Introduction

Ovarian cancer has a poor prognosis; after cervical and uterine cancers, it has the third-highest mortality rate worldwide among gynecological cancers1. The mechanism of action of olaparib entails inhibition of the poly (adenosine diphosphate-ribose) polymerase (PARP) enzyme and PARP trapping; this mechanism of treatment is different from that of conventional cytotoxic chemotherapies, such as platinum and taxanes2. Based on the results of a pivotal multinational phase III trial (SOLO-2 study) and results of a subgroup analysis of the phase II trial (Study-19), olaparib (a small-molecule targeted drug) has become the new standard maintenance therapy after a platinum regimen for patients with platinum-sensitive recurrent ovarian cancer (PSROC)3,4.

Olaparib traps PARP at sites of single-strand DNA breaks and inhibits its release, thereby blocking DNA repair and causing double-strand breaks. Double-strand breaks are not accurately repaired in a homologous recombination defect (HRD), as in tumors with breast cancer susceptibility (BRCA1/2) mutations; PARP inhibition in tumors with HRD results in tumor cell death via synthetic lethality2. Therefore, the BRCA1/2 mutation-positive status is a predictive biomarker of a positive response to olaparib maintenance therapy in patients with PSROC. Furthermore, progression-free survival (PFS) after the initiation of olaparib monotherapy is reported to be longer in patients with PSROC and BRCA1/2 mutations than in patients with PSROC without BRCA1/2 mutations4.

However, PSROC and HRDs occur even without BRCA1/2 mutations5. Moreover, BRCA gene testing is expensive, and a diagnosis of hereditary breast-ovarian cancer causes psychological stress in patients and their families. Thus, it is necessary to identify predictive factors of the response to olaparib maintenance therapy in patients with PSROC but without BRCA1/2 mutations.

Platinum-free interval (PFI), cancer antigen 125 (CA-125) normalization after the last dose of the platinum regimen and response after the last dose of the platinum regimen have been suggested as new predictive factors of response to olaparib maintenance monotherapy in patients with BRCA1/2 mutations and PSROC6. However, there are few reports on the predictors of response to olaparib in patients with recurrent BRCA1/2 mutation-positive ovarian cancer, and the available evidence is insufficient6. The neutrophil-to-lymphocyte ratio (NLR) and the platelet-to-lymphocyte ratio (PLR), indicators of inflammatory response and immune function, have been reported as poor prognostic factors in primary and recurrent ovarian cancer7,8.

A phase II trial on patients with PSROC without BRCA1/2 mutations revealed that compared with the placebo, olaparib markedly prolonged the PFS in those sensitive to previous platinum-based therapy4. Therefore, olaparib has been approved for maintenance treatment, without BRCA gene testing, in patients with PSROC who are sensitive to previous platinum-based therapy in Japan, Europe, and the United States. Recently, post-marketing surveillance showed that the PFS with olaparib maintenance monotherapy in patients with PSROC and the BRCA1/2 mutation-negative status was comparable with that reported in a previous phase II trial9. To the best of our knowledge, however, no studies have explored the factors predicting a positive response to olaparib maintenance monotherapy in patients with PSROC and the BRCA1/2 mutation-negative status.

Platinum-based drugs form DNA interstrand crosslinks, resulting in DNA double-strand breaks. Moreover, HRD tumors may be more susceptible to platinum-based therapy10. Therefore, we expect that platinum sensitivity-associated factors that are suggested to be predictors of olaparib efficacy in patients with PSROC and the BRCA1/2 mutation-positive status would also apply to patients with PSROC and the BRCA1/2 mutation-negative status.

We hypothesized that a combination of clinical factors and peripheral blood markers could help predict the clinical response to olaparib monotherapy in patients with PSROC, regardless of their BRCA1/2 mutation status. Therefore, we performed an exploratory pilot study using real-world data to clarify the patient-associated clinical factors that affected survival in patients treated with olaparib monotherapy for PSROC with or without BRCA1/2 mutations.

Methods

Study design and patients

This multicenter, retrospective, observational study was conducted at three high-volume centers, namely, the National Cancer Center Hospital (Tokyo, Japan), National Cancer Center Hospital East (Chiba, Japan), and Center Hospital of the National Center for Global Health and Medicine (Tokyo, Japan). The surveillance program at each facility is the same. Patient data were extracted from electronic medical records, and data integration and subsequent analyses were performed at the Keio University Graduate School of Pharmaceutical Sciences (Tokyo, Japan). The methodology adopted in this study followed the STROBE statement11.

