Advertisement

Progression-free survival by investigator versus blinded independent central review in newly diagnosed patients with high-grade serous ovarian cancer: Analysis of the VELIA/GOG-3005 trial

Open AccessPublished:June 07, 2021DOI:https://doi.org/10.1016/j.ygyno.2021.05.031

      Highlights

      • Addition of veliparib to chemotherapy and as maintenance improved (INV)-assessed progression-free survival (PFS).
      • Hazard ratios for PFS by blinded independent central review (BICR) were consistent with hazard ratios for PFS per INV.
      • Analyses of overall concordance between INV and BICR assessments indicated a lack of investigator bias.

      Abstract

      Objective

      In the phase 3 VELIA/GOG-3005 trial, veliparib added to carboplatin-paclitaxel and continued as maintenance improved progression-free survival (PFS) compared to carboplatin-paclitaxel alone in patients with newly diagnosed ovarian carcinoma. Primary analysis of PFS was by investigator (INV) assessment, with a supplemental analysis of PFS by blinded independent central review (BICR).

      Methods

      Patients received veliparib or placebo with carboplatin-paclitaxel (6 cycles) and as maintenance (30 additional cycles). The primary analysis compared PFS in the veliparib-throughout arm to the carboplatin-paclitaxel only arm in the BRCA mutation (BRCAm), homologous recombination deficiency (HRD), and intention-to-treat (ITT) populations. Exploratory analyses of PFS in BRCA wildtype (BRCAwt), homologous recombination proficient (HRP), and HRD + BRCAwt populations were also performed. PFS per BICR and overall concordance rates between INV and BICR assessments were analyzed.

      Results

      Hazard ratios for PFS by INV and BICR were consistent in each of the primary analysis and exploratory populations. In the ITT population, median PFS per INV was 23.5 months in the veliparib-throughout arm versus 17.3 months in the control arm (hazard ratio [HR] 0.683, 95% confidence interval [CI] 0.562–0.831; P < 0.001). Median PFS by BICR was 29.3 months versus 19.2 months (HR 0.687, 95% CI 0.504–0.806). In the ITT population, the overall concordance rates between INV and BICR were 78% and 75% for the veliparib-throughout and control arms, respectively.

      Conclusions

      Hazard ratios for PFS per BICR and per INV were consistent, with no suggestion of investigator bias. These findings support the reliability of PFS by INV in ovarian cancer trials.

      Keywords

      1. Introduction

      VELIA/GOG-3005 is a phase 3, randomized, placebo-controlled, global study designed to evaluate the efficacy and safety of the poly (ADP-ribose) polymerase (PARP) inhibitor veliparib added to carboplatin-paclitaxel and continued as maintenance in patients with newly diagnosed, high-grade serous ovarian carcinoma. Patients were eligible regardless of biomarker status or timing or outcome of cytoreductive surgery and were enrolled before response to initial platinum-based chemotherapy was known. The primary endpoint was progression-free survival (PFS) per investigator (INV) assessment, evaluated in the BRCA mutated (BRCAm), homologous recombination deficient (HRD), and ITT populations. Addition of veliparib to chemotherapy and continued as maintenance resulted in a statistically significant and clinically meaningful improvement in PFS per INV in each of these populations [
      • Coleman R.L.
      • et al.
      Veliparib with first-line chemotherapy and as maintenance therapy in ovarian cancer.
      ].
      Imaging-based endpoints such as PFS are commonly used in oncology trials. The Response Evaluation Criteria In Solid Tumors (RECIST) version 1.1 guidelines provide direction on assignment of treatment response based on imaging in trials enrolling patients with solid tumors [
      • Eisenhauer E.A.
      • et al.
      New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).
      ]. However, there can still be disagreement in response between readers, and this has the potential to impact the represented treatment effect determined in a trial overall. In ovarian cancer, disseminated peritoneal spread and ascites are common [
      • Forstner R.
      • Meissnitzer M.
      • Cunha T.M.
      Update on imaging of ovarian cancer.
      ]; this may impact the reliability of imaging assessments and lead to further disagreement between readers. Potential investigator bias is generally mitigated by use of a placebo control, although concerns may persist that the toxicity profile of the investigational agent could be unblinding and influence the investigator's assessment. Presence of bias in the investigator assessment can be evaluated through use of blinded independent central review (BICR) of the imaging-determined endpoints. This exercise is undertaken both to protect against possible investigator bias and to comply with regulatory guidelines. However, a meta-analysis of 27 randomized phase 3 trials (including 5 placebo-controlled trials) showed a strong correlation between treatment effect based on INV versus BICR assessment [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ]. This suggests that it may be possible to avoid the cost and time of a full BICR review of data and either omit, audit, or randomize a subset of patients for BICR provided this is accepted by regulatory authorities.
      Here, we report the results of prespecified analyses of PFS per BICR and compare with the results of PFS per INV in the VELIA/GOG-3005 trial.

