Durvalumab+ Gemcitabine/Cisplatin (Neoadjuvant Treatment) and Durvalumab (Adjuvant Treatment) in Patients With MIBC: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Durvalumab+ Gemcitabine/Cisplatin (Neoadjuvant Treatment) and Durvalumab (Adjuvant Treatment) in Patients With MIBC (NCT03732677), a randomized phase 3 study. The primary outcome is Pathologic Complete Response (pCR) Rates at Time of Cystectomy.

2. Clinical Question

The primary comparison is between treatment strategies assigned at randomization, rather than between participants who did or did not complete a particular treatment. That distinction matters because the primary endpoint denominator is the Full Analysis Set, defined in the registry as all randomized subjects. The observed endpoint counts therefore describe the randomized groups under that analysis population.

The registry also lists outcomes related to event-free survival, cystectomy, overall survival, metastasis-free survival, disease-specific survival, and durvalumab immunogenicity. These give context to the primary pathological response question, but they address different clinical and statistical questions. A pathological response is assessed at surgery; survival outcomes follow participants over time, while the cystectomy outcome describes whether surgery occurred.

3. Trial Design

The registry describes a phase III, randomized, open-label, multicenter, global study with a parallel-group design. It reports enrollment of 1063 participants and an active-not-recruiting status. The recorded start date is 2018-11-16, and the recorded primary completion date is 2024-04-29. These are registry fields; they do not establish whether every follow-up activity or all analyses are complete.

The two groups are durvalumab plus gemcitabine plus cisplatin and gemcitabine plus cisplatin. The trial title describes durvalumab with chemotherapy as neoadjuvant treatment, followed by durvalumab alone as adjuvant treatment. The primary endpoint is assessed at cystectomy, with a timeframe of up to 6 months. The primary analysis population is the Full Analysis Set, defined as all randomized subjects.

Open-label means the registry does not describe masking of participants or study personnel. Random allocation supports comparison of assigned groups, but open-label conduct can matter for outcomes involving clinical decisions or assessments. The primary response assessment is described as central pathology review, which identifies a specified review process for that endpoint; the registry does not report further details about blinding of the reviewers.

The registry lists eight outcome measures and seven analyses, with six secondary outcomes posted. The primary outcome is pathologic complete response, while listed secondary outcomes include survival measures and the proportion undergoing cystectomy. The trial's original sample-size justification, randomization details beyond the stated allocation, and detailed statistical analysis plan are not reported in the registry record.

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: Full Analysis Set (Intention to Treat): includes all randomized subjects

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: The Full analysis set (FAS) included all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Pathologic Complete Response (pCR) Rates at Time of Cystectomy, measured over Up to 6 months.

The primary endpoint is the pathologic complete response rate at cystectomy. The registry defines response as pathological staging T0N0M0, assessed by central pathology review of specimens obtained through radical cystectomy after neoadjuvant treatment. The reported endpoint is a count of participants meeting that definition; the denominator is the Full Analysis Set, meaning all randomized subjects.

This endpoint captures whether no residual tumor is identified in the specified pathological assessment at surgery. It is a binary participant-level outcome: each participant either meets the stated response definition or does not. The registry reports 199 participants in the durvalumab plus gemcitabine plus cisplatin group and 146 in the gemcitabine plus cisplatin group as having the endpoint, among group sizes of 533 and 530, respectively.

Pathological response is used as an outcome after neoadjuvant treatment because it can be assessed from surgical specimens and applies a defined staging criterion. It is distinct from outcomes based on time to an event or survival. A response difference does not, by itself, establish the size or certainty of any longer-term survival difference. The registry separately lists event-free survival and overall survival, among other secondary outcomes, to address additional aspects of outcome.

The registry identifies central pathology review, but does not report further operational details about specimen handling, adjudication, or handling of participants without a cystectomy specimen. Those details would be relevant when evaluating endpoint ascertainment and missing outcome data.

7. Results Posted in the Registry

• Gemcitabine + Cisplatin: 27.5% (146/530) versus 37.3% (199/533) on control; risk difference -9.8 percentage points (95% CI -15.4 to -4.2), risk ratio 0.74 (0.62–0.88); unadjusted two-proportion z-test p = <0.001.

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.

