Phase 2 Durvalumab (Medi4736) for Bacillus Calmette-Guérin (BCG) Refactory Urothelial Carcinoma in Situ of the Bladder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase 2 Durvalumab (Medi4736) for Bacillus Calmette-Guérin (BCG) Refactory Urothelial Carcinoma in Situ of the Bladder (NCT02901548), a na phase 2 study. The primary outcome is Complete Response Rate at 6 Months.

2. Clinical Question

This question is narrower than a broad assessment of treatment benefit. It concerns a specified disease finding at a specified follow-up point, in participants who were evaluable for the endpoint. It does not, by itself, address durability beyond that time, outcomes in participants who were not evaluable, or the comparative effect of durvalumab versus another approach. The registry also lists complete response rate at 24 months as a secondary outcome, but does not report its result. Because there is no comparison group, the record cannot establish how the observed course would differ from what might have happened without the intervention.

3. Trial Design

This was a phase 2, single-group study sponsored by H. Lee Moffitt Cancer Center and Research Institute. The registry lists 17 participants enrolled and describes the study population as people with BCG-refractory urothelial carcinoma in situ of the bladder. The study status is terminated. The listed start date is 2017-02-16, and the primary completion date is 2020-07-24.

The intervention was durvalumab, administered by intravenous infusion. The arm description specifies a fixed dose every 4 weeks for 13 study treatment cycles over 12 months. Cystoscopy with biopsy was planned for monitoring response every 3 months during the first year, with monitoring every 6 months during year 2 described in the record. The record does not give a complete description of the later monitoring schedule.

Allocation is listed as NA, masking as none, and the model as single group. There is therefore no randomized allocation or masked treatment comparison described. All participants are assigned to the Durvalumab Plus Cystoscopy group in the registry's adverse-event information. Cystoscopy with biopsy is a procedure used for monitoring; it is not a separate comparison arm.

The primary outcome is complete response rate at 6 months, based on the week 26 mapping biopsy, among evaluable participants. A secondary outcome is complete response rate at 24 months. The registry lists two outcome measures, but the response results and analysis details are not reported in the registry record. The reason the study was terminated, the planned enrollment target, the stopping rules, and any sample-size justification are not reported in the registry record.

General: A single-group phase 2 study can describe outcomes observed among people who received an intervention and can provide information for subsequent research. Without a concurrent control group, however, the design does not separate an intervention's effect from other explanations for outcomes, such as the course of disease or differences in who remains assessable over follow-up. Those limits matter when interpreting any response percentage, even if it is reported.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Evaluable participants

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 Complete Response Rate at 6 Months, measured over 6 Months.

The primary endpoint is the complete response rate at 6 months among evaluable participants. The registry defines the assessment as the week 26 mapping biopsy in people with BCG-refractory CIS urothelial bladder cancer. After pathological review, absence of bladder CIS on the mapping biopsies is considered a complete response. The endpoint is expressed as a percentage of participants, with a 95% confidence interval specified as the measure of uncertainty.

A biopsy-based endpoint gives the response definition a concrete disease finding rather than relying only on a general impression of improvement. A defined assessment time also makes clear when the response is being evaluated. The endpoint is clinically and statistically distinct from whether response persists at a later point: the registry lists complete response rate at 24 months separately as a secondary outcome.

General: A binary endpoint classifies each assessed participant into one of two outcome categories, such as response or no response under the stated definition. A response rate summarizes the proportion classified as responding in the analysis population. Its interpretation depends on the endpoint definition, the assessment timing, and who is included in that population. A confidence interval conveys uncertainty around the estimated rate; it does not remove design limitations or account automatically for participants whose endpoint status is unavailable.

The registry record does not report the primary response count, percentage, or confidence interval. It also does not report how many participants were evaluable or the detailed rules used to determine evaluability. Consequently, the endpoint can be described from the protocol definition, but its observed result cannot be assessed from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies a binary percentage endpoint: complete response at 6 months. Each evaluable participant is assessed against the stated biopsy definition, and the endpoint is the percentage meeting that definition. The specified assessment is the week 26 mapping biopsy, with absence of CIS on mapping biopsies after pathological review classified as complete response. The registry specifies a 95% confidence interval for the percentage, but neither the estimate nor interval is reported in the registry record.

