PharmacoEconomics

, Volume 33, Issue 11, pp 1155–1185 | Cite as

Cost Effectiveness of Chemotherapeutic Agents and Targeted Biologics in Ovarian Cancer: A Systematic Review

  • Insiya B. Poonawalla
  • Rohan C. Parikh
  • Xianglin L. Du
  • Helena M. VonVille
  • David R. Lairson
Systematic Review

Abstract

Background

Adjuvant chemotherapy is a key component of advanced ovarian cancer treatment, when surgery alone is not sufficient. Recurrence is common in ovarian cancer patients and most women require prolonged second-line and higher-line chemotherapy. With newer targeted therapies, modest improvements in survival and quality of life may be attained at substantial cost, but the relative economic efficiency of these newer agents remains unknown.

Objective

We undertook this systematic review to comprehensively evaluate the cost-effectiveness of various chemotherapeutic and targeted therapy alternatives for ovarian cancer.

Methods

We searched Medline, PubMed, and Embase databases to identify economic evaluations published over the last 18 years (1996–2014). From the 2513 unique papers retrieved, 74 full texts were selected for full-text review based on a priori eligibility criteria. Two authors independently reviewed these articles to determine eligibility for final review. The quality of the included studies was assessed using the Quality of Health Economic Studies (QHES).

Results

A total of 28 studies were included for reporting. Administration of intravenous cisplatin–paclitaxel combination chemotherapy for first-line treatment was the most cost-effective alternative (2014 US dollars [USD] equivalent incremental cost-effectiveness ratio [ICER] ~US$17,000–US$27,000 per life year gained [LYG]), while the use of bevacizumab did not demonstrate similar value for money (2014 USD equivalent ICER was greater than US$200,000 per quality-adjusted life-year [QALY]). For second-line treatment, the use of platinum–paclitaxel combination or platinum monotherapy was cost-effective compared with platinum monotherapy or best supportive care, respectively, in women with recurrent platinum-sensitive disease. For patients with partial platinum sensitivity, pegylated liposomal doxorubicin (PLD) plus trabectedin may be cost-effective (2014 USD equivalent ICER was ~US$57,000–US$62,000 per QALY) compared with PLD alone. For recurrent platinum-resistant cases, there was limited evidence to conclude the most valuable treatment; though one study showed that best supportive care was most cost-effective, while second-line monotherapy with doxorubicin (2014 USD equivalent ICER was ~US$90,000 per LYG) may also be cost-effective compared with best supportive care.

Conclusions

Despite varying methodological approaches and multiple sources for cost and effectiveness inputs, this systematic review demonstrated that standard platinum–taxane combination chemotherapy for first-line treatment was most cost-effective. There was unanimous agreement that bevacizumab was not a cost-effective front-line therapy compared with platinum–taxane combination for the overall ovarian cancer population, though its use in the high-use population may yield better value. For second-line treatment, platinum-based chemotherapy remained cost-effective among patients with recurrent platinum-sensitive disease, while there was limited evidence to conclude the most valuable treatment alternative among patients with recurrent platinum-resistant disease. Future research incorporating real-world data is essential to corroborate findings from trial-based economic evaluations. In addition, for improving consistency in reporting and quality of studies, incorporating QALYs in this population is important, especially since chemotherapy is administered for lengthy periods of time.

Key Points

There is considerable variation in the type of costs included as well as the sources used to estimate the cost of treatment modalities resulting in different cost-effectiveness outcomes.

The review found that the cisplatin–paclitaxel combination was the most cost-effective first-line alternative; however, carboplatin is the preferred platinum-based chemotherapy agent, but cost-effectiveness studies evaluating carboplatin are limited.

Targeted therapies like bevacizumab, although more effective compared with the platinum–taxane combination, were not found to be cost-effective.

For second-line treatment among recurrent platinum-sensitive patients, platinum-based chemotherapy remained cost-effective, while evidence for recurrent platinum-resistant patients was limited.

1 Introduction

An estimated 21,980 new cases of ovarian cancer were diagnosed in the United States in 2014 and 65,500 new cases in Europe in 2012, accounting for 3–4 % of all cancers in women [1, 2]. Despite a decrease in incidence and mortality rates over the last few years, this malignancy continues to cause the highest number of gynecological related cancer deaths [1]. Survival rates in ovarian cancer are highly dependent on patient age and stage at diagnosis; elderly women (65+ years) are half as likely to survive and patients with distant stage tumors are three times more likely to die within 5 years after diagnosis [3, 4]. Primarily diagnosed at an advanced stage (70 % cases), ovarian cancer overall poses a very poor prognosis [5]. Recurrence rates for stages III–IV are as high as 70–90 % [6], making this malignancy almost non-curable, requiring maintenance and palliative chemotherapeutic treatments for lengthy periods of time.

The National Comprehensive Cancer Network (NCCN) guidelines recommend appropriate first-line surgical staging, debulking, and adjuvant systemic chemotherapy for treatment of ovarian cancer [5]. For tumors staged as IA or IB, surgery may be a sufficient treatment approach; however, for patients staged as IC or higher, administration of platinum–taxane combination adjuvant chemotherapy is recommended [3, 5]. Utilization of expensive newer targeted therapies (e.g., bevacizumab) and the prolonged use of chemotherapies may increase the overall economic burden of ovarian cancer. Thus, providing cost-effective first-line treatment and subsequent care is essential from both private and public standpoints. With the introduction of taxane in the late 1990s as part of routine combination treatment, several cost-effectiveness and cost-utility evaluations have been conducted for currently available treatment options. A few studies have previously reviewed the cost-effectiveness literature in ovarian cancer [7, 8, 9, 10]; however, none of those had a formal systematic review approach. In addition, economic evaluations on newer targeted therapies have not been reviewed and summarized in the overall cost-effectiveness literature. In order to better inform decision makers and clinicians, the objective of this study was to synthesize evidence by conducting a systematic review of economic evaluations of chemotherapeutic and targeted treatment alternatives for ovarian cancer over the past 18 years and assess the quality of these studies.

2 Methods

2.1 Eligibility Criteria

The PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines by Moher et al. [11] were followed for review and reporting procedures. Studies eligible for inclusion represented full economic evaluations (i.e., cost-effectiveness analyses [CEA] or cost-utility analyses [CUA]) that evaluated a chemotherapeutic or targeted therapy agent for the treatment of first-line (i.e., primary adjuvant or neoadjuvant) or subsequent-line ovarian cancer. Studies with partial economic evaluations, cost-minimization analyses or cost-benefit analyses were excluded. Studies were required to report both costs and survival outcomes (i.e., overall, progression-free or quality-adjusted survival). Research articles from journals published in the English language were included; comments, editorials, letters, news, correspondence, study protocols, case reports, case series, narrative and systematic reviews were excluded. We also excluded economic evaluations that assessed other aspects of care, such as screening, surgical treatment, or palliative care.

