To Compare Safety and Efficacy of Perrigo's Drug Product Compared to an FDA Approved Drug Product in the Treatment of Actinic Keratosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for To Compare Safety and Efficacy of Perrigo's Drug Product Compared to an FDA Approved Drug Product in the Treatment of Actinic Keratosis (NCT02385318), a randomized phase 3 study. The primary outcome is Number of Participants With Complete Clearance (Absence) of Actinic Keratosis Lesions as Determined by Visual Inspection of the Investigator.

2. Clinical Question

3. Trial Design

The registry identifies this as a phase 3, randomized, double-masked, parallel-group study of treatments for actinic keratosis. It lists three groups: Test Product, Reference Product, and Placebo Product. The test and reference products are both identified as ingenol mebutate; the placebo product is placebo gel. The sponsor is Padagis LLC.

Enrollment is listed as 519. The study start is recorded as 2015-02, primary completion as 2015-12, and status as completed. Those dates describe the registry record and do not supply further detail about follow-up or the timing of individual participant visits. The primary outcome is assessed at Day 57.

Randomization indicates assignment to study groups by a chance-based process, while double masking indicates that masking was used for two groups of people involved in the study. The registry record does not identify who was masked, describe the randomization procedure, or report the intended allocation ratio. Although the groups and their products are identified, details such as treatment administration, concealment procedures, site structure, and protocol deviations are not reported in the registry record.

The primary outcome population is described as the per protocol population. This is not necessarily the same as everyone enrolled or everyone assigned to treatment. Outcome group denominators are recorded separately from safety denominators, so those counts should be read in the context of the specific analysis rather than treated as interchangeable. The original sample-size rationale and the trial's own design assumptions 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

Analysis population for the primary outcome, as defined in the registry: Per protocol 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.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Complete Clearance (Absence) of Actinic Keratosis Lesions as Determined by Visual Inspection of the Investigator, measured over Day 57.

The primary endpoint is the number of participants with complete clearance of actinic keratosis lesions at Day 57. Complete clearance is defined in the registry as absence of lesions, as determined by visual inspection by the investigator. The endpoint is recorded as a participant count, and the population is the per protocol population.

This is a binary outcome at the participant level: each participant either met the stated clearance definition at the assessment or did not. General statistical context: binary outcomes allow a direct comparison of the proportion meeting a defined condition in each group, while retaining the original clinical definition of that condition. The count alone does not describe changes in lesion burden among participants who did not achieve complete clearance.

Investigator visual inspection provides the assessment method specified in the record. The endpoint is tied to a particular assessment time and definition, so it should not be interpreted as a measure of longer-term clearance, recurrence, symptom change, or other outcomes. Those broader outcomes are not part of the primary endpoint. The registry record does not report additional details about assessment procedures or how disagreements in assessment, if any, were handled.

7. Results Posted in the Registry

• Test Product: 48.2% (81/168) versus 43.2% (76/176) on control; risk difference 5.0 percentage points (95% CI -5.5 to 15.6), risk ratio 1.12 (0.89–1.41); unadjusted two-proportion z-test p = 0.349.
• Placebo Product: 12.0% (10/83) versus 43.2% (76/176) on control; risk difference -31.1 percentage points (95% CI -41.3 to -21.0), risk ratio 0.28 (0.15–0.51); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a risk difference describe?

General statistical explanation: a risk difference subtracts the event proportion in one group from the event proportion in another. The sign depends on which group is subtracted from which. In the recalculated comparison of Test Product with Reference Product, the reported order is test minus reference; in the comparison involving placebo, it is placebo minus reference. A positive estimate therefore points in one direction for the first comparison, while a negative estimate in the second indicates a lower observed clearance proportion in placebo than in reference under that subtraction order.

The posted event counts and denominators form the basis for these unadjusted comparisons. The recalculated results do not include adjustment for other participant characteristics. The registry record does not report an adjusted model for the primary endpoint.

How should the confidence intervals be read?

General statistical explanation: a confidence interval expresses the range of values compatible with the observed data under the method's assumptions, and it conveys uncertainty in an estimate. It is not a guarantee that the true value lies within the particular interval. For the test-versus-reference risk difference, the interval spans zero; the recalculation therefore does not narrowly determine the direction of a difference. For placebo versus reference, the reported interval is entirely negative. The intervals are tied to the recalculated risk differences and should not be treated as intervals for a different measure.

What does an unadjusted p-value mean?