The inclusion criteria were as follows: (1) consecutive patients aged ≥ 20 years who were diagnosed with PSROC and (2) patients who had received olaparib monotherapy (300 mg taken orally in tablet form twice daily) as maintenance between April 2018 and December 2020. The treatment schedule and follow-up were modified at the discretion of each clinician according to the efficacy and toxicity profile of each patient.

The exclusion criteria were as follows: (1) consent not provided for the use of medical records for research, (2) insufficient data from the patients’ medical records or lack of baseline laboratory data, (3) lower olaparib dosage at therapy initiation (100–250 mg taken orally twice daily), (4) received olaparib monotherapy once, discontinued olaparib for any reason, and then received olaparib monotherapy again after other chemotherapy, and (5) previously treated with bevacizumab or receiving olaparib and bevacizumab concomitantly.

The study protocol was approved by the Ethics Committees of the National Cancer Center (approval number: 2021-052) and Center Hospital of the National Center for Global Health and Medicine (approval number: NCGM-G-004274-00). The study was conducted in accordance with the principles of the Declaration of Helsinki and Ethical Guidelines for Medical and Health Research involving Human Subjects (Ministry of Education, Culture, Sports, Science, and Technology and Ministry of Health, Labour and Welfare; Japan). The need for written or oral informed consent was waived by the ethics review committees of the National Cancer Center and Center Hospital of the National Center for Global Health and Medicine owing to the retrospective nature of the study. Accordingly, we allowed patients to opt-out using the official website of each institution.

Data collection

Patient data were de-identified and analyzed anonymously. We extracted the necessary baseline clinical and demographic data (last blood counts obtained within 4 weeks prior to treatment initiation and other pre-treatment data). The following data were collected: age, body mass index, cancer type, International Federation of Gynecology and Obstetrics staging, Eastern Cooperative Oncology Group performance status (ECOG PS), medical history of chemotherapy, treatment line, date of progression or death at the time of olaparib initiation, presence or absence of germline BRCA1/2 mutations, PFI, CA-125 level, objective tumor response to the last platinum-based chemotherapy according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.112, and daily available peripheral blood data (including absolute neutrophil, lymphocyte, and platelet counts at baseline). The date of disease progression was defined as the date of the first incidence of disease progression identified on computed tomography scans using RECIST or during a clinical evaluation by each clinician. The PFI was defined as the interval between the completion date of first-line platinum-based chemotherapy and the date of the first relapse. Platinum-sensitive patients were defined as those with a PFI of ≥ 6 months. We calculated the baseline NLR and PLR using routinely available blood cell counts at pre-dose values on the day of olaparib induction. The NLR was calculated by dividing the absolute neutrophil count by the absolute lymphocyte count; the PLR was calculated by dividing the absolute platelet count by the absolute lymphocyte count. The follow-up period ended on March 31, 2021.

Endpoints

The primary endpoint of this study comprised the clinical factors associated with survival after olaparib maintenance therapy. The effectiveness of treatment was evaluated using PFS and overall survival (OS). PFS was defined as the period from the date of olaparib treatment initiation to the date of disease progression or death from any cause. OS was defined as the period from the date of olaparib initiation to the date of death due to any cause. Patients without disease progression and those who survived were defined as censored to PFS and OS, respectively, on the date of the last follow-up.

Statistical analyses

Patients were categorized into the BRCA1/2 mutation-positive and BRCA1/2 mutation-negative groups. Patient characteristics were summarized using descriptive statistics, including frequencies and proportions. The Kaplan–Meier method was used to estimate PFS and OS. The log-rank test was used to compare differences between survival curves. Receiver operating characteristic (ROC) curve analyses and the Youden’s index were used to determine the optimal cutoff values for age, NLR, and PLR associated with PFS13. In the ROC curve analyses, a larger area under the curve (AUC) indicated better predictive ability. Univariable and multivariable Cox proportional hazards models were used to evaluate the association between patient-associated clinical factors and survival endpoints. The proportional hazards assumption was not tested because this was not a prospective study. Hazard ratios (HRs) and 95% confidence intervals (CIs) are presented. Potential explanatory variables reported by several previous studies, specifically the PFI, baseline CA-125 level, and objective tumor response to the last platinum-based chemotherapy according to RECIST, were included as covariates in the univariable and multivariable models3,4,6. Furthermore, we conducted subset analyses according to the BRCA1/2 mutation status, as they are part of the genes responsible for robust olaparib efficacy. We examined whether the clinical factors previously suggested as predictors of BRCA1/2 mutations were applicable in this study3,4,6. In the BRCA1/2 mutation-negative group, we also examined whether clinical factors related to platinum sensitivity were associated with survival endpoints. All statistical analyses were performed using JMP (version 16.2.0; SAS Institute Inc., Cary, NC, USA) and SPSS Statistics version 27 (IBM, Armonk, NY, USA). All P-values were two-sided, and P-values of < 0.05 indicated statistical significance.