Event-free Survival at 24 Months (EFS24) Per Central Review Defined as Time From Randomization to Event

• 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:
  • Gemcitabine + Cisplatin: 59.8% (317/530) versus 67.7% (361/533) on control; risk difference -7.9 percentage points (95% CI -13.7 to -2.2), risk ratio 0.88 (0.81–0.97); unadjusted two-proportion z-test p = 0.007.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Gemcitabine + Cisplatin than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Proportion of Patients Who Undergo Cystectomy

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Gemcitabine + Cisplatin: 83.2% (441/530) versus 88.0% (469/533) on control; risk difference -4.8 percentage points (95% CI -9.0 to -0.6), risk ratio 0.95 (0.90–0.99); unadjusted two-proportion z-test p = 0.026.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Gemcitabine + Cisplatin than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Overall Survival

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Durvalumab+Gemcitabine+Cisplatin vs Gemcitabine+Cisplatin: hazard ratio 0.75 (95% CI 0.56–0.98), reported p = 0.011 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Durvalumab+Gemcitabine+Cisplatin vs Gemcitabine+Cisplatin than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes address several related but distinct questions: event-free survival, whether cystectomy occurred, overall survival, metastasis-free survival, disease-specific survival, and durvalumab immunogenicity. The posted numerical results include analyses for some of these outcomes; six secondary outcomes are posted in the record. These endpoints do not all measure the same event or use the same follow-up structure, so their findings should be considered endpoint by endpoint rather than collapsed into one summary.

Across the posted comparative results, the direction of the reported findings is generally consistent with the primary comparison, including the time-to-event results and the cystectomy outcome. This overall pattern offers context for the pathologic complete response result, but it does not make the endpoints interchangeable. Pathological response at surgery is not itself a survival outcome, and a favorable direction in a secondary estimate does not establish the magnitude or certainty of a longer-term effect.

Secondary analyses carry less weight for confirmatory interpretation when many outcomes are examined and no multiplicity strategy is reported. With multiple tests, some apparently notable findings can occur by chance even when there is no corresponding effect. The registry record does not report whether or how multiplicity was controlled, nor does it provide a hierarchy for interpreting secondary endpoints. Their p-values and intervals should therefore be read as supporting, exploratory, or descriptive evidence according to the analysis context, not as a collection of independent definitive tests.

The record does not report results for every secondary outcome named in the outcome list. Missing posted results should not be interpreted as a null result. They indicate only that the corresponding result is not reported in the registry record.

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

• Event-free Survival at 24 Months (EFS24) Per Central Review Defined as Time From Randomization to Event
• Proportion of Patients Who Undergo Cystectomy
• Overall Survival
• Metastasis-free Survival Per Investigator Assessment or Local Biopsy Review.
• Disease-specific Survival Per Investigator Assessment or Local Biopsy Review.
• Immunogenicity of Durvalumab When Used in Combination With Gemcitabine/Cisplatin as Measured by Presence of Antidrug Antibodies (ADA)

9. Statistical Interpretation

10. Statistical Methods Explained

What does the risk difference describe?

A risk difference compares the proportion of participants with an outcome in one group with the proportion in another. Here, the posted recalculation expresses chemotherapy alone relative to the durvalumab-containing group. Its estimate is -0.098, with a 95% confidence interval from -0.154 to -0.042. The sign depends on the stated comparison direction; reversing the order would reverse the sign without changing the underlying group data. A risk difference is an absolute-scale summary, so it communicates how far apart the observed proportions are rather than comparing odds.

How does the odds ratio differ?

The registry's reported primary analysis uses logistic regression and reports an odds ratio of 1.60, with a 95% confidence interval from 1.227 to 2.084. Odds are not the same as probabilities. An odds ratio summarizes the relative odds of the endpoint between groups, while the risk difference summarizes an absolute difference in probabilities. The two summaries can both describe the randomized comparison, but they have different scales and should not be read as if their estimates were the same quantity.

The registry labels the reported analysis as a superiority analysis. Its reported p-value is 0.001. A p-value describes how compatible the observed result is with a specified null hypothesis under the assumptions of the test; it is not the probability that the null hypothesis is true, nor does it quantify the clinical importance of the finding. The registry record does not report the full model specification, covariates, or all diagnostic checks for the logistic regression.

What does a confidence interval convey?

A confidence interval gives a range of estimates compatible with the data under the analysis method and its assumptions. It communicates precision as well as direction. For a risk difference, zero represents no difference on that scale; for an odds ratio, one represents equal odds. The primary intervals reported in the registry and the unadjusted recalculation lie on one side of their respective null values. That does not mean every value in the interval is equally plausible or that the interval accounts for every source of uncertainty, such as missing data or analysis choices.

What is the distinction between the registry analysis and a recalculation?

