A Study of Rucaparib Versus Physician's Choice of Therapy in Participants With Metastatic Castration-resistant Prostate Cancer and Homologous Recombination Gene Deficiency: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Rucaparib Versus Physician's Choice of Therapy in Participants With Metastatic Castration-resistant Prostate Cancer and Homologous Recombination Gene Deficiency (NCT02975934), a randomized phase 3 study. The primary outcome is Radiographic Progression-free Survival (rPFS) by IRR in Participants With a BRCA Alteration.

2. Clinical Question

This is a time-to-event question: the endpoint incorporates both when an event occurs and the fact that some participants may not have had an event by the end of their observed follow-up. The endpoint does not represent survival alone, because radiographic progression can also count as an event. The population for the primary analysis is narrower than the overall enrollment, so the primary result should be interpreted in relation to participants with a BRCA alteration rather than all participants enrolled.

3. Trial Design

The registry describes a completed, phase 3, randomized, open-label, parallel-group study. The study compared oral rucaparib monotherapy with physician’s choice of abiraterone acetate, enzalutamide, or docetaxel in participants with metastatic castration-resistant prostate cancer and evidence of homologous recombination gene deficiency. The registry reports 405 participants enrolled. It does not report the allocation ratio or the method used to generate or conceal the randomization sequence.

The comparator was a choice among active drug regimens, rather than a single uniform treatment. Abiraterone acetate was paired with prednisone, docetaxel with prednisone or prednisolone, and enzalutamide was used as monotherapy. The registry record does not give the numbers assigned to each physician-selected regimen or explain the rules for making that choice. The study was open label, so participants and treatment teams were not masked to assigned treatment. The primary endpoint was assessed by IRR, as named in the registry record; its full meaning and operational procedures are not reported there.

For the primary endpoint, the posted results describe the ITT population with BRCA-mutated metastatic castration-resistant prostate cancer. The endpoint results list 201 participants for rucaparib and 101 for physician’s choice. The registry also posts results for overall survival in the BRCA subgroup and in the combined BRCA-or-ATM subgroup, as well as objective response rate in the BRCA subgroup. The overall enrollment and the endpoint-specific analysis counts describe different populations or analysis sets and should not be treated as interchangeable.

The registry reports a start date of 2017-06-13, primary completion on 2022-08-25, and completed status. Primary endpoint follow-up is described as extending from enrollment to primary completion, with total follow-up up to approximately 4 years. The planned sample-size justification, original design effect, and protocol-specified statistical analysis details beyond the posted results 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.

Treatment Phase

Reasons for not completing this period:

Cross-Over Phase

Analysis population for the primary outcome, as defined in the registry: ITT Population with BRCA mutated mCRPC.

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: Intent-to-Treat (ITT) Population included all randomized participants (participants with breast cancer gene \[BRCA\] mutated metastatic castration-resistant prostate cancer \[mCRPC\] and participants with ataxia- telangiectasia mutated serine/threonine kinase \[ATM\] mutated mCRPC).

6. Primary Endpoint

The registry lists the primary outcome as Radiographic Progression-free Survival (rPFS) by IRR in Participants With a BRCA Alteration, measured over From enrollment to primary completion of study (Total follow-up was up to approximately 4 years).

The primary endpoint is radiographic progression-free survival by IRR in participants with a BRCA alteration. It measures time from randomization to the first objective evidence of radiographic progression or death due to any cause, whichever occurs first. Thus, a participant can have an endpoint event because imaging shows progression, even if death has not occurred. Conversely, death counts as an event even if radiographic progression was not recorded first.

The registry describes radiographic progression using modified RECIST Version 1.1 for soft tissue disease and PCWG3 criteria for bone disease. The description includes confirmed soft tissue progression, new or unequivocally progressing lesions, and specified patterns of new bone lesions with confirmation. These rules aim to define when imaging changes qualify as progression rather than treating every observed change as an event.

General statistical context: a time-to-event endpoint uses the timing of events and can retain information from participants who have not yet experienced an event at their last observation. This is useful when follow-up lengths differ or when some participants remain event-free at the time of analysis. The registry’s reported medians describe the time by which half the relevant group is estimated to have had an event under the endpoint definition; they are not averages of each participant’s observed time. The hazard ratio summarizes the relative event rate over the analysis period under the model or method used, and is distinct from a ratio of the two medians.

