Introduction The introduction of poly-ADP ribose polymerase inhibitors in ovarian cancer has demonstrated significantly improved progression free survival in four randomized controlled clinical trials in patients with platinum sensitive relapsed ovarian cancer. While overall survival data remain immature, this real world evidence study sets a baseline for future evaluation of poly-ADP ribose polymerase inhibitors.
Methods A retrospective chart review was undertaken to investigate real world survival outcomes across 13 National Health Service Trusts in England, Wales, and Scotland. Patients were included if they had platinum sensitive relapsed high grade serous ovarian cancer and had responded to secondline platinum based chemotherapy. Clinical data were collected retrospectively from electronic prescribing records and chart notes. The index date for overall survival analysis was defined as the later of (1) day 1 of the final secondline platinum based treatment or (2) date of response to secondline treatment. The primary objective was overall survival from the index date. Secondary objectives included progression free survival and overall survival by subsequent line of treatment. BRCA mutation status was collected where available. Quality of life questionnaires were not assessed within this study.
Results 233 patients were identified who met the study inclusion criteria. Patient characteristics were consistent with other published data, with a median age of 61 years (range 35–85). Sensitivity analysis of the primary objective demonstrated that the earliest point poly-ADP ribose polymerase inhibitors may be initiated (following completion of secondline chemotherapy) is associated with a median overall survival of 19.8 months. Secondline median overall survival and progression free survival from the index date were 19.3±2.4 months and 7.3±1.2 months, respectively. 144 patients were treated with thirdline chemotherapy with median overall survival and progression free survival from the index date (either date of last cycle of thirdline treatment or date of response to thirdline treatment) of 8.3±2.6 and 4.4±1.8 months, respectively.
Conclusion Overall survival was shown to be shorter in this real world study compared with randomized clinical trials, and underlines the differences in clinical outcomes of patients in a real life setting. This baseline real world study has demonstrated poor survival outcomes in this patient group prior to availability of poly-ADP ribose polymerase inhibitors.
- ovarian cancer
- fallopian tube neoplasms
- peritoneal neoplasms
Statistics from Altmetric.com
The earliest point poly-ADP ribose polymerase inhibitors may be initiated following completion of secondline chemotherapy was associated with a median overall survival of 19.8 months.
Median progression free survival from completion of secondline platinum based chemotherapy was 7.3 months.
Median progression free survival following thirdline platinum based chemotherapy was 4.4 months.
Randomized controlled trials are regarded as the ‘gold standard’ from the clinical research paradigm. However, tightly controlled eligibility criteria and experimental conditions in randomized clinical trials may differ from real life.1 Less than 5% of adult patients with cancer participate in randomized clinical trials. Randomized clinical trials are not representative of the whole population and their application could be questioned in guiding clinical practice circumstanced.1 Real world evidence complements the findings of randomized clinical trials by providing additional evidence and insight into patient outcomes, treatment patterns, and disease epidemiology. In the United Kingdom, real world evidence is becoming increasingly important in decisions that affect patient access to therapies. In 2017, 37% of National Institute for Health and Care Excellence (NICE) initial submissions incorporated real world data compared with 22% in 2016.2
Ovarian cancer accounts for approximately 4% of all cancers in women and 10% of gynecological cancers worldwide.3 With more than two-thirds of ovarian cancer cases detected at an advanced stage, survival rates are poorest of all gynecological malignancies.3 With the exception of low risk stage I disease, standard treatment for ovarian cancer is based on a combination of surgical resection (with the objective of complete resection of all macroscopic disease) and neoadjuvant or adjuvant platinum based chemotherapy.4
Platinum sensitive relapsed disease is defined as patients who have a treatment free interval ≥6 months between the final cycle of platinum based chemotherapy and time of disease recurrence.5 These patients are considered platinum sensitive and retreated with further platinum based chemotherapy. Although approximately 80% of patients with newly diagnosed ovarian cancer will respond to firstline platinum based chemotherapy,6 the effectiveness of platinum therapy decreases with each subsequent line, and the majority of patients will experience an incremental reduction in benefit due to the onset of platinum resistance and cumulative toxicities.7 As platinum resistance typically develops, new treatments are required to improve the outcomes for patients with platinum sensitive disease.
