Article Text
PDF
Abstract
Objective The PAOLA-1 trial confirmed that adding olaparib to bevacizumab significantly increased clinical benefit following response to platinum-based chemotherapy in homologous recombination deficiency-positive ovarian cancer. The objective of this analysis was to determine the cost-effectiveness of olaparib plus bevacizumab compared with bevacizumab alone as maintenance treatment for patients with homologous recombination deficiency-positive advanced ovarian cancer from the Spanish National Health System perspective.
Methods A lifetime partitioned survival model with four health states (progression-free, post-progression 1, post-progression 2, and death) and monthly cycles was developed. Long-term survival, defined as 60 months, was included as a landmark to extrapolate progression-free survival from PAOLA-1. Weibull distribution was selected as the most accurate survival model for progression-free survival extrapolation. Time to second progression and overall survival were extrapolated using parametric survival models. Mortality was obtained from the overall survival and adjusted by Spanish women mortality rates. Health state utilities and utility decrements for adverse events were included. An expert panel validated data and assumptions. Direct costs (in 2021 euros (€)) were obtained from local sources and included drug acquisition and administration, subsequent therapies, monitoring costs, adverse events, and palliative care. A 3% annual discount rate was applied to costs and outcomes. The incremental cost-effectiveness ratio was calculated as cost per quality-adjusted life-years (QALYs) gained. Deterministic and probabilistic sensitivity analyses were performed.
Results Compared with bevacizumab alone, olaparib plus bevacizumab increased QALYs and life-years by 2.39 and 2.77, respectively, at an incremental cost of €58 295.31, resulting in an incremental cost-effectiveness ratio of €24 371/QALY. Probabilistic sensitivity analysis demonstrated that olaparib plus bevacizumab had a 49.5% and 90.3% probability of being cost-effective versus bevacizumab alone at a willingness-to-pay threshold of €25 000 and €60 000 per QALY gained, respectively.
Conclusion For patients with homologous recombination deficiency-positive advanced ovarian cancer, olaparib plus bevacizumab is a cost-effective maintenance therapy compared with bevacizumab alone in Spain.
- Ovarian Cancer
- Medical Oncology
Data availability statement
Data are available upon reasonable request. Clinical data not published are available from the corresponding author (Sergio Cedillo, spain.healtheconomics@astrazeneca.com, ORCID 0009-0000-1378-1625) upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Olaparib monotherapy as first-line maintenance treatment has been shown to be cost-effective in ovarian cancer patients with BRCA mutations in Spain.
WHAT THIS STUDY ADDS
Olaparib plus bevacizumab achieves clinical benefits and leads to substantial life-years and quality-adjusted life-years gain, compared with bevacizumab alone, providing good economic value as it is cost-effective for homologous recombination deficiency-positive ovarian cancer patients.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The results of this analysis can support health technology assessments and regulatory bodies in decision-making based on efficiency, demonstrating that the olaparib plus bevacizumab combination is a cost-effective option for the Spanish National Health System.