2.2 Information Sources and Search

Medline (Ovid), PubMed (NLM), and Embase (Ovid) were searched with the assistance of a public health librarian (HVV) experienced in developing search strategies for systematic reviews. Search concepts were economic evaluations, treatment (i.e., chemotherapy and targeted agents), and ovarian cancer. The search was last updated on August 31, 2014, and included articles published from 1996 (after the introduction of taxane for ovarian cancer) through 2014. Manufacturers or study authors were not contacted to identify any other unpublished sources of information. RefWorks (ProQuest) was used to store all citations found in the search process, and to check for duplicates. Search strategies and results were tracked using the PRIMARY Excel Workbook for Systematic Reviews [12]. Strategies for each database searched can be found in Appendix A (see electronic supplementary material).

2.3 Study Selection

Two authors (IBP and RCP) independently screened a random sample of 66 titles and abstracts for study eligibility in which they were blinded to authors and journal titles to reduce the potential bias for being influenced by author name or journal title, and they reached a moderate level of agreement on study eligibility (Cohen’s κ = 0.65, SE of κ = 0.24) [13, 14]. Disagreement between the reviewers was resolved through discussion. After initial review, they independently screened all titles and abstracts, still blinded to authors and journal titles, using the Excel Workbook for 2 Screeners [15]. Decision data were compiled in a single Excel workbook and discrepancies were discussed by both reviewers until a consensus was reached regarding eligibility. The full text of the eligible articles was retrieved and reviewed independently by two authors (IBP and RCP) and the study was included for final review if it met all of the eligibility criteria (Fig. 1).
Fig. 1

PRISMA modeled flowchart of the screening and eligibility evaluation phases

2.4 Data Collection Process and Data Items

A data-coding evidence table was developed to abstract citation, study, costs, and effectiveness information from each study. The evidence table was pilot tested on four studies by two coders (IBP and RCP) and evaluated for consistency by the principal investigator (DRL) prior to final use. Abstracted variables included first author, study population, study characteristics (e.g., perspective, time horizon, discount rate, model type, and sensitivity analysis), treatment arms (i.e., drug, dosage, and administration), source of effectiveness data, resource use and source of cost data, cost outcomes, effectiveness outcomes, incremental cost-effectiveness ratio (ICER), author’s conclusion, and limitations. ICER values for all studies were converted to 2014 US dollars. For studies that were reported in US dollars, medical care consumer price index was used to adjust for inflation, and for studies that reported in currencies other than US dollars, purchasing power parity was used [16, 17].

2.5 Quality Assessment

The 16-item validated Quality of Health Economic Studies (QHES) instrument [18], which is based on Drummond’s checklist and the United States Public Health Service Panel on Cost-Effectiveness in Health and Medicine, was used to assess the quality of the included studies. Two authors (IBP and RCP) independently reviewed the studies and scored each question as 1-yes or 0-no. To ensure consistency in the quality assessment, additional decisions similar to those used by Zhang et al. [19] were made and are shown in Table 3. In addition to the objective quality assessment using the QHES instrument, a subjective assessment of each study was also conducted.

3 Results

3.1 Study Selection

A total of 4309 citations were found through the electronic database search process; approximately 40 % (κ = 1796) were duplicate citations and were removed from further review. Of the remaining 2513 citations, 95 % (κ = 2375) of the titles and abstracts were deemed ineligible during the screening process as they were not economic evaluations. Of the 74 papers that qualified for a full-text review, one was not available for review [20] and 45 did not meet the study eligibility criteria. The flowchart in Fig. 1 outlines the search results indicating exclusions at various stages of the review process. Twenty-eight publications were included in this systematic review [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48]. We have summarized the findings of these studies by type of treatment: first-line therapy (18 studies), maintenance therapy (1 study), second-line chemotherapy for recurrent platinum-sensitive disease (7 studies), second-line chemotherapy for recurrent platinum-resistant disease (1 study), multiple-line chemotherapy regimen (1 study), and other treatment regimen (1 study).

3.2 First-Line Therapy

3.2.1 Cisplatin–Paclitaxel Compared with Cisplatin–Cyclophosphamide (κ = 8)

The earliest economic evaluations in ovarian cancer were conducted soon after clinical trial-based evidence by the US Gynaecologic Oncology Group (GOG) supported taxane as an appropriate first-line combination therapy with cisplatin. From 1996 to 1998 a series of cost-effectiveness evaluations compared the standard cisplatin–cyclophosphamide with cisplatin–paclitaxel for treatment of advanced ovarian cancer patients [21, 22, 23, 24, 25]. These studies were conducted in various countries and from varying healthcare perspectives. Results from the GOG-111 trial [49] formed the basis of clinical estimates for all studies except the study by Covens et al. [22] that used patient charts as their source of effectiveness data. Major cost categories included drug acquisition, administration, hospitalizations, adverse event management and follow-up monitoring. Table 1 provides details on specific cost inputs from each study. Two Canadian studies were conducted from the perspective of the Ministry of Health (Ontario) [22, 23] and one study from the Canadian healthcare system perspective [25]. The ICER estimates for these studies were Can$20,355 per life-year gained (LYG), Can$32,213 per LYG, and Can$11,600–Can$24,200 per quality-adjusted progression-free life-year (PFLY), depending on second-line treatment, respectively (see Table 2 for 2014 USD equivalent ICERs). Given that paclitaxel may be infused over either 24 or 4 h, McGuire et al. evaluated the cost-effectiveness of cisplatin–paclitaxel separately for an inpatient and outpatient setting, respectively [24]. From a healthcare provider perspective, the corresponding ICER estimates were US$19,820 per LYG and US$21,222 per LYG. Another study by Messori et al. utilized a lifetime analysis to assess cost-effectiveness of cisplatin–paclitaxel. Despite their social perspective, the investigators only included direct costs related to treatment, with the resultant ICER estimated as US$19,603 per LYG [21]. The study by Berger et al. was the only study from a European perspective [26], with analyses conducted separately for six countries. A major proportion of the total costs in the cisplatin–paclitaxel group was drug-related and in the cisplatin–cyclophosphamide group was hospitalization-related. The ICER estimate per LYG ranged from US$6395 in Spain to US$11,420 in Italy.
Table 1

Overview of the methodology of included studies

Study

Population

Study characteristics

Treatment

Effectiveness data

Resource use/cost data

Primary treatment (κ = 18)

 Cisplatin–paclitaxel compared with cisplatin–cyclophosphamide (κ = 8)

  Messori et al. [21]

Advanced ovarian cancer patients

P: social (only direct costs); TH: lifetime; DR: no discounting in base case; M: empirical study using secondary data; S: multiple one-way sensitivity analyses

