Neoadjuvant Pembrolizumab in Combination With Gemcitabine Therapy in Cis-eligible/Ineligible UC Subjects: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Neoadjuvant Pembrolizumab in Combination With Gemcitabine Therapy in Cis-eligible/Ineligible UC Subjects (NCT02365766), a non-randomized phase 1, phase 2 study. The primary outcome is Rate of Pathologic Muscle Invasive Response (PaIR).

2. Clinical Question

3. Trial Design

NCT02365766 is a completed Phase Ib/II, non-randomized, parallel-group study of neoadjuvant pembrolizumab with gemcitabine-based therapy in urothelial cancer. The record lists an enrollment of 83. The study was conducted in two parts: a one-arm Phase Ib dose-finding portion followed by a two-cohort Phase II portion. Allocation was non-randomized and masking was none.

Treatment was stratified by cisplatin eligibility. The Phase Ib cohort enrolled cisplatin-eligible participants and evaluated pembrolizumab dose levels in combination with gemcitabine and cisplatin. The Phase II cisplatin-eligible cohort received gemcitabine and cisplatin with pembrolizumab at the recommended Phase II dose. The Phase II cisplatin-ineligible cohort received gemcitabine with pembrolizumab at the recommended Phase II dose. The regimen descriptions also specify schedules and cycles, but do not provide a complete account of every treatment detail in the registry record.

The primary endpoint, PaIR, was assessed using the consolidative surgery specimen, including radical cystectomy or nephroureterectomy with lymph node dissection. The stated assessment window was within 2-7 weeks after the last pembrolizumab dose, up to 6 months. Secondary outcomes named in the record were relapse-free survival at 18 months, overall survival at 5 years, and radical cystectomy rate. The record does not report results for these endpoints.

Because treatment assignment was not randomized and the cohorts were defined by cisplatin eligibility, comparisons between cohorts cannot isolate a treatment effect. Eligibility, chemotherapy regimen, and cohort role differ. The registry record does not report a separate comparator group or a randomized treatment contrast.

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.

Before assignment: All participants in Phase 1b received same dose (Pembrolizumab 200 mg).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants in Phase 1b received same dose.

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 Rate of Pathologic Muscle Invasive Response (PaIR), measured over Within 2-7 weeks post last dose of pembrolizumab (up to 6 months).

The primary endpoint was the rate of pathologic muscle invasive response, abbreviated PaIR. It was defined as pathology at or below ypT1N0M0 in the specimen from consolidative surgery. The registry description identifies radical cystectomy or nephroureterectomy with lymph node dissection as the surgical specimens used for assessment. The endpoint was assessed within 2-7 weeks after the last pembrolizumab dose, up to 6 months.

In general, a pathology-based endpoint records what is found in tissue removed during surgery. It can describe the extent of disease at that assessment point, rather than relying only on symptoms or imaging. A response rate expresses the proportion of an analysis population meeting the prespecified pathology definition. The registry states that the endpoint is a percentage of participants and that its dispersion is to be represented by a 95% confidence interval.

The analysis population was defined in accordance with the Statistical Analysis Plan as cisplatin-eligible participants treated at the MTD and cisplatin-ineligible participants treated at the MTD. The record identifies the eligible-cohort MTD regimen as pembrolizumab 200 mg, gemcitabine 1000mg/m2, and cisplatin 70mg/m2. It does not specify the MTD regimen for the cisplatin-ineligible cohort in the endpoint description. Neither the PaIR percentage nor its confidence interval is reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The record characterizes the primary endpoint as a binary percentage: each participant either meets the pathology definition for PaIR or does not. The endpoint is based on the surgical specimen and uses the threshold at or below ypT1N0M0. The stated summary includes a 95% confidence interval. The registry record does not report the observed percentage or interval limits, so no numerical estimate of response or precision can be interpreted.

What does a confidence interval convey?

In general, a confidence interval describes statistical uncertainty around an estimate under the assumptions of the analysis. For a percentage, it helps show how much uncertainty remains about the underlying response rate in the population represented by the analysis. It is not a guarantee that the true value falls within the interval, and it does not address all sources of uncertainty. In particular, it cannot correct biases arising from selection, missing outcomes, or non-random treatment assignment. Since the PaIR interval is not reported in the registry record, its width and implications cannot be assessed.

What does the analysis population mean?

The endpoint description says the analysis population was defined in accordance with the Statistical Analysis Plan. It includes cisplatin-eligible participants treated at the MTD and cisplatin-ineligible participants treated at the MTD. This is a treatment-defined population. It should not be assumed to equal all enrolled participants, or that everyone in the population had an evaluable surgical specimen. The endpoint description gives the MTD regimen for the cisplatin-eligible group but not for the cisplatin-ineligible group. The precise population counts and participant flow are not reported in the registry record.

Why is the non-randomized design important?

