Dose-Finding Study of CD07805/47 Topical Gel in Subjects With Erythematotelangiectatic Rosacea: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Dose-Finding Study of CD07805/47 Topical Gel in Subjects With Erythematotelangiectatic Rosacea (NCT00989014), a randomized phase 2 study. The primary outcome is Response Rate for Achieving a 2 Grade Improvement on Clinician's Erythema Assessment (CEA) and Patient Self Assessment (PSA) Over 12 Hours After Dosing..

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group, vehicle-controlled, phase 2 dose-finding study. The registry record identifies quadruple masking; the study description also calls it double-blind. Participants were assigned in a 1:1:1:1 ratio to CD07805/47 gel at 0.07%, 0.18%, or 0.50%, or to vehicle gel. The intended dosing was a single application, described as once daily for one day.

The study concerned subjects with stable, moderate to severe erythematotelangiectatic rosacea. The registry record lists enrollment of 122 and status as completed. The listed arm counts are 31 for the 0.50% concentration, 31 for the 0.18% concentration, 28 for the 0.07% concentration, and 32 for vehicle. These are the arm denominators reported with the primary outcome; the registry record does not explain any difference between those denominators and the enrollment figure.

The primary outcome was a participant count for achieving a two-grade improvement on both the Clinician's Erythema Assessment and Patient Self Assessment over 12 hours after dosing. Assessment was scheduled at baseline and every hour following application. The registry record lists one outcome measure and no posted secondary outcomes. It does not report a more detailed protocol for handling the repeated hourly assessments, defining the response window beyond the stated period, or selecting the analysis population.

The registry record gives study start and primary completion as 2009-09 and 2009-11, respectively. It does not report the trial's original sample-size justification, detailed analysis plan, or any interim-analysis plan. Those omissions limit what can be inferred about design choices beyond the randomized, masked, parallel-group comparison and the stated outcome.

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.

6. Primary Endpoint

The registry lists the primary outcome as Response Rate for Achieving a 2 Grade Improvement on Clinician's Erythema Assessment (CEA) and Patient Self Assessment (PSA) Over 12 Hours After Dosing., measured over Baseline and every hour for 12 hours following application.

The primary endpoint counts participants who achieved a two-grade improvement on the Clinician's Erythema Assessment (CEA) and the Patient Self Assessment (PSA) during the 12 hours after dosing. Assessments were scheduled at baseline and every hour for 12 hours following application. The endpoint is reported as a participant count in each arm, rather than as a continuous change in an erythema score.

Using both a clinician assessment and a patient assessment makes the response definition depend on improvement recorded by both perspectives, as described in the endpoint title. This kind of combined binary endpoint can provide an interpretable count of participants meeting a prespecified response criterion. In general, a binary endpoint simplifies comparison between randomized groups, but it also reduces a range of possible score changes to whether the response threshold was reached. Participants who improve without reaching the threshold are not counted as responders under that definition.

The registry record does not give further details about the scoring scales, the handling of an assessment that is missing at one or more hourly time points, or whether the criterion had to be met at a particular hour or at any point during follow-up. It also does not describe how the two assessment components were combined operationally. These details matter for reproducing the endpoint definition and interpreting what a recorded response represents.

7. Results Posted in the Registry

• CD07805/47 0.5% Topical Gel: 54.8% (17/31) versus 12.5% (4/32) on control; risk difference 42.3 percentage points (95% CI 21.4 to 63.3), risk ratio 4.39 (1.66–11.58); unadjusted two-proportion z-test p = <0.001.
• CD07805/47 0.18% Topical Gel: 32.3% (10/31) versus 12.5% (4/32) on control; risk difference 19.8 percentage points (95% CI -0.3 to 39.8), risk ratio 2.58 (0.90–7.37); unadjusted two-proportion z-test p = 0.059.
• CD07805/47 0.07% Topical Gel: 25.0% (7/28) versus 12.5% (4/32) on control; risk difference 12.5 percentage points (95% CI -7.2 to 32.2), risk ratio 2.00 (0.65–6.12); unadjusted two-proportion z-test p = 0.212.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant is counted as having or not having met the stated two-grade improvement criterion. The posted results give counts and arm denominators. For comparisons with vehicle, the recalculated measure is the risk difference, expressed as the active-arm response proportion minus the vehicle response proportion. In general, a risk difference describes an absolute contrast between proportions. A value above zero indicates a higher observed response proportion in the active arm; a value below zero indicates a lower observed response proportion. It does not, by itself, explain why the groups differ.

What does a confidence interval add?

