This cohort study explores the long-term outcomes of patients with platinum-sensitive recurrent ovarian cancer who have exceptional response to poly(adenosine diphosphate-ribose) polymerase (PARP) inhibitors.
Key Points
Question
What are the long-term outcomes of patients with platinum-sensitive recurrent ovarian cancer who have an exceptional response to poly(adenosine diphosphate-ribose) polymerase (PARP) inhibitors and a progression-free survival (PFS) of 5 years or longer after starting maintenance PARP inhibitors?
Findings
In this cohort study of 320 patients with platinum-sensitive recurrent ovarian cancer and an exceptional response to PARP inhibitor use, the 10-year PFS was 78.7%, and patients who stopped taking PARP inhibitors for reasons other than progression did not experience worse PFS (10-year PFS, 90.1%). Late-onset myeloid cancers occurred in 5 patients (1.6%).
Meaning
Patients with exceptional response had favorable long-term outcomes, with prolonged PFS and low rates of myeloid cancers, raising questions regarding the optimal duration of PARP inhibitors and the possibility of functional cure.
Abstract
Importance
A subset of patients with platinum-sensitive recurrent ovarian cancer (PS-ROC) treated with maintenance poly(adenosine diphosphate-ribose) polymerase (PARP) inhibitors have exceptional response. Although licensing recommends continuing PARP inhibitors until progression or unacceptable toxic effects, the optimal duration of PARP inhibitors, and the risks of late progression, myelodysplastic syndrome (MDS), or acute myeloid leukemia (AML) in patients with exceptional response are unknown.
Objective
To determine the long-term outcomes of patients with PS-ROC who have exceptional response to PARP inhibitors, and to explore genotype-phenotype associations.
Design, Setting, and Participants
This was an international, multicenter, retrospective cohort study of patients with exceptional response to PARP inhibitors, defined as patients with PS-ROC and progression-free survival (PFS) of 5 years or longer from PARP inhibitor commencement. The study was conducted across 41 sites in 14 countries from January 11, 2023, to November 10, 2025.
Exposures
Treatment with PARP inhibitors.
Main Outcomes and Measures
The primary end point was PFS, and secondary end points included overall survival, toxic effects, and dose reductions.
Results
A total of 320 patients with exceptional response (mean [SD] age, 56.4 [9.4] years) were included, with a median follow-up of 6.8 years (95% CI, 6.6-7.0 years). The median (IQR) PARP inhibitor duration was 75.0 (64.0-91.0) months. Of patients with exceptional response, 211 (65.9%) received continuous PARP inhibitors, but 109 (34.1%) discontinued: 34 (10.6%) due to physician recommendation, 2 (7.5%) had disease progression beyond 5 years, 22 (6.9%) had toxic effects, 17 (5.3%) for patient preference, and 12 (3.8%) for another reason. The 7.5-year and 10-year PFS rates were 88.8% (95% CI, 84.5%-93.3%) and 78.7% (95% CI, 70.5%-87.9%), respectively. Among the patients, 85 (26.6%) discontinued PARP inhibitors for reasons other than disease progression, with a 10-year PFS of 90.1% (95% CI, 80.6%-100%) vs 72.5% (95% CI, 60.3%-87.2%) for those who continued taking PARP inhibitors. Five patients (1.6%) were diagnosed with late-onset MDS/AML. Patients with exceptional response were enriched for variants in the BRCA1 RING domain and the BRCA2 DNA-binding domain.
Conclusions and Relevance
In this cohort study, most patients with exceptional response to PARP inhibitors remained progression free, including those who discontinued PARP inhibitors without progression. The risk of late-onset MDS/AML was low. These results can guide counseling on the duration of maintenance PARP inhibitors in patients with exceptional response and suggest that functional cure may be possible in patients with PS-ROC and exceptional response to PARP inhibitors.
Introduction
When the clinical trials of maintenance poly(adenosine diphosphate-ribose) polymerase (PARP) inhibitors for platinum-sensitive recurrent ovarian cancer (PS-ROC) were designed, long-term progression-free survival (PFS) was not considered likely, as, historically, patients with PS-ROC had a poor prognosis (median PFS, 10-11 months; median survival, 2.5-3 years).1,2,3,4 Clinical trial protocols mandated that treatment with maintenance PARP inhibitors continue until progression or unacceptable toxic effects, which was later mirrored in the licensing of maintenance PARP inhibitors. However, uncertainty regarding the duration of treatment grew as it became evident that a subset of patients experienced exceptional response, with PFS of 5 years or longer, ranging from 6.5% in the NOVA trial (niraparib) to 11% in Study 19 (olaparib) and 22% in the SOLO2 trial (olaparib).5,6,7
Key questions remain for those with exceptional response, including the optimal duration of PARP inhibitor use, the risk of late progression, and whether there is an increased risk of myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) with prolonged exposure.8,9 PARP inhibitors may impair quality of life and are costly, underscoring the importance of investigating treatment duration in patients with exceptional response.10,11 Our previous international survey revealed wide variation in attitudes and practice, with almost 50% of clinicians recommending ceasing PARP inhibitor use after an arbitrary duration.12 Of concern, 22% of clinicians reported having patients with late-onset (>5 years) MDS/AML. We subsequently established the Gynecologic Cancer InterGroup Collaboration into Exceptional Responders to PARP Inhibitors to collect patient-level data.13,14 The aims of this study were to determine the long-term outcomes of patients with PS-ROC who had exceptional response to PARP inhibitors, including risks of late progression and MDS/AML, and to explore genotype-phenotype associations.
