Pembrolizumab in Patients With Locally Advanced or Metastatic Skin Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pembrolizumab in Patients With Locally Advanced or Metastatic Skin Cancer (NCT02964559), a na phase 2 study. The primary outcome is Response Rate of Pembrolizumab.

2. Clinical Question

3. Trial Design

This is a phase II study sponsored by Emory University, registered as NCT02964559. The registry lists a single-group model, allocation as NA, and no masking. It reports enrollment of 11 and a completed status. The study started on 2017-01-27, and primary completion was recorded as 2021-05-18. These record dates describe the trial timeline; they do not establish when individual participants began or ended treatment.

Participants received pembrolizumab by intravenous infusion over 30 minutes on day 1 of a course. Courses were to repeat every 3 weeks while disease progression or unacceptable toxicity had not occurred. The registry does not report individual treatment duration, treatment exposure, adherence, or the number of courses received.

With no randomized or concurrent control group, the design can describe outcomes observed among enrolled participants but cannot directly estimate a treatment contrast against another assigned intervention. The registry lists overall survival and progression-free survival as secondary outcomes, but does not post numerical results for them in the record summarized by the available outcome information. It also does not report a detailed protocol, a sample-size justification, or a statistical analysis plan. Accordingly, design features beyond those explicitly recorded cannot be assumed.

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: all patients except patient diagnosed with covid-19

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: all participants except patient diagnosed with covid

6. Primary Endpoint

The registry lists the primary outcome as Response Rate of Pembrolizumab, measured over Up to 1 year after treatment discontinuation (max treatment duration of 3 years).

The primary outcome is named “Response Rate of Pembrolizumab” and uses a percentage of participants as its unit. In general, a response rate expresses the share of an analysis population classified as having a response under a prespecified assessment rule. It can offer a compact description of tumor response, but its meaning depends on how response is defined, when assessments occur, and how participants without an evaluable assessment are handled.

The registry states that the timeframe extends up to 1 year after treatment discontinuation, with a maximum treatment duration of 3 years. Its accompanying description cites RECIST v1.0 and defines progression using a 20% increase in the sum of the longest diameter of target lesions, a measurable increase in a non-target lesion, or the appearance of new lesions. That wording describes progression, not the criteria for classifying a response. The operational response definition is therefore not clear from the registry record.

The endpoint is listed with a 95% confidence interval as its dispersion measure. A confidence interval would communicate uncertainty around the estimated response percentage, but neither the estimate nor its interval is reported in the available outcome results. The registry identifies the primary population as all patients except a patient diagnosed with covid-19, without clarifying the analysis denominator.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation begins with identifying the estimand being described: the outcome, the population, and the time period. The primary outcome is labeled response rate, uses percentage of participants as its unit, and has a timeframe of up to 1 year after treatment discontinuation, with a maximum treatment duration of 3 years. Its stated population excludes a patient diagnosed with covid-19. The registry does not report the resulting analysis denominator or the response-rate estimate, so neither the event count nor the estimated percentage can be recovered from the posted information.

What does a confidence interval communicate?

In general, a confidence interval describes the statistical uncertainty around an estimated quantity under the assumptions of the method used. A narrow interval indicates greater precision than a wide interval, while an interval that is wide signals substantial uncertainty. An interval is not a guarantee that the true value lies within its bounds, and it does not replace the need to understand the study design or endpoint definition. The registry specifies a 95% confidence interval as the primary outcome’s dispersion measure, but does not post the interval or the estimate. Precision therefore cannot be assessed from the outcome record.

Why does the single-group design matter?

In general, random assignment to a concurrent control group helps balance measured and unmeasured factors between treatment groups, supporting a causal comparison. Here, the registry lists a single-group model, allocation as NA, and no masking. There is no like-for-like control comparison. A response percentage, if reported, would describe participants classified as responding in this trial; it would not by itself quantify how outcomes differ from an alternative treatment. The record does not report randomization or a comparator, so a comparative treatment effect should not be inferred.

What is the distinction between response and progression?

In general, response criteria classify changes in disease according to prespecified rules, while progression criteria identify worsening under a defined rule. These are related but distinct outcome classifications. The registry names the primary measure as response rate, yet its explanatory text describes progression using RECIST v1.0: a 20% increase in the sum of the longest diameter of target lesions, a measurable increase in a non-target lesion, or new lesions. The response-classification rule is not reported in the registry record. This mismatch limits the ability to interpret the intended endpoint operationally, even apart from the lack of a posted response estimate.