The recalculated comparison is a simple, unadjusted analysis from the posted group summaries. It is useful as a transparent check on the direction and approximate size of the group contrast using the displayed counts. It does not reproduce a regression model or any adjustments that may have been used for the sponsor's analysis. The sponsor's reported analysis therefore remains the primary reported inferential analysis, while the recalculation should be understood as a separate unadjusted summary. The registry record does not report evidence of a discrepancy between these results.

Why do endpoint types matter?

Pathologic complete response and cystectomy are binary outcomes: each participant contributes an outcome classification. For these measures, group counts and proportions can be compared using risk differences, odds ratios, or other binary-outcome summaries. Event-free survival and overall survival are time-to-event outcomes. These involve both whether an event occurred and when it occurred, and their analysis can account for follow-up and censoring in ways that a simple binary comparison does not. A hazard ratio is a time-to-event summary, and it should not be interpreted as a risk difference or as a direct comparison of median times.

The registry reports a hazard ratio for event-free survival and a log-rank analysis for overall survival. The event-free survival analysis is labeled Chi-squared despite the reported hazard ratio parameter; the registry record does not provide enough detail to clarify the precise implementation. The overall survival analysis reports a log-rank method. General statistical interpretation of these methods depends on assumptions about follow-up and the event process; further diagnostics or details are not reported in the registry record.

Why can over-dispersion matter?

For binary outcomes, a simple model may assume that the variability in the data follows the model's specified probability structure. Over-dispersion means the observed variability is greater than that structure expects. If it is ignored, uncertainty estimates can be too narrow and tests can appear more decisive than warranted. This is a general statistical consideration, not a finding reported for this trial. The registry record does not report an over-dispersion assessment or adjustment.

Other aspects that commonly affect interpretation include missing outcome handling, analysis population definitions, covariate adjustment, and multiplicity across multiple endpoints. The record defines the primary Full Analysis Set and describes the primary endpoint denominator, but does not report all of these analysis details. Where the registry does not specify the method, the analysis should not be assumed to have used a particular imputation, censoring, or multiplicity procedure.

11. Confidence Intervals and Precision

The primary unadjusted risk difference is -0.098, with a 95% confidence interval from -0.154 to -0.042. Because this interval does not include zero, the observed data are inconsistent with equal group proportions under the assumptions of this calculation at the corresponding confidence level. The interval also indicates that the estimated absolute difference is not known exactly: a range of values remains compatible with the data and method.

The sponsor's logistic regression reports an odds ratio of 1.60, with a 95% confidence interval from 1.227 to 2.084. For an odds ratio, the null value is one rather than zero. The reported interval lies above that value. The odds ratio interval and the risk-difference interval use different effect scales and different analyses, so their endpoints should not be compared numerically as if they were measuring the same quantity.

Confidence intervals describe sampling uncertainty under an analysis and its assumptions. They do not automatically incorporate uncertainty from missing data, endpoint assessment, model choice, or selective reporting. The registry record does not report enough detail to assess all such sources. The interval width is therefore one part of the precision assessment, not a complete summary of every limitation.

For secondary outcomes, the registry also reports confidence intervals, including intervals for event-free survival and overall survival analyses. Each interval pertains to its own endpoint, analysis method, and follow-up context. They should not be combined into a single overall estimate. Interpretation of several intervals at once also needs to consider multiplicity, and the relevant adjustment plan is not reported in the registry record.

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 sample-size figures are planning illustrations based on the observed effect, not the trial's original design justification. For binary outcomes, the displayed quantities are participants per arm. For time-to-event outcomes, the figures are total events, not participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect.

The 80% power column gives the illustrative number of participants per arm, or total events for time-to-event outcomes, under which the calculation would have an 80% probability of detecting the specified effect if the assumptions hold and that effect is true. The 90% power column uses the corresponding 90% target. Power is a property of a planned design under specified assumptions; it is not a measure of how likely the observed finding is to be true.

Rows planning for 100%, 75%, or 50% of the observed effect show how the illustrative requirements change when the target effect is smaller than the observed estimate. Smaller effects need more participants or events to detect with the same target power, all else equal. The figures are not the enrolled sample size and should not be read as a retrospective score for the completed observations. This is not observed power.

The original trial sample-size assumptions, including its own target effect and design calculation, are not reported in the registry record. The posted planning illustrations are therefore useful for understanding how effect size relates to required information, but they do not establish how the trial's enrollment target was chosen.