Methods
Study Design and Population
We conducted an international, multicenter retrospective cohort study of patients with PS-ROC who had exceptional response to treatment with PARP inhibitors. Those with exceptional response were defined as patients with PS-ROC who were progression free for 5 years or longer from the commencement of PARP inhibitor use.
Central ethics approval, including a waiver of consent due to the deidentified, noninterventional nature of the research, was obtained from the South Eastern Sydney Local Health District Human Research Ethics Committee. Each center also obtained local ethics approval or waiver per local requirements. The study complied with the Declaration of Helsinki and multinational ethical guidelines. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
Data Collection and Management
Deidentified data were collected from 41 centers across 14 countries (eTable 1 in Supplement 1) from January 11, 2023, to November 10, 2025. Each center extracted data from medical records into REDCap (Research Electronic Data Capture), hosted by the University of New South Wales and managed by investigators (L.H., Y.C.L., M.F.) at Prince of Wales Hospital in Randwick, Australia (eTable 2 in Supplement 1).15,16 Data included demographics, BRCA1/2 and homologous repair deficiency (HRD) status, prior lines of chemotherapy, PARP inhibitor type and duration of use, toxic effects, other cancers, and survival outcomes.
End Points and Definitions
The primary end point was PFS, defined as the time from PARP inhibitor initiation to progressive disease or death. Secondary end points included overall survival, defined as the time from PARP inhibitor initiation to death, and the incidence and latency of MDS/AML. Patients without an event were censored at last follow-up.
Genetic Analyses
BRCA1/2 variants were annotated for location and predicted molecular impact, and also assigned an allele severity score,17 ranging from 0 (most severe) to 2 (least severe), as described in the eMethods in Supplement 1. Variant characteristics of patients with exceptional response were compared to 2 external ovarian cancer datasets and a dataset of heterozygotes with BRCA1/2 and breast cancer, or without cancer diagnoses.17,18,19
Statistical Analysis
Baseline characteristics and treatments were analyzed descriptively. Survival was estimated using the Kaplan-Meier method, and post hoc subgroup analyses were compared using the log-rank test. The associations of baseline characteristics with PFS were assessed using multivariable Cox proportional-hazards modeling. Differences in variant characteristics were assessed using χ2 or Fisher exact tests. Analyses were performed using RStudio, version 4.3.3 (R Project for Statistical Computing), and Simple Interactive Statistical Analysis.20,21 Two-sided statistical tests were used, with P < .05 deemed statistically significant.
Results
Patient Population and Treatment Characteristics
A total of 526 patients were screened, with 206 patients excluded after screening due to PFS being shorter than 5 years since PARP inhibitor commencement, and 320 eligible patients with exceptional response included (mean [SD] age, 56.4 [9.4] years). The median (range) lines of prior chemotherapy received was 2 (1-8). Of included patients, 263 (82.2%) received PARP inhibitors as routine care (Table 1). The median (IQR) PARP inhibitor duration was 75.0 (64.0-91.0) months (eFigure 1 in Supplement 1). Among the patients with exceptional response, 211 (65.9%) were receiving continuous PARP inhibitors, but 109 (34.1%) discontinued, with 34 (10.6%) discontinuing due to physician recommendation, 24 (7.5%) due to disease progression beyond 5 years, 22 (6.9%) due to toxic effects, 17 (5.3%) due to patient preference, and 12 (3.8%) due to another reason. Of the patients who discontinued PARP inhibitors, the median (range) PARP inhibitor duration was 66 (24-148) months. A total of 29 patients (9.1%) with exceptional response received less than 5 years of PARP inhibitors. Reasons for PARP inhibitor discontinuation within 5 years without disease progression included physician recommendation in 11 patients, patient preference in 9 patients, toxic effects in 7 patients, and another reason in 2 patients. Of the patients who were taking a PARP inhibitor for fewer than 5 years, the median (IQR) PARP inhibitor use duration was 41.0 (28-54) months. Of patients with exceptional response, 17 (5.3%) were diagnosed with a new primary cancer after commencing a PARP inhibitor (Table 1 and eTable 3 in Supplement 1): breast cancer in 5 patients (1.6%), head and neck carcinoma in 3 (0.9%), colorectal cancer in 2 (0.6%), lymphoma in 2 (0.6%), melanoma in 2 (0.6%), urothelial cancer in 2 (0.6%), and lung cancer 1 (0.3%).