Outcomes for patients with ovarian cancer in Europe remain poor, with an overall age standardized 5 year survival rate of 36.1%.3 The United Kingdom demonstrated overall 5 year survival rates below the European average at 30.0% for England, 35.5% for Northern Ireland, 31.5% for Scotland, and 32.3% for Wales.3
The treatment landscape for patients with high grade serous carcinoma has evolved worldwide with the introduction of poly-ADP ribose polymerase inhibitors. Poly-ADP ribose polymerase inhibitors target DNA repair pathways8 and have demonstrated significantly improved progression free survival in four randomized clinical trials of maintenance treatment in patients with platinum sensitive relapsed ovarian cancer (olaparib-Study 19 (Assessment of the efficacy of azd2281 in platinum sensitive relapsed serous ovarian cancer),6 SOLO2 (Olaparib treatment in BRCA mutated ovarian cancer patients after complete or partial response to platinum chemotherapy),9 niraparib-NOVA (A maintenance study with niraparib versus placebo in patients with platinum sensitive ovarian cancer),10 and rucaparib-ARIEL3 (A study of rucaparib as switch maintenance following platinum based chemotherapy in patients with platinum sensitive, high grade serous or endometrioid epithelial ovarian, primary peritoneal, or fallopian tube cancer)).11
The aim of this study was to provide real world clinical outcomes in patients with platinum sensitive relapsed ovarian cancer prior to the introduction of poly-ADP ribose polymerase inhibitors, providing a benchmark for assessing the impact of poly-ADP ribose polymerase inhibitors in standard clinical practice.
This real world, observational, retrospective, multicenter chart review was designed to establish the outcomes from standard of care in the United Kingdom for patients with platinum sensitive relapsed ovarian cancer prior to the introduction of poly-ADP ribose polymerase inhibitor use. Patients were identified from a sample of 13 United Kingdom hospitals (both academic centers and district general hospitals throughout Scotland, Wales, and England). The study collected data from chart notes and electronic prescribing records of patients with platinum sensitive relapsed ovarian cancer treated between January 2007 and December 2014. The study was performed between April 2017 and December 2017.
Because of its retrospective, descriptive nature, the study size was not based on formal statistical considerations. Initially, the aim was for a sample size of a maximum of 470 patients to cover approximately 10% of the eligible United Kingdom population. This was determined by estimating the number of high grade serous platinum sensitive relapsed patients in the United Kingdom who responded to secondline platinum chemotherapy (667 patients per year) over the service evaluation window (2007–2014; 667×7 = 4669 patients who may meet the inclusion criteria).
Patients were eligible if they were 18 years or older and had a primary diagnosis of locally advanced or metastatic high grade serous (grade 2 or 3) ovarian, fallopian tube, or peritoneal cancer. They were required to have platinum sensitive relapsed disease and a minimum of three cycles of platinum based chemotherapy as firstline and secondline treatment. Eligible patients must have completed a secondline of platinum based chemotherapy, with evidence of an objective response (complete or partial response), as defined by the response evaluation criteria in solid tumors, or according to the clinician evaluation. Clinical evaluation was a combination of clinician’s interpretation of computerized tomography reports in conjunction with clinical notes and CA125. Patients treated with poly-ADP ribose polymerase inhibitors in any line were excluded.
Clinical data were collected at each center from the medical chart notes and electronic prescribing records. Data comprising baseline demographics, diagnostic markers, details of chemotherapy treatment by line, response and clinical markers of disease progression, date of death, or last follow-up were collected for all patients. Each record was validated for data gaps and inconsistencies with the site’s principal investigator. BRCA mutation status was collected where available. Quality of life questionnaires were not assessed within this study.
Disease progression was primarily identified by evidence of radiological progression. If radiological progression was not recorded, evidence of progression was based on clinical evaluation. In the absence of either of these, the date of next treatment or date of death was used as a proxy for disease progression. Missing data were not an issue, as records were required to have all data for inclusion.
