Tofacitinib Ointment For Chronic Plaque Psoriasis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tofacitinib Ointment For Chronic Plaque Psoriasis (NCT01831466), a randomized phase 2 study. The primary outcome is Percentage of Participants Achieving a PGA-C Response of Clear (0) or Almost Clear (1) and Greater Than or Equal to (≥) 2 Grade/Point Improvement From Baseline at Week 12.

2. Clinical Question

The registry describes a Phase 2b, randomized, double-blind, vehicle-controlled, parallel-group study. It enrolled 476 participants and is listed as completed. The study's stated aim was to assess efficacy, safety, local tolerability, and pharmacokinetics for two ointment strengths and two application regimens. The primary endpoint focused on a categorical clinical response at Week 12 rather than a continuous change in severity.

The endpoint was defined using the Physician's Global Assessment of Clinical signs (PGA-C). It required both a final category of clear or almost clear and improvement from baseline. Missing primary endpoint values were counted as nonresponses. The registry record does not report the original protocol's detailed statistical plan or the reason this endpoint and assessment time were selected.

3. Trial Design

This was a Phase 2b, multi-site, randomized, double-blind, vehicle-controlled, parallel-group trial of topical tofacitinib for chronic plaque psoriasis. The registry labels masking as quadruple, but does not identify which roles were masked. The study was sponsored by Pfizer and is listed as completed. The recorded start was 2013-05 and primary completion was 2014-09.

The study compared two concentrations, 20 mg/g and 10 mg/g, applied once daily or twice daily, with vehicle applied at the corresponding frequency. Thus, the primary outcome results are reported separately for four tofacitinib regimens and the two vehicle regimens. The posted primary analysis is restricted to the mild/moderate population; the registry also reports safety groups labeled severe.

The study summary describes a 12-week treatment period. The primary endpoint was assessed at Week 12 and concerned the percentage of participants meeting a combined PGA-C response criterion. Other posted outcomes include PGA-C response without the improvement requirement, PGA-G response, and PASI measures at Week 8 or Week 12. The registry lists 18 outcome measures and 8 analyses.

Enrollment was 476, but the primary outcome denominators differ across groups and represent the specific analysis population, not the enrollment figure. The record does not report the original sample-size justification, allocation ratio, details of any stratification factors, interim analyses, or a full protocol-level analysis plan. Those details should not be inferred from the posted endpoint summaries or recalculated estimates.

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.

Only the first 6 of 12 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The mild/moderate full analysis set (FASm) included all participants in the FAS with a baseline PGA-C disease severity of mild (2) or moderate (3). A participant with a missing value was considered a non-responder.

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: Full Analysis Set (FAS) - included all participants who were randomized to the study, received at least one dose of the randomized investigational drug (tofacitinib or vehicle), and were in a baseline PGA-C category of mild (2), moderate (3) or severe (4).

Only the first 6 of 13 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Achieving a PGA-C Response of Clear (0) or Almost Clear (1) and Greater Than or Equal to (≥) 2 Grade/Point Improvement From Baseline at Week 12, measured over Baseline, Week 12.

The primary endpoint was the percentage of participants who, at Week 12, had a PGA-C score of clear or almost clear and at least a 2-point improvement from baseline. PGA-C assesses clinical signs of plaque psoriasis: erythema, induration, and scaling. Each sign is scored on a 5-point scale from 0 to 4. The subscores are combined and rounded to determine the overall PGA-C category; higher scores indicate greater severity. A score of 0 means clear except for possible residual discoloration, while 1 means almost clear.

This endpoint combines the participant's status at Week 12 with change from baseline. Requiring both a low final severity category and improvement can distinguish a participant who reaches a low score after improvement from one who began near that category. It is a binary outcome: each participant either meets the response definition or does not. The reported percentage therefore describes the share meeting the criterion in each group, not the average amount of change.

The primary population included participants in the full analysis set with mild or moderate baseline PGA-C severity. A missing endpoint value was treated as a nonresponse. This rule is important to the interpretation because it specifies how missing observations affect the reported response percentage, although the registry record does not provide the number of missing values or further detail about the full analysis set.

