Pembrolizumab in Treating Patients With Rare Tumors That Cannot Be Removed by Surgery or Are Metastatic: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pembrolizumab in Treating Patients With Rare Tumors That Cannot Be Removed by Surgery or Are Metastatic (NCT02721732), a na phase 2 study. The primary outcome is Non-progression Rate (NPR) at 27 Weeks by irRECIST.

2. Clinical Question

3. Trial Design

This is a phase II study sponsored by M.D. Anderson Cancer Center. The registry describes a single-group model with no allocation to comparative arms and no masking. The listed enrollment is 157. The study began on 2016-08-15, has a primary completion date of 2028-12-31, and is listed as active, not recruiting. These are registry entries, not evidence that the primary endpoint has been fully reported.

All patients in the treatment group received pembrolizumab. The regimen is described as an intravenous infusion over 30 minutes on day 1, repeating every 21 days. Treatment could continue for up to 24 months if progression or toxicity did not occur; patients with clinical response or disease stabilization could continue for up to an additional 12 months. These are the regimen details in the record; further information about dose, treatment adherence, or treatment received after discontinuation is not reported in the registry record.

The study addresses several rare and advanced tumor types rather than one narrowly defined cancer. The listed conditions include unresectable and metastatic diseases, and the record includes both solid neoplasms and named tumor types. This broad scope is relevant when interpreting a single overall endpoint: the registry record does not provide the primary outcome separately by tumor type.

The primary endpoint is non-progression at 27 weeks by irRECIST. Secondary outcomes include response and clinical benefit by RECIST v1.1 and irRECIST, progression-free survival by both criteria, tumor size in relation to PD-L1 status, and treatment-related adverse events. The record lists nine outcome measures and no analysis records. It does not report a randomized comparison, a statistical analysis plan, or results for the primary or secondary outcomes.

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 10 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All patients who have received at least 1 dose of study medication with at least 1on-study tumor assessment were included in the analysis. If a patient discontinued before the first radiological assessment due to progression, they were included in the analysis. If the patient discontinued before 27 weeks due to reasons other than progression or death, they were considered non evaluable for this endpoint. One patient enrolled in the ACC cohort was excluded due to change in diagnosis.

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.

Only the first 6 of 11 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Non-progression Rate (NPR) at 27 Weeks by irRECIST, measured over At 27 weeks.

The primary endpoint is the non-progression rate at 27 weeks by irRECIST. It is defined as the percentage of efficacy-evaluable patients who were alive and progression-free at 27 weeks. It combines survival status and disease progression status at a specified assessment time, rather than measuring tumor response alone.

The registry defines progression using immune-related Response Evaluation Criteria in Solid Tumors (irRECIST). Progression includes an increase of at least 20% (minimum 5 mm) in total measured tumor burden compared with nadir, progression of non-target lesions, or a new lesion. The endpoint therefore depends on tumor assessments interpreted under the specified criteria, as well as whether a patient is alive at the assessment time.

As a general statistical concept, a fixed-time non-progression rate can provide a direct description of the proportion meeting a clinically defined status at a chosen time point. It may be easier to interpret than a time-to-event measure when the question is specifically about status at that time. However, it does not by itself show when progression occurred, describe the course of disease between assessments, or distinguish the effect of treatment from other explanations in a single-group study.

The registry record does not report the observed rate or its precision. The complete operational definition of the efficacy-evaluable population and the handling of every circumstance before the 27-week assessment are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a phase II, single-group study with no masking. These design features determine the scope of statistical interpretation. A single-group study records outcomes in participants receiving the study intervention; it does not create a randomized control group. Open-label means the study is not masked as described in the registry. Neither design label supplies a treatment-effect estimate, and the record does not report a comparative statistical analysis.

What does a non-progression rate represent?

In general, a rate at a fixed time is a proportion: participants meeting the stated condition are considered in relation to the defined analysis population. For this endpoint, the condition is being alive and progression-free at 27 weeks under irRECIST. The registry describes the endpoint as a percentage, but does not post the observed numerator, denominator, or percentage. It is therefore not possible to assess the estimate or its precision from the record.

A confidence interval, in general, describes the uncertainty around an estimate under a specified statistical procedure. It is not a guarantee about the range in which an individual outcome will fall. No confidence interval for the primary endpoint is reported in the registry record. The statistical method for calculating an interval, if one was used, is also not reported.

How does irRECIST define progression?

The record describes progression as an increase of at least 20% (minimum 5 mm) in total measured tumor burden compared with nadir, progression of non-target lesions, or a new lesion. This definition makes the endpoint dependent on tumor measurement and classification rules. As a general matter, response criteria provide a common framework for interpreting assessments, but estimates from different criteria need not be directly interchangeable. RECIST v1.1 outcomes are also listed as secondary outcomes; the record does not report their results.

What is the importance of the analysis population?

The record says that patients who received at least one dose and had at least one on-study tumor assessment were included in the primary analysis. It also says that patients who discontinued before the first radiological assessment because of progression were included. These inclusion rules affect the denominator and interpretation of the rate. The registry does not provide the number included in the primary analysis or the complete rules for handling other situations, such as missing assessments or discontinuation before the endpoint time.