Table 1. Clinical Characteristics of Patients With Platinum-Sensitive Recurrent Ovarian Cancer and Exceptional Response to Poly(Adenosine Diphosphate-Ribose) Polymerase (PARP) Inhibitors (N = 320).
| Characteristic | No. (%) |
|---|---|
| Age, y | |
| ≤40 | 12 (3.6) |
| 41-50 | 82 (25.6) |
| 51-60 | 121 (37.8) |
| 61-70 | 75 (23.4) |
| >70 | 28 (8.8) |
| PARP inhibitor used | |
| Olaparib | 240 (75.0) |
| Niraparib | 61 (19.1) |
| Rucaparib | 17 (5.3) |
| Pamiparib | 1 (0.3) |
| Veliparib | 1 (0.3) |
| BRCA status | |
| BRCA1/2 wild type | 40 (12.5) |
| Germline BRCA1 pathogenic variant | 107 (33.4) |
| Germline BRCA2 pathogenic variant | 100 (31.3) |
| Somatic BRCA1 pathogenic variant | 11 (3.4) |
| Somatic BRCA2 pathogenic variant | 15 (4.7) |
| Unknown | 47 (14.7) |
| Lines of chemotherapy prior to PARP inhibitora | |
| 1 | 32 (10.0) |
| 2 | 217 (67.8) |
| 3 | 54 (16.9) |
| ≥4 | 17 (5.3) |
| Time from initial diagnosis to first recurrence, median (IQR), mo | 27.3 (20.4-42.7) |
| Time from recurrence to PARP inhibitor commencement, median (IQR), mo | 7.5 (6.0-14.1) |
| Country of residence | |
| Spain | 60 (18.8) |
| Germany | 59 (18.4) |
| United Kingdom | 41 (12.8) |
| Canada | 28 (8.8) |
| United States | 24 (7.5) |
| France | 22 (6.9) |
| Italy | 18 (5.6) |
| Israel | 16 (5.0) |
| Australia | 15 (4.7) |
| The Netherlands | 13 (4.0) |
| Norway | 12 (3.8) |
| Republic of Korea | 6 (1.9) |
| Japan | 4 (1.3) |
| Austria | 2 (0.6) |
| Context of PARP inhibitor treatment | |
| Standard of care | 263 (82.2) |
| Clinical trial | 57 (17.8) |
| PARP inhibitor status | |
| Ongoing | 211 (65.9) |
| Ceased | 109 (34.1) |
| Other incident primary cancer after starting PARP inhibitor | |
| Breast cancer | 5 (1.6) |
| Head and neck squamous cell carcinoma | 3 (0.9) |
| Melanoma | 2 (0.6) |
| Colorectal cancer | 2 (0.6) |
| Lymphoma | 2 (0.6) |
| Urothelial cancer | 2 (0.6) |
| Lung cancer | 1 (0.3) |
Patients who had received only 1 line of chemotherapy before commencing PARP inhibitors includes patients receiving PARP inhibitors as treatment either alone or in combination with chemotherapy. Among 87 patients with BRCA1/2 wild type or unknown status, 3 patients had homologous recombination deficiency, and 2 patients had homologous recombination repair proficiency, while 82 patients (94.3%) were untested.
Genomic Status
Variant information was provided for 130 of 233 (55.8%) patients with known BRCA1/2 variants. Germline BRCA1 pathogenic variants and BRCA2 pathogenic variants were reported in 107 patients (33.4%) and 100 patients (31.3%), respectively. Somatic BRCA1 pathogenic variants and BRCA2 pathogenic variants were reported in 11 patients (3.4%) and 15 patients (4.7%), respectively. The remainder were BRCA1/2 wild type (40 [12.5%]) or had unknown status (47 [14.7%]). Variant details were available for 66 (55.9%) BRCA1 variants and 58 (50.4%) BRCA2 variants (eTable 4 in Supplement 1). The comparison of variant characteristics with 2 comparator ovarian cancer cohorts is shown in eTables 5 and 6 in Supplement 1. Most variants were predicted to lead to alleles traditionally considered as loss of function, based on nonsense-mediated messenger RNA decay, with only 2 BRCA1 and 2 BRCA2 alleles designated as missense type. Compared to both the PAOLA-1 and Marchetti et al datasets,18,19 patients with exceptional response were enriched for variants located in the BRCA1 RING domain, depleted for BRCA1 variants in the DNA-binding domain, and enriched for variants in the BRCA2 DNA-binding domain.
The variant characteristics, as captured by allele severity score,17 were compared to those observed in population-based female heterozygotes with breast cancer, or without breast or ovarian cancer (Table 2). Patients with exceptional response showed depletion or absence of the least severe allele score 2, and depletion of allele score 1, compared to the allele score distributions in breast cancer case patients and population controls.
