants that were divided by randomized treatment and included participants taking balovaptan 10 mg equivalent dose and participants from the concurrently randomized placebo group in the corresponding randomization stage. Participants with dose adjustments or interruptions, or who were on a different dose than the final dose for their age group, were excluded.

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: PK Part \& Main Part reported separately. 7 patients ("re-starters") are counted twice. PK review revealed age-adjusted doses of 4 \& 10 mg in 5-17 year olds weren't equivalent to target exposure. For Main Study Part, data was summarised by exposure ranges (tertiles) based on individual patients PK exposure at Week 12, estimated as average plasma concentration since treatment start. To allow clear analysis by exposure tertiles, patients with dose adjustment were excluded from analysis by tertiles.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Vineland™-II Adaptive Behavior Scale Two Domain Composite (2DC) Score at Week 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo, measured over Baseline, Week 24.

The primary endpoint was change from baseline in the Vineland-II Adaptive Behavior Scale Two Domain Composite score at Week 24. The composite is defined as the mean of the Communication and Socialization domain standard scores. If either of those two domain scores is missing, the composite is not computed. The Vineland-II is described as an instrument for assessing communication, daily living skills, socialization, motor skills and maladaptive behavior in people with developmental disabilities. In this study, a reliable caregiver was interviewed by a rater or clinician about the participant’s activities and behavior.

The registry describes the adaptive behavior composite as a standardized score, with higher scores indicating better functioning. A caregiver interview can provide information about behavior across everyday settings, which is relevant to an outcome intended to represent adaptive functioning rather than a single observed task. The primary analysis used assessments at baseline, Week 12 and Week 24, with the Week 24 comparison designated as primary.

The endpoint is a group-level change measure. A difference between treatment-group estimates does not indicate that every participant changed by that amount, and it does not by itself establish that a difference is important to participants or caregivers. The registry does not report a threshold for interpreting a clinically meaningful difference on this endpoint.

7. Results Posted in the Registry

• Balovaptan (RO5285119) 10 mg/d Equivalent: mean 2.17 versus 2.34 on control; difference -0.17 (95% CI -3.30 to 2.96), standardized effect d = -0.02; z-test p = 0.915.

8. Secondary Efficacy Endpoints

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

Change From Baseline in Vineland-II Adaptive Behavior Scale 2DC Score at Week 12 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo

• 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:
  • Balovaptan (RO5285119) 10 mg/d Equivalent: mean 1.88 versus 1.87 on control; difference 0.01 (95% CI -2.72 to 2.74), standardized effect d = 0.00; z-test p = 0.994.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Balovaptan (RO5285119) 10 mg/d Equivalent and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes addressing adaptive behavior, clinical impressions and quality of life. These measures broaden the questions beyond the primary Week 24 composite: they include the Vineland-II composite and its component domains, clinical severity and improvement ratings, and a pediatric quality-of-life scale. They are potentially useful for examining whether patterns in the primary measure are echoed in related domains or in other kinds of assessment.

The posted information does not support a complete directional overview across all secondary endpoints. Although 11 secondary outcomes are recorded, detailed numerical results for the full set are not reported in the registry record in the material available for this analysis. It is therefore not possible to characterize their overall pattern as consistently favoring one group, or to claim that secondary measures confirm or contradict the primary result. The tables on the page should be consulted for the outcome-specific information that is available.

Secondary results are generally interpreted with more caution when many outcomes, time points or comparisons are examined. In general, testing more hypotheses increases the chance that at least one result may appear notable by chance, especially if no adjustment or prespecified hierarchy is used. The registry lists 12 outcome measures and 12 analyses, but the multiplicity procedure, if any, is not reported in the registry record. An isolated secondary result would therefore carry less weight than a prespecified primary result unless its place in the analysis plan and the handling of multiple testing were clear.

Secondary measures can provide context, but they do not change the definition of the primary endpoint or substitute for its result. Here, the primary comparison is imprecise and does not establish a treatment-group difference. Without a complete, interpretable set of secondary estimates and information about multiplicity, the secondary outcomes cannot be used to draw a stronger overall conclusion.

