Veliparib With Carboplatin and Paclitaxel and as Continuation Maintenance Therapy in Adults With Newly Diagnosed Stage III or IV, High-grade Serous, Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Veliparib With Carboplatin and Paclitaxel and as Continuation Maintenance Therapy in Adults With Newly Diagnosed Stage III or IV, High-grade Serous, Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer (NCT02470585), a randomized phase 3 study. The primary outcome is Progression-Free Survival (PFS) in the BRCA-deficient Population (Arm 3 vs Arm 1).

2. Clinical Question

The trial also included a third group that received veliparib with chemotherapy followed by placebo maintenance. This makes the treatment sequence important when reading the comparisons: the primary result concerns veliparib continued as maintenance, not only veliparib given concurrently with chemotherapy. The registry lists progression-free survival in other populations, overall survival, and disease-related symptom measures among the secondary outcomes.

3. Trial Design

The registry describes a randomized, double-masked, parallel-group phase 3 study. It lists enrollment of 1,140 and the status as terminated. The study concerned previously untreated stage III or IV high-grade serous epithelial ovarian, fallopian tube, or primary peritoneal cancer. The registry identifies AbbVie as sponsor. The reason for termination is not reported in the registry record.

Three treatment sequences are described. One group received placebo with carboplatin and paclitaxel, then placebo maintenance. A second received veliparib with chemotherapy, then placebo maintenance. The third received veliparib with chemotherapy and continued veliparib as maintenance. The primary result compares the third sequence with the placebo-throughout sequence; the concurrent-veliparib-then-placebo group is not part of that primary comparison.

The primary endpoint was progression-free survival in the BRCA-deficient population. The registry defines that population as participants with a deleterious or suspected deleterious germline and/or tumor BRCA1 or BRCA2 mutation. It also describes sequentially inclusive analyses in the BRCA-mutation cohort, homologous recombination deficiency cohort, and intention-to-treat population. How the treatment allocation sequence was generated, whether allocation was stratified, and the original design assumptions are not reported in the registry record.

The primary analysis used data through 03 May 2019, with a median follow-up duration of 28 months. The registry says the analysis took place when the protocol-specified number of progression-free survival events was reached. The exact event target and any interim-analysis plan are not reported in the registry record. Termination status alone does not explain why the study ended or establish whether the planned conduct changed.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Participants were randomized in a 1:1:1 ratio to one of three treatment groups. Randomization was stratified according to the timing of surgery and residual disease after primary surgery or interval surgery, the paclitaxel schedule, stage of disease, geographic region, and germline breast cancer susceptibility gene (BRCA) mutation status.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analyses were performed in 3 sequentially inclusive populations. The first analysis was conducted using the BRCA-mutation cohort which included participants with either a gBRCA and/or tBRCA deleterious or suspected deleterious mutation in BRCA1 or BRCA2.

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 intention-to-treat (ITT) population consisted of all randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Progression-Free Survival (PFS) in the BRCA-deficient Population (Arm 3 vs Arm 1), measured over From randomization until the primary analysis data cut-off date of 03 May 2019, the median duration of follow-up was 28 months..

Progression-free survival measures the time from randomization until disease progression or death, whichever occurs first under the registry definition. Progression was assessed using RECIST version 1.1 as determined by the investigator. If neither progression nor death occurred before the analysis cutoff, the participant’s data were censored at the last evaluable disease assessment. Censoring means that the analysis uses the known event-free follow-up up to that assessment without treating the participant as having experienced progression at that time.

The registry estimated progression-free survival with the Kaplan-Meier method and reports medians in months. A median is the time by which the estimated event-free proportion has reached the midpoint; it is not an average duration for every participant. The endpoint is used in time-to-event analyses because participants can have different lengths of follow-up and some have no recorded event by the cutoff. The hazard ratio summarizes the relative event rate over follow-up under the analysis model; it is not a ratio of the reported medians.

The registry defines progressive disease as at least a 20% increase in target-lesion size compared with the smallest size recorded since treatment began, together with an absolute increase of ≥ 5 mm, or unequivocal progression of existing non-target lesions or appearance of new lesions. The results depend on the recorded assessment schedule, investigator assessments, censoring, and event definitions. Further details of assessment frequency and missing-assessment handling are not reported in the registry record.

7. Results Posted in the Registry

• Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Veliparib: hazard ratio 0.43 (95% CI 0.28–0.68), 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.

Progression-Free Survival (PFS) in the BRCA-deficient Population (Arm 2 vs Arm 1)

• 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:
  • Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Placebo: hazard ratio 1.22 (95% CI 0.82–1.80), reported p = 0.335 (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 Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Placebo and the control. The interval gives the range of effects that remain compatible with the data.

