Double-Blind, Placebo-Controlled Study to Evaluate 3 Doses of a Novel Tetracycline in the Treatment of Facial Acne Vulgaris: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Double-Blind, Placebo-Controlled Study to Evaluate 3 Doses of a Novel Tetracycline in the Treatment of Facial Acne Vulgaris (NCT01628549), a randomized phase 2 study. The primary outcome is The Absolute Change From Baseline in the Inflammatory Lesion Count at the Final Visit.

2. Clinical Question

3. Trial Design

This was a randomized, multicenter, double-blind, placebo-controlled, parallel-group phase 2 study. Its title identifies a comparison of three P005672-HCl dose regimens with placebo in facial acne vulgaris. The registry lists enrollment of 285 and status as completed. The recorded study start was 2012-06-30, and primary completion was 2012-11-16.

The four groups received daily oral capsules for 12 weeks. The placebo group took two matching placebo capsules. The approximately 0.75 mg/kg/day group took one 50 mg P005672-HCl capsule and one placebo capsule. The approximately 1.5 mg/kg/day group took two 50 mg capsules, while the approximately 3.0 mg/kg/day group took two 100 mg capsules. These are the regimen descriptions in the record; further dosing or adherence details are not reported in the registry record.

The primary efficacy population was the modified intent-to-treat population: 284 participants from the 285 in the Safety Population who had a post-baseline efficacy assessment. The arm-specific primary outcome counts were 72 for placebo, 76 for the lowest regimen, 70 for the middle regimen, and 66 for the highest regimen. Safety counts use a different population accounting, with the record listing 73, 76, 70, and 66 participants at risk across the respective groups.

The record lists five secondary outcome descriptions, including absolute and percent changes in inflammatory and noninflammatory lesion counts at specified visits and a dichotomized IGA score. The trial's original sample-size justification, planned interim analyses, randomization details, and detailed protocol 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: Of the 285 participants in the Safety Population, 284 received a post-Baseline efficacy assessment, to comprise the mITT (Modified Intent-to-Treat) Population.

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: 285 participants were randomized to double-blind treatment (Safety population)

6. Primary Endpoint

The registry lists the primary outcome as The Absolute Change From Baseline in the Inflammatory Lesion Count at the Final Visit, measured over Baseline (Week 0) to Final Visit (Up to Week 12).

The primary endpoint was the absolute change from baseline in inflammatory lesion count at the final visit, defined as baseline at Week 0 through the final visit, up to Week 12. It is a continuous measure: each participant's inflammatory lesion count is compared with that participant's baseline count, and the group result is expressed as a mean. The record gives the outcome unit as number of inflammatory lesions and specifies the mean and standard deviation as the summary measures.

Inflammatory lesion counts provide a numerical way to describe one aspect of facial acne severity over the study period. Comparing average changes between groups asks whether the observed course differed across randomized assignments, rather than considering only the final count without baseline context. The absolute-change endpoint retains the lesion-count scale, so its units are counts rather than percentages.

The record does not state the precise sign convention used to encode change or provide a separate clinical threshold for an important change. Therefore, the arm means and contrasts should be read as numerical changes as posted, not as an independent judgment about clinical importance. A mean also does not describe how every participant's count changed: the standard deviations indicate substantial participant-to-participant spread around the group averages.

7. Results Posted in the Registry

• P005672-HCl Approximately 0.75 mg/kg/Day: mean 14.40 versus 12.80 on control; difference 1.60 (95% CI -2.79 to 5.99), standardized effect d = 0.12; z-test p = 0.475.
• P005672-HCl Approximately 1.5 mg/kg/Day: mean 17.60 versus 12.80 on control; difference 4.80 (95% CI -0.29 to 9.89), standardized effect d = 0.31; z-test p = 0.065.
• P005672-HCl Approximately 3.0 mg/kg/Day: mean 17.10 versus 12.80 on control; difference 4.30 (95% CI -0.02 to 8.62), standardized effect d = 0.33; z-test p = 0.051.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean change represent?

