This single-center retrospective study collected clinical data from patients with recurrent epithelial ovarian cancer treated at the First Affiliated Hospital of Wenzhou Medical University between December 2012 and December 2023. Clinical data before relapse included age of diagnosis, BMI, CA125 level before initial treatment, BRCA gene status, FIGO staging, initial treatment date, neoadjuvant chemotherapy, postoperative residual disease status, pathological type, postoperative chemotherapy-related data, the efficacy of first-line platinum-containing chemotherapy and use of PARPis during first-line treatment. Clinical data after relapse included the date of initial recurrence, recurrence characteristics, initial treatment time after relapse, the second-line treatment plan, and the use of PARPis in the second-line and last-line treatments. The study was approved by ethics committee for clinical research at the First Affiliated Hospital of Wenzhou Medical University (KY2024-R224).

The inclusion criteria were as follows (1): Histologically confirmed epithelial ovarian cancer (2); Patients received surgery combined with platinum-containing chemotherapy as first-line treatment (3); The first recurrence occurred during the follow-up period, with evidence of tumor recurrence confirmed by imaging, pathology, or tumor markers. The exclusion criteria were as follows (1): Presence of other malignant tumors (2); Complications with hematological diseases, immune system diseases, severe liver or kidney dysfunction, or other serious medical conditions (3); Disease progression occurred during first-line platinum-containing chemotherapy or within one month after the last chemotherapy session (patients with primary platinum-refractory ovarian cancer) (4); Incomplete first-line treatment for ovarian cancer (including surgery and platinum-containing chemotherapy) (5); Incomplete second-line treatment following recurrence during the follow-up period (6); Loss to follow-up or lack of essential clinical data.

The primary clinical data of patients with initial recurrence were collected through electronic medical records. Recurrence and survival outcomes were obtained by reviewing outpatient and inpatient medical record systems and conducting telephone follow-ups, centralized in December 2023.

Overall survival (OS) was defined as the time interval from the date of the first treatment to the time of death or the last follow-up.

Data were analyzed using SPSS 25.0 and MedCalc 18.2.1. The Shapiro-Wilk test was employed to assess the normality of measurement data. Data conforming to a normal distribution were presented as mean ± standard deviation and compared between groups using an independent samples t-test. For data that did not conform to a normal distribution, the median (25th to 75th percentile) was used for description, and the Mann-Whitney U test was applied for intergroup comparisons. Categorical data were described by frequency and percentage, with comparisons using Fisher’s exact test. A Cox regression model was used to analyze OS in patients with recurrent ovarian cancer through univariate analysis, and variables with P < 0.05 were included in a multivariate analysis. Survival curves were generated using the Kaplan-Meier method, and univariate analysis of OS was conducted with the log-rank test. In this study, P < 0.05 was considered statistically significant.

Among the 131 patients with recurrent ovarian cancer included in this study, 83 (63.4%) had received PARPis throughout their treatment, while 48 (36.6%) had not. Patients were divided into two cohorts: the non-PARPis and PARPis cohorts. Patient characteristics are detailed in Table 1 . Significant differences were observed between the two groups in terms of BRCA gene status, neoadjuvant chemotherapy (NACT), postoperative residual condition, and platinum-free interval (PFI) after first-line treatment (P < 0.05).

As shown in Table 2 , 27 patients (20.6%) received PARPis during first-line maintenance treatment, while 104 patients (79.4%) did not. Compared to those who did not receive PARPis in the first line, patients who maintained PARPis during first-line treatment exhibited no significant differences in distant metastasis, number of lesions, or CA125 levels at the recurrence time (P > 0.05). Both groups presented multiple lesions, distant metastasis, and elevated CA125 levels.

All patients were followed up till December 2023. In the non-PARPis cohort, 39 patients (81.3%) died, whereas in the PARPis cohort, 30 patients (36.1%) died. The median follow-up time was 50.3 months (33.0-64.9 months). The median follow-up time for the non-PARPis cohort was 42.7 months (31.2-61.8 months), while for the PARPis cohort, it was 52.5 months (40.8-72.0 months). Kaplan-Meier survival curves were generated, and the log-rank test was used for comparisons. As shown in Figure 2A , the median OS of the non-PARPis cohort was 45.4 months (95% CI: 37.9-52.2 months). The median OS and 95% CI were not calculated for the PARPis cohort. Patients who received PARPis throughout their treatment had significantly better OS than those who did not (P < 0.0001).

In this study, various clinical data were collected and analyzed using univariate and multivariate Cox regression to assess the impact of clinical factors on OS in patients with recurrent ovarian cancer. As shown in Table 3 , after adjusting for multiple factors, the use of PARPis throughout the treatment, CA125 levels before the first treatment, and the PFI after first-line treatment were identified as independent factors influencing OS in these patients (P < 0.05). Patients who received PARPis throughout their treatment, CA125 level ≤ 1000 kU/L before the first treatment and had a longer PFI exhibited better OS.

As shown in Table 4 , of the 83 patients who received PARPis throughout their treatment, 61 used PARPis for a single line during their overall management, while 22 used it in multiple lines. Among those who used PARPis for a single line, 51 patients used it as maintenance treatment and ten as salvage treatment. Specifically, 19 patients received PARPis during first-line treatment, 26 during second-line treatment, and six during last-line treatment.