      2. Methods

      2.1 Patients and trial design

      The VELIA/GOG-3005 study design and eligibility were previously published [
      • Coleman R.L.
      • et al.
      Veliparib with first-line chemotherapy and as maintenance therapy in ovarian cancer.
      ]. Briefly, eligible patients were ≥ 18 years of age with histologically confirmed stage III-IV high-grade serous ovarian cancer, Eastern Cooperative Oncology Group (ECOG) performance status of 0–2, and adequate hematologic, renal, and hepatic function. Patients had no prior radiotherapy or chemotherapy to any portion of the abdominal cavity or pelvis. The BRCAm cohort consisted of patients with deleterious or suspected deleterious germline or somatic mutations in BRCA1 or BRCA2, as determined by the Myriad BRACAnalysis CDx or myChoice HRD CDx assay. The HRD cohort consisted of all patients in the BRCAm cohort plus all patients with HRD tumors according to the myChoice assay (tumors were considered HRD if score was ≥33).
      VELIA/GOG-3005 was an international, phase 3, multicenter, double-blinded, placebo-controlled trial (Fig. 1). Patients received 6 cycles of carboplatin and paclitaxel with either placebo or veliparib at 150 mg BID administered continuously. After the 6 cycles of combination therapy, patients received an additional 30 cycles of maintenance therapy with either placebo or veliparib 300 mg BID increasing to 400 mg BID if tolerated. Paclitaxel could be administered on a weekly or every-3-week schedule, and patients could have primary surgery or interval cytoreductive surgery after cycle 3 of study treatment; selection was at the discretion of the investigator. Patients were randomly assigned in a 1:1:1 ratio to the control group (carboplatin-paclitaxel plus placebo followed by placebo maintenance), the veliparib combination-only group (carboplatin-paclitaxel plus veliparib followed by placebo maintenance), or the veliparib-throughout group (carboplatin-paclitaxel plus veliparib followed by veliparib maintenance). Randomization was stratified by the stage of disease, timing of surgery and residual disease post primary surgery, choice of paclitaxel regimen, region of the world, and germline BRCAm status.
      Fig. 1
      Fig. 1VELIA trial study design.
      Veliparib was administered continuously during both combination and maintenance phases. AUC, area under the curve; BID, twice daily; ECOG PS, Eastern Cooperative Oncology Group performance status; HGSOC, high-grade serous ovarian cancer; R, randomization; Q3W, every 3 weeks; QW, every week.