(1) IV 750 mg/m2 CYC + IV 75 mg/m2 CIS q3w

(2) IV 135 mg/m2 PAC + IV 75 mg/m2 CIS q3w

Clinical trial survival curves

Direct costs related to drug, drug administration, hospitalization and febrile neutropenia. Hospitalization costs and cost to treat febrile neutropenia episode was based on literature

  Covens et al. [22]

Stage IIIc/IV advanced ovarian cancer

P: Ministry of Health (Ontario, Canada); TH: lifetime; DR: no discounting; M: simple linear modeling; S: multiple sensitivity analyses

(1) CYC + CIS

(2) PAC + CIS

Patient charts

Only hospital-based inpatient and outpatient costs were considered (1993 Canadian dollars). Resource use data was collected from patient charts and associated costs were estimated using Ontario Schedule of Benefits and Case Cost System of Sunnybrook Center

  Elit et al. [23]

Suboptimally debulked stage III and IV epithelial ovarian cancer patients

P: Ministry of Health (Ontario, Canada); TH: lifetime; DR: 5 %; M: decision tree analysis; S: multiple one-way sensitivity analyses

(1) IV 750 mg/m2 CYC + IV 75 mg/m2 CIS q3w

(2) IV 135 mg/m2 PAC + IV 75 mg/m2 CIS q3w

GOG-111 survival curves

Costs (1993 Canadian dollars) included chemotherapy, administration, major adverse events and follow-up care. McMaster University Medical Centre model was used for administration costs, chemotherapy costs were based on a hospital, major adverse events costs were derived using resource intensity weights and follow-up care was based on Ontario Health Insurance Plan

  McGuire et al. [24]

Advanced stage ovarian cancer patients

P: healthcare provider; TH: lifetime; DR: 4 % (only costs); M: economic model; S: multiple one-way sensitivity analysis; multivariate Monte Carlo simulation

(1) CIS + CYC for 6 cycles

(2) CIS + PAC (infusion over 24 h in the inpatient setting and 4 h in the outpatient setting—total 6 cycles)

Survival and adverse events data from GOG-111 trial

Cost were included for 6 major categories: drug acquisition, physician visits, laboratory/diagnostic procedures, inpatient/outpatient treatment facilities, adverse event management, and follow-up costs. Cost data was derived from the Oncology Therapeutics Network program, Resource Based Relative Value Schedule and data provided by an oncology practice site. Resource utilization data was derived from the GOG trial and expert opinion (to account for differences between real-world and clinical trial setting)

  Ortega et al. [25]

Stage III or IV ovarian cancer patients

P: Canadian healthcare system; TH: 24 months; DR: no discounting; M: decision tree model; S: multiple one-way sensitivity analysis

(1) 75 mg/m2 at 1 mg/min CIS + 750 mg/m2 CYC for total 6 cycles

(2) 75 mg/m2 at 1 mg/min CIS + 135 mg/m2 over 24 h PAC for 6 cycles

Clinical trial and phase II trial data

Retrospective chart review over a 4-mo time period was conducted to obtain real-life hospital resource utilization data for hospitalization, outpatient clinic visits, chemotherapy drug and administration, antiemetic use, adverse event management, laboratory work, patient monitoring, and physician fees. Cost data (1996 Canadian dollars) were derived from various sources, such as pharmacy ordering catalogs, pharmacy and nursing workload measuring statistics, Princess Margaret Hospital, Ontario Hospital Association, Departments of Biochemistry, Microbiology and Diagnostic Imaging and Schedule of Benefits

  Berger et al. [26]

Stage III ovarian cancer with >1 cm residual after surgery or stage IV

P: national health service; TH: not reported; DR: no discounting; M: empirical study using clinical trial data; S: multiple one-way sensitivity analysis

(1) 75 mg/m2 at 1 mg/min CIS + 750 mg/m2 CYC for total 6 cycles

(2) 75 mg/m2 at 1 mg/min CIS + 135 mg/m2 over 24 h PAC for 6 cycles

Clinical trial data (GOG-111)

Costs included were chemotherapy costs, costs to treat adverse events, hospitalization costs during treatment, consultations, laboratory testing, and costs for all other investigations. Country-specific cost estimates were derived via face-to-face interviews with experts, literature searches or hospital price lists

  Neymark et al. [27]

Women with FIGO stage IIb, IIc, III or IV epithelial ovarian cancer with or without successful debulking surgery

P: Belgian health insurance and financing system; TH: 3.73 years; DR: 3 %; M: empirical study using clinical trial data; S: CEACs using 5000 bootstrap replicates

(1) IV 750 mg/m2 CYC + IV 75 mg/m2 CIS

(2)175 mg/m2 PAC for cycle 1 set to increase to 200 mg/m2 + IV 75 mg/m2 CIS

Clinical trial data

Resource use information collected during the trial was used. Information collected included hospital stay during treatment, cytotoxics and premedication used outpatient consultations, diagnostic tests and concomitant medications. Unit costs (1998 Euros) were obtained from national schedule of tariffs, drug reimbursements and per diem hospital payments

  Limat et al. [28]

Women with FIGO stage IIc, III or IV ovarian cancer. Study excluded patients with localized disease, patients older than 80 years and who received only one cycle of treatment

P: hospital payer; TH: 3 years; DR: no discounting; M: empirical study using secondary data; S: multiple one-way sensitivity analysis

(1) IV 750 mg/m2 CYC as 15-min infusion + 75 mg/m2 CIS as 3-h infusion once q3w

(2) IV 135 mg/m2 PAC as 24-h infusion on day 1 + IV 75 mg/m2 CIS as a 3-h infusion on day 2 once q3w

Observational data from patients who received chemotherapy at a French university hospital

Direct costs (2000 US dollars) related to hospitalization, chemotherapy, supportive drugs and blood products. Per diem hospitalization costs and drug costs were obtained from the accounting system of Besancon University Hospital and wholesale price list of Besancon Hospital Pharmacy

 Intraperitoneal compared with intravenous route of administration (κ = 2)

  Bristow et al. [29]

Stage III ovarian cancer patients who had undergone surgery

P: societal; TH: lifetime; DR: 3 % (only outcome); M: decision analytic model; S: multiple one-way sensitivity analysis and probabilistic sensitivity analysis

(1) IV PAC (3 h) and IV CAR

(2) IV PAC (24 h) and IP CIS + IP PAC

Median survival time from GOG protocol #172 and #158

Costs (2006 US dollars) included were hospital/professional costs related to chemotherapy, intraperitoneal and intravenous procedures, hospitalization for treatment-related toxicity, lost wages, caregiver support. Cost estimates were derived from the Maryland Health Services Cost Review Commission database for hospital charges and common use Johns Hopkins Medical Institutions administrative databases for professional fee charges