General statistical explanation: a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be if that null hypothesis and the statistical assumptions were correct. It is not the probability that the null hypothesis is true, nor a measure of clinical importance. The recalculated p-values are 0.349 for Test Product versus Reference Product and <0.001 for Placebo Product versus Reference Product. They are identified as unadjusted; no multiplicity adjustment is described for these recalculations.

What is the role of the sponsor's equivalence analysis?

The registry's reported sponsor analysis uses Yates correction and identifies its parameter as an equivalence ratio. It reports a value of 1.11 and a 90% confidence interval from -4.38 to 14.44, without a p-value. Equivalence analyses generally require a prespecified margin that defines what range of differences would count as sufficiently similar. General statistical explanation: an equivalence conclusion depends on the interval's relation to that margin, not simply on whether a conventional difference test is statistically significant. The equivalence margin and fuller interpretation rules are not reported in the registry record, so the listed result cannot independently establish equivalence from the information recorded.

The sponsor's measure and the recalculated risk difference are not the same estimand or scale. The registry record does not give enough methodological detail to reconcile them numerically. The record marks no discrepancy between the sponsor analysis and the recalculation. The recalculated results are therefore best read as additional, simple unadjusted comparisons based on the posted summary counts, rather than as a correction to the sponsor's analysis.

Why can over-dispersion matter?

General statistical explanation: over-dispersion occurs when observed variability is greater than a model assumes. If a model assumes less variability than the data contain, standard errors can be too small and confidence intervals too narrow. The primary outcome is a participant-level binary count, and the registry's reported sponsor analysis is identified as using Yates correction rather than a model described in terms of over-dispersion. The registry record does not report an over-dispersion assessment or a model requiring such an assessment; this general issue should not be mistaken for a documented problem in this trial.

What analysis details remain unclear?

The record identifies the endpoint population as per protocol and lists one outcome measure and one reported analysis. It does not report the detailed statistical analysis plan, handling of missing outcome data, rules for protocol deviations, or a prespecified multiplicity strategy. It also does not report the original sample-size justification or interim analysis plans. Those omissions limit assessment of whether the posted analysis follows all intended protocol procedures.

10. Confidence Intervals and Precision

The risk-difference interval for Test Product compared with Reference Product runs from -0.055 to 0.155. Because it includes zero and extends to values on both sides of zero, the recalculated estimate is not precise enough to establish a clear direction for that comparison. The interval also illustrates why a non-small p-value should not be read as proof that the products produce the same outcome: a range of differences remains compatible with the data under the calculation.

For Placebo Product compared with Reference Product, the risk-difference interval runs from -0.413 to -0.21. It lies below zero, but its width still represents uncertainty about the size of the difference. The estimate and interval are unadjusted recalculations based on posted counts and denominators, not sponsor-adjusted estimates.

The sponsor analysis reports a 90% confidence interval from -4.38 to 14.44 for an equivalence ratio of 1.11. Because the record does not state the equivalence margin or provide the interpretation details for that scale, the interval cannot be used from this record alone to establish equivalence. It should also not be conflated with the risk-difference intervals, which answer a different question.

General statistical explanation: interval width reflects both the amount of information and the variability under the analysis. A wider interval permits a broader set of plausible values and indicates less precision. Precision is distinct from whether an estimate is clinically important, and neither a p-value nor an interval alone answers questions beyond the stated Day 57 endpoint and the analyzed population.

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 fractions of the observed effect, not the original trial's design calculation. The assumptions listed for those illustrations 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 to the stated fraction of the observed effect. The trial's own original design assumptions are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for a design that would have an 80% chance of detecting the specified effect under the listed assumptions, if that effect and those assumptions were correct. The 90% power column gives the corresponding participants per arm for a design with a 90% chance under those same assumptions. Power is a property of a proposed design relative to a specified effect and analysis; it is not the probability that a treatment effect is real.

The rows for planning at 100%, 75%, and 50% of the observed effect show how the assumed target effect changes the planning requirement. The figures rise as the assumed effect becomes smaller: a smaller effect is harder to distinguish from random variation, so more participants are needed for the same target power. The illustrations are produced separately for the comparisons and are not a restatement of the actual group enrollment.

These figures are participants per arm, not total events. The planning assumptions explicitly distinguish time-to-event calculations, which would use total events, from other calculations, which use participants per arm; this trial's binary endpoint falls in the latter category. The illustrations are not observed power: they do not describe the trial's achieved probability of detecting an effect after the results were known. They are conditional planning scenarios based on the observed effect and stated assumptions.