7. Results Posted in the Registry

• Rucaparib vs Abiraterone Acetate or Enzalutamide or Docetaxel: hazard ratio 0.50 (95% CI 0.36–0.69), reported p < 0.001 (as reported in the registry).

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.

Overall Survival in Participants With a BRCA Alteration

• 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:
  • Rucaparib vs Abiraterone Acetate or Enzalutamide or Docetaxel: hazard ratio 0.91 (95% CI 0.68–1.20), reported p = 0.504 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Rucaparib vs Abiraterone Acetate or Enzalutamide or Docetaxel and the control. The interval gives the range of effects that remain compatible with the data.

Overall Survival in Participants With a BRCA or ATM Alteration Combined

• 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:
  • Rucaparib vs Abiraterone Acetate or Enzalutamide or Docetaxel: hazard ratio 0.99 (95% CI 0.78–1.26), reported p = 0.937 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval includes no difference (a ratio of 1), so these data do not show a statistically significant difference between Rucaparib vs Abiraterone Acetate or Enzalutamide or Docetaxel and the control. The interval gives the range of effects that remain compatible with the data.

Objective Response Rate (ORR) by IRR in Participants With a BRCA Alteration

• 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:
  • Abiraterone Acetate or Enzalutamide or Docetaxel: 17.1% (7/41) versus 45.1% (37/82) on control; risk difference -28.0 percentage points (95% CI -43.8 to -12.3), risk ratio 0.38 (0.18–0.77); unadjusted two-proportion z-test p = 0.002.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Abiraterone Acetate or Enzalutamide or Docetaxel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary results cover overall survival, objective response, duration of response, and PSA response, including outcomes in BRCA and combined BRCA-or-ATM populations. The posted numerical comparisons available in this record include overall survival for both populations and objective response for the BRCA subgroup. Results for the other listed secondary outcomes are not reported in the registry record summarized for these analyses.

Across the posted overall survival comparisons, the estimated hazard ratios are near 1 and their confidence intervals include 1. This pattern differs from the primary radiographic progression-free survival result, whose confidence interval is below 1. The endpoints measure different things: progression-free survival includes radiographic progression or death, while overall survival measures death. The primary result therefore should not be treated as evidence that the same magnitude or certainty of difference applies to survival.

The response counts show a secondary outcome based on participants with an evaluable response denominator, and the displayed recalculation is unadjusted. It is not a sponsor-reported analysis. The registry does not report a multiplicity adjustment or a hierarchy establishing how confirmatory secondary claims should be handled. General statistical context: when many outcomes, populations, and analyses are examined, the chance of at least one apparently notable result can increase unless the analysis plan accounts for multiple testing. Accordingly, secondary findings generally carry less confirmatory weight than a prespecified primary endpoint, especially when statistical adjustment and analysis hierarchy are not reported. The secondary results can add context, but they do not replace the primary endpoint or resolve the uncertainty in overall survival.

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

• Overall Survival in Participants With a BRCA Alteration
• Overall Survival in Participants With a BRCA or ATM Alteration Combined
• Objective Response Rate (ORR) by IRR in Participants With a BRCA Alteration
• Objective Response Rate (ORR) by IRR in Participants With a BRCA or ATM Alteration Combined
• Duration of Response (DOR) by IRR in Participants With a BRCA Alteration
• Duration of Response (DOR) by IRR in Participants With a BRCA or ATM Alteration Combined
• PSA Response in Participants With a BRCA Alteration
• PSA Response in Participants With a BRCA or ATM Alteration Combined

9. Statistical Interpretation

10. Statistical Methods Explained

The primary comparison is a time-to-event analysis. The registry reports a log-rank method and a hazard ratio for the comparison of rucaparib with physician’s choice. General statistical context: time-to-event methods account for the timing of events and for participants whose event time is not observed during their available follow-up. Such participants are typically censored at their last known observation, but the censoring rules and detailed handling of follow-up are not reported in the registry record.