Ethics approval

This study was conducted in accordance with the principles of the Declaration of Helsinki and Ethical Guidelines for Medical and Health Research involving Human Subjects (Ministry of Education, Culture, Sports, Science, and Technology and Ministry of Health, Labour and Welfare; Japan). The study protocol was approved by the Ethics Committees of the National Cancer Center (approval number: 2021-052) and Center Hospital of the National Center for Global Health and Medicine (approval number: NCGM-G-004274-00).

Consent to participate

The need for written or oral informed consent was waived by the ethics review committees owing to the retrospective nature of the study. Accordingly, we allowed patients to opt-out through the official website of each institution.

Results

Patient characteristics

The patient enrollment flowchart is shown in Fig. 1. Among the 128 patients initially identified, six were excluded based on the exclusion criteria; therefore, data from 122 patients were evaluated. The baseline demographic characteristics of these patients are listed in Table 1. The median age of the patients was 57 years (interquartile range [IQR], 49–68 years). Overall, 91 (74.6%) patients were in good condition with an ECOG PS of 0. The BRCA1/2 mutation-positive group comprised 42 (34.4%) patients, while the BRCA1/2 mutation-negative group comprised 61 (50.0%) patients. The median NLR and PLR were 1.9 (IQR, 1.5–2.5) and 172 (IQR, 139–214), respectively. All patients had a PFI of at least 6 months, and the response to the last platinum dose according to RECIST was either complete response or partial response, indicating that all patients in this study were platinum-sensitive and there were no non-response patients.

Figure 1
figure 1

Flowchart illustrating the patient enrollment process.

Full size image
Table 1 Baseline demographic characteristics of the included patients.
Full size table

Endpoints

The median follow-up period was 13.8 months (95% CI, 12.2–15.5 months). Overall, 73 progressive events and 31 deaths occurred. For all patients, the median PFS and OS were 11.0 months (95% CI, 8.2–13.8 months) and 33.5 months (95% CI, 29.6–37.4 months), respectively (Supplementary Fig. S1). Additionally, the median PFS in the BRCA1/2 mutation-positive and BRCA1/2 mutation-negative groups was 26.1 (95% CI, 10.1–42.0) and 6.8 (95% CI, 5.6–8.0) months, respectively; this difference was statistically significant (Fig. 2, P = 0.001).

Figure 2
figure 2

Kaplan–Meier survival curves for progression-free survival according to the BRCA mutation status-based subgroups. The log-rank test was used to compare the survival curves. BRCA breast cancer susceptibility gene, CI confidence interval, PFS progression-free survival.

Full size image

The optimal cutoff values for age, NLR, and PLR for predicting the onset of progression at the median PFS were initially determined to be 63 years, 3.32, and 210, respectively, with corresponding Youden’s index values of 0.174, 0.101, and 0.199, respectively. The AUC values for age, NLR, and PLR were 0.581, 0.487, and 0.535, respectively. Therefore, we decided that an age of 60 years, NLR ≥ 3.32, and PLR ≥ 210 were appropriate cutoff values for further analyses.

The multivariable Cox proportional hazards model revealed the following as factors that were significantly associated with prolonged PFS (Table 2): a PFI of ≥ 12 months (adjusted HR, 0.38; 95% CI, 0.20–0.70; P = 0.003), CA-125 level within reference values (≤ 35 U/mL [adjusted HR, 0.46; 95% CI, 0.25–0.89; P = 0.016]), and BRCA1/2 mutation-positive status (adjusted HR, 0.34; 95% CI, 0.18–0.61; P = 0.001). Conversely, a high PLR (≥ 210) was significantly associated with worse PFS among all patients (adjusted HR, 3.10; 95% CI, 1.60–5.83; P = 0.001), when adjusted for the selected covariates and the PFI, CA-125 level, PLR, and BRCA1/2 mutation status.

