A Study of Durvalumab Alone and Durvalumab+Olaparib in Advanced, Platinum-Ineligible Bladder Cancer (BAYOU): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Durvalumab Alone and Durvalumab+Olaparib in Advanced, Platinum-Ineligible Bladder Cancer (BAYOU) (NCT03459846), a randomized phase 2 study. The primary outcome is Progression-free Survival (PFS).

2. Clinical Question

3. Trial Design

BAYOU was a phase II, randomized, parallel-group trial in platinum-ineligible patients receiving first-line treatment for unresectable stage IV urothelial cancer. The record describes the study as multicenter and global. It lists quadruple masking, while the official title describes the study as double-blind. The record does not explain the masking terminology further.

The two groups received durvalumab with either olaparib or placebo. Durvalumab was administered at 1500 mg intravenously every 4 weeks, beginning on week 1 day 1. The olaparib group also received 300 mg orally twice daily, adjusted based on creatinine clearance. The comparator group received oral placebo twice daily from week 1 day 1. The registry lists 78 participants in the durvalumab-plus-olaparib group and 76 in the durvalumab-plus-placebo group for the primary outcome.

The record gives enrollment as 154. The study start date is 2018-03-16, and primary completion is listed as 2020-10-15. Its status is active, not recruiting. The primary endpoint is investigator-assessed progression-free survival in the full analysis set. Secondary outcomes listed are overall survival, objective response rate, and duration of response. The registry does not report the original sample-size justification or further operational detail about randomization and masking.

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

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 Progression-free Survival (PFS), measured over Assessments performed at baseline and every 8 weeks from date of randomization until date of objective disease progression or death (by any cause in the absence of progression), assessed up to the data cut-off date (15 Oct 2020), up to a max. of 31 months.

Progression-free survival measures the time from randomization until objective disease progression or death, with death counted when it occurs in the absence of documented progression. In this trial, progression was assessed by investigators according to RECIST 1.1. Assessments were scheduled at baseline and every 8 weeks from randomization, continuing until progression or death, subject to the data cut-off date and a maximum of 31 months.

This endpoint is used in time-to-event analyses because it incorporates both whether an event occurs and how long participants remain without a recorded event. It can therefore summarize the timing of progression or death across the follow-up period, rather than only reporting the proportion with an event at a selected visit. The registry reports median progression-free survival for each group, with a 95% confidence interval, and a Cox regression hazard ratio for the comparison.

These summaries answer related but different questions. A median describes the point in the observed time-to-event distribution at which the event-free proportion reaches its midpoint. A hazard ratio compares event rates over follow-up under the Cox model. Neither summary, by itself, describes every participant’s experience or establishes how the groups would compare beyond the recorded follow-up.

7. Results Posted in the Registry

• Durva + Olaparib vs Durva + Placebo: hazard ratio 0.94 (95% CI 0.64–1.39), reported p = 0.789 (as reported in the registry).

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a time-to-event endpoint. For each participant, the relevant time runs from randomization until the recorded event: objective disease progression or death in the absence of progression. Participants without a recorded event by the applicable observation limit are treated as censored in a conventional time-to-event analysis. That general description explains the method; the registry does not report additional details about its censoring rules.

What does the median represent?

A median time-to-event estimate is the point at which the estimated event-free proportion reaches its midpoint. It is not the average follow-up time, and it does not mean that every participant experienced the event at that time. The registry reports medians of 4.2 months for durvalumab plus olaparib and 3.5 months for durvalumab plus placebo. Their respective 95% confidence intervals are 3.6 to 5.6 months and 1.9 to 5.1 months. The estimates are based on the event and censoring information accumulated during the recorded follow-up.

What does the Cox regression estimate?

The registry reports a Cox regression hazard ratio of 0.94 for durvalumab plus olaparib versus durvalumab plus placebo. In general, a Cox model compares the instantaneous event hazard between groups over time, conditional on the model’s specified structure. A hazard ratio is not a ratio of median survival times, a difference in medians, or a direct probability that an individual will experience an event. The record does not report model covariates or other model specifications, so the estimate should be described as the Cox regression result reported in the registry rather than as an independently reconstructed or adjusted result.

The associated 95% confidence interval is 0.641 to 1.387, and the reported p-value is 0.789. These quantities express uncertainty and statistical compatibility under the analysis; they do not indicate the chance that the finding will replicate. Because the interval includes 1, the data do not pin down a single direction for the hazard contrast. The point estimate is near 1, but precision is better judged from the full interval than from the point estimate alone.