What can the safety denominators tell readers?

In general, an adverse-event count should be read alongside the number of participants at risk and the event definition. The registry lists 11 participants at risk for other adverse events and 11 affected. For serious adverse events, it lists 11 at risk and 6 affected, with zero deaths. These data describe reported counts within the treatment group. They are not adjusted for exposure time, and no comparative risk estimate is available from a single group. The record does not report additional details needed to interpret event timing, duration, or attribution.

Which analysis choices remain unknown?

The registry record does not report the method used to calculate the response-rate confidence interval, the handling of missing or unevaluable assessments, or a detailed statistical analysis plan. It does not report interim analyses, a sample-size justification, or methods for addressing multiplicity. These omissions mean readers should not attribute a particular method or protocol choice to the investigators. General statistical principles can explain why such choices matter, but cannot substitute for a trial-specific description that is not recorded.

Planning calculations are distinct from observed results. In general, a sample-size calculation depends on assumptions about the outcome, the effect to detect, the allocation scheme, and the chosen error rates. Any planning illustration based on an observed effect is not evidence that the study had a particular probability of detecting an effect, and it is not an observed trial result. The registry does not report its own sample-size rationale, so planning assumptions should not be treated as the trial’s original design basis.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Treatment (Pembrolizumab)")
n      <- c(11)
events <- c(8)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Treatment (Pembrolizumab)  11  8  72.7  39.0  94.0  43.4  90.3  42.8  90.1

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary response-rate outcome and the secondary outcomes overall survival and progression-free survival. It also reports three outcome measures and zero analyses in the outcome information available. The record does not describe a multiplicity strategy, such as how interpretation would account for examining more than one outcome. Without such a plan, readers should not assume that reported tests were adjusted, or that a particular hierarchy of outcomes was followed.

Missing-data methods are also not reported in the registry record. This matters because response classification may depend on having an evaluable assessment at the relevant time. The stated primary population excludes a patient diagnosed with covid-19, but the record does not explain the rationale for this exclusion, identify the resulting analysis denominator, or say how other missing or unevaluable assessments were handled. It would be inappropriate to infer that an excluded participant was counted as a nonresponder or was handled by another specific rule.

The registry does not report a detailed analysis population definition beyond the stated covid-19 exclusion, nor does it describe interim analyses or a prespecified statistical analysis plan. These gaps limit reproducibility and make it difficult to judge whether outcome definitions and analysis decisions were fixed before results were known. Such information is not reported in the registry record and should not be guessed. The available information supports a description of the listed outcomes and populations, but not a reconstruction of the analysis procedures.

12. 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: edema cerebral (9.1%), fall (9.1%), general disorders and administration site conditions - other, specify (9.1%), hematoma (9.1%), hydrocephalus (9.1%).

The registry’s pembrolizumab safety group lists 11 participants at risk for serious adverse events, of whom 6 were affected. For other adverse events, 11 were at risk and 11 were affected. Deaths are listed as 0. These are descriptive counts within the study group; there is no concurrent comparator from which to estimate a relative difference in risk.

The listed serious adverse-event terms include Edema cerebral, Fall, General disorders and administration site conditions - Other, specify, Hematoma, Hydrocephalus, and Hypotension. Each is shown with 9.1%. The registry labels these as top serious terms. The percentages do not establish causation, severity beyond the serious-event classification, duration, or whether events occurred in the same or different participants. The record does not provide a detailed event-level narrative or exposure-adjusted rates.

In general, adverse-event reporting captures events observed during a defined collection period and should be interpreted with the reporting rules and denominator in mind. A count among participants receiving the intervention cannot establish whether the intervention caused an event. The registry information does not report an attribution analysis or a controlled safety comparison.

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

The record establishes that a phase II, single-group study of pembrolizumab enrolled 11 participants and is listed as completed. It identifies response rate as the primary outcome and gives a timeframe extending up to 1 year after treatment discontinuation, with a maximum treatment duration of 3 years. It also gives a 95% confidence interval as the intended measure of dispersion. These details describe the planned outcome framework, not the observed primary result.

The primary response-rate estimate and confidence interval are not reported in the registry record. The outcome description cites RECIST v1.0 but states a progression rule rather than a clear response rule. As a result, the posted information does not establish the proportion of participants classified as responding, the uncertainty around that proportion, or how the classification was applied. Numerical results for the listed secondary outcomes are also not present in the available information.