In general, missing outcome information can affect a proportion if participants without an assessment are excluded, counted as not meeting the endpoint, or handled by another method. Those choices can lead to different estimates. The registry record does not report a complete missing-data strategy for this endpoint, so no particular handling method should be assumed.

What do the sample-size assumptions describe?

The planning illustrations are described as calculations based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. These are planning assumptions, not a description of the actual single-group design. The registry does not provide the numerical illustration results in the record, nor does it report that these assumptions formed the trial’s design or sample-size justification. Such planning calculations are not observed power and do not establish the precision of the primary endpoint.

More generally, sample-size planning asks how much information a proposed design might yield under stated assumptions. It does not replace reporting the actual endpoint estimate and its uncertainty. For a single-group study, an illustration assuming two arms does not create a control group or convert the trial into a randomized comparison. The registry record does not report a protocol-based sample-size rationale or the assumptions used to choose enrollment.

What can be concluded from the analysis record?

The record lists nine outcome measures and zero analysis records. This means that the registry entry described does not post analysis records with estimates for those outcomes. It does not establish that no analyses were performed outside the registry. The planned or completed statistical tests, models, confidence-interval methods, covariate adjustments, interim analyses, and prespecified analysis plan are not reported in the registry record. Any interpretation should distinguish the endpoint definition and study design from statistical results that are not posted.

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("Cohort 1: Squamous Cell Carcinoma of the Skin", "Cohort 2: Small Cell Malignancies of Nonpulmonary ", "Cohort 3: Adrenocortical Carcinoma", "Cohort 4: Medullary Renal Cell Carcinoma", "Cohort 5: Carcinoma of Unknown Primary", "Cohort 6: Penile Carcinoma")
n      <- c(17, 12, 20, 5, 26, 2)
events <- c(7, 0, 6, 0, 8, 0)
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  Cohort 1: Squamous Cell Carcinoma of the Skin  17  7  41.2  18.4  67.1  21.6  64.0  21.5  64.3
ARM  Cohort 2: Small Cell Malignancies of Nonpulmonary   12  0  0.0  0.0  26.5  0.0  24.2  0.2  24.7
ARM  Cohort 3: Adrenocortical Carcinoma  20  6  30.0  11.9  54.3  14.5  51.9  14.6  52.2
ARM  Cohort 4: Medullary Renal Cell Carcinoma  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Cohort 5: Carcinoma of Unknown Primary  26  8  30.8  14.3  51.8  16.5  50.0  16.5  50.2
ARM  Cohort 6: Penile Carcinoma  2  0  0.0  0.0  84.2  0.0  65.8  0.8  70.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists nine outcome measures, including the primary non-progression rate and secondary outcomes involving response, clinical benefit, progression-free survival, adverse events, tumor size, and PD-L1 status. When a study evaluates multiple outcomes, general statistical practice considers whether the chance of false-positive findings across tests needs to be addressed. The registry record does not report a multiplicity strategy, such as a testing hierarchy or adjustment, so none should be assumed.

The primary analysis population is described as patients who received at least one dose and had at least one on-study tumor assessment. Patients who discontinued before their first radiological assessment because of progression were also included. The complete population definition is not reported in the registry record, and the record does not give a count for the primary analysis population.

Missing outcome information can influence a fixed-time rate. In general, estimates may differ depending on whether missing assessments are excluded, treated as not meeting the endpoint, or handled using another prespecified method. The registry does not report how other missing assessments, loss to follow-up, or discontinuation before 27 weeks were handled. The record also does not report sensitivity analyses for missing data.

The registry lists zero analysis records. It does not report an analysis plan, interim analysis procedures, or whether the endpoint population and missing-data rules were finalized before results were examined. These omissions limit assessment of how the posted endpoint would have been analyzed, if results become available.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: abdominal pain (5.1%), dyspnea (3.8%), fever (3.2%), lung infection (3.2%), pleural effusion (3.2%).

The adverse-event summary lists 156 participants at risk for serious adverse events and 72 affected. For other adverse events, it lists 156 at risk and 84 affected. The record also lists 114 deaths. These entries are safety summary figures and should not be interpreted as results for the primary non-progression endpoint. The registry record does not specify in this summary the relationship, if any, between the listed deaths and adverse events.

The serious adverse event terms listed most frequently are abdominal pain (5.1%), dyspnea (3.8%), fever (3.2%), lung infection (3.2%), pleural effusion (3.2%), and pneumonitis (2.6%). The record presents these as percentages but does not provide further detail about the calculation or timing of these term-specific percentages in the listed summary. They do not indicate, by themselves, whether an event was caused by pembrolizumab.

In general, an adverse-event count describes events or participants as defined by a study’s reporting rules; it is not automatically evidence of causation. Seriousness and treatment relationship are distinct classifications. The registry lists treatment-related adverse events as a secondary outcome, but does not report a result for that outcome. The safety figures should therefore be read as registry-reported event summaries, not as a comparative estimate of treatment risk.