      2.2 Endpoints and assessments

      The primary endpoint was PFS per INV in the veliparib-throughout arm compared with the control arm, analyzed sequentially in the BRCAm population, the HRD population, and the ITT population. PFS for the veliparib-combination only arm was compared with that of the control arm in secondary analyses. PFS was defined as the number of days from randomization to the date the patient experienced an event of disease progression, according to RECIST version 1.1, or to the date of death (all causes of mortality) if disease progression was not reached. Radiologic tumor assessments were performed at baseline, every 9 weeks during the combination phase, then every 12 weeks for 2 years, then every 6 months for 3 years, then annually until the occurrence of imaging-based progression. Supplemental analysis of PFS by BICR was pre-specified. Scans were prospectively submitted to an independent central review committee for blinded assessment per RECIST version 1.1. Scans were reviewed by two radiologists. In cases of disagreement, a third radiologist served as an adjudicator and determined which of the assessments from the two original radiologists would be used.

      2.3 Statistical analysis

      PFS was compared between the veliparib throughout and control arms in the primary analysis populations (BRCAm, HRD, and ITT). Exploratory analyses of PFS in the BRCA wild type (BRCAwt) population, the HRP population, and in the population defined by both HRD and BRCAwt status (HRD + BRCAwt), were also performed.
      Distributions of PFS were estimated using the Kaplan-Meier (KM) method, and HRs were estimated using a stratified Cox model for the primary analysis populations and an unstratified Cox model for the exploratory populations. The analysis was stratified by residual disease and stage of disease for the BRCAm and HRD populations, and by residual disease, stage of disease, paclitaxel dosing regimen, and BRCA-deficiency status for the ITT population. The same statistical methods were applied to the analyses of PFS by INV and by BICR, although PFS by BICR was not tested in an alpha-controlled manner in either the primary analysis populations or the exploratory populations. Overall concordance rate was defined as the proportion of patients with progressive disease (PD) (regardless of date of event) by both INV and BICR or non-PD by both INV and BICR, and was calculated as:
      number of patients withPDbybothINVand BICR+number of patients with nonPDbybothINVand BICRtotal number of the patients


      The early discrepancy rate (EDR) was defined as the frequency at which INV reported PD earlier than BICR as a proportion of the total number of INV PD events, calculated as [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ]:
      number of patients withPDbyINVbutnot BICR+number of patients withINVPDdate earlier than BICRPDdatenumber of patients withPDbybothINVand BICR+number of patients withPDbyINVbutnot BICR


      The late discrepancy rate (LDR) was defined as the frequency at which INV reported PD later than BICR as a proportion of the total number of discrepancies, calculated as [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ]:
      number of patients withPDbyBICRbutnotINV+number of patients withINVPDdate later than BICRPDdatenumber of patients withPDbyINVbutnot BICR+number of patients withPDbyBICRbutnotINV+number of patients withINVPDdate later than BICRPDdate+number of patients withINVPDdate earlier than BICRPDdate


      Discrepancy index was calculated as [
      • Dello Russo C.
      • Cappoli N.
      • Navarra P.
      A comparison between the assessments of progression-free survival by local investigators versus blinded independent central reviews in phase III oncology trials.
      ]:
      HRperINVHRperBICR


      Only PD by imaging was considered for both INV and BICR calculation of the concordance and discrepancy rates. For calculation of EDR and LDR, PD dates for INV versus BICR were considered earlier/later if they differed by more than 7 days. Death information was used for the derivation of the PFS endpoint for both INV and BICR, even though the ICR did not obtain this information.

      3. Results

      3.1 Patients

      Overall, 1140 patients underwent randomization from July 2015 to July 2017. A total of 298 patients (26%) were included in the BRCAm cohort, and 627 patients (55%) were included in the HRD cohort (Fig. 2). Key demographic and clinical characteristics of the patients and safety data were previously reported [
      • Coleman R.L.
      • et al.
      Veliparib with first-line chemotherapy and as maintenance therapy in ovarian cancer.
      ]. Radiologic scans were performed at 2595 and 2416 post-baseline timepoints in the veliparib-throughout and control arms, respectively. Scans were available for BICR at 2569 (99%) and 2384 (99%) of those timepoints, respectively.
      Fig. 2
      Fig. 2HRD and BRCA status in the ITT population.
      The BRCAm population is inclusive of patients with either germline or tissue BRCA mutations. The subset of patients in the HRD population with BRCAwt status comprised the HRD + BRCAwt population. ITT, intention-to-treat; HRD, homologous recombination deficiency; HRP, homologous recombination proficient; BRCAm, BRCA mutation; BRCAwt, BRCA wild type.