Table 2. BRCA1 and BRCA2 Allele Severity Scores Among Patients With Platinum-Sensitive Recurrent Ovarian Cancer and Exceptional Response to Poly(Adenosine Diphosphate-Ribose) Polymerase Inhibitors vs Case Patients With Breast Cancer and Controlsa.
| Variable | No. (%) | |||||
|---|---|---|---|---|---|---|
| Patients with exceptional response | BRCA1 | BRCA2 | ||||
| BRCA1 | BRCA2 | Breast cancer case dataset as comparison | Control dataset as comparison | Breast cancer case dataset as comparison | Control dataset as comparison | |
| Allele severity score | ||||||
| 0 (Most severe) | 37 (56.1) | 23 (39.7) | 362 (38.5) | 64 (29.9) | 384 (25.9) | 121 (15.3) |
| 1 | 28 (42.4) | 35 (60.3) | 531 (56.4) | 124 (57.9) | 1037 (70.) | 599 (75.8) |
| 2 (Least severe) | 1 (1.5) | 0 | 48 (5.1) | 26 (12.1) | 61 (4.1) | 70 (8.9) |
| Total, No. | 66 | 58 | 941 | 214 | 1482 | 790 |
| P value for comparison | Reference | Reference | <.001 | <.001 | .03 | <.001 |
| Proportion | 0.53 | 0.47 | 0.39 | 0.21 | 0.61 | 0.79 |
Allele severity score has been calculated in accordance with Johnatty et al17 and is described in detail in the eMethods in Supplement 1. Patients with BRCA1 and exceptional response have been compared to case patients with BRCA1 pathogenic variants and breast cancer (BRCA1 breast cancer case dataset) or without cancer (BRCA1 control dataset) and similarly for BRCA2 case patients.
PFS and Overall Survival
After a median follow-up of 6.8 years (95% CI, 6.6-7.0 years), the 7.5-year and 10-year PFS rates were 88.8% (95% CI, 84.5%-93.3%) and 78.7% (95% CI, 70.5%-87.9%), respectively (Figure 1). The latest progression event occurred 14.8 years after commencing PARP inhibitors: this patient had ceased olaparib use after 9.4 years due to comorbidities. The 10-year overall survival rate was 90.5% (95% CI, 84.2%-97.4%). The 85 patients (26.6%) who discontinued PARP inhibitors for reasons other than disease progression had a 10-year PFS rate of 90.1% (95% CI, 80.6%-100%) compared to 72.5% (95% CI, 60.3%-87.2%) for those who continued PARP inhibitors.
Figure 1. Kaplan-Meier Curves of Progression-Free Survival (PFS) in Patients With Exceptional Response to Poly(Adenosine Diphosphate-Ribose) Polymerase (PARP) Inhibitors.
A, PFS in the overall population. B, PFS dependent on whether PARP inhibitors were used continuously or ceased prior to progressive disease (PD). C, PFS dependent on reported BRCA1/2 status.
A total of 34 patients developed disease progression at 5 years or longer: 24 while taking PARP inhibitors and 6 after discontinuing PARP inhibitors, with 4 patients discontinuing due to toxic effects, 1 due to physician recommendation, and 1 due to comorbidities. For those who ceased PARP inhibitors before disease progression, the median (range) PARP inhibitor duration was 65.0 (24-148) months, and progression occurred from 1 to 64 months after ceasing PARP inhibitors (median, 8 months). There was not a difference in PFS based on BRCA1/2 status (Figure 1), PARP inhibitor duration (eFigure 2 in Supplement 1), allele severity score (eFigure 3 in Supplement 1), or the type of BRCA1/2 variant (eFigure 4 in Supplement 1). The only baseline characteristic of those with exceptional response that was associated with prolonged PFS was a long interval between initial diagnosis and recurrence (≥18 months; hazard ratio, 0.28; 95% CI, 0.14-0.58; P < .001; eTable 7 in Supplement 1).
Treatment Modifications and Adverse Events
Dose reductions due to adverse events occurred in 138 patients (43.1%), and 22 (6.9%) ceased treatment due to adverse events (eTable 8 in Supplement 1). The most common reasons for dose reduction were hematologic toxic effects in 79 patients and fatigue in 42.
Incidence of MDS and AML
Five patients (1.6%) developed myeloid cancers at a mean (SD) age of 50.6 (10.3) years (Figure 2). Two patients developed MDS, which transformed to AML; 1 patient developed MDS, and 2 developed isolated AML. Three patients received 2 prior lines of chemotherapy, and 2 patients had 3 prior lines. The median (range) duration of PARP inhibitor exposure was 61 (53-109) months. Three patients received olaparib, 1 received niraparib, and 1 received rucaparib. Three patients had germline BRCA1 pathogenic variants, 1 had a germline BRCA2 pathogenic variant, and 1 had germline BRCA wild type with unknown somatic status.
Figure 2. Bar Graphs Showing Cases of Myelodysplastic Syndrome (MDS) or Acute Myeloid Leukemia (AML) After Starting Poly(Adenosine Diphosphate-Ribose) Polymerase (PARP) Inhibitors.
Time on the x-axis represents time elapsed since the commencement of PARP inhibitors.
Discussion
To our knowledge, this is the largest study of patients with PS-ROC who had exceptional response when treated with maintenance PARP inhibitors. The key finding was that most patients with exceptional response remained progression free long term. Continuous PARP inhibition was common, reflecting licensing and clinical trial recommendations. Of included patients, 109 (34.1%) discontinued maintenance therapy for a range of reasons, and 29 patients (9.1%) received fewer than 5 years of treatment. Importantly, discontinuation of PARP inhibitors due to clinician decision or patient preference did not appear to compromise PFS. Of interest, only 5% of patients with exceptional response were diagnosed with a new primary cancer after commencement of a PARP inhibitor, including 5 (1.6%) who were diagnosed with breast cancer. Finally, the incidence of MDS/AML was low, with 5 (1.6%) patients with exceptional response developing MDS/AML.