  Havrilesky et al. [30]

Women with optimally resected stage III ovarian cancer

P: societal; TH: 7 years (11.5 and 35 years in sensitivity analysis); DR: 3 %; M: Markov model; S: multiple one-way sensitivity analyses and probabilistic sensitivity analysis

(1) 175 mg/m2 PAC over 3 h CAR (7.5 AUC) on day 1 of 21-day cycle (IV-CAR/PAC)

(2) 135 mg/m2 IV PAC over 24 h day 1 + 75 mg/m2 IV CIS day 2 of 21-day cycle (IV-CIS/PAC—GOG-172, GOG-158)

(3) 135 mg/m2 IV PAC over 24 h day 1 + 100 mg/m2 IP CIS on day 2 + 60 mg/m2 IP PAC on day 8 of 21-day cycle (IP-CIS/PAC)

GOG-172, GOG-158, GOG-114 and GOG statistical and data center

Costs (2006 US dollars) related to treatment, inpatient admission, adverse event admission, laboratory work, caregiver, IP port placement/removal were included. Treatment costs were based on 2006 Medicare reimbursement and adverse event charges (converted to costs assuming 60 % of charges) based on AHRQ HCUP Nationwide Inpatient Sample

 Targeted biologics (κ = 4)

  Cohn et al. [31]

Stage III ovarian cancer patients with recently completed cytoreductive surgery

P: third party payer; TH: not applicable; DR: not applicable; M: decision tree analysis; S: multiple one-way sensitivity analyses

(1) PAC (175 mg/m2) + CAR (6 AUC) on day 1 every 21 d for 6 cycles

(2) PAC + CAR + BEV (15 mg/kg on day 1 every 21 d starting at cycle 2)

(3) (PAC + CAR + BEV) + maintenance BEV (15 mg/kg day 1 every 21 d for 16 cycles)

GOG-218 and clinical experience

Medicare reimbursements direct costs (2009 US dollars). Costs for chemotherapy, administration, supportive medications, and intestinal perforation (medical and surgical) were included

  Barnett et al. [32]

Stage III/IV ovarian cancer patients. High-risk cohort included suboptimally debulked stage III and any stage IV disease

P: third-party payer; TH: 39 months (i.e., until disease progression or death in 50 % of cohort); DR: 3 %; M: modified Markov model; SA: Monte-Carlo simulation (1000 trials), multiple one-way sensitivity analyses

Base case:

(1) PAC + CAR for all patients

(2) PAC + CAR + BEV (7.5 mg/m2 for 12 cycles) for all patients

(3) PAC + CAR + BEV (7.5 mg/m2 for 12 cycles) for all high-risk patients

Alternative scenario:

(1–3) Base-case treatment alternatives

(4) Test for PAC + CAR + BEV (BEV administered only if biomarker predictive test is positive in the high risk cohort)

Overall survival and adverse-event outcomes from the ICON-7 trial

Costs (2011 US dollars) were included for treatment and adverse events. Medicare reimbursement data was used to estimate cost of chemotherapy regimen (included routine laboratory work, complete blood count, metabolic panel before each cycle, premedication, chemotherapy drug, and administration fees).AHRQ HCUP Nationwide Inpatient Sample was used to estimate mean/median hospitalization costs for bowel perforation and thromboembolic events. Hypertension costs were considered for 1 y

  Chan et al. [33]

High-risk stage IIIC/IV ovarian cancer patients who completed cytoreductive surgery

P: healthcare system; TH: 46 months; DR: no discounting; M: Markov model; S: multiple one-way sensitivity analysis and probabilistic sensitivity analysis

(1) PAC + CAR q3w for 6 cycles

(2) PAC + CAR + BEV q3w for 6 cycles + maintenance BEV for additional 12 cycles

ICON-7

Drug costs from average Medicare wholesale price. Using the adverse events reported in ICON-7, associated costs were estimated using US costs for major complications

  Mehta and Hay [34]

Previously untreated stage III/IV epithelial ovarian cancer patients who completed cytoreductive surgery (GOG-218 and ICON-7). ICON-7-based model had patients with >1 cm residual disease

P: societal; TH: until death of 99 % of the cohort; DR: 3 %; M: Markov model; S: multiple one-way sensitivity analysis, threshold sensitivity analysis and probabilistic sensitivity analysis

(1) PAC (175 mg/m2) + CAR (6 AUC) on day 1 every 21 d for 6 cycles

(2) PAC + CAR + BEV (15 mg/kg on day 1 every 21 d starting at cycle 2) + maintenance BEV (15 mg/kg day 1 every 21 d for 16/12 cycles for GOG-218/ICON-7)

GOG-218, ICON-7 and clinical experience

Micro-costing and gross-costing approach. Direct costs (2013 US dollars) included drug acquisition cost, laboratory cost, imaging cost, office visit, cost of treating side effects, drug administration, and pharmacist cost. Indirect costs included unpaid patient and caregiver time and costs

 Miscellaneous (κ = 4)

  Messori and Trippoli [35]

Newly diagnosed advanced ovarian cancer patients

P: not stated; TH: lifetime; DR: 5 %; M: empirical study based on literature review; S: multiple one-way sensitivity analyses

(1) CIS at conventional doses

(2) PAC + CIS at conventional doses

(3) High-dose chemotherapy with hematopoietic rescue

Literature review (meta-analysis) of clinical trials

Literature review of pharmacoeconomic studies

  Orr et al. [36]

Women with FIGO stage III ovarian cancer with no previous chemotherapy/radiation therapy, optimal residual disease (<2 cm) and no coexisting neoplasm

P: not stated; TH: 3 years; DR: no discounting; M: empirical study using primary data; S: one-way analysis with chemotherapy cost discounted at 25 %

(1) Platinum + CYC (600 mg/m2)

(2) Platinum + PAC (135 mg/m2)

The doses administered for platinum drugs were: CIS 100 mg/m2; CAR 300 mg/m2

Primary phase II trial data

No details provided

  Dalton et al. [37]

Advanced stage ovarian cancer patients

P: healthcare system; TH: 48 months; DR: no discounting; M: Markov model; S: multiple one-way sensitivity analysis and probabilistic sensitivity analysis

(1) CAR (AUC 6) + PAC (180 mg/m2) q3w

(2) CAR (AUC 6) + dose-dense PAC (80 mg/m2) weekly

Japanese Gynecologic Oncology Group (JGOG) phase III RCT

Treatment costs included physician evaluation, chemotherapy administration, drug costs and laboratory costs. Cost for treating adverse events (including hospitalizations) and discontinuation costs were also considered. Drug costs were derived from Medicare reimbursement data and pharmacy at the Helen Diller Family Comprehensive Cancer Center at the University of California, San Francisco