7. Results Posted in the Registry

• Mild/Moderate: Tofacitinib 20 mg/g BID: 21.1% (15/71) versus 16.9% (12/71) on control; risk difference 4.2 percentage points (95% CI -8.7 to 17.1), risk ratio 1.25 (0.63–2.48); unadjusted two-proportion z-test p = 0.521.
• Mild/Moderate: Tofacitinib 10 mg/g BID: 12.9% (9/70) versus 16.9% (12/71) on control; risk difference -4.0 percentage points (95% CI -15.8 to 7.7), risk ratio 0.76 (0.34–1.69); unadjusted two-proportion z-test p = 0.500.
• Mild/Moderate: Tofacitinib 20 mg/g QD: 20.0% (14/70) versus 16.9% (12/71) on control; risk difference 3.1 percentage points (95% CI -9.7 to 15.9), risk ratio 1.18 (0.59–2.38); unadjusted two-proportion z-test p = 0.635.
• Mild/Moderate: Tofacitinib 10 mg/g QD: 21.6% (16/74) versus 16.9% (12/71) on control; risk difference 4.7 percentage points (95% CI -8.1 to 17.5), risk ratio 1.28 (0.65–2.51); unadjusted two-proportion z-test p = 0.472.

8. Secondary Efficacy Endpoints

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

Percentage of Participants Achieving a PGA-C Response of Clear (0) or Almost Clear (1) at Week 12

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Mild/Moderate: Tofacitinib 20 mg/g BID: 33.8% (24/71) versus 23.9% (17/71) on control; risk difference 9.9 percentage points (95% CI -5.0 to 24.7), risk ratio 1.41 (0.83–2.39); unadjusted two-proportion z-test p = 0.195.
  • Mild/Moderate: Tofacitinib 10 mg/g BID: 25.7% (18/70) versus 23.9% (17/71) on control; risk difference 1.8 percentage points (95% CI -12.5 to 16.0), risk ratio 1.07 (0.60–1.91); unadjusted two-proportion z-test p = 0.808.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mild/Moderate: Tofacitinib 20 mg/g BID and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mild/Moderate: Tofacitinib 10 mg/g BID and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Achieving a PGA-C Response of Clear (0) or Almost Clear (1) at Week 8

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Mild/Moderate: Tofacitinib 20 mg/g BID: 36.6% (26/71) versus 22.5% (16/71) on control; risk difference 14.1 percentage points (95% CI -0.7 to 28.9), risk ratio 1.62 (0.96–2.76); unadjusted two-proportion z-test p = 0.066.
  • Mild/Moderate: Tofacitinib 10 mg/g BID: 20.0% (14/70) versus 22.5% (16/71) on control; risk difference -2.5 percentage points (95% CI -16.0 to 11.0), risk ratio 0.89 (0.47–1.68); unadjusted two-proportion z-test p = 0.713.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mild/Moderate: Tofacitinib 20 mg/g BID and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mild/Moderate: Tofacitinib 10 mg/g BID and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Achieving a Gestalt Physician's Global Assessment (PGA-G) Response of Clear (0) or Almost Clear (1) and ≥2 Grade/Point Improvement From Baseline at Week 12

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Mild/Moderate: Tofacitinib 20 mg/g BID: 18.3% (13/71) versus 14.1% (10/71) on control; risk difference 4.2 percentage points (95% CI -7.9 to 16.3), risk ratio 1.30 (0.61–2.77); unadjusted two-proportion z-test p = 0.494.
  • Mild/Moderate: Tofacitinib 10 mg/g BID: 11.4% (8/70) versus 14.1% (10/71) on control; risk difference -2.7 percentage points (95% CI -13.7 to 8.3), risk ratio 0.81 (0.34–1.94); unadjusted two-proportion z-test p = 0.637.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mild/Moderate: Tofacitinib 20 mg/g BID and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Mild/Moderate: Tofacitinib 10 mg/g BID and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes cover several ways of assessing psoriasis response, including PGA-C categories, PGA-G response, and PASI change or response at Week 8 and Week 12. Taken together, they examine whether the primary endpoint pattern is accompanied by differences in other clinical assessments and at an earlier visit. The registry posts 16 secondary outcomes, but only a subset includes the detailed arm results and recalculated comparisons described in the record.

Across the posted secondary results, the pattern is mixed rather than uniformly favoring one regimen. Some active groups have higher observed response percentages than their corresponding vehicle groups, while other comparisons are smaller or in the opposite direction. The available comparisons are generally imprecise, and their recalculated intervals include zero. This broad pattern is consistent with the uncertainty in the primary outcome; it does not independently establish a clear or consistent between-group separation.