These are the first data, to our knowledge, that describe the long-term outcomes of patients with PS-ROC who had exceptional response to treatment with maintenance PARP inhibitors. This unique population represents an important minority of patients with PS-ROC; for example, 43 patients (22%) in SOLO2 continued taking olaparib beyond 5 years.5 In ARIEL3 (rucaparib), 21%, 13.9%, and 6.9% of participants had a PFS of 2 to 3 years, 3 to 4 years, or more than 4 years, respectively,22 highlighting a substantial risk of progression between 2 and 4 years, supporting the selection of 5 years to define exceptional response in the present study. Furthermore, the seminal trials of maintenance PARP inhibitors have provided limited data on the long-term outcomes of patients with exceptional response. This study adds to these data by documenting a 10-year PFS of 79% and low risk of MDS/AML in this unique subgroup.
Most patients with exceptional response in this study (73%) had deleterious BRCA1/2 pathogenic variants, illustrating the association with long-term PFS and exceptional response to PARP inhibitors. Very few patients had HRD testing performed precluding analysis of HRD status, although all were platinum sensitive, which is considered a functional surrogate of HRD.23 Most studies have reported a survival advantage associated with BRCA1/2 pathogenic variants, but findings are not entirely consistent, with some studies showing no statistically significant difference in overall survival between those with or without BRCA1/2 pathogenic variants.24,25,26,27
A few studies have characterized the genomic profile in patients with ovarian cancer and exceptional response; however, definitions vary across studies. Garsed et al28 documented that 31 of 126 patients with exceptional response (PFS of ≥3 years, overall survival of ≥10 years, or multiple complete responses) had germline BRCA1/2 pathogenic variants, which was lower than the incidence in the present study (65%), likely due to differences in inclusion criteria. In Study 19, 22 of 32 (68.8%) patients with long-term response (>2 years) had BRCA1/2 pathogenic variants compared to 10 of 19 (52.6%) patients with short-term response (<3 months).29 In ARIEL3, patients with exceptional response to rucaparib (PFS of >2 years) were enriched for BRCA1/2 pathogenic variants and RAD51C and RAD51D alterations.22
Recent evidence has shown that the site and type of BRCA1/2 pathogenic variants can also influence prognosis and response to PARP inhibitors.18,19,29 Two studies in the first-line maintenance setting have associated therapeutic benefit with variant location, reporting greater benefit of olaparib in those with pathogenic variants in the DNA-binding domain of BRCA1/2, RING finger and C-terminal domains of BRCA1, and RAD51 binding domain of BRCA2.18,19 In Study 19, which included a population similar to the present study, pathogenic variants in the DNA-binding domain of BRCA1/2 were only observed in patients with long-term response.29 In the present study, patients with exceptional response were enriched for variants in the BRCA1 RING domain and BRCA2 DNA-binding domain, and depleted for variants in the BRCA1 DNA-binding domain. Furthermore, compared to heterozygotes with BRCA1/2, patients with exceptional response were enriched for variants expected to have greatest functional impact based on allele severity score,17 suggesting that various variant molecular features may help predict exceptional response to PARP inhibitors. We hypothesize that the enrichment of patients with exceptional response and severe alleles may be related to a lower likelihood of reversion mutations. Further work is needed to compare the allele severity score in those with exceptional response to PARP inhibitors with other patients with ovarian cancer and, in particular, to patients with short-term response to PARP inhibitors.
Implications of Discontinuing PARP Inhibition
A key observation in this study was that patients who discontinued PARP inhibitors for reasons other than progression did not appear to experience inferior outcomes compared to those who continued PARP inhibitors. This challenges the paradigm of indefinite treatment, as currently recommended. The present results contribute to the debate around fixed-duration maintenance therapy in selected patients with PS-ROC who remain progression free beyond 5 years, and raise the possibility that a subset of patients with exceptional response are functionally cured. These findings are reassuring and suggest that cessation of PARP inhibitors may be considered in selected patients with exceptional response, provided that cancer antigen 125 values are normal and there is no evidence of disease on imaging results. Circulating tumor DNA is a promising approach to identify patients who may be suitable for treatment cessation, although further research is needed.30
Risk of Myeloid Cancers
The risk of MDS/AML with PARP inhibitors and platinum-based chemotherapy is a recognized concern; however, the incidence in patients with exceptional response is unknown. A meta-analysis reported MDS/AML rates of 0.73% with PARP inhibitors vs 0.47% with placebo, with a median latency of 17.8 months, and that the risk was higher in the recurrent than frontline setting.8 However, follow-up was relatively short for the included trials, so it is uncertain whether the risks are cumulative with PARP inhibitor exposure 5 years or longer. In the primary analysis of patients treated with olaparib in the SOLO2 trial, MDS/AML was reported in 4 of 195 (2%) patients31; however, in the final analysis, MDS/AML was reported in 16 of 195 (8%) patients, 9 (5%) of whom developed MDS/AML after the safety follow-up period.5 The risk of MDS/AML among patients with exceptional response in the present study was low (5 [1.6%]), which suggests that long-term PARP inhibition in patients with exceptional response is not associated with an escalating risk of late-onset MDS/AML. The incidence is much lower than the 3.8%, 3.7%, and 8% incidence reported in the NOVA, ARIEL3, and SOLO2 trials, respectively.5,6,32 It is unclear how many patients in these trials developed MDS/AML following subsequent chemotherapy at progression, which is likely to increase risk of MDS/AML.22
Limitations
Patients with exceptional response represent the patients at the tail of the survival distribution in clinical trials, making a prospective study in this subgroup impractical. Consequently, retrospective analysis is the only feasible approach, and although it carries a risk for selection bias, sites were asked to contribute all eligible patients to minimize this risk. Second, genomic data were incomplete for some patients due to site-level privacy restrictions and missing data. This study was not powered to identify predictive factors for benefit from continued vs fixed-duration PARP inhibition. Furthermore, we compared genetic characteristics against patients with ovarian cancer treated with maintenance PARP inhibitors in the first-line setting, as no suitable data were available for comparison with a typical PS-ROC population. Finally, no details were collected regarding the timing or extent of dose reductions or on the proportion who underwent risk reducing bilateral mastectomy.