Table 2 Patient-specific clinical factors associated with a prolonged progression-free survival among all patients (N = 122).
Full size table

The multivariable Cox proportional hazards model revealed that a PFI of ≥ 12 months was significantly associated with prolonged PFS in patients with PSROC and the BRCA1/2 mutation-positive status (adjusted HR, 0.32; 95% CI, 0.10–0.86; P = 0.033) (Table 3), when adjusted for the selected covariates and the PFI and PLR. In contrast, the multivariable Cox proportional hazards model revealed the following as factors that were significantly associated with prolonged PFS (Table 4): a PFI of ≥ 12 months (adjusted HR, 0.44; 95% CI, 0.20–0.89; P = 0.027) and CA-125 level within reference values (adjusted HR, 0.42; 95% CI, 0.20–0.96; P = 0.030). Conversely, a high PLR (≥ 210) was significantly associated with poor PFS in patients with PSROC and the BRCA1/2 mutation-negative status (adjusted HR, 3.05; 95% CI, 1.38–6.58; P = 0.005), when adjusted for the selected covariates and the PFI, CA-125 level, and PLR. Owing to the small number of death events, a multivariable analysis for OS could not be performed.

Table 3 Patient-specific clinical factors associated with progression-free survival in the BRCA1/2 mutation-positive group (N = 42).
Full size table
Table 4 Patient-specific clinical factors associated with progression-free survival in the BRCA1/2 mutation-negative group (N = 61).
Full size table

Discussion

Although previous clinical trials have focused on prolonged survival with olaparib monotherapy, few studies have examined the clinical factors associated with survival in real-world clinical settings6,14,15. Furthermore, studies on clinical response biomarkers for olaparib monotherapy have included only patients with BRCA1/2 mutations. The present study showed that a PFI of ≥ 12 months, CA-125 level within reference values, high PLR, and positive BRCA1/2 mutation status might predict clinical response to olaparib monotherapy in all patients with PSROC. Additionally, a PFI of ≥ 12 months, CA-125 level within reference values, and high PLR may be associated with a prolonged PFS under maintenance therapy with olaparib in patients with PSROC and the BRCA1/2 mutation-negative status. To the best of our knowledge, this is the first report on the clinical factors affecting survival following olaparib monotherapy in patients with PSROC with and without BRCA1/2 mutations in a real-world setting.

Our findings are in line with those of a previous study, in which a PFI of ≥ 12 months was associated with prolonged PFS in patients with PSROC and BRCA1/2 mutations6. Furthermore, previous studies on olaparib effectiveness in Chinese patients suggested that a PFI of ≥ 12 months and an objective tumor response to the last platinum-based chemotherapy (according to RECIST) could be considered clinical response biomarkers14,15. However, the associations between PFS and the objective tumor response to the last platinum-based chemotherapy (according to RECIST) and a CA-125 level within reference values were not significant in this study. This may be due to the small number of disease progression events.

Our findings may be explained by the mechanism of action of olaparib in ovarian cancer. Platinum drug-induced DNA double-strand breaks do not undergo repair in recurrent ovarian cancer cells with HRD; this leads to cell death. PFI is an indicator of platinum sensitivity, and a tumor with a PFI of ≥ 6 months is considered platinum-sensitive3,4,6. Objective tumor response (according to RECIST) and the CA-125 level are also used as clinical response biomarkers for anticancer drug efficacy. Olaparib inhibits DNA single-strand repair, and similar to platinum drugs, results in DNA double-strand breaks. Therefore, olaparib is thought to be effective in patients sensitive to platinum-based therapy.

In this study, the association between a high PLR and PFS was statistically significant, but that between a high NLR and PFS was not. Moreover, re-analyses of the relationship between baseline platelet and neutrophil counts and PFS revealed that only the platelet count was considerably associated with PFS. The platelet count is a poor prognostic factor for ovarian cancer16. Because no control group was included in this study, the baseline platelet count may have influenced the prognosis of the patients. Platelet-derived growth factors are present within the alpha granules of platelets, and the inhibition of platelet-derived growth factor receptors reduces BRCA1/2 and RAD51 expression, thereby decreasing homology-directed DNA repair17. This suggests that an increase in platelet count may have resulted in an increase in the platelet-derived growth factor levels, which in turn may have affected homology-directed DNA repair. However, no previous studies have shown the relationship between platelet counts and the effectiveness of olaparib monotherapy, and further cellular and clinical studies are needed on this.