What does a response percentage describe?

General: A response percentage is a summary of the participants classified as having the endpoint among those included in the relevant analysis population. It is not automatically a description of every person enrolled. In this study, the primary outcome population is called evaluable participants. The registry record does not report the number evaluable, the criteria for evaluability, or how participants without a usable biopsy assessment were handled. Those details are important because a percentage can depend on its denominator and on how unavailable outcomes are treated.

A binary endpoint simplifies a complex clinical course into a defined classification at a defined time. That can make the outcome easier to report and interpret, but it does not capture all aspects of disease status or the path between assessments. The endpoint is specifically biopsy-based and time-specific. The separate listing of a 24-month complete response outcome signals a later assessment question, but its result and analysis are not reported in the registry record.

What does a confidence interval add?

General: A confidence interval expresses the uncertainty associated with an estimated quantity under a statistical method and its assumptions. For a response percentage, it indicates a range of values compatible with the data and method at the stated confidence level. A narrower interval generally indicates greater precision, while a wider interval indicates more uncertainty. The interval is not a guarantee that the true value lies within its endpoints, nor does it address whether the endpoint is clinically sufficient or whether the study design supports a causal claim.

The record specifies a 95% confidence interval for the primary percentage but does not report its endpoints. There is therefore no interval width to assess, and no observed response estimate to compare with a scientific or clinical benchmark. The registry also does not report the statistical method used to construct the interval. That method is not reported in the registry record.

How does the single-group design affect interpretation?

General: In a single-group study, a response estimate describes the observed outcome within the studied group. A control arm in a comparative trial can help distinguish intervention-associated outcomes from outcomes that might occur under a different treatment or without the intervention. No such comparison is described in this record: the model is single group, allocation is NA, and masking is none. Therefore, even a complete outcome analysis would not provide a randomized between-group effect estimate.

The design features are not statistical adjustments that can be repaired by a confidence interval. Precision describes uncertainty in an estimate, while the lack of a comparator limits the kinds of conclusions the design can support. The registry record does not report an adjusted analysis or any other analysis that would provide a comparative estimate.

What do the analysis and planning entries mean?

The registry lists two outcome measures and zero analyses. The primary endpoint and the 24-month secondary endpoint are identified, but analysis results are not reported in the registry record. The reason for the absence of analyses is not reported. It is also not reported whether interim analyses were planned or conducted, whether any statistical analysis plan existed, or how multiplicity was addressed.

The planning illustrations associated with the results are described as based on an observed effect and as using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They are planning illustrations, not observed power. The record does not report the numerical illustration outputs or an effect estimate that can be independently interpreted from the registry results. The illustrations do not describe the actual single-group design, which has no two-arm allocation. No sample-size justification for the trial is reported in the registry record.

General: A sample-size planning calculation asks how many participants might be needed under specified assumptions to detect an effect of a chosen size with a chosen design and analysis. Such an illustration depends on assumptions; it does not establish that a completed study had that power, and it is not evidence that an effect was observed. An observed effect estimate and its uncertainty require outcome data and an analysis. Neither is reported for the primary endpoint in this record.

Over-dispersion is a general concern when outcome variability is greater than a statistical model assumes. If a model underestimates variability, estimated uncertainty can be too small and interval or test results can be misleading. That concept does not establish that over-dispersion occurred in this study. The registry does not report a model-based analysis for the binary outcome or diagnostic checks, so no claim about over-dispersion in this trial can be made.

In short, the registry specifies an endpoint and a confidence level, but not an outcome estimate, a confidence interval, an analysis method, or a detailed analysis population. General statistical concepts help explain what such methods would mean, but they cannot fill in results or methods that are not reported in the registry record.