Progression-Free Survival (PFS) in the Homologous Recombination Deficiency Cohort (Arm 2 vs Arm 1)

• 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:
  • Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Placebo: hazard ratio 1.10 (95% CI 0.85–1.41), reported p = 0.462 (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 Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Placebo and the control. The interval gives the range of effects that remain compatible with the data.

Progression-Free Survival (PFS) in the Intention-to-treat Population (Arm 2 vs Arm 1)

• 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:
  • Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Placebo: hazard ratio 1.07 (95% CI 0.90–1.29), reported p = 0.450 (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 Placebo + Carboplatin + Paclitaxel -> Placebo vs Veliparib + Carboplatin + Paclitaxel -> Placebo and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists secondary outcomes involving progression-free survival, overall survival, and disease-related symptoms. The detailed numeric secondary results include progression-free survival comparisons in the BRCA-deficient population, the homologous recombination deficiency cohort, and the intention-to-treat population for veliparib given with chemotherapy followed by placebo maintenance versus placebo throughout. Across those posted comparisons, the confidence intervals include 1 and the p-values do not fall below 0.05. Numeric results for the listed overall survival and symptom outcomes are not reported in the registry record.

These secondary progression-free survival analyses address a different sequence contrast from the primary result, which compares veliparib continued as maintenance with placebo throughout. Their pattern therefore should not be read as a direct replication of the primary contrast or as a test of maintenance treatment alone. The three populations also differ in their definitions and coverage, so results should be interpreted within each stated population.

Multiple outcomes and populations create opportunities for chance findings. Unless a multiplicity procedure is specified, repeated testing can make isolated nominally significant results less persuasive than a single prespecified primary test. The registry does not report a multiplicity adjustment or hierarchy for interpreting the secondary outcomes. For that reason, secondary findings generally carry less confirmatory weight than the primary endpoint, especially when the analysis plan and adjustment strategy are not available in the registry record. The secondary pattern adds context, but does not replace the primary comparison or establish effects on outcomes without posted numeric results.

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

• Progression-Free Survival (PFS) in the BRCA-deficient Population (Arm 2 vs Arm 1)
• Progression-Free Survival (PFS) in the Homologous Recombination Deficiency Cohort (Arm 2 vs Arm 1)
• Progression-Free Survival (PFS) in the Intention-to-treat Population (Arm 2 vs Arm 1)
• Overall Survival (OS) in the BRCA-deficient Population
• Overall Survival (OS) in the Homologous Recombination Deficiency Population
• Overall Survival (OS) in the Whole Population
• Change From Baseline in Disease Related Symptom (DRS) Score in the BRCA-mutation Population
• Change From Baseline in Disease Related Symptom (DRS) Score in the HRD Population

9. Statistical Interpretation

10. Statistical Methods Explained

Time-to-event analysis

Progression-free survival is a time-to-event outcome: the analysis considers both whether an event occurred and how long participants were followed before an event or censoring. This is useful when follow-up differs among participants and some have not had a recorded event by the cutoff. The registry defines an event as investigator-assessed disease progression under RECIST version 1.1 or death. Participants without an event by the analysis cutoff were censored at their last evaluable disease assessment.

The Kaplan-Meier method estimates the event-free experience over time while accounting for censoring. The reported median is the point at which the estimated progression-free survival curve reaches its midpoint. It is a summary of the curve, not a description of the time for every participant. The registry provides confidence intervals for the arm medians, although an upper confidence limit for the primary treatment arm is not reported in the registry record.

What does the hazard ratio describe?

The registry reports a hazard ratio from a Log Rank analysis. In general, a hazard ratio compares the event rates between groups over follow-up. A value below 1 in the stated comparison indicates a lower estimated event rate in the first-named treatment group relative to the comparator, while a value above 1 indicates a higher estimated rate. It is not a direct comparison of median times, and it does not state the probability that an individual participant will experience progression or death.

The registry reports the primary hazard ratio as 0.435, with a 95% confidence interval from 0.277 to 0.683. It also reports p < 0.001. These are sponsor-reported results, not a newly calculated estimate. The reported method is Log Rank; additional details of the model, including any adjustment or diagnostic checks, are not reported in the registry record.

Why does censoring matter?

Censoring preserves follow-up information up to the last known evaluable assessment without assuming that an event occurred afterward. The Kaplan-Meier approach relies on assumptions about the censoring process, including that censoring does not systematically conceal a different event experience in a way that distorts the estimate. This is a general statistical consideration. The registry explains the censoring date used for participants without an event, but does not report further diagnostics or sensitivity analyses for censoring.