Primary Objective Measures
The primary objective was to estimate overall survival in patients with platinum sensitive relapsed ovarian cancer who responded to secondline platinum based chemotherapy. Due to the variability in patient follow-up schedules and the site’s definitions of progression, the primary endpoint was overall survival to ensure consistency across the study.
Overall survival was defined as time from the index date (date of the last cycle of secondline platinum based chemotherapy or the date of response to this treatment, provided this was ≤8 weeks from the last dose of secondline chemotherapy) until death by any cause or end of follow-up (patients without a recorded date of death were censored at their last available observation date). For patients with a date of response >8 weeks, the date of the last dose of chemotherapy was selected as the index date (online supplementary Figure SF1). This index date was designed to reflect the market authorization for olaparib.12 This allows results to be used as a comparative baseline for poly-ADP ribose polymerase inhibitors.
Sensitivity analyses for overall survival were performed as part of the primary objective using various scenarios to mimic when a patient may be initiated on poly-ADP ribose polymerase inhibitors, including: time from date of last cycle of secondline platinum based chemotherapy until date of death or end of follow-up, time from date of last cycle of secondline platinum based chemotherapy plus 4 weeks until date of death, and time from date of last cycle of secondline platinum based chemotherapy plus 8 weeks until date of death or end of follow-up.
Secondary Objective Measures
Secondary objective measures were progression free survival, defined as the time from the end date of the last cycle of second (or subsequent) line of chemotherapy until the earliest documentation of disease progression, death, or end of follow-up. Other secondary endpoints were defining treatment pathways and outcomes by other lines of therapy.
Of the 345 patients with high grade serous ovarian cancer (167 chart notes and 178 electronic prescribing records), 233 (66%) met the inclusion criteria (Figure 1). Baseline demographics and disease characteristics were recorded (Table 1). Median age was 61 years (range 35–85), 83.7% of patients had International Federation of Gynecology and Obstetrics stage III or IV disease at diagnosis, and 47.2% of patients had an Eastern Cooperative Oncology Group (ECOG) performance status score of 0 or 1 at diagnosis. BRCA mutation status was available for 36 patients (15%); of these, 18 patients had evidence of a deleterious or suspected deleterious BRCA mutation, and 18 patients were BRCA wild-type. Median number of patients contributed by sites was 13 (range 3–55). The median number of lines of chemotherapy was 3.0 (mean±SD, 3.0±1.2; range 2–9). A total of 52 patients (22%) received ≥4 lines of chemotherapy (Table 2). Median follow-up for patients was 23.8 months (range 5.2–131.7) from the start of secondline treatment until death or last recorded follow-up.
Of the 233 patients included in the primary analysis, 197 (85%) had a date of death documented, and 36 (15%) were censored at their last observation date in the absence of evidence of vital status. Median overall survival in this population from the index date was 19.3 months (95% CI ±2.4 months, range 1.5–128.2) (Figure 2). Sensitivity analysis of the primary objective demonstrated that the earliest point poly-ADP ribose polymerase inhibitors may be initiated for a patient (following completion of secondline chemotherapy) was associated with a median overall survival of 19.8 months (Table 2).
All patients included in the study progressed following response to secondline platinum based chemotherapy. Median progression free survival from completion of secondline platinum based chemotherapy was 7.3 months (95% CI ±1.2, range 0.7–53.8). (Figure 3) From the study population, 79 patients (34%) were considered to have developed platinum resistance having progressed or died within 6 months of completion of secondline platinum based therapy. A total of 154 patients had not progressed or died within 6 months of the index date and were therefore defined as platinum sensitive. Of these 154 patients, 10 patients were lost to follow-up, and 144 (94%) patients relapsed, receiving thirdline chemotherapy, with 90 patients (63%) receiving a platinum regimen and 54 patients (38%) receiving non-platinum therapy (online supplementary Table ST1).
Of the 90 patients who received treatment with a thirdline platinum regimen, 53 patients (59%) progressed or died within 6 months, while 37 (41%) patients had not progressed at 6 months. Median progression free survival following thirdline platinum based chemotherapy was 4.4 months (95% CI ±1.8, range 0–115.2) and median overall survival from thirdline therapy was 8.5 months (95% CI ±2.6, range 0–115.2) (Table 2).