Why do time-to-event methods matter?

Time-to-event methods can use information from participants with different follow-up durations and account for observations that end before an event is recorded. Censoring is not the same as observing that progression or death never occurred; it means the event was not recorded within the participant’s available observation period. The validity of these methods depends on assumptions about the observation process and model, but the registry record does not describe the diagnostics or assumptions assessed for this analysis.

The Cox model is commonly used for event-time comparisons because it incorporates event timing rather than reducing all follow-up to a simple event count. A central model consideration is whether the relative hazard is sufficiently stable over time for a single hazard ratio to be an informative summary. The registry does not report a proportional-hazards assessment or time-varying effects. This does not show that an assumption failed; it means the reader cannot evaluate that aspect from the record.

What does the confidence interval add?

A confidence interval conveys uncertainty around an estimate under the method used to calculate it. The reported interval for the hazard ratio spans values below and above 1, so it allows for more than one direction of relative hazard. The median intervals also show uncertainty around each group’s median estimate. They should not be used as if they were a direct interval for the between-group difference.

The registry describes the primary population as the full analysis set and identifies investigator assessments according to RECIST 1.1. Details about the exact analysis population rules, missing data handling, assessment review, and analysis exclusions are not reported in the registry record. Those omissions limit the ability to reproduce or fully audit the reported analysis from registry information alone.

10. Confidence Intervals and Precision

The reported hazard ratio is 0.94, with a 95% confidence interval from 0.641 to 1.387. The interval includes 1 and spans values on either side, indicating that the comparative estimate is not precise enough to establish a single direction of the relative hazard from this result alone. The point estimate is only one part of the result; the interval is important because it shows a range of values compatible with the analysis.

The median progression-free survival estimates also have uncertainty intervals: 3.6 to 5.6 months for the 4.2-month median in the durvalumab-plus-olaparib group, and 1.9 to 5.1 months for the 3.5-month median in the durvalumab-plus-placebo group. These are intervals for the separate group medians, not a confidence interval for a difference between medians. Comparing their endpoints informally does not replace a direct treatment comparison.

Confidence intervals are statistical ranges produced by an analysis method; they do not guarantee that the true value lies within the particular reported range. Their width reflects uncertainty under the model and data used. The registry does not provide further detail on interval construction beyond identifying the median intervals as 95% confidence intervals and reporting a 95% interval for the Cox hazard ratio. The reported p-value is 0.789, but it does not make the estimate more precise or explain the range of plausible comparative effects.

11. 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 planning figures are total event targets for a time-to-event analysis, not participants per arm. They use the observed hazard-ratio effect as a starting point and show planning illustrations for detecting 100%, 75%, or 50% of that effect. For the full observed effect, the figures are 8,201 total events for 80% power and 10,978 total events for 90% power. For 75% of the observed effect, they are 14,579 and 19,517 total events; for 50% of the observed effect, they are 32,802 and 43,912 total events.

In plain terms, the 80% power column gives the event target associated with an 80% chance of detecting the specified effect under the illustration’s assumptions, if that effect is present. The 90% column gives the target for a higher planned chance of detection under those assumptions. Smaller effects require more events to distinguish them from no difference with the same planned power.

The illustration assumes a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; it applies the stated fraction of the observed effect. The registry does not report the trial’s original design assumptions or sample-size justification. These figures are planning illustrations, not observed power.

12. 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

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 0.940, so ln(HR) = -0.0619.
2. d = 4 × (1.96 + 0.842)² / [ln(0.940)]² = 8,200.3, rounded up to 8,201 events in total 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
hr <- 0.94                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(3459846)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 8,201 (80% power), 10,978 (90% power)
 75% of observed effect: events = 14,579 (80% power), 19,517 (90% power)
 50% of observed effect: events = 32,802 (80% power), 43,912 (90% power)
Simulated power with 8,201 events: 0.793 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

13. Group-Sequential Design Analysis

Many time-to-event trials, especially in oncology, stop or report early when interim data are overwhelming, so a single fixed-sample number understates the real design problem. The calculation below is an illustrative group-sequential design built from this trial's posted control median and hazard ratio. The registry record does not state the planned number of looks or the spending function, so the design is a planning illustration, not a reconstruction of the protocol. Assumed: one-sided α = 0.025, 90% power, three equally spaced looks on information, Lan-DeMets O'Brien-Fleming spending, 24 months of accrual, 60 months in total and a 1% monthly dropout hazard.