What does the hazard ratio describe?

General statistical context: a hazard is the rate at which an event occurs among participants who have not yet experienced it, considered over follow-up. A hazard ratio compares these rates between groups under the analysis approach. A value of 1 corresponds to equal estimated hazards; values below or above 1 indicate a lower or higher estimated hazard in the numerator group, respectively. This comparison does not directly state how many months an individual gains, and it is not obtained by dividing the two reported medians. The registry calls the reported value a hazard ratio and names the log-rank method; additional modeling details are not reported.

The primary hazard ratio is reported as 0.50, with a 95% confidence interval of 0.36 to 0.69 and p < 0.001. The interval gives a range of values compatible with the observed data under the statistical procedure and its assumptions. Because the interval does not include 1, it is consistent with a group difference in the direction of a lower estimated event hazard for rucaparib. The log-rank p-value tests a null comparison of the event-time distributions; it is not a measure of effect size and does not quantify the probability that the result will be reproduced.

Why report medians as well?

The median time to event gives a time-based summary that is often easier to interpret than a hazard ratio. It marks the point at which the estimated event-free proportion reaches one half. In these results, the median is 11.2 months for rucaparib and 6.4 months for physician’s choice in the primary population. A median uses only one point on the estimated event-time distribution, whereas the log-rank comparison uses event information across follow-up. The two summaries are related but are not interchangeable. The registry record does not provide the survival curves or enough detail to describe their shape.

How should the response comparison be read?

Objective response rate is a binary outcome: participants are classified according to whether a response was recorded. The posted counts are 37 of 82 for rucaparib and 7 of 41 for physician’s choice in the BRCA subgroup. The calculated comparison is identified as unadjusted and is not listed as a sponsor-reported analysis. General statistical context: an unadjusted comparison does not account for the testing of multiple endpoints or subgroup analyses. Its interpretation should therefore be cautious, particularly when the registry lists many outcomes and analyses. The registry record does not report a sponsor test for this response outcome.

For response outcomes, the planning illustration uses participants per arm. For time-to-event outcomes, the planning illustration uses total events. These are different quantities: a total event target is not a participant count, because participants may be followed for different durations and may not all have an event. The illustration is based on the observed effect rather than the original trial design, and is not observed power. Its assumptions are described with the planning results.

What does the open-label design imply?

General statistical context: in an open-label study, knowledge of treatment assignment can influence care, reporting, follow-up decisions, or subjective assessments. A radiographic endpoint assessed by IRR may reduce some potential influence from treatment knowledge, but the registry record does not specify the assessment process, whether assessors were masked, or how disagreements were resolved. It would therefore be inappropriate to assume a particular protection against assessment bias from the IRR label alone.

The registry reports four analyses and 18 outcome measures, but does not describe a complete statistical analysis plan, multiplicity strategy, interim analyses, or the handling of missing data. It also does not report the assumptions used to justify the trial’s original enrollment target. Those omissions limit what can be concluded about the planned hierarchy of tests or the safeguards used to control the chance of false-positive findings across outcomes.

11. Confidence Intervals and Precision

The primary hazard ratio is 0.50, with a 95% confidence interval from 0.36 to 0.69. This interval is wholly below 1 and is consistent with a lower estimated event hazard for rucaparib in the primary comparison. The interval also shows that the exact size of the relative difference is uncertain: values across a range remain compatible with the analysis. The confidence interval does not mean that 95% of participants’ outcomes fall within its limits.

The median estimates also have uncertainty. The 95% confidence interval for the rucaparib median is 9.2 to 13.8 months, while the interval for physician’s choice is 5.4 to 8.3 months. These intervals describe uncertainty around each median separately; they are not themselves a confidence interval for the difference between medians. The registry does not post a direct confidence interval for that difference, and no difference should be inferred by subtracting the endpoints.

For overall survival in the BRCA subgroup, the hazard ratio confidence interval is 0.68 to 1.20; for the combined BRCA-or-ATM subgroup it is 0.78 to 1.26. Both ranges include 1, so they are compatible with no relative hazard difference as well as with effects in either direction. The respective p-values are 0.504 and 0.937. These estimates are less precise for distinguishing a group difference than the primary result, and the registry’s posted results do not establish a clear overall survival separation.