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 = 0.432; treatment proportion p₁ = 0.482; average p̄ = 0.457.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7045 = 1.3808.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7036 = 0.5922.
4. n = (1.3808 + 0.5922)² / (0.0503)² = 1,536.9, rounded up to 1,537 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 <- 0.431818181818182                          # control proportion
p1 <- 0.482142857142857                          # 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(2385318)
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 = 1,537 (80% power), 2,058 (90% power)
 75% of observed effect: n per arm = 2,727 (80% power), 3,650 (90% power)
 50% of observed effect: n per arm = 6,121 (80% power), 8,194 (90% power)
power.prop.test: 1536.9 
Simulated power at n = 1,537 per arm: 0.798 (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. Analyses Reported by the Sponsor

The registry reports a sponsor analysis for Test Product and Reference Product using Yates correction. Its named parameter is an equivalence ratio, with a value of 1.11 and a 90% confidence interval from -4.38 to 14.44. No p-value is reported for this analysis. The registry record does not provide the equivalence margin or the detailed rules used to interpret the result.

The additional risk-difference calculation is an unadjusted re-calculation from the posted summary counts. It estimates the test-minus-reference risk difference as 0.050, with a confidence interval from -0.055 to 0.155 and p-value 0.349. These analyses use different reported measures, so their numerical estimates and intervals are not directly interchangeable. The record does not provide enough information to reconcile the scales.

The registry record does not flag a discrepancy between the sponsor analysis and the recalculation. The recalculation is useful as a transparent comparison of the posted counts, but it does not supersede or reproduce the sponsor's equivalence analysis. The reported equivalence result cannot be fully evaluated without the missing margin and analysis details.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and one reported analysis. It also records zero posted secondary outcomes. The recalculated results include comparisons involving the active reference product and placebo; their p-values are unadjusted. The registry record does not state whether a multiplicity procedure was prespecified or applied to the sponsor's analysis.

General statistical explanation: when several hypotheses are tested, the chance of at least one apparently positive result can increase if each is evaluated separately at the same threshold. A multiplicity plan can specify how to control that risk, for example through a testing sequence or adjusted thresholds. The absence of a reported plan does not show that no plan existed; it means the plan is not reported in the registry record.

The outcome analysis population is per protocol. The record does not say how missing Day 57 assessments were handled, whether participants with incomplete assessments were excluded, or how protocol deviations affected inclusion. It also does not report an intention-to-treat analysis for this endpoint. These details matter because analysis population and missing-data handling can affect which participants contribute to an estimate and how results should be interpreted.

The registry record does not report the original statistical analysis plan, prespecified interim analyses, or the sample-size justification. It therefore does not permit a full check of whether the posted outcome analysis followed all planned procedures. The observed safety denominators are different from the primary outcome denominators; the record does not explain the detailed population rules behind that difference. These uncertainties should be kept separate from the numerical findings that are explicitly posted.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: acute sigmoid diverticulitis/colitis (1.0%), sudden death, probable myocardial infarction (0.5%), sepsis (0.5%), left breast cancer (0.5%), worsening of right hip osteoarthritis (0.5%).

The serious adverse event counts are listed with separate at-risk denominators for each safety group. The Test Product group lists 4 participants affected among 204 at risk, the Reference Product group lists 1 among 210, and the Placebo Product group lists 1 among 105. The record also lists deaths as 1, 0, and 0, respectively. These counts should be read as registry safety summaries; the record does not state that the events were caused by study treatment.

For other, non-serious adverse events, the record lists 11 participants affected among 204 at risk in Test Product, 14 among 210 in Reference Product, and 0 among 105 in Placebo Product. The listed serious-event terms include Acute Sigmoid Diverticulitis/Colitis at 1.0%, and Sudden Death, Probable Myocardial Infarction; Sepsis; Left Breast Cancer; Worsening of Right Hip Osteoarthritis; and Acute Gastroenteritis/Food Poisoning, each at 0.5%.

The term-level percentages are reported as listed and should not be added together: the record does not establish that terms represent mutually exclusive participants. The safety counts also use different denominators from the per protocol endpoint analysis. No attribution, severity detail beyond the serious-event listing, or comparative safety model is reported in the registry record.

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

The posted primary outcome is a specific measure: complete absence of actinic keratosis lesions by investigator visual inspection at Day 57 in the per protocol population. The counts describe how many participants met that definition in each group. They do not describe partial lesion improvement, durability of clearance, recurrence, or outcomes beyond that assessment.