INTRODUCTION
Ovarian cancer is the eighth leading cancer in incidence and mortality in women worldwide.1 In Spain, data projection for 2023 estimated 3584 new cases,2 and 2036 deaths were recorded due to ovarian cancer in 2020.3 Ovarian cancer is diagnosed in advanced stages (International Federation of Gynecology and Obstetrics (FIGO) III/IV) in 75% of cases.4 Approximately 75% of these cases are high-grade,5 and around 50% of these patients have homologous recombination deficiency.6 The best characterized causes of homologous recombination deficiency are germline or somatic mutations in the BRCA1 and BRCA2 genes. However, it can also be caused by other mutations or methylation of a wider set of homologous recombination repair-related genes.7 These mutations are associated with an increased risk of developing cancer compared with the general population.8
The standard of care for newly diagnosed advanced ovarian cancer is cytoreductive surgery, followed by platinum-based chemotherapy.9–12 However, approximately 70% of patients fail to achieve complete responses or eventually relapse.13 Adding bevacizumab to first-line chemotherapy and continuing with bevacizumab as maintenance therapy has been demonstrated to improve progression-free survival irrespective of disease stage and residual disease after surgery.14 15
As a biomarker, homologous recombination deficiency holds both predictive and prognostic value in high-grade ovarian cancer. Poly-ADP ribose polymerase (PARP) inhibitors were developed based on their predicated synthetic lethality in the context of homologous recombination deficiency-positive cells. The 2016 European Society of Medical Oncology (ESMO) guidelines recommend that maintenance therapy with PARP inhibitors, such as olaparib, could be considered as monotherapy in the first-line maintenance of patients with a BRCA mutation.9 However, patients with a high-risk of progression, regardless of BRCA status, represent a substantial population still in need of more efficacious treatment options following first-line platinum-based therapy. A recent update of the ESMO-European Society of Gynecological Oncology (ESGO) guidelines explicitly recommends the use of homologous recombination deficiency testing for the identification of subpopulations of women without BRCA1/BRCA2 mutations and may benefit from PARP inhibitor-based therapies.16
The recent European Medicines Agency approval (September 17, 2020) of olaparib in combination with bevacizumab for patients newly diagnosed with homologous recombination deficiency-positive high-grade ovarian cancer provides a major challenge for personalized medicine. The clinical efficacy and safety of olaparib plus bevacizumab has been demonstrated in the pivotal PAOLA-1 trial, where maintenance therapy with olaparib plus bevacizumab demonstrated a clinically significant benefit in progression-free survival compared with bevacizumab monotherapy (37.2 months with olaparib plus bevacizumab and 17.7 months with bevacizumab monotherapy; hazard ratio for disease progression or death 0.33, 95% confidence interval (CI) 0.25 to 0.45).17
In this context, this study aimed to estimate the cost-effectiveness of olaparib plus bevacizumab compared with bevacizumab monotherapy from the Spanish National Health System perspective as maintenance treatment for patients with homologous recombination deficiency-positive advanced ovarian cancer.
METHODS
Model Structure
A partitioned survival model with four health states (progression-free, post-progression 1, post-progression 2, and death) was developed based on independent curves of overall survival and progression-free survival for each comparator, in line with previous appraisals of maintenance treatment in ovarian cancer.18 19
The model evaluated costs and outcomes of newly diagnosed advanced high-grade ovarian cancer (including patients with primary peritoneal and/or fallopian tube cancer) who responded after first-line platinum/taxane-based chemotherapy plus bevacizumab,17 receiving: olaparib (300 mg) twice daily (for up to 2 years) plus bevacizumab (15 mg/kg) every 3 weeks (for up to 15 months); or bevacizumab (15 mg/kg) as maintenance monotherapy every 3 weeks (for up to 15 months).
A 1 month cycle length (with half-cycle correction) and a lifetime horizon were considered. The baseline patient characteristics (mean age 66.5 years; weight 65.5 kg; height 159.5 cm; body surface 1.7 m2) were extracted from a Spanish observational study (OvarCostStudy20) except for serum creatinine (72.6 µmol/L), which was obtained from the PAOLA-1 trial21 due to the absence of data.
Patients entered the model in the progression-free state and began treatment until disease progression (as defined by Response Evaluation Criteria In Solid Tumors 1.1) or treatment discontinuation. Once patients experienced progression, they entered the ‘post-progression 1’ health state and started receiving subsequent treatments. Patients remained in that state until they experienced a second disease progression and entered the ‘post-progression 2’ state, where they remained until death. Transition to absorbing ‘death’ state was possible from any health state (Figure 1).
Model structure.
The model used first and second progression-free survival, and overall survival curves from the PAOLA-1 trial to directly estimate the proportion of patients occupying each state over time (Online supplemental figure 1).