Conclusions
This cohort study describes the long-term safety and efficacy of PARP inhibitors in patients with PS-ROC and exceptional response to PARP inhibitors. Late progression beyond 5 years was infrequent, and the incidence of late-onset MDS/AML was low. Although the optimal duration of PARP inhibitors in patients with exceptional response remains uncertain, patients who ceased maintenance treatment after long-term use for reasons other than progression appeared to have outcomes comparable to those who continued indefinitely. These results provide, to our knowledge, previously unavailable data to inform counseling of patients with PS-ROC and exceptional response to PARP inhibitors. It is possible that many patients with exceptional response are functionally cured, although longer follow-up is needed for confirmation.
eMethods
eTable 1. Included Sites
eTable 2. Electronic Database Questionnaire
eTable 3. Second Primary Cancers
eTable 4. Overview of Characteristics of BRCA1 and BRCA2 Pathogenic Variants Documented in Exceptional Responders
eTable 5. Variant Characteristics for Exceptional Responders Compared to Those for PAOLA-1 Trial Participants
eTable 6. Variant Characteristics for Exceptional Responders Compared to Those Reported by Marchetti et al., (2025)
eTable 7. Multivariable Model of Baseline Characteristics Associated with Progression Free Survival
eTable 8. Treatment Modifications and Adverse Events in Exceptional Responders with Platinum Sensitive Recurrent Ovarian Cancer
eFigure 1. Progression Free Survival in Exceptional Responders Based on Variant Type
eFigure 2. Progression Free Survival in Exceptional Responders Based on Duration of PARP Inhibitor Received
eFigure 3. Progression Free Survival in Exceptional Responders Based on Allele Severity Score
eFigure 4. Progression Free Survival in Exceptional Responders Based on Variant Type
eReferences
Nonauthor Collaborators. Gynecologic Cancer InterGroup Collaboration Into Exceptional Responders to PARP Inhibitors (GCIG-CERP) members
Data Sharing Statement
References
- 1.Aghajanian C, Goff B, Nycum LR, Wang YV, Husain A, Blank SV. Final overall survival and safety analysis of OCEANS, a phase 3 trial of chemotherapy with or without bevacizumab in patients with platinum-sensitive recurrent ovarian cancer. Gynecol Oncol. 2015;139(1):10-16. doi: 10.1016/j.ygyno.2015.08.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Aghajanian C, Blank SV, Goff BA, et al. OCEANS: a randomized, double-blind, placebo-controlled phase III trial of chemotherapy with or without bevacizumab in patients with platinum-sensitive recurrent epithelial ovarian, primary peritoneal, or fallopian tube cancer. J Clin Oncol. 2012;30(17):2039-2045. doi: 10.1200/JCO.2012.42.0505 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wagner U, Marth C, Largillier R, et al. Final overall survival results of phase III GCIG CALYPSO trial of pegylated liposomal doxorubicin and carboplatin vs paclitaxel and carboplatin in platinum-sensitive ovarian cancer patients. Br J Cancer. 2012;107(4):588-591. doi: 10.1038/bjc.2012.307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pujade-Lauraine E, Wagner U, Aavall-Lundqvist E, et al. Pegylated liposomal doxorubicin and carboplatin compared with paclitaxel and carboplatin for patients with platinum-sensitive ovarian cancer in late relapse. J Clin Oncol. 2010;28(20):3323-3329. doi: 10.1200/JCO.2009.25.7519 [DOI] [PubMed] [Google Scholar]
- 5.Poveda A, Floquet A, Ledermann JA, et al. ; SOLO2/ENGOT-Ov21 investigators . Olaparib tablets as maintenance therapy in patients with platinum-sensitive relapsed ovarian cancer and a BRCA1/2 mutation (SOLO2/ENGOT-Ov21): a final analysis of a double-blind, randomised, placebo-controlled, phase 3 trial. Lancet Oncol. 2021;22(5):620-631. doi: 10.1016/S1470-2045(21)00073-5 [DOI] [PubMed] [Google Scholar]
- 6.Matulonis UA, Herrstedt J, Oza A, et al. ENGOT-OV16/NOVA trial of niraparib in recurrent ovarian cancer: survival and long-term safety. Gynecol Oncol. 2025;195:192-199. doi: 10.1016/j.ygyno.2025.03.018 [DOI] [PubMed] [Google Scholar]