The presence of BRCA1/2 mutations is known to affect patient response to olaparib. However, patients with HRDs and the BRCA1/2 mutation-negative status may also benefit from olaparib. Additionally, some patients have difficulty undergoing BRCA gene testing owing to the high cost of the test and the psychological stress associated with a hereditary breast and ovarian cancer diagnosis. Through subgroup analyses, phase II clinical trials overseas have shown that olaparib is effective in treating patients with PSROC and the BRCA1/2 mutation-negative status, although the PFS after olaparib monotherapy initiation in these patients is shorter than that in patients with PSROC and the BRCA1/2 mutation-positive status4. Moreover, patients who cannot undergo BRCA gene testing may also benefit from olaparib therapy, e.g., patients in Japan, who cannot undergo BRCA1/2 somatic mutation testing or HRD testing under insurance coverage despite the fact that approximately 30% of all Japanese patients with recurrent ovarian cancer have BRCA1/2 mutations and that somatic BRCA1/2 mutations account for 5% of all patients with recurrent ovarian cancer overall. Subgroup analyses in phase II trials have suggested no differences in the efficacy of olaparib between patients with germline and somatic mutations4. In Japan, though comprehensive genomic profiling can be conducted to test for somatic BRCA1/2 mutations, its wide application in clinical practice is hampered by its high cost and the reservation of testing in patients with primary ovarian cancer who cannot be treated with standard therapy. If insurance coverage changes, it may be possible to test for somatic BRCA1/2 mutations.

This study has two strengths. First, this was a multicenter study that involved three national institutions with several patients with cancer in Japan. Therefore, our data may be generalizable to similar populations in clinical settings. Moreover, patient populations in randomized controlled trials differ from those encountered in daily practice. We believe that evidence from real-world data is important to support shared decision-making between clinicians and patients. Second, to the best of our knowledge, this is the first study to evaluate the clinical response biomarkers of olaparib monotherapy in patients with PSROC with or without BRCA1/2 mutations.

This study has a few limitations. First, this was a retrospective, observational study; therefore, the effect of information bias cannot be ignored. However, we performed a multivariable analysis to reduce the influence of confounding factors related to observational studies and patient characteristics. Nevertheless, we could not control for confounders that were not measured in the multivariable analysis. Second, the sample size was relatively small; therefore, the number of progression events was insufficient. This may have influenced the difference in the detected associated factors between this study and the previous studies. In particular, the number of progression events in the BRCA1/2 mutation-positive group was very small (18 cases). Thus, our findings should be validated by increasing the number of such cases. Third, owing to the retrospective nature of the study, we methodologically did not calculate the sample size. Fourth, the median follow-up period of 13.8 months was very short compared with the 37.3-month follow-up duration reported in a previous phase II trial4. This accounted for the small number of death events in this study; therefore, a multivariable analysis for OS could not be performed. Thus, investigations must be performed with a longer follow-up period in the future. Finally, our data was only olaparib. In Japan, olaparib and niraparib were approved in April 2018 and November 2020, respectively. Thus, we will validate niraparib in a future study.

The U.S. Food and Drug Administration has already limited the indication for PARP maintenance therapy for recurrent ovarian cancer to BRCA1/2 mutation-positive cases, and we believe that it is important. On the other hand, this study was conducted with clinical factors that would be applicable to the patient population that cannot undergo BRCA mutation testing due to economic reasons or concerns about HBOCs. The results of this study suggest that a PFI of ≥ 12 months may be a clinical factor regardless of BRCA mutation.

In conclusion, our findings suggest that a PFI of ≥ 12 months may be associated with prolonged survival in patients with PSROC, irrespective of their BRCA1/2 mutation status. Moreover, a PFI of ≥ 12 months and CA-125 level within reference values may be linked to prolonged survival in patients who have received olaparib monotherapy for PSROC and have the BRCA1/2 mutation-negative status. Testing for these predictive factors is routinely accessible in clinical practice and can help classify and screen for high responders to olaparib monotherapy. Furthermore, the findings of this multicenter study can be extrapolated to the entire population of Japan.