The safety record reports counts within the pembrolizumab group, including serious adverse events in 6 of 11 participants at risk and other adverse events in 11 of 11 participants at risk. It lists zero deaths and several serious-event terms. These descriptions do not show whether the intervention caused an event or whether event frequency differed from what would have occurred under another treatment.

More generally, a single-group trial can document outcomes in its participants, but without a comparator it cannot provide a direct randomized estimate of a treatment contrast. The record also does not report key analysis details for missing data, multiplicity, interim analyses, or sample-size justification. The available information therefore supports a description of the trial design and recorded safety counts, but not a conclusion about the size or certainty of a treatment effect.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of why endpoint labels and endpoint definitions need to be read together. The primary outcome is called response rate, but the accompanying RECIST v1.0 description specifies progression criteria. Because the registry does not report the response result or clarify the operational response definition, readers cannot connect the stated endpoint to a numerical estimate or assess its precision.

The study also illustrates the limits of interpreting a single-group design. A study can describe outcomes among participants who receive an intervention, but a concurrent randomized comparator is needed for a direct randomized contrast. Here, allocation is listed as NA and the model as single-group. Any outcomes should therefore be understood as observations in the trial cohort, not as an isolated estimate of what changed because of treatment.

The safety information highlights the importance of reporting denominators and event definitions. The registry gives the numbers at risk and affected for serious and other adverse events, alongside listed serious-event terms. Those details allow a basic description of the recorded safety counts, while the absence of a comparator and further event-level context limits causal and comparative interpretation.

Finally, the record shows why analysis transparency matters. It does not report the response estimate, confidence interval, missing-data rules, multiplicity approach, interim analyses, or trial-specific sample-size justification. These are not details that can be reconstructed from enrollment or outcome labels. A careful reading distinguishes what the registry records from what remains unknown, avoiding a stronger statistical conclusion than the available data support.

15. 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.

Several limitations specific to this trial record constrain interpretation. Enrollment was 11, and the trial had a single-group design with no comparator. The study therefore provides no randomized between-group contrast. The registry does not report the primary response-rate estimate or its 95% confidence interval, and it does not give numerical results for the listed secondary outcomes in the available outcome information.

The primary-outcome description is not fully aligned: the title identifies response rate, while the accompanying RECIST v1.0 description specifies progression criteria. The record does not state the response rule that was applied. It also describes the primary population as all patients except a patient diagnosed with covid-19, without clarifying the analysis denominator or the reason for the exclusion.

The registry does not report methods for missing or unevaluable outcomes, a detailed statistical analysis plan, multiplicity handling, interim analyses, or a sample-size justification. The treatment schedule is stated, but individual treatment exposure and follow-up details are not provided in the available information. Safety counts are posted without a control group or event-level detail sufficient to determine attribution or comparative risk.

These are limitations of what can be interpreted from the registry record; they do not establish that a method was or was not used outside the information recorded. The missing details should be treated as unknown rather than filled in by assumption.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pembrolizumab in Patients With Locally Advanced or Metastatic Skin Cancer — NCT02964559.
• Registry record last updated 2024-06-26; status last verified 2024-06.
• Results first posted 2024-06-26.
• 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.

19. Record Summary

NCT02964559 is a completed phase II, single-group trial sponsored by Emory University, with enrollment of 11. It studied pembrolizumab in metastatic cutaneous squamous cell carcinoma. The intervention was an intravenous infusion over 30 minutes on day 1, with courses repeating every 3 weeks in the absence of progression or unacceptable toxicity. The registry lists no masking and allocation as NA.

The primary outcome is response rate, expressed as a percentage of participants, with a 95% confidence interval specified. Its stated timeframe is up to 1 year after treatment discontinuation, with a maximum treatment duration of 3 years. The registry describes an analysis population excluding a patient diagnosed with covid-19, but does not report the primary result, interval, or analysis denominator. The outcome description gives progression criteria rather than a clear response definition.

Overall survival and progression-free survival are listed as secondary outcomes. Safety reporting lists 6 of 11 participants affected by serious adverse events, 11 of 11 affected by other adverse events, and zero deaths. The registry record does not provide a randomized comparison or the statistical details needed to resolve the missing outcome information.