  Lairson et al. [38]

Stage I–IV elderly ovarian cancer patients

P: payer; TH: lifetime; DR: 3 %; M: empirical study using secondary data; S: utility measurements were varied under a best- and worst-case scenario

(1) No chemotherapy

(2) Other non-platinum regimen

(3) Platinum-based regimen

(4) Platinum-taxane regimen

Patient-level overall survival data obtained from the Surveillance, Epidemiology, and End Results (SEER) Medicare data set

By using the amount Medicare paid for each claim, total healthcare costs (2009 US dollars) were estimated by including costs for various Part A and Part B services (i.e., inpatient services, outpatient visits and procedures, physician fees, skilled nursing facility, hospice care, costs for devices and medical equipment) starting from treatment until death or end of study. Costs were adjusted for inflation and geographic differences

 Maintenance chemotherapy (κ = 1)

  Lesnock et al. [39]

Patients with stage III or greater disease; undergone surgery with residual disease either optimal or sub-optimal

P: healthcare system; TH: 10 years; DR: 3 %; M: Markov model; S: multiple one-way or two-way sensitivity analysis

(1) CAR AUC 7.5 + 135 mg/m2 PAC over 3 h—21-day cycle for 6 cycles

(2) CAR AUC 7.5 + 135 mg/m2 PAC over 3 h—28-day cycle + maintenance PAC for 12 cycles

(3) CAR + PAC (AUC 6 + 175 mg/m2 for 1 cycle, AUC 6 + 175 mg/m2 + 15 mg/kg for 5 cycles) + 15 mg/kg BEV + maintenance BEV for 16 cycles

Data for PFS, overall survival were obtained from GOG protocol 158, 178, 218 or from the literature

Costs (2009 US dollars) included were for drug and administration, treatment complications and surveillance. Cost estimates were derived from hospital costs, Medicare reimbursement rates, the AHRQ database, the AHRQ HCUP database, the American Medical Association database, the CMS Physician Payment database, or Red Book AWP medication costs. For BEV, the cost was derived from the author’s home institution

 Recurrent platinum-sensitive (κ = 7)

  Fisher and Gore [40]

Patients who had received one prior platinum-based chemotherapy regimen and had experienced recurrence

P: National Health Service (UK) and Personal Social Services; TH: lifetime; DR: 3.5 %; M: decision analysis model based on NICE Multiple Technology Appraisal; S: one-way and probabilistic sensitivity analysis

(1) Three 0.25-mg vials and one 1-mg vial of TRA + one 50-mg vial of PLD for 6.9 months

(2) Two 50-mg vials of PLD alone

Extrapolated survival data of the platinum-sensitive sub-group from the OVA-301 trial were used

Costs (2011 UK pounds) for drug acquisition, administration, medical management, and treating adverse events were estimated from the British National Formulary and appropriate UK Health Resource Group codes

  Lee et al. [41]

Women with platinum-sensitive ovarian cancer

P: societal; TH: 10 years; DR: no discounting; M: Markov model; S: multiple one-way sensitivity analyses and probabilistic sensitivity analysis

(1) PAC + CAR

(2) PLD + CAR

Clinical trial data

Direct medical and nonmedical costs (2011 US dollars) were included. Costs associated with drug, drug administration, adverse events, chemotherapy response, BSC and follow-up tests were estimated through micro-costing using Korean fee schedule. Hospice costs and transportation costs were also estimated

  Havrilesky et al. [42]

Patients with recurrent platinum-sensitive ovarian, peritoneal or tubal cancer

P: third-party payer; TH: 24 months; DR: no discounting; M: Markov model; S: multiple one-way and Monte Carlo probabilistic sensitivity analysis

(1) 30 mg/m2 DOC on d 1,8 q3w followed by CAR AUC 6 q3w at first progression or after 6 cycles DOC for stable disease or a partial response

(2) 30 mg/m2 DOC on d 1,8 + CAR AUC 6 mg/mL per min on day 1 q3w

Phase II clinical trial data

Medicare reimbursement data (2010 US dollars) was used to estimate costs per cycle, which included one physician visit, chemotherapy drugs, infusion/treatment and pre-treatment medications. Cost of adverse effects (neutropenia, neurotoxicity) and cost of administering granulocyte–colony-stimulating factor or erythropoietin was included. Costs that were assumed to be similar between treatment groups were not included

  Montalar et al. [43]

Women with platinum-sensitive relapsed ovarian cancer

P: Spanish national health system; TH: lifetime; DR: 3.0 %; M: Markov model; S: multiple one-way sensitivity analyses and probabilistic sensitivity analysis

(1) PLD alone

(2) TRA + PLD

OVA-301 trial

Direct costs (2011 Euros) related to drug, drug administration, medical management and adverse events were included. Unit costs were obtained from a healthcare cost database and drug costs were obtained from General Council of Official Pharmaceutical Colleges. Adverse event management cost was obtained from literature

  Case et al. [44]

Advanced-stage ovarian cancer patients who underwent surgery, received combination platinum–taxane adjuvant chemotherapy and recurrence was experienced after 6 months of completing primary chemotherapy

P: third-party payer; TH: not reported; DR: no discounting; M: decision analysis model; S: one-way (costs and effectiveness were varied)

(1) BSC

(2) Second-line monotherapy (CAR)

(3) Second-line combination (CAR + PAC

(4) Third-line previous monotherapy (DOX)

(5) Third-line previous combination (DOX)

(6) Fourth-line previous monotherapy (GEM)

(7) Fourth-line previous combination (GEM)

A review of published phase II, III trials and clinical experience

BSC costs (2004 US dollars) included outpatient office visits, emergency department visits, hospitalizations, home health care, medications and palliative procedures. Costs for other treatment strategies included chemotherapy drug and other related costs. Direct costs were estimated by adjusting local charges using a cost-to-charge ratio of 60 %. Laboratory and procedure costs were estimated in consultation with the University of Alabama at Birmingham. Pharmacy costs were calculated using average wholesale drug costs

  Havrilesky et al. [45]

Recurrence was experienced after 6 months of completing primary chemotherapy

P: payer; TH: 42 months; DR: no discounting; M: Markov model; S: multiple one-way sensitivity analysis

(1) CAR AUC 5 on day 1 of a 21-day cycle for a total 6 cycles

(2) 175 mg/m2 PAC + CAR AUC 5 on day 1 of a 21-day cycle for a total 6 cycles

(3) IV 1000 mg/m2 GEM on day 1 and 8 + CAR AUC 4 on day 1 of a 21-day cycle for total 6 cycles

PFS data was obtained from 2 clinical trials

Costs (2006 US dollars) for treatment (including drug cost, infusion, laboratory work, and professional fees) and management of adverse effects were estimated using data from the national 2007 physician fee schedules, and Medicare reimbursement data