Progression assessments are also part of the endpoint definition. The registry states that progression was determined by investigators using RECIST version 1.1, and gives criteria involving target lesions, non-target lesions, and new lesions. It does not report the assessment schedule, independent review procedures, or handling rules for missed assessments. Those details can affect when progression is recorded and how much follow-up contributes to the analysis.

How should confidence intervals and p-values be read?

In general, a confidence interval describes the range of estimates compatible with the data under the method’s assumptions, at the stated confidence level. It conveys precision as well as the direction and scale of an estimate. The primary hazard-ratio interval lies below 1, while intervals for the posted secondary comparisons of veliparib during chemotherapy followed by placebo maintenance versus placebo throughout include 1. A p-value describes how unusual the observed data, or more extreme data, would be under a specified null model. It is not the probability that the null hypothesis is true.

These quantities do not compensate for design or reporting details that are absent. The registry record does not report the trial’s original sample-size justification, the prespecified multiplicity approach, or any interim-analysis plan. The recorded statement that the analysis occurred after a protocol-specified number of events was reached does not give that number.

What are the populations and contrasts?

The primary analysis concerns the BRCA-deficient population, described as participants with a deleterious or suspected deleterious germline and/or tumor BRCA1 or BRCA2 mutation. The registry says analyses were conducted in three sequentially inclusive populations: the BRCA-mutation cohort, the homologous recombination deficiency cohort, and the intention-to-treat population. The primary comparison is veliparib with chemotherapy followed by veliparib maintenance versus placebo throughout. The posted secondary comparisons described in detail compare veliparib with chemotherapy followed by placebo maintenance versus placebo throughout.

Because the treatment sequences differ, each comparison answers a distinct question. An effect estimated for the full veliparib-throughout sequence cannot by itself identify the effect of maintenance alone. Nor should a result in a selected molecular cohort automatically be generalized to the intention-to-treat population. The registry record does not report additional subgroup definitions or adjustment methods beyond the population descriptions and analyses stated.

11. Confidence Intervals and Precision

The primary hazard-ratio estimate is 0.435, with a 95% confidence interval from 0.277 to 0.683. The interval gives a sense of the estimate’s precision under the reported analysis: its values are below 1, but the range still allows for different magnitudes of relative event-rate reduction. A confidence interval is not a guarantee that the true value lies in its bounds, nor does it capture every source of uncertainty, such as endpoint assessment or unreported design details.

The primary arm medians are 34.7 months and 22.0 months. Their reported 95% confidence intervals are 31.8 months to an upper limit not reported in the registry record, and 17.8 to 29.1 months, respectively. A missing upper limit means that the registry does not give a complete interval for that median; it should not be inferred from the other results.

The median follow-up was 28 months at the primary analysis cutoff. Follow-up duration and the precision of a time-to-event estimate are related but not identical: the number and timing of events, censoring, and the shape of follow-up all matter. The registry record does not provide a complete account of event counts for the primary analysis or the distribution of follow-up by group. Accordingly, the intervals should be interpreted alongside the reported methods and the limitations of the registry record, rather than as a complete description of uncertainty.

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 planning illustrations are based on the observed primary effect and give total event counts for a time-to-event analysis, not participants per arm. They are not a reconstruction of the trial’s original design. The calculation assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The target effect in each illustration is the stated fraction of the observed effect.

The 80% power column describes the total events that the planning calculation estimates would be needed to have an 80% chance of detecting the specified effect under those assumptions, if that effect were the true one. The 90% power column gives the corresponding event target for a 90% chance. These are planning targets, not observed counts or guarantees about the outcome of a future study. They are expressed as total events because this is a time-to-event endpoint; they should not be read as participants per arm.

The rows considering 100% of the observed effect, 75% of the observed effect, and 50% of the observed effect illustrate the consequence of planning for effects smaller than the one observed. Smaller effects are harder to distinguish from random variation, so they require more events under the stated assumptions. The figures are planning illustrations, not observed power. The original sample-size assumptions and justification for this trial are not reported in the registry record.

