A Safety and Efficacy Study of Setipiprant Tablets in Androgenetic Alopecia in Males: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Safety and Efficacy Study of Setipiprant Tablets in Androgenetic Alopecia in Males (NCT02781311), a randomized phase 2 study. The primary outcome is Change From Baseline in Target Area Hair Count (TAHC) at Week 24.

2. Clinical Question

3. Trial Design

This was a multicenter, randomized, double-blind, placebo-controlled Phase 2A study of setipiprant tablets in males with androgenetic alopecia. The registry lists the allocation as randomized, the masking as quadruple, and the model as parallel. The study is marked completed.

The registry reports enrollment of 169 participants. The study arms included setipiprant, placebo and finasteride. Setipiprant was given orally at 1000 mg twice daily for 24 weeks; placebo was given twice daily for the same period; and finasteride was given orally at 1 mg once daily for 24 weeks. The primary reported comparison is setipiprant versus placebo. The finasteride group was not included in the analysis model, as described in the primary outcome record.

The primary outcome was change from baseline to Week 24 in target area hair count (TAHC), expressed as terminal hairs per cm². The primary analysis used the mITT population, and the numbers analyzed were 61 for placebo and 70 for setipiprant. These are the analysis counts for the reported comparison, not the full enrollment figure. The registry’s description of the mITT population requires receipt of study intervention and a baseline plus at least one postbaseline measurement for one of the coprimary efficacy measures.

The registry lists the study start as 2016-07-14 and primary completion as 2018-03-15. The trial’s original sample-size assumptions and the reason for the planned enrollment 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: mITT population included all randomized participants who received study intervention and had a baseline and at least 1 postbaseline measurement for 1 of the coprimary efficacy measures. As per protocol Amendment 1, Finasteride arm group was not included in the analysis model. Number analyzed is participants with data available for analysis.

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: Modified Intent-to-Treat (mITT) population included all randomized participants who received study intervention in the study and had a baseline and at least 1 postbaseline measurement for one of the coprimary efficacy measures.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Target Area Hair Count (TAHC) at Week 24, measured over Baseline (Day 1) to Week 24.

TAHC is a count of terminal hairs within a standardized target area of the scalp. The registry describes the target as a circular area of 1 cm², clipped to approximately 1 mm of hair and located at the anterior leading edge of the vertex thinning area. A semi-permanent microdot tattoo was used to reproduce the same location at each visit. Digital macrophotographs were used to count terminal hairs, defined in the outcome description as hairs with a width of at least 30 μm.

The endpoint is the change from baseline to Week 24. A positive change indicates an increase in the number of counted terminal hairs. Measuring a defined area and using digital imaging makes the outcome an objective, standardized count rather than a subjective assessment of overall hair appearance. Repeating measurement at the same marked location is intended to make visits comparable.

The outcome is reported as a least-squares mean, with standard error as its dispersion measure. The registry states that missing data were imputed through Week 24 using last observation carried forward (LOCF). The endpoint description identifies TAHC as one of the coprimary efficacy measures; the other coprimary measure’s results are not reported in the registry record.

7. Results Posted in the Registry

• Setipiprant: mean 7.10 versus 6.70 on control; difference 0.40 (95% CI -8.31 to 9.11), standardized effect d = 0.02; z-test p = 0.928.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the least-squares mean difference represent?

A mean difference expresses the contrast between the two groups’ average outcome changes. The registry reports an ANCOVA analysis and labels its estimate a least-squares mean difference. In general, ANCOVA estimates a group contrast while accounting for variables included in the model. The registry record does not report the full model specification or list the covariates used, so those details cannot be inferred. The reported contrast for setipiprant versus placebo is 0.4 terminal hairs/cm².

The arm-level reported means are 6.7 for placebo and 7.1 for setipiprant. The difference is not itself a count from an individual participant; it is a group-level estimate. Its interpretation depends on the outcome definition, analysis population and statistical model.

What does the confidence interval communicate?

A confidence interval describes uncertainty around an estimated group difference. In general, a 95% confidence interval is produced by a method designed to cover the true value in 95% of repeated samples under its assumptions. It is not a statement that there is a 95% probability that this particular interval contains the true effect. For this result, the sponsor-reported interval is -8.38 to 9.23. The recalculated unadjusted interval is -8.309 to 9.109. Both extend across zero and include values on either side of the point estimate.