In general, random allocation helps balance measured and unmeasured participant characteristics across treatment groups. This study did not use random allocation. Instead, the Phase II groups were based on cisplatin eligibility, and the regimens differed accordingly. The Phase Ib cohort had a dose-finding purpose. Thus, any cross-cohort difference could reflect differences in eligibility, clinical characteristics, treatment, or other factors, as well as chance. A direct comparison would not have the same interpretation as a randomized treatment contrast. The record itself notes that the arms are not a like-for-like control comparison.

The analysis description does not identify a statistical test, a regression model, covariate adjustment, or a procedure for comparing the cohorts. Those methods are not reported in the registry record. General methods for comparing binary outcomes include estimating group-specific proportions and their uncertainty, but the exact analysis method used for PaIR cannot be inferred from the endpoint definition alone.

How should planning illustrations be read?

The stated sample-size illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The illustration describes detectable effects as a stated fraction of the observed effect. These are planning assumptions, not the trial's own design justification and not evidence about the probability that a result is true. The record does not report the trial's sample-size justification, and no numerical planning estimates are reported in the registry record. Such estimates are not observed power; observed power is not a substitute for the result estimate and its uncertainty.

In general, an illustration based on an observed effect can be sensitive to how large or small that initial estimate is. Planning for a fraction of an effect asks what sample size would correspond to a smaller assumed effect under the stated assumptions. It does not establish that the smaller effect is clinically important, nor that a future study would reproduce the assumed effect. No additional effect size or sample-size quantity can be interpreted from the registry record.

What can be said about outcome timing?

The PaIR assessment window is tied to the last pembrolizumab dose and the consolidative surgery specimen. The registry also names relapse-free survival at 18 months and overall survival at 5 years as secondary outcomes. In general, time-to-event outcomes require a definition of the event, follow-up duration, and methods for handling participants whose event status is not known at the analysis time. Those analytic details and the numerical secondary-outcome findings are not reported in the registry record. The names and time points alone do not show how many events occurred or how precise any estimate was.

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 I: Cisplatin Eligible - Pembrolizumab 200 m", "Cohort II: Cisplatin Ineligible - Pembrolizumab 20")
n      <- c(38, 33)
events <- c(23, 17)
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 I: Cisplatin Eligible - Pembrolizumab 200 m  38  23  60.5  43.4  76.0  44.7  74.4  44.6  74.4
ARM  Cohort II: Cisplatin Ineligible - Pembrolizumab 20  33  17  51.5  33.5  69.2  35.2  67.5  35.1  67.6

11. Multiplicity, Missing Data and Analysis Populations

The registry record names one primary endpoint, PaIR, and three secondary outcomes: relapse-free survival at 18 months, overall survival at 5 years, and radical cystectomy rate. It reports five outcome measures but no analyses. It does not describe a method for controlling the chance of false-positive findings across multiple endpoints, time points, cohorts, or analyses. A multiplicity strategy is not reported in the registry record.

In general, testing many outcomes or making many comparisons can increase the chance that at least one apparently notable result arises by chance. Methods such as prespecifying a hierarchy or adjusting inference can address that issue, but no such method should be assumed for this study. The registry record does not report whether any adjustment was planned or used.

The record identifies an analysis population for PaIR in accordance with the Statistical Analysis Plan: eligible participants treated at the MTD in each cisplatin-eligibility category. It does not provide the numbers included in that population, participant disposition, or how participants without an assessable surgical specimen were handled. The handling of missing outcome data, withdrawals, and other unavailable observations is not reported in the registry record.

These gaps limit assessment of whether the reported endpoint population would represent all enrolled participants and whether missingness could affect a response percentage. The Statistical Analysis Plan is referenced for the population definition, but the analysis procedures themselves—including missing-data methods, sensitivity analyses, and multiplicity handling—are not reported in the registry record. No assumptions about these procedures can be made from the endpoint label alone.

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: acidosis (16.7%), acute kidney injury (16.7%), fever (16.7%), hyponatremia (16.7%), sepsis (16.7%).

The registry lists adverse-event figures separately for the Phase Ib cisplatin-eligible cohort, the Phase II cisplatin-eligible cohort, and the Phase II cisplatin-ineligible cohort. For each, it reports counts at risk and affected for serious adverse events and for the category labelled “other,” along with a death count. These are registry counts, not estimates of comparative risk. The record does not establish that reported events or deaths were caused by study treatment.

The serious-adverse-event counts are 1 affected among 6 at risk in the Phase Ib cohort, 17 among 37 in Phase II Cohort I, and 17 among 38 in Phase II Cohort II. The death counts are 1, 11, and 16, respectively. These groups were not randomized comparators and differ by cohort and cisplatin eligibility, so their counts should not be read as a treatment comparison. The registry also lists six serious adverse-event terms, each at 16.7%. The denominator and the basis for those term percentages are not reported in the registry record. Counts and percentages alone do not describe timing, severity distribution, duration, or attribution.