- 7.Friedlander M, Matulonis U, Gourley C, et al. Long-term efficacy, tolerability and overall survival in patients with platinum-sensitive, recurrent high-grade serous ovarian cancer treated with maintenance olaparib capsules following response to chemotherapy. Br J Cancer. 2018;119(9):1075-1085. doi: 10.1038/s41416-018-0271-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Morice PM, Leary A, Dolladille C, et al. Myelodysplastic syndrome and acute myeloid leukaemia in patients treated with PARP inhibitors: a safety meta-analysis of randomised controlled trials and a retrospective study of the WHO pharmacovigilance database. Lancet Haematol. 2021;8(2):e122-e134. doi: 10.1016/S2352-3026(20)30360-4 [DOI] [PubMed] [Google Scholar]
- 9.Moss HA, Perhanidis JA, Havrilesky LJ, Secord AA. Real-world treatment patterns of maintenance therapy in platinum-sensitive recurrent ovarian cancer. Gynecol Oncol. 2021;163(1):50-56. doi: 10.1016/j.ygyno.2021.07.026 [DOI] [PubMed] [Google Scholar]
- 10.Philp L, Eisenhauer E, Bregar A, et al. Patient-reported quality of life during PARP-inhibitor treatment for high-grade serous ovarian cancer. Gynecol Oncol. 2022;166(suppl 1):S232. doi: 10.1016/S0090-8258(22)01689-4 [DOI] [Google Scholar]
- 11.Goldsberry WN, Summerlin SS, Guyton A, et al. The financial burden of PARP inhibitors on patients, payors, and financial assistance programs: who bears the cost? Gynecol Oncol. 2021;160(3):800-804. doi: 10.1016/j.ygyno.2020.12.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Haggstrom L, Lee YC, Scott C, et al. How long is long enough? an international survey exploring practice variations on the recommended duration of maintenance therapy with PARP inhibitors in patients with platinum sensitive recurrent ovarian cancer and long-term outcomes. Int J Gynecol Cancer. 2024;34(12):1932-1939. doi: 10.1136/ijgc-2024-005976 [DOI] [PubMed] [Google Scholar]
- 13.Rosenman R, Tennekoon V, Hill LG. Measuring bias in self-reported data. Int J Behav Healthc Res. 2011;2(4):320-332. doi: 10.1504/IJBHR.2011.043414 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Albarqouni LN, López-López JA, Higgins JPT. Indirect evidence of reporting biases was found in a survey of medical research studies. J Clin Epidemiol. 2017;83:57-64. doi: 10.1016/j.jclinepi.2016.11.013 [DOI] [PubMed] [Google Scholar]
- 15.Harris PA, Taylor R, Minor BL, et al. ; REDCap Consortium . The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95:103208. doi: 10.1016/j.jbi.2019.103208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-381. doi: 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Johnatty SE, Tudini E, Parsons MT, et al. Analysis of BRCA1, BRCA2 and PALB2 related Fanconi anemia identifies scope to expand disease phenotypic features and predict breast cancer risk in heterozygotes. medRxiv. Preprint posted online May 26, 2025. doi: 10.1101/2025.05.25.25327887 [DOI] [PMC free article] [PubMed]
- 18.Marchetti C, Fagotti A, Fruscio R, et al. Benefit from maintenance with PARP inhibitor in newly diagnosed ovarian cancer according to BRCA1/2 mutation type and site: a multicenter real-world study. ESMO Open. 2025;10(4):104533. doi: 10.1016/j.esmoop.2025.104533 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Labidi-Galy SI, Rodrigues M, Sandoval JL, et al. Association of location of BRCA1 and BRCA2 mutations with benefit from olaparib and bevacizumab maintenance in high-grade ovarian cancer: phase III PAOLA-1/ENGOT-ov25 trial subgroup exploratory analysis. Ann Oncol. 2023;34(2):152-162. doi: 10.1016/j.annonc.2022.11.003 [DOI] [PubMed] [Google Scholar]
- 20.The R Project for Statistical Computing website. Accessed May 30, 2026. https://www.R-project.org/
- 21.Uitenbroek D. Simple Interactive Statistical Analysis (SISA). Sociedade Galega para a Promoción da Estatística e Investigación de Operacións . Accessed June 5, 2026. https://www.sgapeio.gal/INFORMEST/VICongreso/taller/applets/Sisa/about.htm