      3.2 PFS analyses

      Per INV-assessment, in the ITT population there were 191 PFS events in the veliparib-throughout arm and 237 in the control arm. Veliparib-throughout demonstrated a statistically significant effect compared to the control. Median PFS was 23.5 months in the veliparib-throughout arm compared to 17.3 months in the control arm (HR 0.683, 95% confidence interval [CI], 0.562–0.831; P < 0.001) (Table 1; Fig. 3). The results of the analysis of PFS per BICR assessment were consistent with the primary analysis of PFS per INV. Per BICR, in the ITT population there were 130 PFS events in the veliparib-throughout arm and 171 events in the control arm. Median PFS was 29.3 months in the veliparib-throughout arm and 19.2 months in the control arm (HR 0.637, 95% CI 0.504–0.806).
      Table 1PFS assessment by INV and BICR.
      INV assessmentBICR assessment
      VEL-throughoutControlVEL-throughoutControl
      ITTEvents, n/N191/382237/375130/382171/375
      Median PFS, mo (95% CI)23.5 (19.3–26.3)17.3 (15.1–19.1)29.3 (24.8–37.5)19.2 (16.7–23.5)
      Stratified HR (95% CI); p value0.683 (0.562–0.831); p < 0.0010.637 (0.504–0.806)
      BRCAmEvents, n/N34/10851/9227/10838/92
      Median PFS, mo (95% CI)34.7 (31.8–NR)22.0 (17.8–29.1)NR (31.9–NR)28.8 (18.0–NR)
      Stratified HR (95% CI); p value0.435 (0.277–0.683); p < 0.0010.436 (0.261–0.728)
      BRCAwtEvents, n/N142/245171/25493/245124/254
      Median PFS, mo (95% CI)18.2 (15.9–21.7)15.1 (12.6–17.5)23.6 (18.2–29.3)17.1 (15.1–20.9)
      Unstratified HR (95% CI)0.798 (0.639–0.997)0.734 (0.561–0.961)
      HRD + BRCAwtEvents, n/N53/10673/11539/10652/115
      Median PFS, mo (95% CI)22.9 (18.2–37.5)19.8 (16.7–22.2)28.7 (18.2–NR)20.9 (17.1–26.0)
      Unstratified HR (95% CI)0.743 (0.521–1.059)0.779 (0.513–1.182)
      HRDEvents, n/N87/214124/20766/21490/207
      Median PFS, mo (95% CI)31.9 (25.8–38.0)20.5 (17.8–22.8)34.7 (29.0–NR)22.7 (18.4–30.0)
      Stratified HR (95% CI); p value0.572 (0.433–0.756); p < 0.0010.597 (0.433–0.825)
      HRPEvents, n/N80/12589/12448/12567/124
      Median PFS, mo (95% CI)15.0 (12.7–18.0)11.5 (10.1–14.9)21.1 (17.0–26.5)13.1 (10.0–15.9)
      Unstratified HR (95% CI)0.805 (0.595–1.090)0.647 (0.446–0.938)
      BICR, blinded independent central review; BRCAm, BRCA mutation; BRCAwt, BRCA wild type; CI, confidence interval; HR, hazard ratio; HRD, homologous recombination deficiency; ITT, intention-to-treat; INV, investigator-assessed; NR, not reached; PFS, progression-free survival; VEL, veliparib; HRP, homologous recombination proficiency.
      Fig. 3
      Fig. 3KM estimates of PFS in the veliparib-throughout arm and control arm in the ITT population.
      Distributions were estimated by means of the KM method in the ITT population per INV (Panel A) and per BICR (Panel B), with the veliparib-throughout group compared with the control group (primary end point). PFS was compared between the arms by the stratified log-rank test. HRs were estimated by a Cox model with stratification according to the same factors as were used in the log-rank test. The dashed line indicates the median, and tick marks indicate censored data.