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

The record describes a clinical study of neoadjuvant pembrolizumab combined with gemcitabine-based treatment, with regimens organized by cisplatin eligibility. It specifies a pathology-based primary endpoint and the population intended for its analysis. These details explain what the study set out to measure; they do not establish an outcome. The PaIR rate and its 95% confidence interval are not reported in the registry record.

The record also does not report numerical findings for the named secondary outcomes. Therefore, the listed outcomes cannot be used to infer relapse-free survival, overall survival, or radical cystectomy results. A completed status indicates the study is recorded as completed, but it does not supply those missing outcome estimates.

In general, a response rate is descriptive of the analyzed participants and depends on the endpoint definition, analysis population, and handling of participants whose outcome is unavailable. The registry defines PaIR using a surgical pathology threshold and specifies a treated-at-MTD population. It does not report the participant flow or missing-data approach needed to understand how these elements were applied in practice.

The non-randomized design also limits causal interpretation. Cisplatin eligibility determined cohort assignment and was linked to different chemotherapy regimens; the Phase Ib cohort had a separate dose-finding purpose. Even if results were available, differences across these cohorts would not on their own show that one regimen caused a different response. The record does not describe a like-for-like randomized control comparison.

Nothing in the available registry information establishes benefit or lack of benefit. It supports a description of the study design, outcome definitions, and reported safety counts, while leaving the efficacy estimates, their precision, and relevant analysis details unreported. No treatment recommendation follows from these registry entries.

14. Why This Trial Matters Statistically

This study raises statistical questions that are important when interpreting early-phase and non-randomized clinical research. It combines a Phase Ib dose-finding portion with Phase II cohorts and distinguishes participants by cisplatin eligibility. That structure can address different questions within one study, but it does not create randomized, directly comparable treatment groups. Keeping cohort purpose and eligibility context in view is essential when reading any outcome contrast.

The primary endpoint is based on surgical pathology rather than a symptom measure: PaIR is defined as pathology at or below ypT1N0M0 in a consolidative surgery specimen. Such an endpoint depends on reaching the specified assessment and having an evaluable specimen. The record identifies a treated-at-MTD analysis population, making the population definition especially relevant to interpretation. The participant flow and handling of unavailable specimens are not reported in the registry record.

Precision also matters. A percentage without its confidence interval can give an incomplete picture of how uncertain the response estimate is. The registry specifies a 95% confidence interval but does not report either the response estimate or interval. Consequently, readers cannot assess the magnitude or precision of the primary finding from the registry record.

The study also names longer-term secondary outcomes, but provides no outcome estimates or analysis details for them. Separating endpoint definitions from endpoint results prevents planned measures from being mistaken for evidence. Finally, the safety counts are grouped by cohort and do not establish comparative risk or treatment attribution. These features make transparent reporting of populations, missing outcomes, uncertainty, and analysis methods central to any fuller statistical interpretation.

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 key limitation for interpreting efficacy is that the registry record does not report the primary PaIR result or its 95% confidence interval. It also does not report numerical results for the secondary outcomes. Without estimates and uncertainty, the magnitude and precision of outcomes cannot be assessed.

The study is non-randomized, with cohorts defined in part by cisplatin eligibility and treatment regimen. The Phase Ib portion is a dose-finding cohort, not a control group. Thus, the cohorts are not like-for-like comparisons, and differences between them cannot be attributed to regimen alone. The record does not describe a randomized comparator.

The endpoint population is defined by treatment at the MTD, but the record does not report its size, participant flow, or how missing pathology assessments were handled. The eligible-cohort MTD regimen is specified, while the endpoint description does not state the MTD regimen for the cisplatin-ineligible group. These details limit assessment of exactly who contributed to the endpoint analysis.

The registry does not report the sample-size justification, statistical analysis methods, multiplicity approach, missing-data strategy, or sensitivity analyses. It lists secondary endpoints but does not give their analytical definitions or results. Safety counts are reported by cohort, but event attribution and the denominator for the listed serious-event term percentages are not reported in the registry record. These limitations constrain both outcome interpretation and comparisons across cohorts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Neoadjuvant Pembrolizumab in Combination With Gemcitabine Therapy in Cis-eligible/Ineligible UC Subjects — NCT02365766.
• Registry record last updated 2025-03-18; status last verified 2025-02.
• Results first posted 2025-03-18.
• 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

NCT02365766 is a completed, non-randomized Phase Ib/II study of neoadjuvant pembrolizumab with gemcitabine-based therapy in urothelial cancer. The registry lists enrollment of 83 and a parallel design without masking. Treatment cohorts were separated by cisplatin eligibility, with a Phase Ib dose-finding portion followed by Phase II cohorts.

The primary endpoint was the rate of pathologic muscle invasive response, defined as pathology at or below ypT1N0M0 in a consolidative surgery specimen. The registry specifies a 95% confidence interval but does not report the result or interval. Secondary outcomes named include relapse-free survival at 18 months, overall survival at 5 years, and radical cystectomy rate; their results are also not reported in the registry record. The record includes adverse-event counts by cohort, but the non-randomized design does not support a direct like-for-like comparison between those cohorts.