- 22.O’Malley DM, Oza AM, Lorusso D, et al. Clinical and molecular characteristics of ARIEL3 patients who derived exceptional benefit from rucaparib maintenance treatment for high-grade ovarian carcinoma. Gynecol Oncol. 2022;167(3):404-413. doi: 10.1016/j.ygyno.2022.08.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Feng Z, Shao D, Cai Y, et al. Homologous recombination deficiency status predicts response to platinum-based chemotherapy in Chinese patients with high-grade serous ovarian carcinoma. J Ovarian Res. 2023;16(1):53. doi: 10.1186/s13048-023-01129-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Yang D, Khan S, Sun Y, et al. Association of BRCA1 and BRCA2 mutations with survival, chemotherapy sensitivity, and gene mutator phenotype in patients with ovarian cancer. JAMA. 2011;306(14):1557-1565. doi: 10.1001/jama.2011.1456 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kotsopoulos J, Rosen B, Fan I, et al. Ten-year survival after epithelial ovarian cancer is not associated with BRCA mutation status. Gynecol Oncol. 2016;140(1):42-47. doi: 10.1016/j.ygyno.2015.11.009 [DOI] [PubMed] [Google Scholar]
- 26.Hyman DM, Zhou Q, Iasonos A, et al. Improved survival for BRCA2-associated serous ovarian cancer compared with both BRCA-negative and BRCA1-associated serous ovarian cancer. Cancer. 2012;118(15):3703-3709. doi: 10.1002/cncr.26655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Nahshon C, Barnett-Griness O, Segev Y, Schmidt M, Ostrovsky L, Lavie O. Five-year survival decreases over time in patients with BRCA-mutated ovarian cancer: a systemic review and meta-analysis. Int J Gynecol Cancer. 2022;32(1):48-54. doi: 10.1136/ijgc-2020-001392 [DOI] [PubMed] [Google Scholar]
- 28.Garsed DW, Alsop K, Fereday S, et al. ; Nadia Traficante, for the Australian Ovarian Cancer Study Group . Homologous recombination DNA repair pathway disruption and retinoblastoma protein loss are associated with exceptional survival in high-grade serous ovarian cancer. Clin Cancer Res. 2018;24(3):569-580. doi: 10.1158/1078-0432.CCR-17-1621 [DOI] [PubMed] [Google Scholar]
- 29.Lheureux S, Lai Z, Dougherty BA, et al. Long-term responders on olaparib maintenance in high-grade serous ovarian cancer: clinical and molecular characterization. Clin Cancer Res. 2017;23(15):4086-4094. doi: 10.1158/1078-0432.CCR-16-2615 [DOI] [PubMed] [Google Scholar]
- 30.Lee JY, Lee D, Kim YN, Won D, Shin S, Kim SW. Detection of minimal residual disease using circulating tumor DNA in patients treated with long-term PARP inhibitors maintenance for epithelial ovarian cancer. Abstract presented at: 39th Annual Meeting of the Korean Society of Gynecologic Oncology; April 26, 2024; Busan, South Korea. [Google Scholar]
- 31.Pujade-Lauraine E, Ledermann JA, Selle F, et al. ; SOLO2/ENGOT-Ov21 investigators . Olaparib tablets as maintenance therapy in patients with platinum-sensitive, relapsed ovarian cancer and a BRCA1/2 mutation (SOLO2/ENGOT-Ov21): a double-blind, randomised, placebo-controlled, phase 3 trial. Lancet Oncol. 2017;18(9):1274-1284. doi: 10.1016/S1470-2045(17)30469-2 [DOI] [PubMed] [Google Scholar]
- 32.Ledermann JA, Oza AM, Lorusso D, et al. Rucaparib for maintenance treatment of platinum-sensitive, recurrent ovarian carcinoma: final results of the phase 3, randomized, placebo-controlled ARIEL3 trial. Eur J Cancer. 2025;225:115584. doi: 10.1016/j.ejca.2025.115584 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods
eTable 1. Included Sites
eTable 2. Electronic Database Questionnaire
eTable 3. Second Primary Cancers
eTable 4. Overview of Characteristics of BRCA1 and BRCA2 Pathogenic Variants Documented in Exceptional Responders
eTable 5. Variant Characteristics for Exceptional Responders Compared to Those for PAOLA-1 Trial Participants
eTable 6. Variant Characteristics for Exceptional Responders Compared to Those Reported by Marchetti et al., (2025)
eTable 7. Multivariable Model of Baseline Characteristics Associated with Progression Free Survival
eTable 8. Treatment Modifications and Adverse Events in Exceptional Responders with Platinum Sensitive Recurrent Ovarian Cancer
eFigure 1. Progression Free Survival in Exceptional Responders Based on Variant Type
eFigure 2. Progression Free Survival in Exceptional Responders Based on Duration of PARP Inhibitor Received
eFigure 3. Progression Free Survival in Exceptional Responders Based on Allele Severity Score
eFigure 4. Progression Free Survival in Exceptional Responders Based on Variant Type
eReferences
Nonauthor Collaborators. Gynecologic Cancer InterGroup Collaboration Into Exceptional Responders to PARP Inhibitors (GCIG-CERP) members
Data Sharing Statement