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 = 0.373; treatment proportion p₁ = 0.275; average p̄ = 0.324.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6621 = 1.2976.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6584 = 0.5542.
4. n = (1.2976 + 0.5542)² / (-0.0979)² = 357.9, rounded up to 358 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 <- 0.373358348968105                          # control proportion
p1 <- 0.275471698113208                          # 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(3732677)
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 = 358 (80% power), 479 (90% power)
 75% of observed effect: n per arm = 650 (80% power), 869 (90% power)
 50% of observed effect: n per arm = 1,488 (80% power), 1,991 (90% power)
power.prop.test: 357.9 
Simulated power at n = 358 per arm: 0.803 (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 the sponsor's primary analysis as logistic regression, with an odds ratio of 1.60, a 95% confidence interval from 1.227 to 2.084, and p< 0.001. The separate recalculation uses only the posted group counts and denominators to produce an unadjusted risk difference of -0.098, with a 95% confidence interval from -0.154 to -0.042 and an unadjusted p-value of <0.001. The negative risk difference is expressed as chemotherapy alone relative to durvalumab plus chemotherapy; the sponsor's odds ratio uses the registry's listed group ordering.

These analyses use different effect measures and methods. The odds ratio is not expected to equal the risk difference, and the registry's regression analysis need not be reproduced by a calculation from summary counts. The posted estimates are directionally consistent once their comparison order and scale are taken into account. The registry record does not indicate a discrepancy between them.

The sponsor's own analysis is the reported inferential analysis for the trial. The recalculation is a simple unadjusted check of the group summaries, not a substitute for the sponsor's model or a reconstruction of its analysis plan. The record does not provide the full regression specification or enough information to assess whether additional adjustments were used.

15. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is clearly identified as pathologic complete response at cystectomy, and the primary population is defined as the Full Analysis Set, including all randomized subjects. The registry also lists multiple secondary outcomes and reports analyses for several endpoints. When many hypotheses are tested, the chance of at least one apparently unusual result can increase. A prespecified hierarchy or multiplicity adjustment can address that issue, but the multiplicity plan is not reported in the registry record.

The record does not report whether secondary outcomes were intended as confirmatory or exploratory, whether testing followed a fixed sequence, or whether p-values were adjusted across outcomes. Accordingly, secondary p-values should not automatically be interpreted as if each were the sole primary test. The primary endpoint remains the central result identified by the registry; secondary results provide context with a more cautious evidential status.

The primary denominator is defined as all randomized subjects, but details about missing endpoint measurements are not reported in the registry record. The record does not specify how participants without an evaluable cystectomy specimen were handled, whether missing values were imputed, or whether sensitivity analyses were conducted. The same absence of detail applies to missingness and censoring procedures for the time-to-event outcomes.

These omissions matter because missingness can affect the interpretation of an estimate if the chance of an unavailable outcome is related to treatment assignment or outcome status. Randomization protects the comparison at allocation, but it cannot by itself resolve the consequences of unobserved endpoint data. No particular missing-data assumption or imputation method should be inferred without a registry-reported description.

The registry record also does not report the full statistical analysis plan, interim-analysis procedures, stopping rules, or detailed model covariates. Their absence from the record is not evidence that such procedures were or were not used. It limits what can be concluded about analysis choices beyond the methods explicitly reported.

16. 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: urinary tract infection (13.1%), prostate cancer (6.6%), acute kidney injury (4.7%), pyelonephritis (4.4%), prostate cancer stage ii (4.0%).

The registry reports serious adverse events among 326 of 530 participants in the durvalumab plus gemcitabine plus cisplatin group and 287 of 526 participants in the gemcitabine plus cisplatin group. Other adverse events are reported for 522 of 530 and 511 of 526 participants, respectively. Deaths are reported as 27 and 29 in those groups. These are descriptive counts; the record does not provide an adjusted comparison of safety risk or enough detail to attribute events to a particular treatment component.

The serious-event groups use denominators of 530 and 526, which differ from the primary efficacy group sizes. Safety counts should therefore be read using their stated denominators rather than the primary endpoint populations. The registry lists urinary tract infection, prostate cancer, acute kidney injury, pyelonephritis, prostate cancer stage ii, and urosepsis among the top serious terms, with the posted percentages ranging from 3.8% to 13.1%.

These summaries do not describe event timing, duration, recurrence, exposure-adjusted rates, or causal attribution. The registry record does not report those details in the information presented. The counts can describe what was recorded in each group but cannot, on their own, establish why an event occurred or provide individualized medical guidance.

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

The primary result is a comparison of observed pathologic complete response proportions in two randomized groups. It shows more participants meeting the stated pathological definition in the durvalumab-containing group than in the chemotherapy-alone group. The registry's logistic regression and the unadjusted risk-difference calculation both support a difference in the direction of the durvalumab-containing group, while quantifying it on different scales.