      BICR, blinded independent central review; CI, confidence interval; HR, hazard ratio; INV, investigator-assessed; ITT, intention-to-treat; PFS, progression-free survival; VEL, veliparib.
      The veliparib-throughout arm also demonstrated statistically significant improvements in PFS per INV-assessment in the BRCAm and HRD primary analysis populations. Results of the PFS analysis per BICR were consistent (Table 1, Fig. 4). Exploratory analyses of PFS per INV in the BRCAwt, HRD + BRCAwt, and HRP populations favored the veliparib-throughout arm. When PFS per BICR was analyzed in these populations, the results were consistent with PFS per INV (Table 1, Fig. 4).
      Fig. 4
      Fig. 4KM estimates of PFS in the veliparib-throughout arm and control arm in the A) BRCAm, B) BRCAwt, C) HRD + BRCAwt, D) HRD, and E) HRP populations.
      Distributions were estimated by means of the KM method. PFS was compared between the trial treatment groups by the stratified log-rank test. HRs were estimated by a Cox model with stratification according to the same factors as were used in the log-rank test. The dashed line indicates the median, and tick marks indicate censored data. BICR, blinded independent central review; BRCAm, BRCA mutation; BRCAwt, BRCA wild type; CI, confidence interval; HR, hazard ratio; HRD, homologous recombination deficiency; HRP, homologous recombination proficient.
      PFS per INV for the veliparib-combination only arm was not statistically different from the control arm in the BRCAm, HRD, or ITT populations [
      • Coleman R.L.
      • et al.
      Veliparib with first-line chemotherapy and as maintenance therapy in ovarian cancer.
      ]. Results per BICR were consistent (data not shown).
      Table 2Concordance rates between PFS per INV and BICR in the ITT population.
      Veliparib-throughout arm n/N (%)Control arm n/N (%)
      PD by both INV and BICR122/382 (32)161/375 (43)
       INV PD date = BICR PD date76/382 (20)96/375 (26)
       INV PD date later than BICR PD date45/382 (12)59/375 (16)
       INV PD date earlier than BICR PD date1/382 (<1)6/375 (2)
      PD by INV but not BICR69/382 (18)78/375 (21)
      PD by BICR but not INV14/382 (4)15/375 (4)
      Non-PD by both INV and BICR177/382 (46)121/375 (32)
      Concordance rate299/382 (78)282/375 (75)
      Early discrepancy rate
      The frequency at which INV reported PD earlier than BICR as a proportion of the total number of PD events by INV [4].
      70/191 (37)84/239 (35)
      Late discrepancy rate
      The frequency at which INV reported PD later than BICR as a proportion of the total number of discrepancies [4].
      59/129 (46)74/158 (47)
      Discrepancy index1.072
      BICR, blinded independent central review; INV, investigator-assessed; ITT, intention-to-treat; PD, progressive disease; PFS, progression-free survival.
      a The frequency at which INV reported PD earlier than BICR as a proportion of the total number of PD events by INV [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ].
      b The frequency at which INV reported PD later than BICR as a proportion of the total number of discrepancies [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ].