  Main et al.: analysis 2 [46]

Platinum-sensitive ovarian cancer patients

P: National Health Service (UK); TH: lifetime; DR: 1.5 % (only for outcomes); M: probabilistic decision analysis; S: CEAC were computed

(1) TOP

(2) PAC

(3) PLD

(4) PAC + platinum

(5) platinum-based

(6) CAP

Systematic review of multiple trials

Costs (2003–04 UK pounds) included were drug acquisition and administration, premedication, monitoring and managing of adverse events. Unit costs were obtained from British National Formulary, national cost databases, literature and industry submission data

 Recurrent platinum-resistant (κ = 1)

  Rocconi et al. [47]

Advanced (stage III/IV) epithelial ovarian cancer patients who underwent cytoreductive surgery and platinum–taxane chemotherapy with recurrence within 6 months

P: US—third party payer; TH: 1 year; DR: not applicable; M: Markov hypothetical cohort of 4000 patients; S: multiple one-way sensitivity analyses

(1) BSC

(2) Second-line monotherapy (DOX 40 mg/m2 for 4 mo)

(3) Second-line combination (GEM 750 mg/m2 + CIS 30 mg/m2 on d 1,8 every 21 d for 4 mo)

(4) Third-line previous monotherapy (TOP 1.5 mg/m2 for 5 d every 21 d for 3 cycles)

(5) Third-line previous combination (TOP 1.5 mg/m2 for 5 d every 21 d for 3 cycles)

Multiple clinical trials and clinical judgment

Direct costs (2004 US dollars) only estimated by adjusting local charges with cost-to-charge ratio (60 %). Laboratory and procedure estimates from University of Alabama at Birmingham, drug costs based on average wholesale price

 Multiple lines (κ = 1)

  Fedder et al. [48]

Women with epithelial ovarian carcinoma, FIGO stage I–IV

P: not stated; TH: 5 years; DR: 3 %; M: Markov cohort model; S: none

(1) First-line CAR followed by second-line TOP

(2) First-line CAR followed by second-line DOX

(3) First-line CAR + PAC followed by second-line DOX

(4) First-line CAR + PAC followed by second-line TOP

Multiple phase II and phase III clinical trials

Costs (2002 Euros) included treatment costs (including antiemetic therapy), inpatient care, outpatient aftercare. Costs were estimated from department of gynecology, hospital pharmacy and finance department of Friedrich Schiller University Hospital Jena

 Other (κ = 1)

  Main et al.: analysis 1 [46]

Advanced ovarian cancer patients receiving subsequent/second-line treatment (includes both platinum-sensitive and platinum-resistant patients)

P: National Health Service (UK); TH: lifetime; DR: 1.5 % (only for outcomes); M: probabilistic decision analysis; SA: subgroup analysis, additional cost and effectiveness information considered and CEAC were computed

(1) PAC

(2) PLD

(3) TOP

Systematic review of multiple trials

Costs (2003–04 UK pounds) included were drug acquisition and administration, premedication, monitoring and managing of adverse events. Unit costs were obtained from British National Formulary, national cost databases, literature and industry submission data

AHRQ HCUP Agency for Healthcare Research and Quality—Healthcare Cost and Utilization Project, AUC area under the concentration–time curve, AWP average wholesale price, BEV bevacizumab, BSC best supportive care, CAP cyclophosphamide–doxorubicin–cisplatin, CAR carboplatin, CEAC cost-effectiveness acceptability curve, CIS cisplatin, CMS Centers for Medicare and Medicaid Services, CYC cyclophosphamide, DOC docetaxel, DOX doxorubicin, DR discount rate, FIGO International Federation of Gynecology and Obstetrics, GEM gemcitabine, GOG US Gynaecologic Oncology Group, ICON-7 Gynecologic Cancer Intergroup International Collaboration on Ovarian Neoplasms 7, IP intraperitoneal, IV intravenous, M model, P perspective, PAC paclitaxel, PFS progression-free survival, PLD pegylated liposomal doxorubicin, q3w every 3 weeks, RCT randomized controlled trial, S sensitivity analysis, TH time horizon, TOP topotecan, TRA trabectedin

Table 2

Overview of the outcomes of included studies

Study

Treatment

Total costs/incremental costs

Effectiveness/incremental effectiveness

Incremental cost-effectiveness ratio (ICER)

Authors’ conclusion (AC) and limitations (AL)

Primary treatment (k = 18)

 Cisplatin–paclitaxel compared with cisplatin–cyclophosphamide (κ = 8)

  Messori et al. [21]

(1) IV 750 mg/m2 CYC + IV 75 mg/m2 CIS q3w

(2) IV 135 mg/m2 PAC + IV 75 mg/m2 CIS q3w

Total cost per 100 patients:

(1) US$252,279

(2) US$986,002

Incremental overall survival for 100 patients:

(1) Referent

(2) 40 y of life gained

(1) Referent

(2) US$19,603 per LYG

AC: PAC is expensive but PAC-based chemotherapy as first-line treatment was found to be cost-effective using undiscounted and discounted life-years gained

AL: No quality-of-life adjustments were conducted

  Covens et al. [22]

(1) CYC + CIS

(2) PAC + CIS

Total cost per patient:

(1) Can$36,837

(2) Can$50,054

Incremental cost per patient:

(1) Referent

(2) Can$13,217

Overall weighted survival:

(1) 15.6 mo

(2) 23.4 mo

Incremental weighted survival:

(1) Referent

(2) 7.8 mo

(1) Referent

(2) Can$20,355 per LYG (2014 USD equivalent 16,948 per LYG)

AC: PAC along with CIS was found to be cost-effective first-line treatment for advanced ovarian cancer patients

AL: Data based on a small number of patients, surgery cost was not included and limited generalizability

  Elit et al. [23]

(1) IV 750 mg/m2 CYC + IV 75 mg/m2 CIS q3w

(2) IV 135 mg/m2 PAC + IV 75 mg/m2 CIS q3w

Total cost per patient:

(1) US$5,228

(2) US$17,469

Mean overall survival:

(1) 2.06 y

(2) 2.44 y

(1) Referent

(2) US$32,213 per LYG (2014 USD equivalent 26,822 per LYG)

AC: PAC-based therapy extends survival but with increased costs and the government may or may not be able to cover PAC-based therapy

AL: Effectiveness data is based on one clinical trial, cost data was obtained from multiple sources retrospectively, capital costs were not included, QALYs were not evaluated and a broader societal perspective was not adopted

  McGuire et al. [24]

(1) CIS + CYC for 6 cycles

(2) CIS + PAC (infusion over 24 h in the inpatient setting and 4 h in the outpatient setting—total 6 cycles)

Cost per patient:

Inpatient: (1) US$27,320; (2) US$29,824

Outpatient: (1) US$17,964; (2) $21,086

Incremental cost:

Inpatient: (1) Referent; (2) US$8737.34

Outpatient: (1) Referent; (2) US$9,335.21

Mean overall survival:

(1) 2.32 y

(2) 2.76 y

Inpatient setting:

(1) Referent

(2) US$19,820

Outpatient setting:

(1) Referent

(2) US$21,222

AC: PAC-CIS is cost-effective for the treatment of advanced ovarian cancer in the inpatient as well as outpatient settings

AL: Comparative costs across treatment arms were limited to cost of therapy

  Ortega et al. [25]

(1) 75 mg/m2 at 1 mg/min CIS + 750 mg/m2 CYC for total 6 cycles

(2) 75 mg/m2 at 1 mg/min CIS + 135 mg/m2 over 24 h PAC for 6 cycles

Cost per patient:

(1) Can$7,300

(2) Can$11,700–Can$15,000 (depending on second line treatment)

Quality-adjusted PFS:

(1) 6.1 mo

(2) 10–10.6 mo (depending on second-line treatment)

(1) Referent

(2) Can$11,600–Can$24,200 per quality-adjusted PFLY (2014 USD equivalent 9,659–20,150 per quality-adjusted PFLY)

AC: PAC-CIS provides substantial quality-adjusted PFS benefit at an affordable cost to the Canadian healthcare system

AL: Use of data from non-randomized phase II trials for second-line therapies may compromise internal validity of model; assumptions were made regarding second-line therapies and their activity in PAC-resistant disease; PAC infusion time was considered to be 3 h compared with 24 h from the GOG trial; treatment costs and patient preferences were not considered for third- or fourth-line protocols

  Berger et al. [26]

(1) 75 mg/m2 at 1 mg/min CIS + 750 mg/m2 CYC for total 6 cycles

(2) 75 mg/m2 at 1 mg/min CIS + 135 mg/m2 over 24 h PAC for 6 cycles

Incremental cost for CIS + PAC [2]:

Germany: US$11,909

Spain: US$8,230

France: US$8,648

Italy: US$14,652

The Netherlands: US$10,010

UK: US$8,112

Incremental life years for CIS + PAC [2]:

Germany: 1.27 y

Spain: 1.29 y

France: 1.30 y

Italy: 1.28 y

The Netherlands: 1.28 y

UK: 1.27 y

ICER for 2 versus 1

Germany: US$9362 per LYG

Spain: US$6395 per LYG

France: US$6642 per LYG

Italy: US$11,420 per LYG

The Netherlands: US$7796 per LYG

UK: US$6403 per LYG

AC: PAC-CIS is cost-effective for the treatment of advanced ovarian cancer. Depending on the country, the ICER ranged from US$6403 to US$11,420, which compares favorably to other oncologic and non-oncologic interventions

AL: Retrospective in nature; cost estimates were based on expert opinion; quality of life was not assessed

  Neymark et al. [27]

(1) IV 750 mg/m2 CYC + IV 75 mg/m2 CIS

(2)175 mg/m2 PAC for cycle 1 set to increase to 200 mg/m2 + IV 75 mg/m2 CIS

Mean total cost per patient:

(1) €16,529

(2) €23,324

Median overall survival:

(1) 25.8 mo

(2) 35.6 mo

Mean overall survival:

(1) 26.6 mo

(2) 30.6 mo

(1) Referent

(2) €20,385 per LYG (2014 USD equivalent 23,594 per LYG)

AC: PAC-based regimen showed significant improvement in patient survival which was associated with increased cost primarily due to high acquisition cost of PAC

AL: Resource use data from several sites was pooled assuming the study was conducted in a single healthcare setting. Societal perspective was not used

  Limat et al. [28]

(1) IV 750 mg/m2 CYC as 15-min infusion + 75 mg/m2 CIS as 3-h infusion once q3w

(2) IV 135 mg/m2 PAC as 24-h infusion on day 1 + IV 75 mg/m2 CIS as a 3-h infusion on day 2 once q3w

Mean total cost per patient:

(1) US$6798

(2) US$17,514

Median PFS:

(1) 11.2 mo

(2) 14.2 mo

Median overall survival:

(1) 21.2 mo

(2) 32 mo

Median quality-adjusted life months:

(1) 12.3 mo

(2) 21.6 mo

(1) Referent

(2) US$11,907 per LYG (2014 USD equivalent 19,874 per LYG)

(1) Referent

(2) US$13,827 per QALY gained (2014 USD equivalent 23,078 per QALY gained)

AC: From a French university hospital perspective PAC + CIS regimen could be considered as cost-effective for first-line therapy of advanced ovarian cancer patients

AL: Single center study with limited sample size (lack of statistical power). Cost and quality of life effect of chronic adverse events were not accounted

 Intraperitoneal compared with intravenous route of administration (κ = 2)

  Bristow et al. [29]

(1) IV PAC (3 h) and IV CAR

(2) IV PAC (24 h) and IP CIS + IP PAC

Total cost per patient:

(1) US$18,823

(2) US$39,861

Incremental cost:

(1) Referent

(2) US$21,038

Overall survival (discounted):

(1) 4.06 QALYs

(2) 4.41 QALYs

Incremental overall survival:

(1) Referent

(2) 0.35 QALYs

ICER:

(1) Referent

(2) US$60,976 per QALY gained (2014 USD equivalent 78,948 per QALY gained)

AC: Compared with IV/IV chemotherapy, administration of IP/IV chemotherapy was associated with slightly improved quality-adjusted survival; although it was accompanied by higher costs

AL: Two different studies provided survival estimates; use of median survival rather than mean survival time; use of FACT-O and EORTC QLC-C30 as an approximation for utility functions; cost of ambulatory management of treatment-related toxicity was not considered; cost estimates may not be generalized to other practice settings or regions; underestimate or overestimate of QALY as adjustments did not extend beyond 1 y

  Havrilesky et al. [30]

(1) 175 mg/m2 PAC over 3 h CAR (7.5 AUC) on day 1 of 21-day cycle (IV-CAR/PAC)

(2) 135 mg/m2 IV PAC over 24-h day 1 + 75 mg/m2 IV CIS day 2 of 21-day cycle (IV-CIS/PAC—GOG-172, GOG-158)

(3) 135 mg/m2 IV PAC over 24 h day 1 + 100 mg/m2 IP CIS on day 2 + 60 mg/m2 IP PAC on day 8 of 21-day cycle (IP-CIS/PAC)

Total cost per patient for 7 years:

(1) US$18,213

(2) US$59,851 (GOG-172), US$61,093 (GOG-158)

(3) US$63,114

Mean overall survival:

(1) 4.20 y

(2) 3.99 y (GOG-172), 3.92 y (GOG-158)