How do standard errors relate to the reported means?

The primary outcome record gives standard errors of 3.25 for placebo and 3.03 for setipiprant. A standard error describes the estimated sampling variability of a group mean; it is not the spread of individual participant measurements. General statistical interpretation: greater standard error means less precision in estimating the mean, while the standard deviation describes variation among individual observations. The recalculation uses the posted summary statistics and an approximate pooled standard deviation; the record labels that value as approximate. The registry’s ANCOVA model and the unadjusted recalculation are not identical analyses.

What does the p-value say?

The sponsor-reported analysis gives p = 0.924, and the unadjusted recalculation gives p = 0.928. A p-value is a measure, under a specified model and null hypothesis, of how unusual data at least as incompatible with that null would be. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of a difference. The p-values are consistent with the broad uncertainty expressed by the intervals, but neither one supplies evidence about outcomes beyond the analysis and endpoint being tested.

Why can two estimates differ slightly?

The registry’s result is based on ANCOVA and reports a least-squares mean difference. The recalculation is an unadjusted calculation from posted summary statistics. Adjusted and unadjusted analyses can yield slightly different intervals and p-values because they use different information and assumptions. The point estimate is reported as 0.4 in both. The record does not flag a discrepancy, and the sponsor’s reported analysis remains the registry’s primary inferential result.

General statistical context: estimates from a randomized trial are subject to sampling variability, and modeling choices affect the way uncertainty is quantified. An analysis should therefore be read with its population, endpoint, missing-data method and model in view. Here, the endpoint description specifies an mITT population and LOCF imputation; the full ANCOVA specification and the trial’s original statistical plan are not reported in the registry record.

10. Confidence Intervals and Precision

The registry-reported ANCOVA estimate is a least-squares mean difference of 0.4 terminal hairs/cm², with a 95% confidence interval from -8.38 to 9.23. The unadjusted recalculation gives the same displayed point estimate, with a 95% confidence interval from -8.309 to 9.109. The small difference between the interval limits reflects that the sponsor analysis and the summary-statistic recalculation use different methods.

Both intervals cover zero and extend in both directions. That means the data do not estimate the group difference precisely enough to exclude either a negative or a positive difference within the intervals. The point estimate alone should not be mistaken for a precise estimate: uncertainty is represented by the interval, not only by the central value.

The reported standard errors are 3.25 for placebo and 3.03 for setipiprant. Standard errors quantify uncertainty in estimated means, whereas participant-level variation is described by a standard deviation. The recalculation’s pooled standard deviation is approximate because it is derived from posted summary information. The registry record does not report the participant-level data needed to independently reproduce the ANCOVA model.

A wide interval does not prove that the groups are equivalent, and an interval spanning zero does not establish that there is no difference. Equivalence requires a prespecified margin and an appropriate design and analysis; such a margin is not reported in the registry record. The precision of this estimate is best described by the reported interval and the analysis details, rather than by a binary reading of statistical significance.

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 for 80% power and participants per arm for 90% power. They are not total event counts. They are illustrative calculations based on the observed effect and are not the trial’s original enrollment justification; the trial’s own design assumptions are not reported in the registry record.

The stated planning 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 at the stated fraction of the observed effect. For planning for 100% of the observed effect, the illustration gives 63,128 participants per arm for 80% power and 84,510 participants per arm for 90% power. For planning for 75% of the observed effect, it gives 112,227 participants per arm for 80% power and 150,240 participants per arm for 90% power. For planning for 50% of the observed effect, it gives 252,511 participants per arm for 80% power and 338,040 participants per arm for 90% power.

In general, 80% power means that a study designed under the stated assumptions would have an 80% chance of detecting the specified effect at the stated significance level if that effect were the true one. The 90% column represents a higher target chance under the same assumptions. Smaller effects require more participants to detect reliably, which is why the planning figures rise as the target effect is reduced. These are planning illustrations, not observed power.