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("Durvalumab Plus Cystoscopy")
n      <- c(13)
events <- c(3)
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  Durvalumab Plus Cystoscopy  13  3  23.1  5.0  53.8  8.2  50.3  8.4  50.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures: complete response rate at 6 months as the primary outcome and complete response rate at 24 months as a secondary outcome. It lists zero analyses. The primary and secondary outcome results, and any planned comparisons or tests, are not reported in the registry record. The record does not say whether multiplicity was addressed or whether any formal adjustment was planned. It also does not report whether interim analyses were planned or conducted.

General: When a study assesses multiple outcomes or makes multiple statistical comparisons, the chance of at least one apparently notable result can increase if each is considered separately. Multiplicity methods may be used to control or describe that issue, depending on the study's goals and analysis plan. The number of listed outcomes alone does not reveal whether a correction is needed or was used. The registry record does not give enough analysis information to assess multiplicity handling.

The primary analysis population is described as evaluable participants. The record does not define evaluability, report the number of evaluable participants, or explain how missing biopsy assessments or other unavailable outcomes were handled. Missing-data methods, including any imputation or sensitivity analyses, are not reported in the registry record. It is also not reported whether all enrolled participants were included in any analysis.

These omissions affect interpretation because the response rate depends on which participants contribute an outcome and how unavailable outcomes are treated. No particular missing-data procedure or analysis population should be inferred. The registry record does not report an analysis plan, a prespecified hierarchy among outcomes, or a rationale for any analysis choices.

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: dysuria (5.9%), urinary tract infection (5.9%), pain in extremity (5.9%), rash muculo-papular (5.9%), anemia (5.9%).

The registry lists 17 participants at risk for serious adverse events and 5 affected. For other adverse events, it lists 17 at risk and 17 affected. The safety information also lists 3 deaths. These are counts in the registry record; the timing, event-level details, relationship to durvalumab, and relationship of deaths to treatment are not reported in the registry record.

The serious adverse event terms listed with reported frequencies are dysuria, urinary tract infection, pain in extremity, rash muculo-papular, anemia, and thromboembolic event. Each is shown with a frequency of 5.9%. The registry does not provide further event descriptions in the registry record for these terms, so the listed label and frequency should not be expanded into assumptions about severity, duration, recurrence, or cause.

General: The number at risk is the group against which an adverse-event count is reported, while the number affected indicates participants recorded as having an event under that category. Counts do not by themselves establish that the intervention caused an event. A single-group safety summary also does not provide a controlled comparison of event rates. The listed categories should be read as registry-reported safety information, not as a causal or comparative estimate.

The 3 deaths are reported without details that would establish their causes or timing. The registry record does not report whether they were considered treatment-related. No causal conclusion should be drawn from the count alone.

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

The record defines a study of durvalumab for BCG-refractory urothelial carcinoma in situ of the bladder and specifies a biopsy-based complete response endpoint at 6 months. It reports 17 participants enrolled and marks the study terminated. It also reports selected adverse-event counts and lists a 24-month complete response outcome as secondary. These registry entries describe the study and the outcomes it was designed to assess.

The record does not report the primary response percentage, its 95% confidence interval, or the 24-month result. It therefore does not establish how many evaluable participants met the response definition, how precise an estimate might be, or whether response was sustained to the later assessment. The lack of those results is distinct from an observed result showing no response; no such conclusion can be drawn from the missing analysis.

The single-group design also limits the meaning of any within-group outcome. There is no randomized comparator in the registry record, so the study cannot, on the information recorded, estimate a comparative effect against another treatment or a control condition. A response observed in a group is a descriptive finding; it is not by itself proof that the intervention caused that response. The design also does not answer what outcomes would have occurred under an alternative approach.

Safety counts should be interpreted with similar care. The record reports 5 participants affected by serious adverse events among 17 at risk, 17 affected by other adverse events among 17 at risk, and 3 deaths. It does not report event timing, causal attribution, or a control-group comparison. These entries describe recorded events and do not establish that the intervention caused them or that the intervention did not cause them.