General statistical context: a confidence interval reflects sampling uncertainty under the method and assumptions used. It does not capture every source of uncertainty, such as bias from study conduct, endpoint assessment, missing information, or choice among multiple analyses. The registry record does not provide enough detail to assess all such sources. Precision is also specific to the endpoint population: the primary result is for participants with a BRCA alteration, not automatically for the full enrollment or the combined BRCA-or-ATM population.

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 original design assumptions. The registry record does not report the assumptions used for the trial’s own sample-size justification. The illustration uses a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. It considers effects equal to the observed effect and fractions of that effect. The smaller the effect targeted, the more information is generally needed to detect it; the planning rows reflect that principle.

For time-to-event comparisons, the figures are total events required, not total participants and not participants per arm. The number of events a future study accumulates depends on event occurrence and follow-up. For the binary response comparison, the figures are participants per arm. The units differ, so the two kinds of planning values should not be compared as if they counted the same thing.

The 80% power column describes the planned chance of detecting the stated effect under the illustration’s assumptions if that effect is real. The 90% power column describes a higher target chance of detection under those same assumptions. These are design targets for hypothetical planning scenarios, not probabilities that the observed result is correct. The sample-size figures are not observed power and do not provide a retrospective measure of the trial’s success. They should be interpreted as illustrations of how required information changes with the assumed effect, rather than as evidence about the study’s original planning or conduct.

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

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 = 0.500, so ln(HR) = -0.6931.
2. d = 4 × (1.96 + 0.842)² / [ln(0.500)]² = 65.3, rounded up to 66 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 <- 0.5                              # 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(2975934)
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 = 66 (80% power), 88 (90% power)
 75% of observed effect: events = 117 (80% power), 156 (90% power)
 50% of observed effect: events = 262 (80% power), 350 (90% power)
Simulated power with 66 events: 0.823 (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.

14. 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 103 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 405 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 <- 6.4          # control-arm median, months (converted from the posted unit)
hr          <- 0.5          # 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: 70        p (1-sided)   0.0001
 Events: 30       ~HR at bound   0.2529
  Month: 16   P(Cross) if HR=1   0.0001
            P(Cross) if HR=0.5   0.0338
  IA 2: 67%                  Z   2.5114
     N: 104        p (1-sided)   0.0060
 Events: 59       ~HR at bound   0.5179
  Month: 26   P(Cross) if HR=1   0.0060
            P(Cross) if HR=0.5   0.5603
      Final                  Z   1.9930
     N: 104        p (1-sided)   0.0231
 Events: 88       ~HR at bound   0.6529
  Month: 60   P(Cross) if HR=1   0.0250
            P(Cross) if HR=0.5   0.9000



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

15. Analyses Reported by the Sponsor

For the primary endpoint, the sponsor-reported analysis uses the log-rank method and reports a hazard ratio of 0.50, a 95% confidence interval of 0.36 to 0.69, and p < 0.001. The recalculated display does not identify a discrepancy for this endpoint. The sponsor’s reported result is therefore the principal inferential result to use when describing the primary comparison.

The same approach applies to the posted overall survival comparisons: the registry reports hazard ratios and log-rank p-values for the BRCA subgroup and the combined BRCA-or-ATM subgroup. The recalculated material does not identify a discrepancy for these comparisons. For the response outcome, the registry posts participant counts but no sponsor-reported analysis in the reported analyses. Any recalculated response comparison is an unadjusted analysis of the posted counts, not a substitute for a sponsor analysis.

General statistical context: a recalculation from summary statistics can be useful for orientation, but it may not reproduce an analysis based on participant-level data, prespecified stratification, censoring rules, or covariate adjustments. The registry does not report enough detail to establish whether such features were used beyond the named method. Where a sponsor analysis is posted, it takes precedence over a simple unadjusted recalculation; where none is posted, the recalculated comparison should be treated as descriptive and qualified accordingly.