This real world evidence study was intended to be representative of patients with platinum sensitive relapsed ovarian cancer who responded to secondline platinum based chemotherapy in order to understand the outcomes of such patients prior to the introduction of maintenance treatment with poly-ADP ribose polymerase inhibitors. While previous observational studies in patients with epithelial ovarian cancer and known germline BRCA mutation have been performed,13 14 to our knowledge this is the first in this specific patient population.
The majority of baseline demographics and disease characteristics of the 233 patients were as expected and consistent with other published data.6 10 15 One difference noted was the majority of patients in this real world evidence study had an ECOG score of 1 compared with randomized clinical trial data where the majority of patients presented with an ECOG score of 0.6 10 15 This is as expected, as the stringent eligibility criteria of randomized clinical trials select the healthiest of the diseased population. In the real world setting, patients are likely to be affected by comorbidities and other complexities. The sample size for the BRCA+ subgroup was too small to formulate any robust conclusions. BRCA testing was not part of routine clinical practice within the study period as poly-ADP ribose polymerase inhibitors were not recommended by NICE.
Median overall survival from the time of response to secondline platinum based chemotherapy in this retrospective study was 19.3 months (95% CI ±2.4, range 1.5–128.2). There are currently only limited overall survival data available from poly-ADP ribose polymerase inhibitor randomized clinical trials to draw comparisons. In Study 19, median overall survival from the time of response to secondline or later-line platinum based chemotherapy was 27.8 months but this figure was confounded by post-progression poly-ADP ribose polymerase inhibitor use in 13% of patients.16 Mature overall survival data from other poly-ADP ribose polymerase inhibitor trials (SOLO2, NOVA, and ARIEL3) are not yet available.8–10 Other non-poly-ADP ribose polymerase inhibitor, platinum sensitive randomized clinical trials have demonstrated overall survival data of 35.2 months,17 33.0 months,18 and 21.0 months.15 However, the index date differed in these studies as it was from randomization before chemotherapy. On the other hand, it should also be considered that the population selected for this real world evidence study may show a survival bias; in other words, the population were a group of patients with ‘good prognosis’ as they had responded to platinum again after a >6 month interval from firstline therapy and therefore may have prolonged overall survival compared with those who had relapsed over 6 months from firstline platinum based chemotherapy but where the response to secondline platinum was not yet known.
Median progression free survival decreased from 7.3 months (n=233) following secondline therapy to 4.4 months (n=144) for thirdline treatment. Although these findings are consistent with investigator assessed progression free survival, as defined by the response evaluation criteria in solid tumors in the control (placebo) arms of Study 19, SOLO2, NOVA, and ARIEL3 (4.8, 5.5, 5.5, and 5.4 months, respectively)6 9–11 it highlights the very poor prognosis associated with multiple relapses. It also emphasizes the small patient group that reach thirdline treatment. The median number of lines of treatment in this real world evidence study was three. This is relatively low, considering women in randomized clinical trials often receive multiple lines of therapy, reiterating the difference in the real world and randomized clinical trial population, and further emphasizing the need for early effective therapies.
It is widely recognized that imaging assessments are not performed as frequently in routine clinical practice as in randomized clinical trials. Randomized clinical trial estimations of progression free survival are determined by time to the response evaluation criteria in solid tumors, radiological progression, whereas real world estimations are commonly based on time to next treatment. In the absence of radiological progression data, start date of the first cycle of subsequent treatment was used as a proxy. This is likely to be a conservative approach as the date of actual progression may have been several weeks or months prior to the start of treatment. This difference in clinical practice may explain the longer progression free survival observed in the real world setting compared with randomized clinical trial data and must be considered when interpreting the results.