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

• Change From Baseline in Vineland™-II Composite Standard Score After 12 Weeks and 24 Weeks of Treatment for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Change From Baseline in Vineland™-II Adaptive Behavior Scale Communication, Socialization, and Daily Living Skills Domain Standard Scores at Weeks 12 and 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Proportion of Subjects With >=6 Points Improvement in the Vineland-II 2DC Score for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Change From Baseline in Clinical Global Impressions-Severity (CGI-S) Score at Weeks 12 and 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Change From Baseline in Ohio Autism Clinical Impressions Scale-Severity (OACIS-S) Score at Weeks 12 and 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Clinical Global Impressions- Improvement (CGI-I) Score at Weeks 12 and 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Ohio Autism Clinical Impressions Scale- Improvement (OACIS-I) Score at Weeks 12 and 24 for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo
• Change From Baseline in Patient-Reported Pediatric Quality of Life (PedsQL) v4.0 Generic Core Scale After 12 Weeks and 24 Weeks of Treatment for Balovaptan (R05285119) 10 mg Equivalent Compared to Placebo

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a score that can take a range of values rather than a count of events. The registry reports adjusted least-squares means and their difference for the primary comparison. A least-squares mean is a model-based group estimate. In general, model-based estimates can account for the structure of repeated observations and other terms included in the fitted model; the registry does not report the complete model specification, so the exact adjustment terms cannot be described.

The stated method was a mixed model with repeated measures, or MMRM, using assessments at baseline, Week 12 and Week 24. In general, an MMRM uses the repeated outcome measurements together rather than treating each visit as an unrelated analysis. This approach can represent the fact that observations from the same participant over time are related. It can estimate group differences at a particular visit while using the repeated-measure structure specified by the model. The covariance structure, covariates, estimation details and other model settings are not reported in the registry record.

The sponsor’s reported parameter is the difference in adjusted least-squares means. The posted estimate is -0.16, with a 90% confidence interval from -2.56 to 2.23 and a p-value of 0.911. The confidence interval describes a range of values compatible with the fitted model and the observed data under its assumptions. It is not a range containing a stated proportion of individual participants’ scores. The p-value summarizes how compatible the data are with a null hypothesis of no group difference under the analysis model; it does not give the probability that the null hypothesis is true.

The comparison table also includes a recalculated unadjusted result based on the posted group summaries. The arm means are 2.34 for placebo and 2.17 for balovaptan, and the recalculated mean difference is -0.17. The recalculated interval is -3.303 to 2.963, and the unadjusted p-value is 0.915. This re-calculation uses summary statistics and is not a replication of the fitted MMRM. In particular, the standard deviations used for the simple calculation are approximated from the reported standard errors, as indicated in the posted calculation. It therefore cannot reproduce adjustments or repeated-measures handling that may be part of the sponsor’s analysis.

In general, over-dispersion means that observed variability is greater than a model assumes. If it is ignored, standard errors and inferential results can be too optimistic. This is a general statistical issue, not a reported feature of this trial; the registry does not report an over-dispersion assessment. Likewise, the record does not describe a test of model assumptions, a sensitivity analysis, or an alternative covariance structure.

The main analysis population is described as an efficacy inferential population containing randomized participants assigned to balovaptan 10 mg equivalent and participants in the concurrently randomized placebo group for the corresponding randomization stage. The displayed description then ends before the population definition is complete. The complete inclusion rules, handling of dose adjustments and any exclusions are not reported in the registry record. The term randomized does not, by itself, establish that every randomized participant contributed to this specific analysis.

General interpretation of the p-value and interval requires attention to the endpoint, model and analysis population. A p-value above a conventional threshold is not evidence that the treatments are equivalent. Equivalence requires an appropriately designed analysis with a prespecified margin; that design and margin are not reported in the registry record. Similarly, the registry does not report whether the analysis was adjusted for testing multiple outcomes or time points.

The primary reported analysis is distinct from planning illustrations based on the observed effect. Those illustrations use a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation, and consider fractions of the observed effect. They are not the trial’s original design justification and do not describe the fitted MMRM. Details of the trial’s original statistical analysis plan, interim analyses and sample-size justification are not reported in the registry record.