      3.3 Concordance

      In the ITT population, the overall rate of concordance between INV and BICR assessment of PFS was 78% for the veliparib-throughout arm and 75% for the control arm. Differences between arms in EDR and LDR were small (Table 2). Concordance rates between INV and BICR PFS for the veliparib-throughout and control arms ranged from 68% to 85% in the primary and exploratory populations analyzed (Table S1). Discrepancy index for the ITT population was 1.072 (Table 2). In primary and exploratory populations, discrepancy index ranged from 0.954–1.244 (Table S1).

      4. Discussion

      In the primary analyses of the VELIA/GOG-3005 trial, the addition of veliparib to carboplatin-paclitaxel and continuation as maintenance therapy resulted in statistically significant improvements in PFS per INV assessment in patients with newly diagnosed ovarian carcinoma in the BRCAm, HRD, and ITT populations. In supplementary analyses, hazard ratios for PFS per BICR were consistent with the primary analyses in all populations. In the ITT population, overall concordance rates between INV and BICR were high (≥75%) and comparable between treatment arms. Differences between treatment arms in EDR and LDR were small. There was no investigator bias in favor of the veliparib-throughout arm.
      In studies of PARP inhibitors for the treatment of ovarian cancer, the observed magnitude of benefit has been larger in populations of patients with tumors that are BRCAm or HRD [
      • Mirza M.R.
      • et al.
      Niraparib maintenance therapy in platinum-sensitive, recurrent ovarian cancer.
      ,
      • Gonzalez-Martin A.
      • et al.
      Niraparib in patients with newly diagnosed advanced ovarian cancer.
      ,
      • Coleman R.L.
      • et al.
      Rucaparib maintenance treatment for recurrent ovarian carcinoma after response to platinum therapy (ARIEL3): a randomised, double-blind, placebo-controlled, phase 3 trial.
      ]. Exploratory analyses have also suggested benefit in populations of patients not carrying one of these biomarkers, although there appears to be a smaller magnitude of benefit. Consistent with these previous reports, exploratory analyses of PFS per investigator for the veliparib-throughout versus control arm in the BRCAwt and non-HRD populations in VELIA favored the veliparib-throughout arm. Importantly, PFS per BICR was consistent in favoring the veliparib-throughout arm in these populations despite the smaller magnitude of benefit observed.
      While the hazard ratios were comparable for PFS per INV and per BICR in this analysis, in all populations and in both treatment arms the median PFS per BICR was consistently longer compared to the median PFS per INV. The phenomenon of longer PFS per BICR compared to per INV has been previously described [
      • Stone A.
      • et al.
      Exaggeration of PFS by blinded, independent, central review (BICR).
      ]. When the INV but not BICR determines progressive disease for a patient, protocol-defined scanning is often terminated. With no additional scans available, data for the patient is then censored for BICR analysis at the time of progression per INV. This handling of data for patients with INV-only progression may result in exaggerated median PFS per BICR. This observation may be important to consider when interpreting median PFS reported among studies that differ in their use of INV-assessed or BICR-assessed PFS as the primary endpoint, and highlights the importance of the hazard ratio for assessing the treatment effect.
      Discrepancy index provides a measure of whether the INV or BICR had a more optimistic assessment of the treatment effect [
      • Dello Russo C.
      • Cappoli N.
      • Navarra P.
      A comparison between the assessments of progression-free survival by local investigators versus blinded independent central reviews in phase III oncology trials.
      ]. The discrepancy index tended toward 1 in the primary and exploratory populations evaluated in the current study, indicating that overall there was a lack of bias of the INV. In the ITT population, the discrepancy index was slightly greater than 1, consistent with a trend toward more optimistic BICR compared to INV among double-blind trials reported in a recent meta-analysis [
      • Dello Russo C.
      • Cappoli N.
      • Navarra P.
      A comparison between the assessments of progression-free survival by local investigators versus blinded independent central reviews in phase III oncology trials.
      ]. The discrepancy indices were under 1 in the biomarker positive populations evaluated in the current study and over 1 in the biomarker positive populations. This may suggest a trend toward greater INV optimism in the evaluation of the biomarker positive populations. Treatment assignment was blinded in the study but the investigators had access to the results of the biomarker analyses. Regardless, any investigator bias in the treatment effect within biomarker positive populations was miniminal as discrepancy indices were > 0.95 in each of the biomarker positive populations.
      Consistency between PFS per INV and BICR has been noted in large trials enrolling patients with ovarian cancer specifically and enrolling patients with solid tumors more generally [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ,
      • Floquet A.
      • et al.
      Progression-free survival by local investigator versus independent central review: comparative analysis of the AGO-OVAR16 trial.
      ,
      • Burger R.A.
      • et al.
      Independent radiologic review of the gynecologic oncology group study 0218, a phase III trial of bevacizumab in the primary treatment of advanced epithelial ovarian, primary peritoneal, or fallopian tube cancer.
      ]. BICR adds substantially to the financial costs associated with a trial and increases the workload of participating sites. Audit strategies have been proposed that may eliminate the need for a full case BICR [
      • Amit O.
      • et al.
      Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
      ]. The consistency in observed treatment effect between INV and BICR results in the current study further support the reliability of PFS by INV in trials enrolling patients with ovarian cancer, and suggest that employing audit strategies may be appropriate to reduce the burden associated with BICR.