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 = 0.435, so ln(HR) = -0.8324.
2. d = 4 × (1.96 + 0.842)² / [ln(0.435)]² = 45.3, rounded up to 46 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 <- 0.435                              # 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(2470585)
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 = 46 (80% power), 61 (90% power)
 75% of observed effect: events = 81 (80% power), 108 (90% power)
 50% of observed effect: events = 182 (80% power), 243 (90% power)
Simulated power with 46 events: 0.789 (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 118 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 1,140 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 <- 22          # control-arm median, months (converted from the posted unit)
hr          <- 0.435          # 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: 108         p (1-sided)   0.0001
 Events: 21        ~HR at bound   0.1937
  Month: 22    P(Cross) if HR=1   0.0001
            P(Cross) if HR=0.44   0.0338
  IA 2: 67%                   Z   2.5114
     N: 118         p (1-sided)   0.0060
 Events: 41        ~HR at bound   0.4559
  Month: 35    P(Cross) if HR=1   0.0060
            P(Cross) if HR=0.44   0.5603
      Final                   Z   1.9930
     N: 118         p (1-sided)   0.0231
 Events: 62        ~HR at bound   0.6011
  Month: 60    P(Cross) if HR=1   0.0250
            P(Cross) if HR=0.44   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

The hazard ratio, confidence interval, p-value, and Log Rank method for the primary comparison are reported by the sponsor in the registry record. The analysis shown is therefore the sponsor’s reported analysis, rather than an independently recalculated hazard ratio from arm-level summary statistics. The registry record does not flag a discrepancy for this result.

That distinction matters because a hazard ratio cannot be recovered reliably from the reported medians alone. The time-to-event comparison depends on event timing and censoring across follow-up, information not contained in a pair of median estimates. The sponsor’s reported hazard ratio is the appropriate estimate to cite for this primary analysis; the median progression-free survival values provide complementary summaries of the arm-specific curves.

The record identifies Log Rank as the method and reports a 95% confidence interval and p-value, but does not provide all model details or the full analysis dataset. No independent recalculation should be treated as replacing the sponsor’s analysis. Further information about the sponsor’s detailed statistical procedures is not reported in the registry record.

16. Multiplicity, Missing Data and Analysis Populations

The registry names a primary progression-free survival analysis in the BRCA-deficient population and lists several secondary outcomes and populations. It does not report whether testing across these endpoints, populations, and treatment comparisons was controlled using a multiplicity adjustment, a testing hierarchy, or another prespecified procedure. Without that information, the degree to which secondary p-values were protected against multiple testing is not reported in the registry record. This limits confirmatory interpretation of secondary findings.

For the primary endpoint, the registry states that participants without progression or death before the cutoff were censored at their last evaluable disease assessment. It also describes progression and death as events. This is the reported handling relevant to the endpoint definition. Further missing-data procedures, sensitivity analyses, and rules for missed or unevaluable assessments are not reported in the registry record. The record also does not report whether additional methods were used to assess the impact of censoring.

The analysis populations are described as sequentially inclusive, beginning with the BRCA-mutation cohort and then including the homologous recombination deficiency cohort and intention-to-treat population. The primary result is for the BRCA-deficient population. The registry’s population description refers to deleterious or suspected deleterious germline and/or tumor BRCA1 or BRCA2 mutations. Other details of population construction, including handling of unknown or incomplete status, are not reported in the registry record.

The registry states that the progression-free survival analysis occurred when a protocol-specified number of events was reached, but does not give that event target. The original sample-size justification, interim-analysis plan, allocation-generation details, and analysis diagnostics are also not reported in the registry record. These omissions do not change the numbers reported, but they limit how fully the analysis plan and its safeguards can be evaluated.

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: small intestinal obstruction (5.1%), febrile neutropenia (5.0%), neutropenia (4.2%), malignant neoplasm progression (3.7%), anaemia (3.7%).

The safety summary reports serious adverse events for each treatment group using the number at risk and the number affected. The placebo-throughout group has 143 affected among 375 at risk; the veliparib-with-chemotherapy then placebo group has 130 among 383 at risk; and the veliparib-throughout group has 146 among 382 at risk. These are group-level registry counts. They do not by themselves establish that a treatment caused an event or allow an assessment of event severity beyond the serious-adverse-event classification.

The registry also lists the leading serious adverse event terms and their reported percentages: small intestinal obstruction, febrile neutropenia, neutropenia, malignant neoplasm progression, anaemia, and nausea. The percentages are registry-reported; their denominator details are not reported in the registry record. The record separately reports deaths by group, but does not clarify in this summary whether these are all-cause deaths or deaths attributed to treatment. Safety comparisons require care because the posted counts do not describe timing, exposure duration, event-level overlap, or causal attribution.

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

The posted primary result describes a statistically significant progression-free survival comparison in the BRCA-deficient population between veliparib given with carboplatin and paclitaxel and continued as maintenance, and placebo given with chemotherapy and continued as placebo. The reported hazard ratio is 0.435, its 95% confidence interval is 0.277 to 0.683, and the reported p-value is p < 0.001. The arm-specific median estimates provide another description of the progression-free survival experience.