In Figure 2B , Kaplan-Meier survival analysis was performed on the patients who used PARPis in the second-line and last-line maintenance therapy. The median OS and 95% CI for the second-line and last-line groups were not calculated. Compared with the patients who had not received PARPis treatment, the patients who used PARPis in the second-line (P<0.0001) and last-line (P=0.0063) maintenance treatment had better OS. There was no significant difference in OS between the second-line and last-line groups (P=0.8608). In Figure 2C , Kaplan-Meier survival analysis was conducted on patients who used PARPis repeatedly in multiple lines and those who used PARPis in only a single line. The median OS in the multi-line group was 69.5 months (95% CI: 60.3-83.5 months). The median OS and 95% CI for the single-line group were not calculated. There was no significant difference in OS between patients who used PARPis in multiple lines and those who used PARPis in only a single line (P=0.3919).

A total of 131 patients with recurrent ovarian cancer were included in this study. Of these, 83 patients (63.4%) received PARPis throughout their treatment, while 48 (36.6%) did not. Significant differences were observed between the two groups regarding BRCA gene status, NACT, postoperative residual condition, and PFI after first-line treatment (P < 0.05). As of the last follow-up, patients who received PARPis throughout their treatment had better OS than those who did not receive PARPis (P < 0.05). The use of PARPis throughout treatment, a CA125 level ≤ 1000 kU/L before the first treatment, and a longer PFI were identified as independent factors associated with improved OS in patients with recurrent ovarian cancer. First-line maintenance use of PARPis did not alter disease recurrence characteristics. PARPis administered in second-line and last-line treatments was associated with improved OS, although no significant difference in OS was found between these two stages of treatment. No evidence was found that patients who had previously used PARPis could derive additional OS benefits from reusing them again.

This study demonstrated that disease progression patterns were similar in patients receiving either PARPis or placebo during first-line maintenance treatment. This finding aligns with the results of the SOLO-2 study presented at the ASCO meeting in 2022, which reported no significant difference in disease progression patterns between the PARPis and placebo groups (19). However, the SOLO-2 study focused on recurrence characteristics in patients with platinum-sensitive recurrent ovarian cancer who continued PARPis maintenance after recurrence. In contrast, our study examined patients with recurrent ovarian cancer more broadly. Additionally, SOLO-2 investigated a more comprehensive range of recurrence characteristics, including lesion sites categorized as target, non-target, and new lesions, and analyzed the distribution of specific progressive sites such as the liver, lung, and central nervous system.

The subgroup analysis in our study indicated that the use of PARPis as maintenance therapy, whether in second-line or last-line treatment, was associated with improved OS. However, no significant difference in OS was observed between these two groups. The L-MOCA subgroup analysis reported that the median progression-free survival (PFS) for patients receiving second-line treatment was 18.0 months, which was notably higher than that for patients receiving third-line treatment (≥third-line) (20). Although our study did not find a statistical difference in OS between the second-line and last-line treatment groups, the median OS for both groups had not been reached by the follow-up time, and the small sample size limits this conclusion.

Our study’s analysis of another subgroup did not find that multiple uses of PARPis improved OS compared to single-line use. The OReO study (21), a randomized, placebo-controlled Phase IIIB trial, included 220 patients with non-mucinous epithelial ovarian cancer who had previously received at least one line of PARPis maintenance treatment and had responded to the latest platinum-based chemotherapy. The study found better PFS through repeated maintenance treatment with Olaparib. Conversely, a real-world multi-center study in China (22) indicated no significant difference in PFS among patients with varying BRCA mutation statuses after receiving PARPis in a second time. This real-world study also included patients who received PARPis for salvage rather than maintenance. While our findings contrast with those of the OReO study, it is essential to note that our research focused on OS rather than PFS and involved a different patient population. In our study, some patients who received PARPis in multiple lines belonged to maintenance therapy, others had salvage therapy, and some had maintenance therapy followed by salvage therapy. Additionally, the control group of patients who received PARPis in a single line did not uniformly receive maintenance therapy.

This study divides recurrent ovarian cancer patients who have used PARPis into first-line, second-line, and last-line groups based on their medication courses and into single-line and multi-line groups based on their medication frequency. This is an innovative approach in retrospective researches. This study provides novel insights into the recurrence characteristics of ovarian cancer patients following first-line maintenance treatment with PARPis, an area with limited previous research. However, the study’s conclusions are constrained by its small sample size, being a single-center, retrospective analysis with only 27 patients who received PARPis during first-line maintenance. Further large-scale studies are needed to validate these findings. Additionally, the study’s limitations include the lack of BRCA gene detection data, which prevented an analysis of the benefit of BRCA mutations in patients receiving PARPis. The study also did not examine the recurrence characteristics of patients who maintained PARPis treatment after disease recurrence, representing another limitation.

This study aids in classifying disease risk for patients, developing individualized follow-up plans, and demonstrating the potential survival benefits of PARPis for patients with recurrent ovarian cancer. It suggests that PARPis may extend OS in previously used patients. However, the study did not find a survival advantage from repeated use of PARPis, indicating that further exploration through large-scale randomized studies is needed in this area.