16. Multiplicity, Missing Data and Analysis Populations

The registry lists 18 outcome measures and four analyses. It identifies radiographic progression-free survival in participants with a BRCA alteration as the primary efficacy endpoint, but does not report a complete testing hierarchy for the secondary outcomes or subgroup analyses. A multiplicity adjustment, such as a prespecified procedure to control the overall chance of false-positive findings across tests, is not reported in the registry record.

This omission matters when interpreting secondary results. A nominal p-value for one secondary analysis does not by itself show that the result would remain statistically persuasive after accounting for the full set of outcomes and comparisons. The objective response comparison is presented as an unadjusted recalculation and has no sponsor-reported analysis in the posted analyses. It should not be treated as confirmatory evidence. The overall survival analyses are sponsor-reported, but their placement in a broader testing strategy is not described.

The primary endpoint population is identified as the ITT population with BRCA-mutated metastatic castration-resistant prostate cancer. The registry does not provide detailed definitions of the ITT set, exclusions after randomization, or the analysis populations used for every secondary outcome. It also does not report how missing endpoint data, incomplete imaging, withdrawals, or censoring were handled. Those details affect the interpretation of time-to-event estimates and response denominators, but cannot be inferred from the posted counts alone.

Interim analyses and any associated adjustment to statistical thresholds are not reported in the registry record. The original statistical analysis plan and sample-size justification are also not reported. These gaps do not invalidate the posted results, but they limit independent assessment of prespecification, missing-data handling, and control of false-positive conclusions across the full set of analyses.

17. 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: febrile neutropenia (6.2%), pneumonia (4.3%), anaemia (3.3%), pulmonary embolism (3.1%), acute kidney injury (3.0%).

The registry lists serious adverse event counts by treatment phase and group. During the treatment phase, 81 of 270 participants in the rucaparib group and 36 of 130 in the physician’s-choice group are listed as having serious adverse events. In the crossover phase, 18 of 70 participants receiving rucaparib are listed. These are phase-specific figures; the crossover group should not be read as another randomized comparison group. The registry also reports deaths of 6, 3, and 1 in those respective groups, but does not provide enough context in these figures to attribute deaths to treatment.

Among the listed serious adverse event terms, the registry reports febrile neutropenia at 6.2%, pneumonia at 4.3%, anaemia at 3.3%, pulmonary embolism at 3.1%, acute kidney injury at 3.0%, and cerebrovascular accident at 2.9%. These terms describe reported serious events, not all adverse events or their duration, severity distribution, or causal relationship to study treatment.

General statistical context: safety counts depend on who was at risk, exposure time, event definitions, and how repeated events are counted. The registry record does not describe those details for these summaries, nor does it provide a randomized treatment-phase risk comparison adjusted for exposure. The figures should therefore be read as descriptive reporting, not as proof that one regimen caused more or fewer serious events.

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

The primary result describes a randomized comparison of rucaparib with a physician-selected active therapy in participants with BRCA-altered metastatic castration-resistant prostate cancer. The registry reports a lower estimated hazard of radiographic progression or death for rucaparib in this analysis, with a hazard ratio of 0.50 and a 95% confidence interval of 0.36 to 0.69. The reported p-value is p < 0.001. The median endpoint estimates also differ, with 11.2 months for rucaparib and 6.4 months for physician’s choice.

This result does not mean every participant experienced a longer time without progression or death, and it does not predict an individual outcome. A group-level hazard ratio is not a direct measure of months gained by each participant. The endpoint also combines radiographic progression and death, so it is not an overall survival result.

The posted overall survival analyses have confidence intervals that include 1 and p-values of 0.504 in the BRCA subgroup and 0.937 in the combined BRCA-or-ATM subgroup. They do not provide clear evidence of a difference in overall survival. This does not establish that the groups are identical; the intervals retain uncertainty about the size and direction of a possible difference.

The findings are specific to the study population, comparator choices, endpoint definitions, and analyses reported in the registry. The primary result should not automatically be generalized to participants without a BRCA alteration, to other disease settings, or to unreported treatment options. The open-label design, unreported analysis details, and lack of a reported multiplicity strategy are relevant when assessing the secondary results. The safety summaries are descriptive and do not establish comparative causality.