      Data sharing statement

      AbbVie is committed to responsible data sharing regarding the clinical trials we sponsor. This includes access to anonymized, individual and trial-level data (analysis data sets), as well as other information (e.g., protocols and Clinical Study Reports), as long as the trials are not part of an ongoing or planned regulatory submission. This includes requests for clinical trial data for unlicensed products and indications. This clinical trial data can be requested by any qualified researchers who engage in rigorous, independent scientific research, and will be provided following review and approval of a research proposal and Statistical Analysis Plan (SAP) and execution of a Data Sharing Agreement (DSA). Data requests can be submitted at any time and the data will be accessible for 12 months, with possible extensions considered. For more information on the process, or to submit a request, visit the following link: https://www.abbvie.com/our–science/clinical–trials/clinical–trials–data–and–information–sharing/data–and–information–sharing–with–qualified–researchers.html.

      Funding

      AbbVie contributed to the analysis and interpretation of the data and participated in the writing, review and approval of the manuscript All authors had full access to all of the data and vouch for its integrity and completeness and were responsible for writing the manuscript, with editorial assistance funded by AbbVie. All authors reviewed draft and final versions of the manuscript before submission and approved the final version. The corresponding author had full access to all of the data and had final responsibility for the decision to submit for publication.

      Declaration of interests

      C Aghajanian: Consulting/advisory role: Clovis Oncology, Tesaro, Mateon Therapeutics, Immunogen, Cerulean Pharma, Eisai/Merck, and Mersana Therapeutics. Research funding (to institution) from Genentech/Roche, AbbVie, Clovis Oncology, and AstraZeneca.
      MA Bookman: Member, international protocol steering committee: AbbVie GOG3005. Member, protocol steering committees: Genentech-Roche and Aravive. Participant, advisory boards (Clovis, Tesaro, Merck, and AstraZeneca).
      GF Fleming: Author fees: UpToDate. Speaker: Curio Science. Research funding (to institution): Corcept Therapeutics, Genentech, Tesaro, Syndax, Forty Seven, Iovance, Syros, Astex, Merck, Incyte, Leap Therapeutics, Hoffman La Roche, Sanofi, and Sermonix. Advisory board: GSK. Investigator/Steering committee: VELIA (uncompensated): AbbVie.
      EM Swisher: Research funding to institution: Tesaro and Clovis Oncology.
      KD Steffensen: Consulting/advisory role: AbbVie.
      M Friedlander: Consulting/advisory role: AstraZeneca, MSD, AbbVie, Lilly, Takeda, Novartis, GSK. Speakers' bureau: AstraZeneca. Honoraria: AstraZeneca, MSD, Lilly, Takeda,GSK,Novartis. Research funding: BeiGene, Astra Zeneca, Novartis.
      A Okamoto: Consulting/advisory role: AstraZeneca and Chugai. Honoraria: AstraZeneca, MSD and Chugai. Grant/Research funding: Kaken, Mochida, Kissei and Pfizer.
      C Gunderson Jackson: Consulting/advisory role: Clovis Oncology, Tesaro/GSK, LEAP. Research funding to institution: Lilly, Genentech, Clovis Oncology.
      RL Coleman: Consulting/advisory role: Clovis Oncology, Genentech/Roche, Esperance, NCCN, AstraZeneca/MedImmune, Genmab, GamaMabs Pharma, Tesaro, OncoMed, Sotio, Oncolytics and AbbVie/Stemcentrx. Travel/accommodations/expenses: Merck, AstraZeneca/MedImmune, Array Biopharma, Clovis, Roche/Genentech, Research to Practice, GOG, Sotio, and Vaniam Group. Research funding: AstraZeneca/MedImmune, Esperance, OncoMed, Array, Clovis, Johnson & Johnson, Merck, Roche/Genentech, and Abbott/AbbVie.
      DM Sullivan, CK Ratajczak: AbbVie employees and may own stock.