In general, a confidence interval displays uncertainty around an estimate using a specified statistical procedure. A narrower interval indicates greater precision under that procedure; a wider interval indicates less precision. When an interval for a risk difference includes zero, the data are compatible with no difference as well as with values on either side of zero within the interval. When it excludes zero, the interval does not include no difference. Neither feature alone establishes the clinical importance of the contrast. The registry record does not report the confidence level or the method used to calculate these intervals, so those details cannot be inferred from the interval endpoints.

What does the p-value describe?

In general, a p-value measures how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the assumptions of the statistical test were true. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The posted p-values are explicitly unadjusted. The registry record does not report which test produced them, whether tests were one-sided or two-sided, or any further test assumptions. The comparisons should therefore be described using the values posted, without attributing an unreported test procedure.

Why does the outcome form matter?

For a binary response, participants just above and just below the two-grade threshold are classified differently even if their underlying score changes are similar. This is a general feature of threshold-based outcomes. The chosen criterion offers a straightforward responder count, but the count does not show the full distribution of CEA or PSA changes. The registry record does not post score-level results or a separate analysis of the hourly measurements. It also does not explain whether response at any scheduled assessment, a particular assessment, or another rule determined the final responder count beyond the endpoint title's description.

What can randomization support?

In general, randomized assignment helps make groups comparable on average at the start of a study, subject to the conduct of randomization and the analysis. It does not remove uncertainty in a study with finite group sizes, and it does not resolve missing outcome data or other limitations unless those are addressed in the analysis. The registry record identifies randomized allocation and gives the assigned arms, but does not describe sequence generation, concealment, adherence, or analysis by assigned group. Those methods are not reported in the registry record.

The posted risk differences and intervals are simple unadjusted recalculations from the summary counts. The registry record lists zero posted analyses, and it does not provide a sponsor analysis for comparison. As a result, the recalculated contrasts should not be represented as a complete account of a prespecified inferential analysis. The record does not report a formal trend test across concentrations, direct active-versus-active contrasts, covariate adjustment, or procedures for missing values.

Comparing several active concentrations with one vehicle group also raises a general multiplicity issue: considering several comparisons can increase the chance that at least one result appears unusual by chance. Whether and how the original study addressed that issue is not reported in the registry record. The p-values are unadjusted, so they do not account for the set of comparisons as a family. This point is separate from the size of any individual risk difference or the width of its interval.

10. Confidence Intervals and Precision

The precision of each comparison is conveyed by its confidence interval, which shows a range of risk differences compatible with the estimate under the interval procedure. The registry record does not report the confidence level or interval method. The intervals should therefore be read as posted, without assuming a particular confidence level.

For the 0.50% gel versus vehicle, the estimated risk difference is 0.423 and the interval extends from 0.214 to 0.633. This interval is entirely above zero. It is narrower than the intervals for the other active-versus-vehicle comparisons, while still allowing a range of possible effect sizes. The estimate is not an exact value for the difference that would be expected in every setting.

For the 0.18% gel, the estimate is 0.198 and the interval extends from -0.003 to 0.398. Its lower endpoint is close to zero and the interval includes zero, leaving uncertainty about the direction and size of the underlying difference. For the 0.07% gel, the estimate is 0.125 and the interval extends from -0.072 to 0.322. This interval also includes zero and spans values below and above it.

In general, interval width reflects both the observed data and the method used to estimate uncertainty. Smaller groups and less frequent outcomes can contribute to less precise estimates, though the registry record does not report a formal precision analysis. The arm denominators range from 28 to 32 participants in the posted outcome results, and the responder counts are 4 in vehicle and 7, 10, and 17 in the active arms. These counts help place the estimates in context but do not eliminate uncertainty.

Confidence intervals are not guarantees that the true effect lies within the displayed range, nor are they a measure of clinical relevance. A point estimate gives one summary of the observed contrast; its interval indicates how much uncertainty remains. The registry record does not state a clinically important difference threshold, so the intervals cannot determine whether any particular value would be meaningful to patients or clinicians.

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 sample-size figures are planning illustrations based on the observed effect, not the trial's own design calculation. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row asks what would be needed to detect the stated fraction of the observed effect. The figures are participants per arm for 80% power and participants per arm for 90% power; they are not total sample sizes or total event counts. The curve presents the same planning relationship visually.

For the 0.50% gel comparison, planning for 100% of the observed effect gives 19 participants per arm for 80% power and 25 participants per arm for 90% power. Planning for 75% of the observed effect gives 31 and 41 participants per arm, respectively; planning for 50% gives 62 and 82.

For the 0.18% gel comparison, the corresponding figures are 69 and 92 participants per arm for 100% of the observed effect; 113 and 151 for 75%; and 231 and 308 for 50%. For the 0.07% gel comparison, they are 152 and 203 for 100%; 254 and 339 for 75%; and 529 and 708 for 50%.