Supplemental material
The inputs were validated by a panel of five Spanish experts (three oncologists and two hospital pharmacists).
Efficacy
First and second progression-free survival, overall survival, and time on treatment were modeled using data from the primary analysis of the PAOLA-1 trial.21 First and second progression-free survival and overall survival data were extrapolated to estimate the likely outcomes beyond the observed duration of the clinical trial.
In the case of first progression-free survival, a Kaplan-Meier curve approximation was performed using parametric mixture survival modeling to capture the heterogeneity in the population’s survival, and the fact that survival is a mixture of short-term and long-term survival. Since not all women with newly diagnosed ovarian cancer have had the same susceptibility to relapse or disease progression in the PAOLA-1 trial, parametric mixture survival modeling allowed patients’ long-term survivorship to be captured more accurately. To distinguish between short- and long-term survival, a cut point at 5 years was set, since approximately 20–25% of patients with stage III–IV epithelial ovarian cancer on first-line chemotherapy remain progression-free for longer than 5 years after completion of first-line treatment (classified as long-term survivors).22 23
As long-term survival data were unavailable for olaparib plus bevacizumab at the time of this analysis, long-term estimations were performed using extrapolation methods and predicting events with mathematical techniques and parametric distributions, which are in line with recently published final data.24 In addition, the National Institute for Health and Care Excellence recommendations25 for approximating two Kaplan-Meier curves drawn from the same clinical trial were followed. A weighted AIC (Akaike information criterion) in the two arms was considered to determine which distribution performs a better approximation. The results were assessed to align with clinical practice expectations and literature data. As a result, the best-fitting model for the combined dataset was the Weibull distribution, which was selected to model the first progression in both treatment arms (Online supplemental figure 2, Online supplemental table 1).
For second progression-free survival and overall survival, standard parametric curves were used to extrapolate outcomes beyond the observed clinical trial duration at the data cut-off time used for the analysis and were modeled in both treatment arms using generalized gamma distribution, considered the best approximation for the curves (lowest AIC/BIC (Bayesian information criterion); Online supplemental figures 3 and 4, Online supplemental tables 2 and 3).
Additionally, the model considered age-dependent mortality data of Spanish women obtained from National Institute of Statistics mortality tables.26 In each cycle, the highest probability of death between the overall survival curve and the mortality rate of the general population was considered to reduce a potential underestimation of mortality.
The Kaplan-Meier curves for the time from randomization to discontinuation or death were used to derive the time-on-treatment in the model. No extrapolation was performed as the curves cover the entire treatment duration, capped at 24 months for olaparib and 15 months for bevacizumab, in line with the trial protocol and olaparib’s Summary of product characteristics.
Costs
Costs included drug-related costs (acquisition and administration), subsequent therapies, monitoring costs (office visits, blood count, and CT scans), adverse events management costs, and palliative care. All costs were presented in 2021 euros (€).
Drug-related costs were calculated using list prices obtained from the database of the Official College of Pharmacists27 (the mandatory rebate established in Spanish Royal Decree-Law 8/2010 was applied following standardized working procedures for clinical evaluation in Spain.28 29 The cost of bevacizumab was calculated as the product of the dose (mg) and the cost per mg. The lowest cost per mg of the biosimilar drug was assumed (400 mg), and wastage was not considered. In addition, drug administration cost (nursing cost) was included for bevacizumab (Table 1).30
Model parameters
Bevacizumab and olaparib dose reductions and interruptions were incorporated in the cost per model cycle using the relative dose intensity, which captures the percentage of the actual dose intensity delivered relative to the intended dose until treatment discontinuation, obtained from the PAOLA-1 trial (Table 1).21
For subsequent treatment, chemotherapy and PARP inhibitors (only in the bevacizumab monotherapy arm) were considered. The proportion of patients receiving each subsequent line of treatment (second-line, third-line, fourth-line, and subsequent) and the distribution of subsequent therapies by line of treatment for both arms was based on expert opinion (Online supplemental table 4 and Online supplemental table 5) as well as the proportion of patients receiving each PARP inhibitor and chemotherapy combination (Online supplemental table 6). Based on the dosing schedule, the median duration of treatment (Online supplemental table 6), and the average cost per mg,27 a single cost per treatment group and comparator was derived and applied when a patient progressed (Table 1).