      Financial support

      AbbVie provided financial support for the study and participated in the design, study conduct, analysis and interpretation of data as well as the writing, review, and approval of this manuscript. No honoraria or payments were made for authorship.

      Acknowledgments

      We thank the patients and their families, investigators, study coordinators, and support staff; Matthias Fällström of AbbVie for project management. AbbVie provided financial support for the study and participated in the design, study conduct, and data analysis and interpretation. Fatemeh Atashi, PhD, medical writer at AbbVie Inc., provided writing and editorial assistance based on specific direction from the authors.

      Appendix A. Supplementary data

      The following are the supplementary data related to this article.

      References

        • Coleman R.L.
        • et al.
        Veliparib with first-line chemotherapy and as maintenance therapy in ovarian cancer.
        N. Engl. J. Med. 2019; 381: 2403-2415
        • Eisenhauer E.A.
        • et al.
        New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).
        Eur. J. Cancer. 2009; 45: 228-247
        • Forstner R.
        • Meissnitzer M.
        • Cunha T.M.
        Update on imaging of ovarian cancer.
        Curr. Radiol. Rep. 2016; 4: 31
        • Amit O.
        • et al.
        Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.
        Eur. J. Cancer. 2011; 47: 1772-1778
        • Dello Russo C.
        • Cappoli N.
        • Navarra P.
        A comparison between the assessments of progression-free survival by local investigators versus blinded independent central reviews in phase III oncology trials.
        Eur. J. Clin. Pharmacol. 2020; 76: 1083-1092
        • Mirza M.R.
        • et al.
        Niraparib maintenance therapy in platinum-sensitive, recurrent ovarian cancer.
        N. Engl. J. Med. 2016; 375: 2154-2164
        • Gonzalez-Martin A.
        • et al.
        Niraparib in patients with newly diagnosed advanced ovarian cancer.
        N. Engl. J. Med. 2019; 381: 2391-2402
        • Coleman R.L.
        • et al.
        Rucaparib maintenance treatment for recurrent ovarian carcinoma after response to platinum therapy (ARIEL3): a randomised, double-blind, placebo-controlled, phase 3 trial.
        Lancet. 2017; 390: 1949-1961
        • Stone A.
        • et al.
        Exaggeration of PFS by blinded, independent, central review (BICR).
        Ann. Oncol. 2019; 30: 332-338
        • Floquet A.
        • et al.
        Progression-free survival by local investigator versus independent central review: comparative analysis of the AGO-OVAR16 trial.
        Gynecol. Oncol. 2015; 136: 37-42
        • Burger R.A.
        • et al.
        Independent radiologic review of the gynecologic oncology group study 0218, a phase III trial of bevacizumab in the primary treatment of advanced epithelial ovarian, primary peritoneal, or fallopian tube cancer.
        Gynecol. Oncol. 2013; 131: 21-26