Under these assumptions the design needs about 11,688 participants in total, with the boundaries below.

Fixed design versus group-sequential alternatives

Group-sequential designs cost a small increase in maximum sample size and events compared with the fixed design, in exchange for the chance to stop early for overwhelming efficacy. O'Brien-Fleming spending keeps early boundaries very conservative and the final boundary close to the fixed-design value, whereas Pocock spending spends more alpha early.

Sensitivity to a smaller true effect

Planning for exactly the observed effect is optimistic, because observed effects in published trials tend to be somewhat larger than the true effect. The table repeats the main design (three looks, O'Brien-Fleming spending, 90% power) for smaller true hazard ratios.

For context, this trial enrolled 154 participants.

Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993. Agreement between the two packages confirms the boundary calculation.

R code and output

suppressWarnings(suppressMessages({ library(gsDesign); library(rpact) }))
med_control <- 3.5          # control-arm median, months (converted from the posted unit)
hr          <- 0.94          # hazard ratio used for planning
lambdaC     <- log(2) / med_control
eta         <- 0.01               # dropout hazard per month (assumption)
acc_months  <- 24                 # accrual duration (assumption)
study_months<- 60                 # total study duration (assumption)

design <- function(k, sfu, beta, h = hr) gsSurv(k = k, test.type = 1, alpha = 0.025, beta = beta, sfu = sfu,
  lambdaC = lambdaC, hr = h, hr0 = 1, eta = eta, gamma = 1, R = acc_months, T = study_months, minfup = study_months - acc_months)

# Main design: 3 looks, Lan-DeMets O'Brien-Fleming spending, one-sided alpha 0.025, 90% power
x <- design(3, sfLDOF, 0.10)
cat(sprintf("N_TOTAL|%s\n", format(ceiling(sum(x$gamma * x$R)), big.mark = ",")))
for (i in 1:3) cat(sprintf("GS|%d|%.1f|%.0f|%.3f|%.5f|%.5f\n", i, x$T[i], x$n.I[i], x$upper$bound[i], pnorm(-x$upper$bound[i]), cumsum(x$upper$spend)[i]))
cat("\n"); print(gsBoundSummary(x))

# Comparison of designs
cmp <- list(list("Fixed design, one analysis", 1, sfLDOF), list("Two looks, O'Brien-Fleming", 2, sfLDOF),
            list("Three looks, O'Brien-Fleming", 3, sfLDOF), list("Three looks, Pocock", 3, sfLDPocock))
cat("\n")
for (b in c(0.20, 0.10)) for (d in cmp) {
  y <- design(d[[2]], d[[3]], b)
  cat(sprintf("DES|%s|%d|%.0f|%s|%.1f\n", d[[1]], round(100 * (1 - b)), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ","), tail(y$T, 1)))
}

# Sensitivity: what if the true effect is smaller than the observed one? (3 looks, O'Brien-Fleming, 90% power)
cat("\n")
for (fr in c(1, 0.75, 0.5)) {
  y <- design(3, sfLDOF, 0.10, exp(log(hr) * fr))
  cat(sprintf("SENS|%.2f|%.3f|%.0f|%s\n", fr, exp(log(hr) * fr), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ",")))
}

# Independent cross-check of the boundaries with rpact
r <- getDesignGroupSequential(kMax = 3, alpha = 0.025, beta = 0.10, sided = 1, typeOfDesign = "asOF", informationRates = x$timing)
cat(sprintf("\nCross-check, upper boundaries: gsDesign %s | rpact %s\n", paste(round(x$upper$bound, 3), collapse = ", "), paste(round(r$criticalValues, 3), collapse = ", ")))

Method: LachinFoulkes 
      Analysis               Value Efficacy
     IA 1: 33%                   Z   3.7103
       N: 6116         p (1-sided)   0.0001
  Events: 3703        ~HR at bound   0.8852
     Month: 13    P(Cross) if HR=1   0.0001
               P(Cross) if HR=0.94   0.0338
     IA 2: 67%                   Z   2.5114
      N: 10168         p (1-sided)   0.0060
  Events: 7405        ~HR at bound   0.9433
     Month: 21    P(Cross) if HR=1   0.0060
               P(Cross) if HR=0.94   0.5603
         Final                   Z   1.9930
      N: 11688         p (1-sided)   0.0231
 Events: 11108        ~HR at bound   0.9629
     Month: 60    P(Cross) if HR=1   0.0250
               P(Cross) if HR=0.94   0.9000



Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993

14. Analyses Reported by the Sponsor

The comparative estimate reported for the primary outcome is the sponsor’s Cox regression hazard ratio: 0.94, with a 95% confidence interval from 0.641 to 1.387 and a p-value of 0.789. This is the result to use when describing the sponsor’s analysis of progression-free survival. The registry record does not identify a discrepancy between that analysis and the reported result.

For a time-to-event outcome, the group medians and their confidence intervals do not contain all the information used by the Cox regression. Reconstructing a hazard ratio from those summaries alone would not reproduce the participant-level event-time analysis. Accordingly, the sponsor’s reported hazard ratio is the relevant comparative statistic; no independently recalculated hazard ratio is used in this interpretation.

The record names Cox regression as the method but does not give further model details, such as covariates or checks of model assumptions. The reported estimate should therefore be attributed to the analysis as documented, without extending it to claims about unreported model specifications or other endpoints.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists overall survival, objective response rate, and duration of response as secondary outcomes, in addition to the primary progression-free survival outcome. It does not report numerical results for those secondary outcomes in the information summarized for this analysis. The record also does not describe a multiplicity adjustment, a hierarchy of hypotheses, or how confirmatory conclusions across outcomes were to be controlled. Without those details, the p-value for progression-free survival should be read as the reported result for that endpoint, not as a complete account of the trial’s inference across all outcomes.

The record identifies the primary analysis population as the full analysis set. It does not explain the operational definition of that set, the handling of participants with missing assessments, or the rules for censoring and analysis exclusions. These details matter because missing observations and censoring can affect the information available for a time-to-event estimate. The registry’s assessment schedule and endpoint definition do not by themselves establish how every incomplete observation was handled.

Progression was assessed by investigators according to RECIST 1.1. The record does not report whether independent central review was used, how disagreements were resolved, or whether assessment procedures were audited. It also does not give details of the statistical analysis plan, interim analyses, stopping rules, or adjustments for repeated testing. Those matters are not reported in the registry record, so no assumptions about them should be made.

The absence of these details is a limit on what can be learned from the registry summary, not evidence that a particular procedure was or was not used. The reported outcome estimates remain interpretable as the results recorded for the stated endpoint and population, but a fuller assessment of analysis conduct would require the relevant study documentation.

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 (7.9%), anaemia (6.6%), renal failure (5.3%), haematuria (3.9%), hydronephrosis (3.9%).

The posted safety figures give serious adverse events and other adverse events by group. For durvalumab plus olaparib, serious adverse events affected 37 participants among 76 at risk, and other adverse events affected 65 among 76 at risk. For durvalumab plus placebo, serious adverse events affected 26 among 76 at risk, and other adverse events affected 63 among 76 at risk. These are participant counts, not event counts; a participant may experience more than one event.

The record lists deaths as 52 in the durvalumab-plus-olaparib group and 46 in the durvalumab-plus-placebo group. It does not specify in these figures whether deaths were attributed to treatment, disease, or another cause. They should not be interpreted as treatment-related deaths without that information.

The listed serious adverse event terms and percentages are urinary tract infection at 7.9%, anaemia at 6.6%, renal failure at 5.3%, haematuria at 3.9%, hydronephrosis at 3.9%, and pneumonia at 3.9%. The registry summary does not assign these percentages to a specific group in the listed terms. Safety counts describe reported events in this trial and do not, on their own, establish causation or comparative risk.

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

The posted primary result is a comparison of investigator-assessed progression-free survival between durvalumab plus olaparib and durvalumab plus placebo in the trial’s full analysis set. The median estimates were 4.2 months and 3.5 months, respectively. The reported Cox regression hazard ratio was 0.94, with a 95% confidence interval from 0.641 to 1.387 and p-value 0.789. Together, these figures describe the observed analysis and its uncertainty; they do not establish a clear direction or precise magnitude for the relative hazard.