Secondary outcomes can provide context by showing whether conclusions are similar across different definitions, assessments, or time points. They do not replace the primary endpoint, which was specified as the Week 12 PGA-C response with improvement from baseline in the mild/moderate population. Differences in endpoint definition and time of assessment mean that secondary results address related but not identical questions.

Unadjusted secondary results carry less weight when many outcomes and comparisons are examined. In general, the more comparisons considered, the greater the chance that some may appear notable by chance. The registry record does not report whether multiplicity adjustments were applied to secondary outcomes. Their results are therefore best treated as descriptive or supportive evidence, with attention to the consistency of the overall pattern rather than selective emphasis on an individual comparison.

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

• Percentage of Participants Achieving a PGA-C Response of Clear (0) or Almost Clear (1) at Week 12
• Percentage of Participants Achieving a PGA-C Response of Clear (0) or Almost Clear (1) at Week 8
• Percentage of Participants Achieving a Gestalt Physician's Global Assessment (PGA-G) Response of Clear (0) or Almost Clear (1) and ≥2 Grade/Point Improvement From Baseline at Week 12
• Percentage of Participants Achieving a PGA-G Response of Clear (0) or Almost Clear (1) and ≥2 Grade/Point Improvement From Baseline at Week 8
• Percent Change From Baseline to Week 12 in Psoriasis Area and Severity Index (PASI)
• Percent Change From Baseline to Week 8 in PASI
• Percentage of Participants Achieving at Least a 75% Reduction in PASI Response (PASI75), Relative to Baseline at Week 12
• Percentage of Participants Achieving PASI75, Relative to Baseline at Week 8

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary: a participant either meets the Week 12 response definition or does not. A response percentage is the proportion of participants in a group who meet that definition. For a two-group comparison, a risk difference expresses the difference between those proportions. A value above zero means the proportion was higher in the active-treatment group; a value below zero means it was lower. The simple recalculations use the posted response counts and denominators and are unadjusted comparisons.

The primary response counts in the mild/moderate groups were 15 of 71 for tofacitinib 20 mg/g twice daily and 12 of 71 for twice-daily vehicle; 9 of 70 for 10 mg/g twice daily; 14 of 70 for 20 mg/g once daily; 16 of 74 for 10 mg/g once daily; and 13 of 74 for once-daily vehicle. These counts describe the posted binary endpoint. They do not convey the degree of improvement among participants who did not meet the full response definition.

The registry's own inferential analysis is identified as Cochran-Mantel-Haenszel. In general, this method compares categorical outcomes while accounting for defined strata. The registry record does not identify the strata or give enough detail to reconstruct the sponsor's analysis. Its reported contrast is a difference in response rates, accompanied by an 80% interval and a p-value. Those sponsor results should be distinguished from the recalculated unadjusted risk differences and their intervals.

A p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of the test; it is not the probability that the null hypothesis is true. A larger p-value does not establish equivalence, and a smaller p-value alone does not show that a difference is important. Here, the sponsor p-values and the recalculated p-values all need to be read alongside the estimated differences and interval widths. The record does not report a prespecified threshold for clinical importance.

An interval gives a range of values compatible with the data under the analysis method and its assumptions. If an interval for a risk difference includes zero, the data are compatible with effects in either direction as well as no difference. The recalculated intervals include zero for each primary comparison. The sponsor reports 80% intervals, which are narrower than intervals at a higher confidence level would generally be, all else equal. The exact confidence level for the recalculated intervals is not stated in the registry record, so it should not be assumed from the sponsor's interval level.

The primary analysis population was the mild/moderate full analysis set, defined as participants in the full analysis set with baseline PGA-C severity of mild or moderate. The registry explicitly states that a missing value was classified as a nonresponse. This is a conservative-looking rule in the limited sense that missing observations do not count as successes, but it does not eliminate uncertainty from missing data. The number and reasons for missing observations, and any sensitivity analyses, are not reported in the registry record.

The study has multiple treatment regimens and multiple reported outcomes. In general, examining many comparisons increases the chance that some results will appear notable by chance. Adjustments for multiplicity can address that issue, but whether an adjustment was planned or applied is not reported in the registry record. The sponsor's stated analysis method should not be taken to imply an unreported multiplicity strategy.