In general, smaller effects require more participants to distinguish them from no difference at a given power and significance level. The figures depend on the observed effect used as a planning reference and should not be treated as a confirmed requirement for a future study. This is not observed power. The trial's original sample-size assumptions and justification are not reported in the registry record.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.125; treatment proportion p₁ = 0.548; average p̄ = 0.337.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6683 = 1.3099.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5975 = 0.5029.
4. n = (1.3099 + 0.5029)² / (0.4234)² = 18.3, rounded up to 19 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
p0 <- 0.125                          # control proportion
p1 <- 0.548387096774194                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(989014)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 19 (80% power), 25 (90% power)
 75% of observed effect: n per arm = 31 (80% power), 41 (90% power)
 50% of observed effect: n per arm = 62 (80% power), 82 (90% power)
power.prop.test: 18.3 
Simulated power at n = 19 per arm: 0.839 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 reports three active-versus-vehicle comparisons for the primary binary outcome. The associated p-values are unadjusted. No family-wise error procedure, multiplicity adjustment, or hierarchical testing strategy is reported in the registry record. Consequently, the values should be read as separate unadjusted comparisons rather than as a set whose overall false-positive rate is known to have been controlled.

The registry record lists one outcome measure and zero posted secondary outcomes. That does not establish whether other outcomes were collected; it describes what is posted in the record. The endpoint involves assessments at baseline and every hour for 12 hours, but no multiplicity handling for those repeated assessments or rule for consolidating them into a participant-level response is reported beyond the endpoint description.

The available record does not describe missing-data rules, imputation, exclusions, or sensitivity analyses. It does not state whether all randomized participants were included in the outcome analysis or define an intention-to-treat, per-protocol, or safety population. Although arm denominators are reported for the primary outcome, reasons for any differences among those denominators and the enrollment figure are not reported in the registry record. It is not possible to infer from the posted counts whether there were missing assessments, post-randomization exclusions, or another explanation.

Analysis populations and missing-data methods can affect how an outcome estimate should be interpreted. In general, excluding participants with unavailable outcomes may introduce bias if availability is related to response, while imputation requires assumptions about the unobserved outcomes. No particular approach should be attributed to this study because its missing-data and population definitions are not reported in the registry record.

The registry record also does not report a detailed statistical analysis plan, model specification, covariate adjustment, or a formal comparison among the three concentrations. The presented risk differences are unadjusted recalculations from posted counts. They answer the separate active-versus-vehicle comparisons as calculated, but do not provide an adjusted overall test or establish a dose-response pattern.

14. 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 posted serious adverse event results list zero participants affected in each arm. The corresponding numbers at risk are 31 for the 0.50% gel, 31 for the 0.18% gel, 28 for the 0.07% gel, and 32 for vehicle. No serious adverse event terms are listed. Deaths are not reported in the registry record; the death entries are not presented as zero.

These figures describe the serious adverse events captured in the posted safety results for the listed groups. A count of zero affected participants in a group means no serious event is recorded for that group in this result, not that the possibility of such an event is ruled out. Counts from a study with these arm sizes and a single application provide limited information about uncommon events or effects outside the recorded follow-up.

The registry record also lists other adverse events, with 4 participants affected in the 0.50% arm, 1 in the 0.18% arm, 3 in the 0.07% arm, and 5 in the vehicle arm. These are separate from the serious adverse event counts and should not be conflated with them. Details of event types, severity, timing, and relatedness are not reported in the registry record.

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

The results describe a short-term, participant-level response outcome after one application in a randomized, vehicle-controlled study. In the posted counts, 17 of 31 participants in the 0.50% arm met the response definition, compared with 4 of 32 in the vehicle arm. The recalculated risk difference for that comparison is 0.423, and its confidence interval is above zero. The 0.18% and 0.07% comparisons have positive point estimates, but their intervals include zero and are wider relative to their estimates.

These results do not establish that the same differences would occur in other populations, with different dosing schedules, or over longer periods. The study description specifies one application and assessment over 12 hours; the registry record does not provide longer-term outcome results. The posted endpoint also does not give the full distribution of clinician or patient assessment scores, only the count meeting the response criterion.

The findings do not by themselves establish clinical importance. The registry record does not report a threshold for a clinically meaningful change, additional patient-centered outcomes, or a detailed interpretation of the assessment scales. Nor do the active-versus-vehicle contrasts directly compare concentrations with one another. A sequence of point estimates across concentrations is not equivalent to a formal dose-response analysis.