Monitoring costs considered different resource use rates for progression-free patients and progressed disease to reflect clinical practice, obtained from the literature31 and expert consensus (Online supplemental table 7). A maximum follow-up of 5 years was established in progression-free patients.
The cost of treatment-related adverse events (Table 1) was based on the incidence of grade 3–4 adverse events (>2%) in the intention-to-treat population of PAOLA-1,21 and the unitary cost obtained from the literature17 31 and eSalud database30 (Online supplemental table 8).
To approximate patients receiving palliative care, an estimation was made from the extrapolation of the Kaplan-Meier curve of the progression-free survival, considering the average number of patients progressing at 5 years in both arms. A weighted cost between hospital and home palliative care was assumed (Table 1).
Utility Values
Utility values for progression-free, post-progression 1, and post-progression 2 health states and utility decrements for adverse events were included. Progression-free and post-progression 1 utility values were obtained from PAOLA-1,21 and post-progression 2 utility value was based on data from the OVA-301 study32 due to the absence of long-term data in PAOLA-1 (Table 1). Age-related utility decrements were also included in the model to reflect the natural decline in health status associated with age.33
Adverse events disutilities were calculated considering the incidence of adverse events from the PAOLA-1 trial,21 the duration (years),18 and the disutility values (Online supplemental table 8).18 34–36
Model Outputs
The model reported outputs including total costs, life-years, and quality-adjusted life-years (QALYs). The incremental cost-per-LY gained and the incremental cost-per-QALY gained were calculated. Willingness-to-pay thresholds of €25 000 and €60 000 per QALY gained37 were considered in line with the Spanish guidelines. Both costs and outcomes were discounted at a rate of 3% per year according to the local recommendations for economic evaluation of health technologies.28
Sensitivity Analyses
One-way and probabilistic sensitivity analyses were carried out to identify the most influential parameters and test the robustness of the results.
One-way sensitivity analysis was performed by varying each parameter in isolation between its upper and lower limits taken from their 95% CI or applying a standard error (SE) of 10% of the mean when the CI was unknown. Finally, the results were compared with the base case in a tornado diagram.
The probabilistic sensitivity analysis was conducted using a Montecarlo simulation with 10 000 iterations. A probabilistic sensitivity analysis simultaneously sets all inputs to a value randomly sampled from the appropriate distribution. When uncertainty data were not reported, the SE was assumed to be 10% of the mean.
RESULTS
Base-Case Analysis
Olaparib plus bevacizumab was a cost-effective maintenance therapy compared with bevacizumab alone in patients with advanced positive homologous recombination deficiency ovarian cancer in Spain, considering a willingness-to-pay threshold of €25 000 and €60 000 per QALY.37 Olaparib plus bevacizumab generated €58 295.31 incremental costs, 2.39 incremental QALYs, and 2.77 incremental life-years over a lifetime horizon, compared with bevacizumab maintenance alone, resulting in an incremental cost-effectiveness ratio of €24 371.00 per QALY gained and €21 013.41 per life-years gained (Table 2).
Deterministic cost-effectiveness results
Sensitivity Analyses
In the one-way sensitivity analysis, the parameter with the most significant impact on the incremental cost-effectiveness ratio was the discount rate applied to QALYs. Even in the less favorable upper limit scenario, olaparib plus bevacizumab generated an incremental cost-effectiveness ratio of around €30 000 per QALY compared with bevacizumab maintenance alone, confirming the efficiency of this therapy (Figure 2).