The primary endpoint is continuous, and the registry identifies the reported summary as a mean. For each group, the mean describes the average posted absolute change in inflammatory lesion count. The standard deviation describes how spread out individual observations are around that average. A mean can be useful for comparing groups, but it does not show the full distribution or how many participants had a particular response.

The posted arm means are 12.8 for placebo, 14.4 for approximately 0.75 mg/kg/day, 17.6 for approximately 1.5 mg/kg/day, and 17.1 for approximately 3.0 mg/kg/day. Their associated standard deviations are 14.18, 13.01, 16.65, and 11.69. The spread is appreciable relative to the mean values. This means that individual participants' changes varied, and the group average should not be read as the change experienced by every participant.

Why compare means between groups?

A mean difference expresses how far apart two group averages are on the endpoint's original scale. The unadjusted recalculations compare each active arm's mean with the placebo mean using the posted summary statistics. They do not adjust for baseline values or other participant characteristics. Their estimates are 1.6, 4.8, and 4.3 for the approximately 0.75, 1.5, and 3.0 mg/kg/day arms, respectively. The record describes these as simple recalculations; they are not the sponsor's primary analysis.

The sponsor reports ANCOVA, or analysis of covariance, and labels its parameter a difference in least-squares means. In general, ANCOVA estimates a group contrast while accounting for covariates included in the model. The specific covariates, model details, and missing-data procedures are not reported in the registry record. Thus, the adjusted contrast cannot be reconstructed fully from arm means and standard deviations alone.

How should the ANCOVA results be read?

The sponsor's estimated differences versus placebo were 1.53 for approximately 0.75 mg/kg/day, 4.46 for approximately 1.5 mg/kg/day, and 4.37 for approximately 3.0 mg/kg/day. The reported 95% confidence intervals were -2.32 to 5.38, 0.52 to 8.40, and 0.39 to 8.36, with p-values 0.434, 0.027, and 0.032. A confidence interval describes uncertainty in the estimated contrast under the analysis model. When an interval includes zero, it is compatible with no group difference under that model; intervals that exclude zero indicate estimates less compatible with zero at the stated confidence level. These statements do not resolve clinical importance or multiplicity.

A p-value measures how incompatible the observed data are with a specified null hypothesis under the analysis assumptions. It is not the probability that the null hypothesis is true, nor does it convey the magnitude of the effect. The sponsor labels the comparisons as superiority analyses. The registry record does not report whether the p-values were adjusted across the three dose comparisons or other outcomes.

Why do the recalculations differ?

The recalculated comparisons are unadjusted, while the registry's sponsor analyses use ANCOVA and differences in least-squares means. The estimates, intervals, and p-values therefore need not match. For example, the recalculated estimate for the approximately 1.5 mg/kg/day arm is 4.8, while the sponsor's ANCOVA estimate is 4.46. The discrepancy flag is true, so the sponsor's own analysis takes precedence. The recalculations provide a transparent check based on posted summaries, not a replacement analysis.

For a continuous outcome, standard errors reflect both within-group variation and the amount of information in each group. Greater variability or fewer observations generally makes an estimate less precise. The reported group standard deviations and efficacy counts therefore matter to uncertainty, but the sponsor's adjusted standard errors and model structure are not provided in full. No inference about individual response can be derived solely from the group means.

What is not specified?

The registry record does not report the sponsor's detailed statistical analysis plan, the covariates in ANCOVA, assumptions checks, handling of missing endpoint values, or any interim analysis plan. It also does not report whether a multiplicity procedure was applied. These omissions limit reconstruction and evaluation of the full analysis strategy. General statistical concepts explain how these methods work, but they cannot fill in details not reported for this trial.

10. Confidence Intervals and Precision

The sponsor's 95% confidence intervals convey the uncertainty around each adjusted mean contrast under the reported ANCOVA. For the approximately 0.75 mg/kg/day regimen, the estimate was 1.53 and the interval ranged from -2.32 to 5.38. Because that interval includes zero, the reported data are compatible with a range of contrasts spanning both sides of zero under the model.