11. Confidence Intervals and Precision

The sponsor-reported 90% confidence interval for the primary adjusted difference extends from -2.56 to 2.23. A confidence interval communicates uncertainty in the estimated group difference: in general, a method that repeatedly produced intervals in the same way would cover the underlying parameter at the stated rate, assuming the model and its assumptions are appropriate. It does not mean there is a 90% probability that this particular fixed parameter lies within the observed interval.

Here, the interval includes zero and reaches both below and above it. The data are therefore compatible with a range of possible directions for the group difference. Its width also means the point estimate alone should not be read as a precise measure of the size of any difference. The registry does not report a clinical-importance threshold for the score, so the interval cannot be classified against a stated threshold for meaningful benefit or harm.

The unadjusted summary-based re-calculation has an interval from -3.303 to 2.963. It is not the same interval as the sponsor’s model-based 90% interval; the calculations use different methods and should not be interchanged. The sponsor’s reported interval corresponds to the repeated-measures model, while the re-calculation is based on posted summaries with an approximate standard deviation.

Precision depends on factors such as the amount of information, outcome variability and analysis model. The enrollment figure for the entire study is not the same as the number contributing to the primary comparison: the posted primary arm summaries include 81 placebo participants and 86 balovaptan participants. The registry does not report a complete account of missing endpoint observations or the reasons for any missing data, limiting a fuller assessment of how missingness may affect precision.

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 table and curve are planning illustrations based on the observed effect, not the trial’s original design calculation. The figures are participants per arm, not total participants or total events. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms and a normal approximation. The table shows planning for 100%, 75% and 50% of the observed effect.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the stated effect under those planning assumptions, if that effect were the true effect and the assumptions held. The 90% power column gives the corresponding number per arm for a 90% chance. These are prospective planning concepts, not probabilities that the trial result is correct. This is not observed power.

The listed estimates are very large because the observed primary estimate is small relative to the variability used in the simple calculation. Planning for 100% of the observed effect gives 57,862 participants per arm for 80% power and 77,460 participants per arm for 90% power. Planning for 75% of the observed effect gives 102,865 and 137,707 participants per arm, respectively; planning for 50% gives 231,445 and 309,839 participants per arm, respectively. These values are illustrations from the specified calculation, not a recommendation for a future study.

Smaller effects require more participants to detect under otherwise unchanged assumptions. The curve should be read as displaying that relationship, rather than as evidence that an effect exists or will be observed. The registry’s original sample-size justification, target power and design assumptions are not reported in the registry record.

13. How the Sample Size Is Derived

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

Formula

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

Worked calculation for this trial

1. Observed difference Δ = -0.170; pooled standard deviation σ = 10.321.
2. n = 2 × (1.96 + 0.842)² × 10.321² / -0.170² = 57,861.2, rounded up to 57,862 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 <- -0.17                      # observed difference in means
sigma <- 10.3210438161334                     # 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(2901431)
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 = 57,862 (80% power), 77,460 (90% power)
 75% of observed effect: n per arm = 102,865 (80% power), 137,707 (90% power)
 50% of observed effect: n per arm = 231,445 (80% power), 309,839 (90% power)
power.t.test: 57862.1 
Simulated power at n = 57,862 per arm: 0.800 (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 result comes from a mixed-model analysis of repeated measurements. It estimates a difference in adjusted least-squares means of -0.16, with a 90% confidence interval from -2.56 to 2.23 and a p-value of 0.911. This is the trial’s reported inferential result for the primary endpoint and is the result to use when describing the sponsor’s analysis.

The separate re-calculation uses the posted group summaries rather than the fitted repeated-measures model. It gives a difference of -0.17, an interval from -3.303 to 2.963 and an unadjusted p-value of 0.915. The close point estimates and p-values indicate that the simple summary calculation gives a similar broad reading, but the two analyses are not interchangeable. The sponsor analysis uses the model specified for the repeated visits; the re-calculation uses an approximate standard deviation derived from the posted standard errors and does not reproduce the model.

The registry does not indicate a discrepancy between the sponsor analysis and the re-calculation. Neither result should be interpreted as demonstrating equivalence. The sponsor’s interval remains compatible with estimates on either side of zero, and no equivalence margin is reported. The sponsor’s result also does not resolve questions about unreported model details, missing-data assumptions or multiplicity.