ICER tornado diagram. 2 L, second-line; 3 L, third-line; 4 L, fourth-line; ICER, incremental cost-effectiveness ratio; PARP, poly-ADP ribose polymerase; PD2; post-progression 2; PFS, progression-free survival; QALY, quality-adjusted life-year.
The probabilistic sensitivity analyses showed that olaparib plus bevacizumab had a 49.5% and 90.3% probability of being cost-effective versus bevacizumab alone at a willingness-to-pay threshold of €25 000 and €60 000 per QALY gained, respectively (Figure 3).
Probabilistic sensitivity analysis (PSA): incremental cost-effectiveness scatter plot (A) and cost-effectiveness acceptability curve (B). ICER, incremental cost-effectiveness ratio; WTP, willingness-to-pay.
DISCUSSION
Summary of Main Results
Our analysis showed that treatment with olaparib plus bevacizumab led to incremental costs of €58 295.31, a 2.77 increase in life-years, and 2.39 additional QALYs, being cost-effective (incremental cost-effectiveness ratio €24 371/QALY) considering a cost-effectiveness threshold between €25 000–€60 000/QALY.37 The cost difference was mainly explained by the pharmacological cost of olaparib in the olaparib plus bevacizumab arm and the subsequent treatment costs, primarily due to use of PARP inhibitors as a subsequent therapy in the bevacizumab arm. In terms of clinical benefit, olaparib plus bevacizumab significantly increased patients’ progression-free survival compared with bevacizumab monotherapy, resulting in significant improvement in life-years and QALYs.
Results in the Context of Published Literature
Our results were comparable with a US-based model that similarly evaluated the cost-effectiveness of olaparib plus bevacizumab compared with bevacizumab alone as a maintenance treatment among positive homologous recombination deficiency patients with advanced ovarian cancer, using the same four health state model structure.19 This study determined that olaparib plus bevacizumab was cost-effective ($56 836/QALY) at a willingness-to-pay threshold of $100 000/QALY. Probabilistic sensitivity analysis results demonstrated that olaparib plus bevacizumab was associated with a 97.0% probability of being cost-effective compared with bevacizumab alone.19 Another cost-effectiveness analysis has been recently published based on the final survival results from the PAOLA-1 trial.38 The analysis determined that at a US willingness-to-pay threshold of $150 000/QALY, olaparib plus bevacizumab was cost-effective relative to bevacizumab monotherapy for homologous recombination deficiency-positive patients, with an incremental cost-effectiveness ratio of $141 636/QALY; 74.3% of the simulations fell below the considered threshold.38 Our results align with these conclusions, as 49.5–90.3% of the simulations from our probabilistic sensitivity analysis fell below the €25 000–€60 000/QALY Spanish threshold.37 Moreover, several economic evaluations have been carried out regarding the indication of olaparib monotherapy in ovarian cancer. A Spanish economic evaluation31 analyzed the cost-effectiveness of olaparib as first-line maintenance therapy compared with no maintenance therapy in patients with advanced high-grade serous ovarian carcinoma and BRCA mutations from the Spanish National Health System perspective. This study determined that olaparib was cost-effective (€14 653.2/QALY) as first-line maintenance monotherapy in patients with advanced BRCA1/2-mutated high-grade serous ovarian carcinoma.31
The results of this economic evaluation are aligned with the previous olaparib monotherapy cost-effectiveness model, confirming the great value of the drug in terms of efficiency and being the only PARP inhibitor with cost-effectiveness evidence from the Spanish payer perspective.