For the approximately 1.5 mg/kg/day regimen, the estimate was 4.46, with a 95% confidence interval from 0.52 to 8.40. For the approximately 3.0 mg/kg/day regimen, the estimate was 4.37, with an interval from 0.39 to 8.36. Both intervals exclude zero, but they remain broad enough to include a range of possible contrast sizes. The intervals describe statistical uncertainty, not a range containing a fixed proportion of individual participants' outcomes.

Precision depends on the variation in outcomes, the number of observations, and the analysis model. The arm-level standard deviations vary, and the efficacy counts are 72, 76, 70, and 66. The record does not provide enough detail to reproduce the ANCOVA intervals or determine how covariate adjustment changed their precision. The simple recalculated intervals differ from the sponsor's intervals; because the discrepancy flag is true, the sponsor's results take precedence.

Confidence intervals do not establish whether a contrast is important to patients, and they do not account automatically for multiple comparisons. The registry record does not report a clinical-importance threshold or the multiplicity procedure, if any. The intervals should be read as uncertainty summaries for the specified comparisons, not as a complete assessment of all trial uncertainty.

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 each pairwise comparison; they are not total event counts. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect equal to the stated fraction of the observed effect. Time-to-event planning figures, when relevant, would instead be total events, but this trial's primary endpoint is continuous and its figures are per arm.

The 80% power column gives the illustrative number of participants per arm that would be needed, under those assumptions, for an 80% chance of detecting the chosen effect if that effect were the true one. The 90% power column gives the analogous planning count for a 90% chance. These are design illustrations based on the observed effect, not observed power and not the trial's original sample-size justification.

For example, planning for 100% of the observed effect gives 1,133 participants per arm for 80% power and 1,517 for 90% power for the approximately 0.75 mg/kg/day comparison. The corresponding counts are 163 and 218 for approximately 1.5 mg/kg/day, and 145 and 194 for approximately 3.0 mg/kg/day. Planning for 75% or 50% of the observed effect requires larger counts in each comparison. In general, smaller effects are harder to distinguish from random variation and therefore require more participants under the same planning assumptions.

The figures are hypothetical planning calculations, not evidence about the probability that the observed findings are true. The registry record's own design assumptions and original sample-size justification are not reported. The displayed curve should be interpreted in the same planning context: it illustrates how required sample size changes as the target effect changes, rather than documenting trial enrollment or a further observed result.

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 = 1.600; pooled standard deviation σ = 13.592.
2. n = 2 × (1.96 + 0.842)² × 13.592² / 1.600² = 1,132.8, rounded up to 1,133 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 <- 1.6                      # observed difference in means
sigma <- 13.5915585959997                     # 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(1628549)
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 = 1,133 (80% power), 1,517 (90% power)
 75% of observed effect: n per arm = 2,014 (80% power), 2,696 (90% power)
 50% of observed effect: n per arm = 4,532 (80% power), 6,066 (90% power)
power.t.test: 1133.7 
Simulated power at n = 1,133 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 reports the sponsor's primary comparisons as ANCOVA differences in least-squares means. The adjusted estimates versus placebo were 1.53 for approximately 0.75 mg/kg/day, 4.46 for approximately 1.5 mg/kg/day, and 4.37 for approximately 3.0 mg/kg/day. Their p-values were 0.434, 0.027, and 0.032, respectively. The sponsor's 95% confidence intervals were -2.32 to 5.38, 0.52 to 8.40, and 0.39 to 8.36.

The simple recalculations from arm summaries are unadjusted and produce estimates of 1.6, 4.8, and 4.3. Their contrasts, intervals, and p-values do not exactly match the sponsor's adjusted analyses. That difference is expected when the analyses use different methods, although the registry does not provide the ANCOVA covariates and details needed to reproduce the sponsor results.

Because the discrepancy flag is true, the sponsor's own analysis takes precedence. The unadjusted calculations should be treated as checks based on the posted group summaries, not as a correction to the reported analysis. The registry record does not report whether sponsor p-values were adjusted for multiple comparisons, so they should be read as reported rather than as confirmed multiplicity-adjusted evidence.