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

The registry establishes that this phase II study evaluates pembrolizumab in patients with rare tumors that cannot be removed by surgery or have spread, and that its primary endpoint is non-progression at 27 weeks by irRECIST. It also describes the treatment schedule, the intended endpoint definition, a broad set of eligible conditions, and summary safety figures. Those details explain what the study is designed to assess.

The record does not report the primary endpoint result. Without the observed count or rate and its uncertainty, the size and precision of the endpoint estimate cannot be evaluated. No conclusion about the frequency of non-progression in the analyzed population can be drawn from the endpoint definition alone. The secondary outcome names likewise do not constitute outcome results.

Even when results are reported, a single-group study without a randomized comparator cannot by itself determine what would have happened to comparable participants under another treatment or no study treatment. An observed outcome in this group would be descriptive; attributing it to the intervention requires caution. The registry record does not specify a historical benchmark or other comparator for interpreting the primary endpoint.

The outcome is defined at a particular time and uses immune-related response criteria. It does not, on its own, describe the full timing of progression, the experience of each participant, or outcomes within each rare tumor type. The listed safety counts also require separate interpretation and do not establish treatment causation or a comparative risk.

In short, the record communicates the question and design, but does not post the primary result or enough analysis detail to assess its precision, missing-data handling, or multiplicity approach. A statistical conclusion about the primary endpoint would require reported results and their analysis context.

14. Why This Trial Matters Statistically

Rare tumor studies face a statistical challenge: the population of interest is defined around uncommon diseases, while individual tumor types may differ. This study’s registry record spans a range of rare, advanced conditions and uses a single-group phase II design. Such a design can document outcomes in the enrolled group, while the absence of a randomized comparator limits direct causal comparison.

The endpoint illustrates another important issue. Non-progression at 27 weeks is not simply tumor shrinkage: it combines being alive with being progression-free at a fixed time, using specified immune-related criteria. The interpretation depends on both the response rules and the analysis population. Clear reporting of the observed rate, its uncertainty, and the handling of participants without an assessment would be important for judging what the endpoint estimates.

The record also lists outcomes based on both irRECIST and RECIST v1.1, as well as response, clinical benefit, progression-free survival, and safety outcomes. These measures address different aspects of disease and treatment experience. If several outcomes are examined, readers need to know which analyses were prespecified and how multiplicity was considered; those details are not reported in the registry record.

Finally, the entry illustrates why a trial’s design and a trial’s results must be distinguished. The registry describes a treatment regimen and a planned primary question, but lists no analysis records and does not report the primary outcome. Statistical interpretation should therefore remain focused on what is documented, without treating a study objective as an observed finding.

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.

The primary limitation for interpreting efficacy is that the registry record does not report the primary endpoint result. It gives no observed non-progression count or percentage, confidence interval, or statistical test. The record lists enrollment of 157, but does not state how many participants contributed to the primary endpoint analysis.

The study has a single-group, open-label design and no randomized comparator. The resulting observations cannot provide a concurrent treatment contrast. The registry does not report a benchmark for interpreting the primary rate, and the record does not provide results separately by the listed tumor types.

The population description includes a range of rare malignancies. This breadth may make an overall result difficult to interpret as representative of any one tumor type; the registry record does not provide tumor-specific primary endpoint estimates. The full operational definition of the efficacy-evaluable population is also not reported, beyond the stated inclusion rules.

The record does not report how all missing assessments, discontinuations, or incomplete follow-up were handled for the 27-week endpoint. It also does not report the statistical analysis plan, the method for estimating uncertainty, interim analyses, or multiplicity procedures. These gaps make it difficult to judge the robustness of any future posted estimate from the registry entry alone.

The study is listed as active, not recruiting, with a primary completion date of 2028-12-31. The registry record does not report primary or secondary outcome analyses. The listed serious adverse event and death counts are not a substitute for those efficacy results, and the adverse-event summary does not establish causation.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pembrolizumab in Treating Patients With Rare Tumors That Cannot Be Removed by Surgery or Are Metastatic — NCT02721732.
• Registry record last updated 2026-07-17; status last verified 2026-07.
• Results first posted 2024-06-18.
• 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

NCT02721732 is a phase II, single-group, open-label study sponsored by M.D. Anderson Cancer Center. It evaluates pembrolizumab in patients with rare tumors that are unresectable or metastatic. The registry lists enrollment of 157, a start date of 2016-08-15, a primary completion date of 2028-12-31, and a status of active, not recruiting.

The primary endpoint is non-progression at 27 weeks by irRECIST, defined as the percentage of efficacy-evaluable patients alive and progression-free at that time. The record describes progression criteria and a partial analysis-population definition, but does not report the primary endpoint result, its precision, or a complete analysis plan. It lists nine outcome measures and zero analysis records.

The adverse-event summary lists 72 participants affected by serious adverse events among 156 at risk, 84 affected by other adverse events among 156 at risk, and 114 deaths. The record does not report the primary efficacy outcome or a randomized comparison.