15. Multiplicity, Missing Data and Analysis Populations

The study has a designated primary endpoint and multiple secondary outcomes. The registry records 12 outcome measures and 12 analyses, including the Week 24 Vineland-II two-domain composite as the primary outcome. The secondary list covers additional visits, score domains, clinical impression measures and quality of life. Testing multiple outcomes or time points can increase the chance of false-positive findings if results are interpreted without accounting for the number of tests. The registry does not report a multiplicity adjustment, a hierarchy of tests, or whether secondary analyses were intended to be confirmatory or exploratory.

The primary analysis used a mixed model with repeated measures at baseline, Week 12 and Week 24. The registry does not report the model’s full specification, the covariance structure, covariates or the handling of departures from model assumptions. It also does not report a detailed missing-data strategy, the amount or reasons for missing outcome data, or sensitivity analyses examining the effect of missingness. In general, repeated-measures models rely on assumptions about the observed data and missing observations; the assumptions used are not reported in the registry record.

The primary population description identifies randomized participants in the 10 mg equivalent balovaptan group and the corresponding concurrently randomized placebo group. The displayed definition is incomplete after its reference to dose adjustments. Full eligibility for the efficacy analysis, exclusions, and the treatment of dose-adjusted participants are not reported in the registry record. Consequently, the enrollment count should not be treated as the primary analysis denominator.

The registry does not report whether interim efficacy analyses were conducted, whether any stopping rules were used, or how the study’s termination related to its statistical plan. The reason for termination is also not reported in the registry record. These absences limit assessment of the complete analysis process but do not alter the numerical primary result posted for the comparison.

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: pneumonia (9.1%), dehydration (2.2%), acute lymphocytic leukaemia (2.2%), anxiety (2.2%), suicidal ideation (1.8%).

The safety summary separates participants by study part and, in the main study and extension, by placebo, exposure tertile or dose-adjuster group. Those groups have different denominators and should be read separately rather than combined into a single treatment comparison. In the main study part, serious adverse events were recorded for 4 of 112 participants in the placebo group, 1 of 57 in the low exposure tertile, 0 of 66 in the medium exposure tertile, 1 of 66 in the high exposure tertile and 0 of 7 dose-adjusters. The pharmacokinetic part and open-label extension have their own group counts, shown in the safety table.

The registry reports zero deaths in each listed group. It also lists serious-event terms with percentages, including Pneumonia at 9.1%, Dehydration at 2.2%, Acute lymphocytic leukaemia at 2.2%, Anxiety at 2.2%, Suicidal ideation at 1.8% and Agitation at 1.5%. These terms should be read as the record presents them; the summary does not establish a causal relationship to treatment or provide enough context to compare those term percentages across groups.

These figures describe events recorded in the registry, not a complete assessment of comparative safety. The differing study parts, exposure groups and denominators matter when interpreting the counts. The registry summary alone does not establish whether an event was caused by balovaptan.

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

The primary result estimates the difference in Week 24 change in the Vineland-II two-domain composite between the balovaptan 10 mg equivalent group and its randomized placebo comparator. The sponsor-reported estimate is close to zero on the score scale, and its 90% confidence interval includes values below and above zero. The posted data therefore leave substantial uncertainty about the size and direction of the group difference.

This is not evidence that the groups are equivalent. Equivalence is a specific statistical claim that requires a prespecified margin and an analysis designed to evaluate whether the plausible difference lies within that margin. No such margin is reported in the registry record. Nor does a p-value of 0.911 mean there is a 91.1% probability that the treatments are the same. A p-value is a measure tied to a statistical model and a null hypothesis, not a probability that a treatment effect exists.

The endpoint has a defined direction: higher scores indicate better functioning. The numerical estimate is lower for balovaptan than placebo, but its uncertainty prevents a firm conclusion about the true group difference. The registry does not report a clinical-importance threshold, so the score difference cannot be judged against a threshold for meaningful change from the information.

The result applies to the recorded randomized comparison and this endpoint at this visit. It does not answer every question about other outcomes, doses, study parts or longer-term treatment. The extension was open-label, so it is not the same randomized comparison as the primary analysis. Secondary measures may provide additional context, but their overall pattern and multiplicity handling are not fully.