For binary outcomes, over-dispersion is generally not the central concern in the same way it can be for count data; the main considerations include the observed event frequency, group sizes, missing outcome handling, and the model or comparison method. More generally, over-dispersion means that observed variability exceeds what a statistical model assumes. When relevant, ignoring it can make uncertainty appear too small. The registry record does not report an over-dispersion model for this binary endpoint, and no such model should be inferred.

Finally, the recalculations are simple unadjusted re-calculations from posted summary statistics, not substitutes for the sponsor's full analysis. They do not account for any stratification or other adjustments that may be part of the sponsor method. No discrepancy is flagged in the registry record, but the methods still differ: the sponsor reports Cochran-Mantel-Haenszel analyses, while the recalculated comparisons are unadjusted. The sponsor's reported analysis remains the primary source for the trial's inferential claims.

11. Confidence Intervals and Precision

The observed response percentages alone do not show how precisely the underlying group rates or between-group differences are estimated. Precision depends in part on the number of participants and responses in each group. The primary groups had denominators from 70 to 74, and the response counts were modest. With this amount of information, uncertainty around a difference can remain substantial even when the displayed percentages differ.

For the recalculated risk differences, every interval includes zero. The intervals also extend to values on either side of zero, indicating that the data are compatible with a higher response percentage in either group and with little or no difference. The intervals are not proof that there is no effect; rather, they show that the estimate is not precise enough to rule out a range of possible differences. The interval level for these recalculations is not reported in the registry record.

The sponsor reported 80% intervals for its response-rate differences. Those intervals likewise include zero. An 80% interval is constructed to have less coverage than a 95% interval under its method and assumptions, so the interval level matters when judging apparent precision. The sponsor's estimates and p-values are based on its Cochran-Mantel-Haenszel analysis, whereas the recalculated estimates use simple unadjusted comparisons. Differences between these results can reflect analytical approach as well as rounding.

None of these intervals captures every source of uncertainty. They do not resolve uncertainty about missing data, the clinical importance of the response threshold, multiplicity, or generalizability beyond the analyzed population and follow-up period. The registry record does not report a prespecified clinically important difference for this endpoint. Consequently, statistical precision should be considered together with the endpoint definition and the limits of the reported analysis, rather than inferred from whether a single interval includes zero.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm for the binary outcomes, not total participants across a comparison. The planning assumptions stated for these illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. The original trial's own sample-size assumptions and justification are not reported in the registry record.

The 80% power column represents the planned number of participants per arm under those assumptions for a design that would have an 80% chance of detecting the specified effect if that effect were the true effect and the assumptions held. The 90% power column uses a higher chance of detection and therefore generally requires more participants for the same target effect. These numbers are planning illustrations, not observed power.

Each set of planning figures uses an observed difference as its starting point. Planning for 75% or 50% of that observed effect means targeting a smaller difference than the full observed estimate. Smaller effects are harder to distinguish from random variation and require more participants. When the observed difference is close to zero, the resulting planning requirements can become very large; such figures should not be read as evidence that a future study needs exactly that size.

A sample-size curve expresses the same relationship visually: under the stated assumptions, larger target effects generally require fewer participants, while smaller target effects require more. The curve is conditional on its assumptions and is not a prediction of what a new trial will observe. It does not establish that the observed effect is a reliable estimate of the effect to plan around. The planning numbers also do not incorporate any unreported protocol features, adjustment strategy, or missing-data allowance.

13. How the Sample Size Is Derived

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

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.169; treatment proportion p₁ = 0.211; average p̄ = 0.190.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5550 = 1.0877.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5541 = 0.4664.
4. n = (1.0877 + 0.4664)² / (0.0423)² = 1,352.8, rounded up to 1,353 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 <- 0.169014084507042                          # control proportion
p1 <- 0.211267605633803                          # 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(1831466)
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 = 1,353 (80% power), 1,811 (90% power)
 75% of observed effect: n per arm = 2,355 (80% power), 3,152 (90% power)
 50% of observed effect: n per arm = 5,181 (80% power), 6,935 (90% power)
power.prop.test: 1352.8 
Simulated power at n = 1,353 per arm: 0.805 (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.