14. Multiplicity, Missing Data and Analysis Populations

The trial compares three active dose regimens with placebo on the primary endpoint, and the registry also lists five secondary outcome descriptions. Those secondary outcomes cover inflammatory and noninflammatory lesion counts, absolute and percent changes, multiple assessment visits, and a dichotomized IGA score. Testing several doses and outcomes creates a multiplicity issue: when many hypotheses are examined, the chance of at least one apparently unusual result can rise if no adjustment is made.

The sponsor reports p-values of 0.434, 0.027, and 0.032 for the three primary endpoint comparisons. The record does not report whether these p-values were adjusted for multiple dose comparisons, whether the primary comparisons followed a prespecified hierarchy, or whether a family-wise error or false-discovery procedure was used. No multiplicity adjustment should therefore be assumed. The estimates and confidence intervals remain interpretable as the reported individual comparisons, but they do not by themselves settle the broader multiplicity question.

The efficacy population is described as the modified intent-to-treat population: 284 participants from the 285 in the Safety Population who had a post-baseline efficacy assessment. The primary outcome counts by arm are 72, 76, 70, and 66. The record does not explain the exclusions or give detailed participant flow, so the difference between the Safety Population and efficacy population cannot be characterized further.

Missing-data methods are also not reported in the registry record. It is not stated whether missing final-visit values were imputed, whether other available visits contributed to the endpoint analysis, or how the ANCOVA handled incomplete covariate or outcome data. These details can affect both the analysis population and estimated uncertainty. The reported population definition does establish that a post-baseline efficacy assessment was required, but it does not describe all analysis-set rules.

The detailed analysis plan, ANCOVA covariates, sensitivity analyses, and interim-analysis plan are not reported in the registry record. Without these details, the posted results permit a description of the reported comparisons but not a complete reconstruction of the analysis strategy or its safeguards against multiplicity and missing-data bias.

15. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent non-serious adverse event terms

The registry lists serious adverse events as affecting 0 participants in each of the four groups. The corresponding numbers at risk were 73 for placebo, 76 for approximately 0.75 mg/kg/day, 70 for approximately 1.5 mg/kg/day, and 66 for approximately 3.0 mg/kg/day. These are participant counts, not estimates of comparative risk adjusted for follow-up or other factors. No serious adverse event terms are listed.

The record also lists participants affected by other adverse events: 13 for placebo, 15 for approximately 0.75 mg/kg/day, 8 for approximately 1.5 mg/kg/day, and 7 for approximately 3.0 mg/kg/day. The registry does not provide further event descriptions in these summary counts, so they should not be interpreted as a list of specific symptoms or as a complete characterization of severity and timing.

Death counts are not reported in the registry record. The adverse-event figures are descriptive and do not establish that groups had equivalent safety, particularly because the record's summary does not provide event-level details, duration-specific rates, or comparative safety analyses.

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

The posted results describe average changes in inflammatory lesion count in four randomized groups and provide sponsor ANCOVA comparisons of each active regimen with placebo. The adjusted estimates and confidence intervals quantify the differences reported by that analysis. The intervals for the approximately 1.5 mg/kg/day and approximately 3.0 mg/kg/day comparisons exclude zero; the interval for approximately 0.75 mg/kg/day includes zero. These findings apply to the stated endpoint and the sponsor's model, not to every outcome or participant.

The results do not, on their own, establish clinical importance. The registry record does not report a threshold for a meaningful change, and the meaning of the positive sign is not fully specified. The p-values are not measures of effect size, and the confidence intervals do not describe individual participant responses. Neither should be used as a substitute for a clinical interpretation of the endpoint and its context.

The posted comparisons also do not settle the question of multiple testing. The registry lists several dose comparisons and secondary outcomes, while multiplicity adjustment is not reported. Missing-data handling and ANCOVA covariates are also not reported, limiting the ability to assess whether alternative analysis choices would change the estimates or their uncertainty.