Real world evidence studies are best designed to complement randomized clinical trials. Both study designs have inherent limitations due to their distinct methodologies. Unlike randomized clinical trials, real world evidence lacks standardized guidelines (such as Good Clinical Practice) potentially leading to poor quality of data and limited transparency in the methods.19 The most significant concern for real world evidence studies is the potential for bias, and caution is required when interpreting results. To avoid bias in this study, all electronic prescribing records at each site were reviewed and checked for study eligibility. Another limitation to consider is that the definition of response in this study differed from that of published trials (using a combination of computerized tomography, CA125, and clinical response in the absence of response evaluation criteria in solid tumors); however, this is more representative of the population who are now receiving poly-ADP ribose polymerase inhibitors.
Patient notes are often the most complete and robust source of retrospective patient data, however both data sources contained data gaps. To mitigate data gaps, each record met the minimum data required to be included and was validated by the principal investigator.
This study was undertaken to assess real world data on the treatment pathways and prognosis of platinum sensitive relapsed ovarian cancer patients in the United Kingdom, regardless of BRCA mutation status. These findings support the implementation of using real world evidence to complement randomized clinical trial data, in particular demonstrating median overall survival to be shorter in the real world setting compared with previously reported trial data. These results reiterate the unmet need for novel treatment options in platinum sensitive relapsed ovarian cancer. It highlights that earlier intervention is key in order to extend time to progression, delay the onset of platinum resistance, and ultimately improve response rates and overall survival.
In conclusion, national healthcare systems should not rely solely on randomized clinical trial data to inform healthcare and health economic decisions. This study in platinum sensitive relapsed patients has proven the value of real world data that may compliment or address some of the clinical data gaps in randomized clinical trials. The findings set a baseline of clinical outcomes for future studies to be compared against. It is recommended that this real world evidence study is repeated in the future when patients have had routine access to poly-ADP ribose polymerase inhibitors, to re-assess overall survival.
Medical writing was provided by SVMPharma Ltd UK, funded by AstraZeneca. The authors would like to thank all the investigators that contributed data to this study: Dr Sarah Williams (Birmingham), Dr Clare Green (Southampton), Dr Dan Lee (Airedale), Dr Chit Cheng Yeoh (Portsmouth), Dr Andrew Proctor (York), and Dr Simon Pledge (Sheffield).
Contributors All authors were involved in the development of the study protocol and objectives, and review and writing of the manuscript.
Funding This work was supported by AstraZeneca.
Competing interests TM, JR, and OC are employees of AstraZeneca UK Ltd, a biopharmaceutical company who manufactures treatments for ovarian cancer. RL declares advisory board work for and travel grants from AstraZeneca and Tesaro. RM reports consultancy work for Merck, Tesaro, and Clovis Oncology, serving on speaker’s bureau for Roche and Tesaro, and has received travel grants from AstraZeneca and Tesaro. RJ reports advisory board work for Tesaro and Clovis Oncology and travel grants from AstraZeneca, Pharma Mar, and Tesaro. RMG reports advisory board work for AstraZeneca, Tesaro, Clovis, Immunogen, and Sotio; funding to attend conferences from AstraZeneca, Tesaro, and Roche; and research funding from Boehringer Ingelheim and Lilly/Ignyta. RMG is also the national coordinating investigator for the OREO trial (AstraZeneca) and First trial (Tesaro), and is the site principal investigator for the ORZORA trial (AstraZeneca) as well as other trials sponsored by Tesaro, Clovis, Immunogen, and Pfizer. EH reports advisory board work for Roche and Tesaro. MH has undertaken advisory work for Roche, Clovis Oncology, Amgen, Tesaro, and AstraZeneca.
Patient consent for publication Not required.
Ethics approval Ethics approval came through the National Health Service ‘Service Evaluation’ from each individual National Health Service trust. Data were pseudonymised at source and gathered retrospectively, meaning no patient consent was required. This study was performed in accordance with the Declaration of Helsinki and the Good Clinical Practice guidelines.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data may be obtained from a third party and are not publicly available. Pseudonymised chart notes (ie, the data) are the property of the relevant hospital trust, and may be requested from them through liaison with the corresponding author, RL (ORCID iD: 0000-0003-1544-8396). Conditions for reuse are subject to the request. Additional information (eg, protocols, statistical analysis plans) may be requested from firstname.lastname@example.org.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.