Strengths and Weaknesses
The main strength of the analysis is the structure, a partitioned survival model with four health states. Partitioned survival models are directly linked to trial outcomes such as progression-free survival and overall survival, and are one of the most common cost-effectiveness model structures for advanced cancers. Furthermore, this model structure has recently been used in health technology assessments of treatments for advanced ovarian cancers.18 19
Our model is also subject to limitations. The progression-free survival and overall survival data from the PAOLA-1 trial were still immature at the model adaptation time; therefore, extrapolations were needed, as is usual in economic models evaluating oncological therapies. Final 5-year overall survival and progression-free survival results have been recently published,24 which are in line with the Kaplan-Meier estimates included in the model, which are more conservative (extrapolated overall survival: 63.5% olaparib and bevacizumab, 51.0% bevacizumab; updated overall survival: 65.5% olaparib and bevacizumab, 48.4% bevacizumab; extrapolated progression-free survival: 43.5% olaparib and bevacizumab, 20.3% bevacizumab; updated progression-free survival: 46.1% olaparib and bevacizumab, 19.2% bevacizumab), demonstrating its robustness. These results reinforce the efficacy of olaparib plus bevacizumab versus bevacizumab in a homologous recombination deficiency-positive population.24 Additionally, in the absence of data, it has been assumed that 68.1% of patients will receive palliative care based on the extrapolation of the Kaplan-Meier curve of the progression-free survival, considering the average number of patients progressing at 5 years in both arms. However, variations in base-case values, including the proportion and costs of patients in palliative care, were analyzed in the one-way sensitivity analysis and changes did not significantly affect the results.
Implications for Practice and Future Research
Understanding the cost-effectiveness of olaparib plus bevacizumab in the homologous recombination deficiency-positive population is therefore important for payers and healthcare decision-makers. Our analysis supports the use of the combination as a maintenance therapy in this population as it is not only highly efficacious but also has strong potential of being a cost-effective treatment option versus bevacizumab monotherapy in Spain. Olaparib plus bevacizumab would potentially improve patient outcomes and decrease the cost of illness management in subsequent years. Furthermore, failure to administer olaparib in this setting could deprive many patients of the benefit of a potential cure.
Finally, this study is focused on cost-effectiveness comparisons between olaparib and bevacizumab within the label indications. Although beyond the scope of this initial economic evaluation, further research is warranted to evaluate retreatment with PARP inhibitors from a clinical practice perspective.
CONCLUSION
This partitioned survival model, developed to assess the efficiency of olaparib in combination with bevacizumab compared with bevacizumab for maintenance treatment of patients with homologous recombination deficiency-positive advanced ovarian cancer, showed that olaparib plus bevacizumab is cost-effective from the payer perspective in Spain.
Data availability statement
Data are available upon reasonable request. Clinical data not published are available from the corresponding author (Sergio Cedillo, spain.healtheconomics@astrazeneca.com, ORCID 0009-0000-1378-1625) upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
Acknowledgments
The authors would like to thank Alba Bellmunt (Outcomes’10) for medical writing support.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors CG performed the analysis and interpretation of the data, critically revised the manuscript, approved the final version, and agrees to be accountable for all aspects of the work. SA reviewed the analysis and interpretation of the data, drafted the manuscript, approved the final version, and agrees to be accountable for all aspects of the work. SC, LM, ACCL, PVB, AC, and JAPF reviewed the analysis and interpretation of the data, critically revised the manuscript, approved the final version, and agree to be accountable for all aspects of the work. SC is the guarantor, accepting full responsibility for the work of the study, having access to the data, and controlling the decision to publish.
Funding The study was funded by Astrazeneca Farmacéutica Spain S.A.
Competing interests SC is an employee of AstraZeneca. CG and SA are employees of Outcomes’10, an independent consultancy that received fees from AstraZeneca for medical writing. Authors LM, ACCL, PVB, AC, and JAPF report personal fees from AstraZeneca, during the conduct of the study.
Provenance and peer review Not commissioned; externally peer reviewed.
Author note All data inputs were validated by an expert panel of five Spanish Healthcare professionals (three Oncologists and two hospital pharmacists) in two virtual meetings in 2021 to ensure that the study perspective is consistent with current Spanish clinical practice.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.