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 = 0.400; pooled standard deviation σ = 25.366.
2. n = 2 × (1.96 + 0.842)² × 25.366² / 0.400² = 63,127.6, rounded up to 63,128 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 <- 0.399999999999999                      # observed difference in means
sigma <- 25.3659261471432                     # 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(2781311)
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 = 63,128 (80% power), 84,510 (90% power)
 75% of observed effect: n per arm = 112,227 (80% power), 150,240 (90% power)
 50% of observed effect: n per arm = 252,511 (80% power), 338,040 (90% power)
power.t.test: 63128.5 
Simulated power at n = 63,128 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. Analyses Reported by the Sponsor

The registry’s sponsor-reported analysis is an ANCOVA with a least-squares mean difference of 0.4 terminal hairs/cm². It reports a 95% confidence interval from -8.38 to 9.23 and p = 0.924. The unadjusted recalculation from posted summary statistics also gives a mean difference of 0.4, with a 95% confidence interval from -8.309 to 9.109 and p = 0.928.

The small interval and p-value differences are consistent with the different approaches: the sponsor analysis is model-based and the recalculation is unadjusted. The record does not flag a discrepancy. The sponsor’s analysis is the registered analysis to use when describing the trial’s reported inferential result; the recalculation is a separate summary-statistic check, not a replacement for that analysis.

The registry does not report the full ANCOVA covariate specification, so the adjustment cannot be reconstructed from the record. The results should be interpreted using the sponsor’s analysis population and endpoint definition, with the recalculation serving only as a transparent unadjusted comparison.

14. Multiplicity, Missing Data and Analysis Populations

The primary outcome description calls TAHC one of the coprimary efficacy measures. The registry reports two outcome measures, but results for the other coprimary measure are not reported in the registry record. No secondary outcomes are posted. The multiplicity strategy—such as whether or how the analysis controlled the overall chance of a false-positive conclusion across outcomes—is not reported in the registry record. The p-value for TAHC should therefore be read as the result for this endpoint, not as a documented multiplicity-adjusted test across all efficacy outcomes.

The primary outcome record defines the mITT population as randomized participants who received study intervention and had a baseline and at least one postbaseline measurement for one of the coprimary efficacy measures. The analysis counts were 61 for placebo and 70 for setipiprant. The registry says the finasteride arm was not included in the analysis model, consistent with protocol Amendment 1. Reasons for the difference between overall enrollment and the analyzed groups are not fully reported in the registry record.

Missing data were imputed through Week 24 using last observation carried forward (LOCF). In general, LOCF carries a participant’s last observed value forward to fill a later missing measurement. This method relies on assumptions about what would have happened after the last observation; if those assumptions are not suitable, estimates can be affected. The registry does not report sensitivity analyses using alternative missing-data methods, so the robustness of the result to other approaches is not known.

The full statistical analysis plan, any interim analyses, the covariates in the ANCOVA model, and the trial’s original sample-size justification are not reported in the registry record. Their absence limits assessment of how the posted analysis relates to all planned analyses, but it does not change the numerical result reported for TAHC.

15. 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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cellulitis (1.4%), multiple sclerosis (1.4%).

The registry reports serious adverse events among 73 participants at risk in the placebo group, 81 in the setipiprant group, and 12 in the finasteride group. The reported numbers affected were 2, 0 and 0, respectively. The registry also reports 0 deaths in each group.

For other adverse events, the record lists 26 participants affected in the placebo group, 28 in the setipiprant group and 6 in the finasteride group. These figures are separate from the serious adverse event counts. The most frequently listed serious adverse event terms are cellulitis and multiple sclerosis, each reported as 1.4%.

These are descriptive registry figures. The numbers at risk differ among arms, and the safety populations are not the same as the analyzed groups for the primary efficacy comparison. The registry summary does not provide event timing, duration, severity details beyond the serious-event classification, or a comparative statistical analysis. These counts alone do not establish a causal relationship between treatment and an event, or provide a complete assessment of safety.

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

The primary result is a Week 24 TAHC mean difference of 0.4 terminal hairs/cm² for setipiprant compared with placebo in the reported analysis. The sponsor’s 95% confidence interval extends from -8.38 to 9.23, and the unadjusted recalculation produces an interval from -8.309 to 9.109. The results therefore leave substantial uncertainty about the size and direction of the group difference.

The result does not establish equivalence between setipiprant and placebo, because no equivalence margin or equivalence analysis is reported in the registry record. Nor does it establish that a difference is absent. It is a comparison of the two analyzed groups for one defined measure and timeframe, using the stated mITT population and LOCF method.