14. Analyses Reported by the Sponsor

The registry reports four sponsor analyses of the primary endpoint, each comparing an active regimen with the vehicle regimen of the same frequency. The method is Cochran-Mantel-Haenszel. The sponsor's reported response-rate differences were 3.9 for 20 mg/g twice daily, -4.0 for 10 mg/g twice daily, 3.3 for 20 mg/g once daily, and 4.0 for 10 mg/g once daily. The reported intervals are 80% intervals, and each includes zero. The corresponding sponsor p-values were 0.542, 0.498, 0.604, and 0.528.

The recalculated figures are simple unadjusted risk differences derived from posted counts and denominators. They are useful as a transparent check on the direction and approximate size of the comparisons, but they do not reproduce the sponsor's method. The sponsor analysis may account for strata or other features that are not described in enough detail in the registry record to reconstruct. Therefore, the sponsor's own analysis takes precedence when describing the trial's reported inferential findings.

The sponsor estimates and recalculated estimates are broadly similar in direction for each corresponding comparison, but they should not be treated as identical methods or interchangeable results. The sponsor's intervals are explicitly 80%; the confidence level for the recalculated intervals is not reported in the registry record. Neither analysis, as posted, resolves the uncertainty around the group differences.

15. Multiplicity, Missing Data and Analysis Populations

The record lists 18 outcome measures and 8 analyses, alongside several dose and schedule comparisons. When multiple hypotheses are tested, the chance of at least one apparently unusual result can increase even if no true differences exist. A prespecified multiplicity strategy can control that risk, but whether one was planned or applied is not reported in the registry record. The posted p-values should therefore not be interpreted as if each comparison were the only test conducted.

The primary endpoint's missing-data rule is reported: participants with a missing value were considered nonresponders. This approach makes the handling of missing primary outcomes explicit, but the number of missing values and reasons for missingness are not reported. The registry also does not report sensitivity analyses assessing alternative missing-data assumptions. The rule may affect response rates, especially if missingness differs across groups, and the available record does not allow that influence to be evaluated.

The primary analysis population is described as the mild/moderate full analysis set, comprising participants in the full analysis set with mild or moderate baseline PGA-C severity. The registry record does not provide a fuller definition of the full analysis set or clarify the analysis populations used for all secondary outcomes. It also does not report a detailed protocol or statistical analysis plan, so further population rules and analysis decisions should not be inferred.

In general, an interim analysis can affect statistical interpretation if it changes when or how results are evaluated. Interim analyses are not reported in the registry record. Nor does the record report whether the primary comparison was adjusted for multiple active regimens. These omissions do not demonstrate that such procedures were absent; they mean their role cannot be assessed from the registry record.

16. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: transient ischaemic attack (14.3%), cardiac failure congestive (1.4%), myocardial infarction (1.4%), non-cardiac chest pain (1.4%), abdominal wound dehiscence (1.4%).

The registry lists serious adverse event counts by treatment group for mild/moderate and severe participants. In the mild/moderate groups, the recorded counts were 0 for 20 mg/g twice daily, 5 for 10 mg/g twice daily, 2 for twice-daily vehicle, 0 for 20 mg/g once daily, 2 for 10 mg/g once daily, and 1 for once-daily vehicle. The severe-group counts were 0, 0, 0, 0, 0, and 1 in the corresponding sequence of regimens. These are counts, not estimates of comparative risk, and the group sizes vary.

The registry also lists serious-event terms with percentages, including transient ischaemic attack at 14.3% and several other terms at 1.4%. It does not provide enough accompanying detail in this record to interpret those listed percentages as group-specific rates or to reconstruct their denominators. Death counts are not reported in the registry record. Serious adverse event summaries alone cannot establish whether groups differ in safety, especially without fuller event definitions, follow-up context, and statistical comparisons. They should be read as descriptive registry data, not as a complete safety assessment.

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

The posted results describe the proportion of participants in the mild/moderate analysis population who met a particular Week 12 clinical response definition. They show that response percentages varied across the tofacitinib regimens and vehicle groups. The sponsor's reported comparisons and the unadjusted recalculations do not show a precise separation between the active and corresponding vehicle groups: the reported sponsor intervals and recalculated intervals include zero.

This does not demonstrate that all regimens have identical effects. A result that is uncertain can be compatible with no difference and with differences in either direction. It also does not establish that a regimen is beneficial or ineffective. The registry outcome is one endpoint at one assessment time in a defined population, and the available results do not settle questions about other outcomes, populations, or durations.