The result does not show that one regimen works or does not work. A hazard ratio is a comparative time-to-event statistic, not a direct measure of an individual participant’s outcome or a guarantee about future experience. The confidence interval includes 1, and the p-value is a summary of compatibility under the reported analysis, not the probability that a treatment is effective. The registry does not report the analysis threshold or complete statistical plan, so the p-value should not be used to infer unreported decision rules.

These findings also do not answer every question about the regimens. The registry lists overall survival, objective response rate, and duration of response as secondary outcomes, but their numerical results are not reported in the registry record. The posted safety counts are descriptive and do not establish that an event was caused by a study drug. Nor do the results necessarily generalize beyond the stated population, setting, and follow-up.

A careful reading keeps the point estimate, its uncertainty, the endpoint definition, and the analysis population together. It also distinguishes reported trial findings from planning illustrations. The event targets in the planning material are hypothetical design calculations based on the observed effect, not evidence about the trial’s achieved power or its original design rationale.

18. Why This Trial Matters Statistically

This trial is statistically useful for illustrating why a time-to-event result should not be reduced to a comparison of medians alone. The group medians are 4.2 months and 3.5 months, while the comparative result is a Cox regression hazard ratio of 0.94. These statistics summarize different features of the event-time data. The hazard ratio uses the timing of events over follow-up, whereas the median identifies a point in each group’s estimated event-time distribution.

The confidence interval is central to the interpretation. The 95% interval for the hazard ratio ranges from 0.641 to 1.387, which includes values on either side of 1. That breadth makes the point estimate less definitive than its single value may suggest. The reported p-value of 0.789 likewise needs to be read as an inferential summary, not as a measure of effect size or proof of equivalence.

The registry also illustrates the distinction between an observed result and a planning calculation. The event targets are hypothetical illustrations that vary the effect to be detected and the desired power. They do not describe the number of events observed in the trial, and they do not reveal the original sample-size justification, which is not reported in the registry record.

Finally, this record makes several limits visible: details of the analysis plan, missing-data handling, multiplicity, and secondary outcome results are not reported in the registry record. Recognizing what is documented and what remains unspecified is part of sound statistical interpretation.

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 gives a concise account of the primary endpoint and reported comparison, but important details needed for independent evaluation are not reported in the registry record. These include the original sample-size justification, the complete statistical analysis plan, the exact analysis-set rules, missing-data procedures, censoring details, and model covariates. The record names Cox regression but does not report diagnostics or checks of model assumptions.

The primary endpoint relied on investigator assessments according to RECIST 1.1. The registry does not report whether central review was also conducted or how assessment differences were handled. It specifies assessments at baseline and every 8 weeks through the stated follow-up limits, but does not give further information about missed assessments or the timing of individual events.

The record lists secondary outcomes but does not include their numerical results in the information available for this analysis. It does not report a multiplicity strategy, interim analyses, stopping rules, or the trial’s own decision threshold. These gaps limit assessment of the broader inferential context and prevent conclusions about unreported endpoints.

The safety summary provides counts for serious and other adverse events and lists serious adverse event terms with percentages, but it does not give detailed event definitions, exposure-adjusted rates, or causal attribution in the figures described. The listed terms are not assigned to a specific treatment group in the summary. Death counts are reported without a stated cause attribution.

These limitations do not alter the values reported for progression-free survival. They define the boundary of what can be concluded from the registry summary and caution against reconstructing unreported design or analysis choices.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Durvalumab Alone and Durvalumab+Olaparib in Advanced, Platinum-Ineligible Bladder Cancer (BAYOU) — NCT03459846.
• Registry record last updated 2026-08-05; status last verified 2026-08.
• Results first posted 2022-10-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

BAYOU is a randomized, parallel-group phase II trial in platinum-ineligible patients with unresectable stage IV urothelial cancer. It compared durvalumab plus olaparib with durvalumab plus placebo. The registry records 154 participants enrolled and identifies the study as active, not recruiting.

The primary outcome was investigator-assessed progression-free survival according to RECIST 1.1 in the full analysis set. Median estimates were 4.2 months with durvalumab plus olaparib and 3.5 months with durvalumab plus placebo. The reported Cox regression hazard ratio was 0.94, with a 95% confidence interval from 0.641 to 1.387 and p-value 0.789.

The record also lists overall survival, objective response rate, and duration of response as secondary outcomes. Further details on the original design assumptions, missing-data procedures, multiplicity, and analysis plan are not reported in the registry record.