The unadjusted p-values do not account for the set of three comparisons, and the record does not report a multiplicity strategy. Missing-data methods, analysis populations, and the original statistical plan are also not reported in the registry record. These omissions make it difficult to know how the posted results relate to a prespecified analysis or how sensitive they are to alternative assumptions.

Accordingly, the most defensible reading is descriptive: the posted response counts and unadjusted contrasts differ across the dose-versus-vehicle comparisons, with the 0.50% comparison having the estimate whose interval excludes zero. The results alone should not be treated as a complete assessment of benefit, harm, durability, or clinical relevance.

16. Why This Trial Matters Statistically

This trial illustrates how a dose-finding study can generate several related comparisons from one randomized experiment. Each concentration is compared with a shared vehicle group, and the resulting estimates answer separate questions. Because there are multiple comparisons, the unadjusted p-values need to be interpreted with awareness that the registry record does not report a multiplicity adjustment.

The endpoint also shows why the definition of a response matters. A two-grade improvement on both the clinician's assessment and the patient's self-assessment creates a clear participant count, but it compresses repeated measurements and score changes into a binary outcome. Knowing the count is useful for comparing proportions, while the absence of score-level results limits understanding of the range of observed changes and their timing during follow-up.

Precision differs across the comparisons. The 0.50% estimate has an interval entirely above zero, while the intervals for the other concentrations include zero and cover a range on both sides. Reporting estimates together with intervals helps distinguish the size of an observed contrast from the uncertainty around it. The p-value alone would not communicate that range or the magnitude of the difference.

The record also demonstrates the boundary between reported results and analysis details. Counts and unadjusted recalculations are available, but the original sample-size justification, confidence-interval method, analysis populations, missing-data approach, and multiplicity plan are not reported. Statistical interpretation is stronger when those design and analysis decisions are transparent. Without them, readers can describe the posted comparisons but should avoid assuming how the original investigators planned or conducted every analysis.

Finally, planning illustrations based on observed effects are not a substitute for a prospective design rationale. They show how participant requirements change under stated assumptions and effect scenarios, but do not reveal the study's original planning process or guarantee that a future study would have the same results.

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 study examined a single application and the primary outcome covered the 12 hours following dosing. The registry record does not report longer-term outcomes, so durability cannot be assessed from these results. Enrollment was 122, while the primary outcome denominators listed by arm are 31, 31, 28, and 32; the reason for the difference is not reported in the registry record.

The endpoint combines a two-grade improvement on clinician and patient assessments into a participant-level response. The registry record does not specify the scoring details, explain how hourly observations were consolidated, or give score-level results. That limits assessment of the size and timing of changes among participants who did not meet the response threshold, as well as those who did.

Only one outcome measure is listed and no secondary outcomes are posted. The record does not describe a detailed statistical analysis plan, the method for calculating confidence intervals, the tests behind the p-values, multiplicity control, analysis populations, or missing-data procedures. The posted p-values are unadjusted, and the recalculated comparisons are based on summary counts rather than a reported adjusted analysis.

The study has small arm denominators, and the response counts are also modest. This makes the precision of the estimates important, particularly where confidence intervals span values on both sides of zero. The planning illustrations rely on observed effects and stated assumptions; the original design assumptions are not reported in the registry record.

Safety interpretation is limited by the short single-application setting and the absence of reported event details. Serious adverse event counts are zero in the posted results, but deaths are not reported, and the types, severity, timing, and relatedness of other adverse events are not described. These limits mean the registry record supports a focused description of the posted short-term outcomes, not a complete account of longer-term efficacy or safety.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Dose-Finding Study of CD07805/47 Topical Gel in Subjects With Erythematotelangiectatic Rosacea — NCT00989014.
• Registry record last updated 2021-02-18; status last verified 2013-09.
• Results first posted 2013-11-21.
• 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

NCT00989014 is a completed phase 2 randomized, quadruple-masked, parallel-group, vehicle-controlled dose-finding study sponsored by Galderma R&D. It enrolled subjects with stable, moderate to severe erythematotelangiectatic rosacea and compared single applications of CD07805/47 topical gel at 0.07%, 0.18%, and 0.50% with vehicle.

The primary outcome was the count of participants achieving a two-grade improvement on the Clinician's Erythema Assessment and Patient Self Assessment over 12 hours after dosing. Posted counts were 17 of 31, 10 of 31, and 7 of 28 in the active arms, and 4 of 32 in vehicle. The recalculated risk difference for the 0.50% comparison had a confidence interval above zero; the intervals for the other two comparisons included zero.

The registry record does not report the confidence level, detailed analysis plan, missing-data approach, analysis populations, or multiplicity strategy. Its posted serious adverse event counts are zero in each arm; deaths are not reported in the registry record.