This result does not by itself establish the effect on overall survival, metastasis-free survival, or disease-specific survival. Those are separate outcomes with different definitions and follow-up. Nor does a response at cystectomy establish how any particular participant will fare. Group-level estimates describe the trial groups, not an individual prognosis.

Interpretation also depends on how completely the primary endpoint was assessed and how missing outcomes were handled. The registry specifies the Full Analysis Set and central pathology review, but does not report the missing-data procedures or all operational details of assessment. The estimates should be understood within the limits of those available descriptions.

The observed association should not be expanded into a claim that treatment works or does not work in every setting. This record concerns a particular randomized comparison in muscle-invasive bladder cancer, with the treatment strategies and endpoints described in the registry. It does not answer questions about other regimens, populations, durations, or clinical decisions.

The safety figures are also descriptive rather than a complete benefit-harm evaluation. They do not show event attribution or time at risk in detail. Overall, the posted results provide statistical evidence about the specified group comparison, but they do not settle every question about durability, safety, treatment delivery, or applicability.

18. Why This Trial Matters Statistically

This trial is statistically notable because it evaluates a clearly defined binary pathological endpoint in a randomized, parallel-group phase III study. The primary response definition uses a specific stage, T0N0M0, and central pathology review. Those features make the endpoint and assessment framework explicit in the registry record, while leaving important implementation details unreported.

The comparison illustrates why effect measures need to be read on their own scales. The primary registry analysis reports an odds ratio, while a simple recalculation from posted counts gives a risk difference. Both can describe the same group contrast, yet they answer the quantitative question differently. Confidence intervals add information about precision that a p-value alone cannot convey.

The record also demonstrates the distinction between a primary endpoint and several secondary outcomes. A pathological response, cystectomy, event-free survival, and overall survival are related but not equivalent outcomes. A consistent direction across some results can strengthen the overall context, but it does not remove the need to account for multiple testing or the different endpoint definitions.

Finally, the sample-size illustrations show how planning requirements depend on the target effect and whether the calculation is based on participants per arm or total events. They are not the trial's original design justification. Keeping that distinction clear helps prevent a retrospective planning calculation from being mistaken for a report of observed power or protocol intent.

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.

The registry record does not report the original sample-size justification or the full statistical analysis plan. It therefore does not establish the assumptions that guided enrollment, the planned handling of all endpoints, or the complete set of model specifications. The planning illustrations based on observed effects cannot fill that gap.

The primary endpoint requires assessment of a cystectomy specimen, yet details about participants without an evaluable specimen and missing endpoint data are not reported in the registry record. Although the Full Analysis Set is defined as all randomized subjects, the record does not explain in detail how unavailable pathological assessments were incorporated into the reported endpoint counts.

The study is open-label according to the registry. Central pathology review is specified for the primary endpoint, but further information about reviewer masking, adjudication, and assessment procedures is not reported in the registry record. Open-label conduct may be more consequential for some secondary outcomes involving clinical decisions or investigator assessment than for a centrally reviewed pathological classification.

The registry lists multiple secondary outcomes, but a multiplicity plan is not reported. The record also does not report results for every listed secondary outcome. Consequently, the strength of evidence across the full outcome set cannot be assessed from the posted results alone, and absent results should not be treated as evidence of no effect.

The safety summaries give participant counts and selected serious-event terms but do not provide detailed attribution, timing, exposure-adjusted rates, or event-level context. The registry status is active not recruiting; the record's dates and status should not be taken as confirmation that all follow-up and analyses are complete. These limitations constrain how broadly the results can be interpreted.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Durvalumab+ Gemcitabine/Cisplatin (Neoadjuvant Treatment) and Durvalumab (Adjuvant Treatment) in Patients With MIBC — NCT03732677.
• Registry record last updated 2026-10-15; status last verified 2026-10.
• Results first posted 2025-07-03.
• 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

The registry describes NCT03732677 as a randomized, open-label, parallel-group phase III global study in muscle-invasive bladder cancer. It compares durvalumab with gemcitabine and cisplatin against gemcitabine and cisplatin alone, with durvalumab alone described for adjuvant treatment in the trial title. Enrollment is 1063, and the status is active not recruiting.

The primary outcome is pathologic complete response at cystectomy, defined as T0N0M0 by central pathology review, with the Full Analysis Set as the denominator. The posted primary counts and analyses favor the durvalumab-containing group on the reported response comparison. Secondary outcomes address survival, cystectomy, and other measures; their interpretation is limited by unreported multiplicity and missing-data procedures. The registry also posts descriptive adverse-event counts. The original design assumptions and full statistical analysis plan are not reported in the registry record.