General: Statistical findings are most informative when the endpoint, analysis population, uncertainty, and design are considered together. A point estimate without its precision can be difficult to judge; a precise estimate from a single group still does not supply a controlled comparison. Here, the outcome analysis is not reported, so interpretation must remain at the level of the study design, endpoint definition, enrollment, and available safety counts.

14. Why This Trial Matters Statistically

This study illustrates why a clearly defined endpoint and a reported analysis are both needed to interpret a clinical trial. The primary outcome is tied to a specific assessment: the week 26 mapping biopsy, with complete response defined by absence of bladder CIS after pathological review. That definition clarifies what the trial intended to count as a response. Yet the response rate and its 95% confidence interval are not reported in the registry record, so the outcome itself cannot be evaluated statistically.

The design also illustrates the difference between describing an outcome and estimating a comparative effect. This was a single-group study with no randomized comparator. The record can describe the intervention, endpoint, enrollment, and selected safety counts, but it does not provide a controlled comparison. That distinction matters when readers encounter a response endpoint: an outcome within a treated group is not the same as evidence of a difference between treatment strategies.

The label evaluable participants raises another important statistical issue. The denominator of a response percentage depends on who is included, and the registry does not report the number evaluable or the rules used to determine evaluability. Without those details and a missing-data account, a reader cannot assess whether the endpoint analysis represents all enrolled participants or only a subset with available assessments.

Finally, the record lists a primary and a secondary response outcome, but no analyses. This makes it impossible to assess the reported precision, analytic choices, or handling of multiple outcomes. The study is therefore useful as an example of how registry design information can be informative while still leaving key statistical evidence unavailable. Transparent reporting of outcome estimates, uncertainty, populations, and analysis methods is necessary for readers to distinguish what a study set out to measure from what its data establish.

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 most direct limitation for interpreting this study is that the registry record does not report the primary complete response rate or its 95% confidence interval. It also does not report the secondary complete response result at 24 months. Although the record lists two outcome measures, it lists zero analyses. Without outcome results, neither the magnitude nor the precision of response can be assessed.

The study is single group, with allocation listed as NA and masking as none. No randomized comparator is described. This limits causal and comparative interpretation, irrespective of whether an outcome estimate might otherwise be precise. The record also does not report why the study was terminated, so termination status cannot be used to infer an outcome or reason for stopping.

The primary endpoint applies to evaluable participants, but the criteria for evaluability and the number of participants in that population are not reported in the registry record. The handling of missing biopsies and other unavailable outcomes is also not reported. This prevents assessment of how the analysis denominator relates to the 17 participants enrolled.

Further details that would help evaluate the results are not reported in the registry record: the statistical analysis method, any analysis plan, sample-size justification, multiplicity strategy, and whether interim analyses were planned or conducted. The record provides adverse-event counts and selected serious adverse event terms, but does not report timing, causal attribution, or detailed outcome context for the deaths. These limits are specific to the information recorded and should not be replaced with assumptions about the study procedures or results.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Phase 2 Durvalumab (Medi4736) for Bacillus Calmette-Guérin (BCG) Refactory Urothelial Carcinoma in Situ of the Bladder — NCT02901548.
• Registry record last updated 2022-07-29; status last verified 2022-07.
• Results first posted 2021-06-11.
• 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

This is a phase 2, single-group study sponsored by H. Lee Moffitt Cancer Center and Research Institute, investigating durvalumab for BCG-refractory urothelial carcinoma in situ of the bladder. The registry lists 17 participants enrolled, no masking, and allocation as NA. The study is marked terminated; the listed start date is 2017-02-16 and the primary completion date is 2020-07-24.

The primary outcome is complete response rate at 6 months among evaluable participants, based on the week 26 mapping biopsy. Absence of CIS on mapping biopsies after pathological review is defined as complete response. A 24-month complete response rate is listed as a secondary outcome. Neither outcome result is reported in the registry record, which lists zero analyses.

Safety entries list 5 of 17 participants affected by serious adverse events, 17 of 17 affected by other adverse events, and 3 deaths. The record does not report the primary response estimate, its confidence interval, detailed analysis methods, or a reason for termination.