The safety summary records no serious adverse events in any group, but it contains only aggregate counts and does not report deaths. It cannot establish that there are no safety risks or that risks are equivalent across groups. The findings are confined to this completed phase 2 study, its 12-week regimen descriptions, and the populations and analyses reported in the registry. Further claims about other settings or unreported outcomes are not established by these results.

17. Why This Trial Matters Statistically

This trial is statistically instructive because it reports both arm-level summaries and sponsor-adjusted comparisons for a continuous endpoint. The mean absolute changes and standard deviations make the scale and variability visible, while the ANCOVA results show how an adjusted analysis can yield estimates that differ from simple calculations based on group summaries.

The discrepancy between the unadjusted recalculations and sponsor estimates illustrates why analysis methods matter. Recalculating from means and standard deviations can be transparent and useful for checking the broad pattern, but it cannot recreate an ANCOVA model when covariates and other model details are unavailable. In this record, the sponsor analysis takes precedence because the discrepancy flag is true.

The study also illustrates the need to read estimates together with uncertainty. Two sponsor intervals exclude zero and one includes zero, but all intervals span a range of contrast sizes. Statistical evidence does not answer whether a difference is clinically important, and the record supplies no threshold for that judgment. The sign convention is also not stated in enough detail to assign direction confidently from the sign alone.

Finally, several doses and outcomes raise questions about multiplicity, while missing-data procedures and original design assumptions are not reported. Planning illustrations based on observed effects can help readers understand how target effects relate to sample-size requirements, but they are not the trial's original design rationale or evidence of achieved power. Taken together, these issues show why a result is best read as a model-specific estimate with stated limits, rather than as a single p-value.

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 registry record does not report the ANCOVA covariates, detailed model specification, or the sign convention for the change endpoint. The sponsor results therefore cannot be independently reproduced from the posted arm means and standard deviations alone. The reported contrasts are adjusted estimates, whereas the recalculations are unadjusted; their difference should not be treated as an arithmetic error.

Multiplicity adjustment is not reported despite comparisons of three active regimens with placebo and several secondary outcomes at multiple visits. Missing-data handling, sensitivity analyses, and the detailed criteria for the modified intent-to-treat population are also not reported. The record states that 284 participants with a post-baseline efficacy assessment comprised that population, but it does not explain participant exclusions or incomplete final-visit observations.

The posted efficacy summaries concern the final visit, up to Week 12, and do not give the underlying participant-level distribution. Group means and standard deviations cannot show how many participants had particular changes or whether outcomes were concentrated in a subset. The record does not report a clinical-importance threshold, so statistical estimates cannot be translated into a judgment about practical importance from these data alone.

The safety summary is aggregate: serious adverse event counts are listed, other adverse event counts are listed, and deaths are not reported in the registry record. It does not provide event terms, timing, duration, or comparative statistical analyses. The trial's original sample-size justification, interim analyses, and full statistical plan are not reported, limiting evaluation of how the study was planned and how all analyses were selected.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Double-Blind, Placebo-Controlled Study to Evaluate 3 Doses of a Novel Tetracycline in the Treatment of Facial Acne Vulgaris — NCT01628549.
• Registry record last updated 2019-02-01; status last verified 2019-01.
• Results first posted 2019-01-08.
• 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.

22. Record Summary

This completed randomized, multicenter, double-blind, placebo-controlled phase 2 trial evaluated three daily oral P005672-HCl regimens against placebo for moderate to severe facial acne vulgaris. The study enrolled 285 participants, and each regimen was described as lasting 12 weeks.

The primary endpoint was the absolute change from baseline in inflammatory lesion count at the final visit. The sponsor reported ANCOVA comparisons versus placebo, with estimates of 1.53, 4.46, and 4.37 for the approximately 0.75, 1.5, and 3.0 mg/kg/day regimens. The corresponding p-values were 0.434, 0.027, and 0.032. Because the discrepancy flag is true, the sponsor's analysis takes precedence over unadjusted recalculations.

The registry reports no serious adverse events in any group; death counts are not reported in the registry record. Multiplicity adjustment, missing-data procedures, detailed ANCOVA specifications, and the original sample-size justification are not reported.