General statistical context: statistical evidence describes uncertainty about group-level effects under specified assumptions. It does not by itself establish clinical importance, individual benefit, or the balance of outcomes that matters to a particular person. Those judgments require evidence and context beyond the endpoint statistics reported in the registry.

19. Why This Trial Matters Statistically

This trial illustrates why endpoint definition and analysis population matter. The primary result is not a general comparison across every enrolled participant: it concerns radiographic progression-free survival in the ITT population with a BRCA alteration. The registry also reports an overall survival analysis for a broader combined BRCA-or-ATM subgroup. Keeping the populations distinct prevents an estimate from one subgroup being treated as if it applied to another.

It also demonstrates the different roles of time-to-event summaries. The median gives an interpretable time point, while the hazard ratio and log-rank test compare event experience over follow-up. These measures answer related but non-identical questions. Their uncertainty should be considered alongside the point estimates, rather than interpreting a p-value or median alone.

The posted results further show why secondary outcomes need a clear testing plan. Several outcomes and populations are listed, but the registry record does not report a multiplicity strategy or a complete hierarchy of tests. A nominal secondary comparison therefore cannot be given the same confirmatory weight as a clearly designated primary endpoint. The response result also distinguishes sponsor-reported analyses from unadjusted recalculations based on posted counts.

Finally, the comparison is between rucaparib and a set of physician-selected regimens, and the study was open label. These features shape what the results directly address and what information is needed to interpret them fully. The statistical record provides important estimates, but transparent reporting of analysis details, missing-data handling, and testing strategy is needed to judge the full strength and limits of the evidence.

20. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary analysis concerns participants with a BRCA alteration, while the overall enrollment is 405. The endpoint results list 201 participants for rucaparib and 101 for physician’s choice. The registry does not fully explain the relationship between these counts and the overall enrollment, or give detailed accounting of every analysis population. The primary result should not be assumed to represent all enrolled participants.

The control group includes abiraterone acetate, enzalutamide, or docetaxel, but the registry does not report the numbers receiving each option or the criteria for selecting among them. The comparison is therefore against a physician-choice strategy rather than a single uniform comparator regimen. The record also does not report allocation concealment or the randomization sequence generation method.

The study is open label. Although the primary endpoint is described as assessed by IRR, the registry does not provide details about assessor masking, procedures for resolving discrepancies, or imaging review. It does not report the detailed analysis plan, censoring rules, handling of missing data, interim analyses, or multiplicity adjustment. These gaps restrict independent evaluation of potential bias and the interpretation of secondary results.

The posted overall survival follow-up is described as up to approximately 7 years, but the record does not provide detailed information about subsequent therapies or their handling in the analyses. The serious adverse event summaries are grouped by treatment and crossover phase and do not provide exposure-adjusted comparisons or full causal assessments. The registry record also does not report the study’s original sample-size justification. These are reporting limitations; they do not by themselves establish that the analyses were conducted improperly.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: A Study of Rucaparib Versus Physician's Choice of Therapy in Participants With Metastatic Castration-resistant Prostate Cancer and Homologous Recombination Gene Deficiency — NCT02975934.
• Registry record last updated 2025-07-25; status last verified 2025-07.
• Results first posted 2023-08-16.
• 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.

24. Record Summary

The registry record describes a completed, randomized, open-label, phase 3, parallel-group study of rucaparib versus physician’s choice of abiraterone acetate, enzalutamide, or docetaxel in metastatic castration-resistant prostate cancer associated with homologous recombination gene deficiency. It reports 405 participants enrolled.

The primary endpoint is radiographic progression-free survival by IRR in participants with a BRCA alteration. Median estimates are 11.2 months for rucaparib and 6.4 months for physician’s choice. The sponsor-reported hazard ratio is 0.50, with a 95% confidence interval of 0.36 to 0.69 and p < 0.001.

Posted overall survival analyses in the BRCA and combined BRCA-or-ATM populations have confidence intervals that include 1. The response counts are posted for the BRCA subgroup, but no sponsor-reported response analysis is listed. Details about multiplicity, missing-data handling, and the original sample-size justification are not reported in the registry record.