The primary measure is a categorical threshold combining clear or almost clear status with improvement from baseline. It does not convey the full range of changes in signs, nor does it establish how participants valued those changes. The registry record does not report a clinically important difference threshold for interpreting the size of the risk differences.

These findings also should not be taken as a complete appraisal of safety. The registry provides adverse-event counts and selected serious-event terms, but not a full clinical interpretation of event timing, severity, relatedness, or comparative uncertainty. The trial's posted results are one source of evidence, and the record does not report enough detail to answer every design and analysis question. No treatment decision follows from these statistics alone.

18. Why This Trial Matters Statistically

This trial illustrates why response percentages need to be interpreted alongside their comparison estimates and uncertainty. Looking only at the observed percentages can make differences appear more conclusive than the data support. The recalculated risk differences and their intervals show that the primary comparisons remain compatible with a range of possible effects. The sponsor's own analysis similarly reports 80% intervals that include zero.

It also demonstrates how a clinical endpoint definition shapes the statistical question. The primary outcome is not simply a low PGA-C score; it requires both a clear or almost-clear category and improvement from baseline. The result therefore concerns the probability of meeting that combined criterion, rather than average symptom change or every aspect of disease severity.

Multiple regimens and multiple endpoints create an interpretation challenge. A pattern across outcomes can be informative, but isolated comparisons deserve caution when multiplicity handling is not reported. The distinction between sponsor analyses and simple unadjusted recalculations is also important: transparent recalculation can help readers understand the posted counts, while the sponsor method remains the primary reported inference.

Finally, planning figures based on observed effects are not evidence of the trial's original design rationale or an estimate of achieved power. They illustrate how a future planning target changes required sample size under stated assumptions. Keeping the observed results, inferential uncertainty, and hypothetical planning exercise separate is essential for sound statistical interpretation.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the original sample-size justification, the full statistical analysis plan, details of any stratification used in the sponsor's Cochran-Mantel-Haenszel analysis, or whether multiplicity adjustments were planned. These omissions limit independent reconstruction of the reported inferential results. The sponsor's reported analysis is therefore the appropriate reference for the trial's own stated comparisons.

The primary endpoint is restricted to the mild/moderate full analysis set, although safety results also include groups labeled severe. The primary efficacy results should not be generalized to severe baseline disease on the basis of these data. In addition, the record's definition of the full analysis set is incomplete, and outcome population details for secondary analyses are not fully specified.

The registry states that missing primary endpoint values were counted as nonresponses, but does not give the number or distribution of missing values, reasons for missingness, or sensitivity analyses. This prevents assessment of how robust the response comparisons are to alternative assumptions. The posted primary outcome is also a binary threshold, which omits information about changes that do not cross the response cutoff.

Comparisons involve several concentrations, dosing schedules, and endpoints. The record does not report a multiplicity approach, and many secondary outcomes are not accompanied by detailed analyses in the posted information. The sponsor intervals are 80%, while the level for the recalculated intervals is not stated. These factors, together with the modest group denominators, constrain precision and make individual differences difficult to interpret as definitive.

Safety information is also limited in this record. Serious adverse event counts are provided, but the listed serious-event terms and percentages lack enough context to infer comparative rates. Death counts are not reported in the registry record. The trial covers the stated 12-week period, and longer-term findings are not reported in the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Tofacitinib Ointment For Chronic Plaque Psoriasis — NCT01831466.
• Registry record last updated 2015-11-25; status last verified 2015-10.
• Results first posted 2015-11-25.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

This completed Phase 2b randomized, double-blind, vehicle-controlled, parallel-group study evaluated tofacitinib ointment at 20 mg/g or 10 mg/g, applied once or twice daily, in chronic plaque psoriasis. Enrollment was 476. The primary outcome was the Week 12 percentage meeting a PGA-C clear or almost-clear response with at least a 2-point improvement from baseline, among participants with mild or moderate baseline severity.

The primary response percentages differed across regimens, but the sponsor's reported intervals and the simple unadjusted recalculated intervals include zero for each active-versus-vehicle comparison. The findings are therefore uncertain rather than definitive evidence of a between-group difference. Missing primary outcomes were counted as nonresponses. The registry record does not report the original sample-size justification, full statistical plan, or multiplicity approach. Safety counts are descriptive, and death counts are not reported in the registry record.