The recalculated test-versus-reference interval includes zero, and its unadjusted p-value is 0.349. This means the posted summary data do not give a precise estimate that clearly determines the direction of the difference between those products. It does not prove that their outcomes are identical, and it does not establish equivalence. The sponsor's own equivalence analysis reports a separate measure, but its margin and fuller interpretation rules are not reported in the registry record.

The placebo-versus-reference recalculation has a negative risk-difference estimate and an interval entirely below zero, with p <0.001. This describes the observed comparison for the specified endpoint and population. It should not be extended automatically to other outcomes or groups, or treated as a complete account of benefit and harm.

The registry's safety counts provide a separate descriptive record of serious and other adverse events. They do not, in the absence of reported attribution and fuller analysis details, establish that an event was caused by a product or settle comparative safety. The endpoint and safety summaries also use different analysis denominators.

Overall, the evidence described by the registry is bounded by the Day 57 visual-clearance endpoint, the per protocol analysis, the reported group counts, and the methods actually recorded. The registry record does not include enough detail to answer questions about missing outcomes, multiplicity, or the trial's original design rationale.

17. Why This Trial Matters Statistically

This trial illustrates why a posted event count, a recalculated comparison, and a sponsor-reported analysis should be distinguished. The primary outcome is a binary participant-level measure, while the sponsor analysis is labeled an equivalence ratio and the recalculation is a risk difference. Those measures answer related but non-identical statistical questions. Reading them as interchangeable could obscure what each estimate represents.

The contrast between the active-product comparison and the placebo comparison also demonstrates the importance of precision. A confidence interval that spans zero leaves uncertainty about the direction of the recalculated active comparison, while an interval that lies below zero conveys a different pattern for the placebo comparison. Neither interval by itself establishes clinical importance or answers questions beyond the recorded endpoint.

The trial also highlights how analysis populations matter. The primary outcome is reported for the per protocol population, whereas safety counts use their own at-risk denominators. Those figures should not be combined or treated as if they came from one shared analysis set. Clear reporting of population definitions and missing-data handling is necessary to assess how well each result represents the randomized groups.

Finally, planning illustrations based on an observed effect are not the same as the original trial design justification or observed power. They show how assumptions about the target effect influence participant requirements, but the registry record does not report the original rationale. Keeping those distinctions explicit helps readers interpret the evidence without making unsupported claims about equivalence, efficacy, safety, or design adequacy.

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 gives limited detail about the trial procedures and analysis. It identifies randomization, double masking, a parallel model, and three study groups, but does not describe the randomization method, allocation ratio, or who was masked. The protocol details needed to understand treatment administration and assessment procedures are not reported in the registry record.

The primary result is based on the per protocol population. The record does not report the criteria for that population, the handling of missing Day 57 assessments, or the treatment of protocol deviations. It also does not report an intention-to-treat primary outcome analysis. Those omissions constrain assessment of how participant inclusion may have influenced the posted estimates.

The sponsor's equivalence analysis is identified by its method and parameter, but the equivalence margin and fuller interpretation rules are not reported. The unadjusted recalculations use a different measure, so they cannot fill that gap. A detailed multiplicity plan, original sample-size justification, interim analysis plan, and full statistical analysis plan are also not reported in the registry record.

The primary endpoint captures complete clearance by visual inspection at Day 57, not other measures of disease status or longer-term outcomes. Safety findings are posted as descriptive counts and terms, without enough information to establish treatment attribution or to support a detailed comparative safety analysis. These limitations apply to interpretation of the registry record and do not imply that the missing information was absent from study documents outside the record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: To Compare Safety and Efficacy of Perrigo's Drug Product Compared to an FDA Approved Drug Product in the Treatment of Actinic Keratosis — NCT02385318.
• Registry record last updated 2021-11-02; status last verified 2021-10.
• Results first posted 2021-01-22.
• 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

NCT02385318 is a completed, randomized, double-masked, parallel phase 3 study sponsored by Padagis LLC. It compared a Perrigo ingenol mebutate test product with a reference ingenol mebutate product and placebo gel in actinic keratosis. Enrollment is listed as 519; the recorded study start is 2015-02 and primary completion is 2015-12.

The primary outcome was the number of participants in the per protocol population with complete lesion clearance at Day 57 by investigator visual inspection. The registry reports group counts and a sponsor equivalence analysis for test versus reference; recalculated unadjusted risk-difference comparisons provide additional context. Safety summaries report serious and other adverse-event counts using separate denominators.

The registry record does not report the original sample-size justification, full analysis plan, multiplicity strategy, or missing-data handling. Interpretation is therefore limited to the endpoint, populations, counts, and analysis details explicitly recorded.