The reported outcome is an objective count from a marked scalp target area, not a measure of every aspect of hair growth or of participants’ experience. The registry describes TAHC as one of the coprimary efficacy measures, but the other coprimary outcome result is not reported in the registry record. The posted TAHC estimate cannot stand in for the unreported outcome or for an overall conclusion across all study endpoints.

The finasteride group was part of the study but was excluded from this primary analysis model. Thus, the primary estimate does not compare setipiprant with finasteride. The safety figures are descriptive and include different at-risk counts; they should not be treated as proof of a causal effect or as a complete safety comparison.

In short, the registry reports a small point estimate for the setipiprant-placebo contrast, but the confidence interval is broad relative to that estimate. The data as posted do not give a precise estimate of the difference in Week 24 target area hair count.

17. Why This Trial Matters Statistically

This trial illustrates why a group mean and its uncertainty need to be read together. The least-squares mean difference is 0.4 terminal hairs/cm², while the 95% confidence interval ranges from -8.38 to 9.23. The point estimate alone would obscure the uncertainty in the comparison; the interval shows that the estimate is not precise.

It also illustrates that the name of an analysis matters. The registered result comes from ANCOVA, whereas the recalculation is unadjusted and based on summary statistics. Their point estimates match as reported, but the confidence limits and p-values differ slightly. Distinguishing the two avoids treating a simple recalculation as if it reproduced the sponsor’s full model.

The analysis population and endpoint construction are equally important. The efficacy comparison uses the mITT population, excludes the finasteride group from the model, and applies LOCF for missing values through Week 24. The result is therefore tied to a particular population, method of handling missing measurements and objectively defined target area, rather than being a general measure of all possible outcomes.

Finally, the planning illustrations show how a small observed effect can imply very large participant requirements under stated assumptions, and how targeting a smaller effect increases those requirements. They are not evidence about the trial’s original design or its observed power. These distinctions help readers separate the trial’s posted finding from uncertainty, planning calculations and unreported design choices.

18. 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 efficacy comparison covers the mITT population rather than every enrolled participant. The record defines this population using receipt of study intervention and availability of baseline and postbaseline measurement; the reasons some enrolled or randomized participants were not included in the reported analysis are not fully described. The analyzed groups contain 61 placebo and 70 setipiprant participants, while the registry reports enrollment of 169 overall.

The endpoint description specifies LOCF imputation through Week 24. This approach may not capture the course of outcomes after a participant’s last observed measurement if the outcome would have changed. Alternative missing-data analyses are not reported in the registry record, so the sensitivity of the estimate to the imputation method cannot be assessed from the posted results.

The registry provides the ANCOVA result but not the complete model specification, including the covariates used. It also does not report the original sample-size justification, any interim analyses, or the full statistical analysis plan. The presence of coprimary efficacy measures and the absence of results for the other coprimary measure leave the broader multiplicity context unresolved.

The finasteride arm was excluded from the primary analysis model, so the main result applies only to the placebo-setipiprant comparison. The safety summary has different at-risk counts across arms and is descriptive; it does not provide a full comparative safety analysis. These limitations constrain interpretation of the posted estimate, without changing what the registry reports for the TAHC comparison.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Safety and Efficacy Study of Setipiprant Tablets in Androgenetic Alopecia in Males — NCT02781311.
• Registry record last updated 2019-04-05; status last verified 2019-03.
• Results first posted 2019-04-05.
• 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.

22. Record Summary

The registry describes a completed, multicenter, randomized, double-blind, placebo-controlled Phase 2A parallel study of setipiprant in males with androgenetic alopecia. Overall enrollment was 169. Participants received setipiprant, placebo or finasteride for 24 weeks.

The primary reported outcome was change from baseline to Week 24 in target area hair count. In the mITT analysis, the placebo and setipiprant groups had reported mean changes of 6.7 and 7.1 terminal hairs/cm², with a sponsor-reported least-squares mean difference of 0.4. The sponsor’s 95% confidence interval was -8.38 to 9.23.

The analysis excluded the finasteride arm and used LOCF for missing data. Results for the other coprimary efficacy measure and details of the original sample-size justification are not reported in the registry record.