These statistical results are a description of the posted trial data, not medical advice. They do not determine what treatment choice is appropriate for any individual. A full interpretation would also consider the complete study report, analysis plan, missing-data information, safety data and clinical context, several of which are not reported in the registry record.

18. Why This Trial Matters Statistically

This trial illustrates why an estimated effect and its uncertainty need to be read together. The primary point estimate is small in relation to the breadth of its confidence interval. Reporting only the point estimate could make the result seem more definite than it is; reporting only the p-value would omit the range of differences compatible with the analysis. The interval shows that the posted result does not precisely settle the direction or size of the group difference.

It also illustrates the distinction between a sponsor’s fitted analysis and a simple re-calculation from group summaries. The two calculations produce similar point estimates here, but the mixed model uses the repeated-measures framework described for the primary endpoint, whereas the re-calculation is unadjusted and relies on an approximate standard deviation. A summary-based calculation can help readers understand the scale of a result, but it cannot replace the model-based analysis.

The changing protocol description matters for interpretation. The study was described as having three arms before Version 6 and two arms for participants enrolled under Version 6; the 4 mg equivalent group was restricted to earlier enrollment. The primary comparison concerns the 10 mg equivalent dose and placebo, not every group or study part. Keeping the randomized comparison separate from pharmacokinetic and open-label extension data avoids treating unlike groups as if they formed one trial contrast.

Finally, the large planning illustrations show how a very small observed effect translates into sample-size calculations under specified assumptions. They are not a verdict about the study’s original design or the value of the trial. The original planning rationale, multiplicity strategy and several analysis details are not reported, so the posted results are best read with those uncertainties in view.

19. Limitations

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

The registry record gives a limited account of the statistical analysis. It names an MMRM for the primary endpoint and identifies the assessment visits, but does not provide the complete model specification, covariates, covariance structure or a full account of diagnostics and sensitivity analyses. These details would help readers assess how the adjusted estimate was obtained and how dependent it is on modeling choices.

The primary population description is incomplete in the record. It identifies randomized balovaptan 10 mg equivalent participants and the corresponding placebo group, then refers to dose adjustments without completing the definition. The full analysis denominator and the treatment of exclusions or dose changes are therefore not clear from this description. Enrollment of 339 applies to the study as a whole and does not identify how many participants contributed to each analysis.

Protocol versions describe different numbers of arms, and the earlier 4 mg equivalent arm is not part of the primary comparison. The registry does not explain the reason for the protocol change or provide the original sample-size rationale. The study is marked terminated, but the reason is not reported in the registry record. Neither fact alone establishes why the study ended or how the result should be interpreted.

Information about missing outcome data, missing-data assumptions, multiplicity control and interim analyses is not reported in the registry record. These omissions limit assessment of the robustness and confirmatory status of the findings. The posted safety summary also divides participants across study parts and exposure categories, so its counts should not be treated as a single randomized comparison. The open-label extension lacks the masking and concurrent randomized comparison of the main study.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study to Investigate the Efficacy and Safety of Balovaptan (RO5285119) in Participants With Autism Spectrum Disorder (ASD) — NCT02901431.
• Registry record last updated 2021-02-08; status last verified 2021-01.
• Results first posted 2021-02-08.
• 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

This Phase II study evaluated balovaptan against placebo in children and adolescents aged 5–17 years with autism spectrum disorder. The registry describes randomized allocation, double masking and a parallel-group design, with enrollment of 339 and a terminated status. The primary endpoint was Week 24 change from baseline in the Vineland-II Two Domain Composite score.

The sponsor reported an adjusted mean difference of -0.16, with a 90% confidence interval from -2.56 to 2.23 and a p-value of 0.911. The interval spans both sides of zero and leaves uncertainty about the direction and size of the group difference. A separate summary-based re-calculation gives a similar point estimate but is not the sponsor’s repeated-measures analysis.

The record also describes secondary outcomes, multiple study parts and an optional open-label extension. Important details about multiplicity, missing data, the complete analysis population, the original sample-size rationale and the reason for termination are not reported in the registry record.