This result does not identify the separate contribution of veliparib during chemotherapy and veliparib during maintenance, because the primary contrast changes both treatment phases. The trial includes a different treatment sequence with veliparib during chemotherapy followed by placebo, but that is a distinct comparison. The primary result also pertains to the stated BRCA-deficient population, and should not automatically be generalized to every participant or population.

A group-level estimate does not predict what will happen to an individual. It does not show that every participant experienced the same outcome, and it does not settle questions about overall survival, symptoms, or other outcomes for which numeric results are not reported in the registry record. The reported follow-up and cutoff define the period of data used for this analysis.

The registry status is terminated, but the reason for termination is not reported in the registry record. That status alone does not indicate why the study stopped or how termination affected the analyses. The record also does not report multiplicity procedures, the original sample-size justification, or full details of missing-data sensitivity analyses. Those gaps matter when judging the strength and scope of conclusions beyond the reported primary comparison.

19. Why This Trial Matters Statistically

This trial illustrates why treatment sequences and analysis populations must be read precisely. The primary estimate is not simply a comparison of veliparib versus placebo: it contrasts veliparib during chemotherapy followed by veliparib maintenance with placebo during chemotherapy followed by placebo maintenance. The registry also includes a sequence that stops veliparib after chemotherapy, which answers a different question. Keeping those contrasts separate prevents an overall sequence effect from being attributed to one treatment phase alone.

The endpoint also demonstrates the value and limits of time-to-event methods. Progression-free survival can use follow-up up to progression, death, or censoring, rather than requiring every participant to have the same observation duration. Kaplan-Meier estimates and a hazard ratio summarize different aspects of that experience, and neither replaces the other. Their interpretation depends on event definitions, assessments, and censoring.

Finally, the record shows why a significant primary result should be considered alongside precision, population, secondary analyses, and reporting gaps. The primary hazard-ratio interval is below 1, while the posted secondary comparisons concern another treatment sequence and have intervals that include 1. The absence of reported multiplicity details and some analysis-plan information limits assessment of the wider set of results. Statistical interpretation is strongest when the treatment contrast, endpoint, population, uncertainty, and limits of the record are considered together.

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.

Several limitations are specific to the registry record. The trial is marked terminated, but the reason is not reported in the registry record. The primary analysis is tied to a cutoff of 03 May 2019 and median follow-up of 28 months; longer-term results are not reported in the registry record. The upper confidence limit for the primary treatment arm’s median is also not reported.

The primary analysis concerns the BRCA-deficient population, while the registry describes additional, sequentially inclusive populations. The available summary does not give further details on how molecular status was established or how incomplete status was handled. The primary treatment contrast includes both concurrent and maintenance veliparib, so it does not isolate the effect of either phase.

The registry does not report the original sample-size justification, a multiplicity strategy, the interim-analysis plan, or detailed sensitivity analyses for censoring and missing assessments. It also does not provide the assessment schedule or independent review details for progression. These omissions make it harder to assess all assumptions behind the reported time-to-event results.

Safety information is presented as group-level counts and selected event terms. The record does not provide the full context needed to interpret event timing, exposure, attribution, or overlap among event categories. Numeric results for listed overall survival and disease-related symptom outcomes are not reported in the registry record. These limits call for a focused interpretation of the posted primary result rather than broad conclusions across all outcomes.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: Veliparib With Carboplatin and Paclitaxel and as Continuation Maintenance Therapy in Adults With Newly Diagnosed Stage III or IV, High-grade Serous, Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer — NCT02470585.
• Registry record last updated 2024-10-26; status last verified 2024-10.
• Results first posted 2020-09-14.
• 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 describes a randomized, double-masked, parallel-group phase 3 study of veliparib with carboplatin and paclitaxel, with or without continuation maintenance, in adults with previously untreated stage III or IV high-grade serous epithelial ovarian, fallopian tube, or primary peritoneal cancer. Enrollment is listed as 1,140, and the study status is terminated.

The primary endpoint was progression-free survival in the BRCA-deficient population. The primary comparison was veliparib with chemotherapy followed by veliparib maintenance versus placebo with chemotherapy followed by placebo. The reported hazard ratio is 0.435, with a 95% confidence interval from 0.277 to 0.683 and p < 0.001. Median follow-up was 28 months at the primary analysis cutoff of 03 May 2019.

The registry also lists secondary progression-free survival, overall survival, and symptom outcomes. Important details, including the reason for termination, multiplicity procedures, and the original design assumptions, are not reported in the registry record.