(3) 4.45 y

Median overall survival:

(1) 57 mo

(2) 51 mo (GOG-172), 48 mo (GOG-158)

(3) 66 mo

(1) Referent

(2) Dominated

(3) US$180,022 per QALY gained (2014 USD equivalent 233,082 per QALY gained)

AC: Intraperitoneal chemotherapy was not cost-effective in base-case analysis (7-y time horizon) but was cost-effective over longer time horizon

AL: Survival data used for effectiveness was based on separate phase III clinical trials (GOG-158 and GOG-172)

 Targeted biologics (κ = 4)

  Cohn et al. [31]

(1) PAC (175 mg/m2) + CAR (6 AUC) on day 1 every 21 d for 6 cycles

(2) PAC + CAR + BEV (15 mg/kg on day 1 every 21 d starting at cycle 2)

(3) (PAC + CAR + BEV) + maintenance BEV (15 mg/kg day 1 every 21 d for 16 cycles)

Total cost for 600 patients:

(1) US$2.5 million

(2) US$21.4 million

(3) US$78.3 million

Median PFS:

(1) 10.3 mo

(2) 11.2 mo

(3) 14.1 mo

(1) Referent

(2) US$479,712 per PFLY gained (2014 USD equivalent 555,931 per PFLY gained)

(3) US$401,088 per PFLY gained (2014 USD equivalent 464,815 per PFLY gained)

AC: Addition of BEV and maintenance BEV was not cost-effective. The results were sensitive to the cost of drugs

AL: Effectiveness was based on abstract results presented at 46th ASCO meeting; underestimation of survival outcomes since they used PFS; lacks toxicity-related costs and quality-of-life measures

  Barnett et al. [32]

Base case:

(1) PAC + CAR for all patients

(2) PAC + CAR + BEV (7.5 mg/m2 for 12 cycles) for all patients

(3) PAC + CAR + BEV (7.5 mg/m2 for 12 cycles) for all high-risk patients

Alternative scenario:

(1–4) Base-case treatment alternatives

(5) Test for PAC + CAR + BEV (BEV administered only if biomarker predictive test is positive in the high risk cohort)

Mean costs for base case:

(1) US$6220

(2) US$20,751

(3) $56,351

Mean costs for alternative scenario:

(1) US$6211

(2) US$19,605

(3) US$20,758

(4) US$56,334

QALY for base case:

(1) 2.80 QALY

(2) 2.89 QALY

(3) 2.88 QALY

QALY for alternative case:

(1) 2.80 QALY

(2) 2.90 QALY

(3) 2.89 QALY

(4) 2.89 QALY

ICER for base case:

(1) Referent

(2) US$168,610 per QALY (2014 USD equivalent 183,368 per QALY)

(3) Dominated

ICER for alternative case:

(1) Referent

(2) US$128,928 per QALY (2014 USD equivalent 140,213 per QALY)

(3) Dominated

(4) Dominated

AC: Use of BEV together with standard first-line platinum–taxane is not a cost-effective approach for all stage III/IV ovarian cancer patients. Using this approach selectively in high-risk patients may be more preferable, although the ICER exceeds the common willingness-to-pay thresholds. Further research on alternative cases such as the predictive test is necessitated.

AL: Cost data was not prospectively collected as part of the ICON trial; BEV regimen with lower dosing and frequency (based on the ICON trial) is not utilized in the US; equivalent quality of life was assumed for all groups

  Chan et al. [33]

(1) PAC + CAR q3w for 6 cycles

(2) PAC + CAR + BEV q3w for 6 cycles + maintenance BEV for additional 12 cycles

Total costs per cycle:

(1) US$535

(2) US$3760 (for first 6 cycles) + US$3225 (for 12 maintenance cycles)

Median PFS:

(1) 10.5 mo

(2) 15.9 mo

Median overall survival:

(1) 28.8 mo

(2) 36.6 mo

(1) Referent

(2) US$167,771 per LYG

AC: For high-risk advanced ovarian cancer patients the ICER was nearly US$170,000 per life-year saved

AL: Simplified simulation model—subset from single trial; hematologic and non-hematologic complications were not considered; indirect, societal, patient-related costs as well as cost of recurrent disease was not considered; assumption of a constant hazard rate; timing of BEV administration was not considered

  Mehta and Hay [34]

(1) PAC (175 mg/m2) + CAR (6 AUC) on day 1 every 21 d for 6 cycles

(2) PAC + CAR + BEV (15 mg/kg on day 1 every 21 d starting at cycle 2) + maintenance BEV (15 mg/kg day 1 every 21 d for 16/12 cycles for GOG-218/ICON-7)

Total costs:

GOG-218:

(1) US$70,158

(2) US$201,423

ICON-7:

(1) US$63,311

(2) US$125,114

Incremental costs:

GOG-218:

(1) Referent

(2) US$131,265

ICON-7:

(1) Referent

(2) US$61,803

Median overall survival:

GOG-218:

(1) 1.78 y; (2) 1.97 y

ICON-7:

(1) 1.67 y; (2) 2.01 y

Incremental overall survival:

GOG-218:

(1) Referent; (2) 0.19 y

ICON 7:

(1) Referent; (2) 0.34 y

Incremental QALY:

GOG-218:

(1) Referent; (2) 0.05 QALY

ICON 7:

(1) Referent; (2) 0.27 QALY

GOG-218 trial:

(1) Referent

(2) US$2,420,691 per QALY gained (2014 USD equivalent 2,478,530 per QALY gained)

ICON 7 trial:

(1) Referent

(2) US$225,515 per QALY gained (2014 USD equivalent 230,903 per QALY gained)

AC: Addition of BEV and maintenance BEV was not cost-effective. Bevacizumab may be cost-effective for high-risk ICON-7 population when bio-similar BEV is used

AL: Assumption of a constant proportional hazard for mortality and progression; lacks a perspective if multiple lines of therapy are used in clinical practice; costs related to end-of-life palliative care were not considered

 Miscellaneous (κ = 4)

  Messori and Trippoli [35]

(1) CIS at conventional doses

(2) PAC + CIS at conventional doses

(3) high-dose chemotherapy with hematopoietic rescue

Incremental cost per patient:

High-dose chemotherapy with hematopoietic rescue—US$60,000; cost not reported for PAC + CIS at conventional doses as survival advantage was observed

Mean overall survival:

(1) 2.58 y

(2) 2.74 y

(3) 4.92 y

Incremental LYG:

(1) Referent

(2) PAC + CIS was considered not to have any survival advantage over CIS

(3) 2.34 LYG

(1) Referent

(2) Dominated

(3) US$25,641 per LYG

AC: High-dose chemotherapy with hematopoietic rescue was more effective and cost-effective than conventional treatments with ICER below US$50,000 in sensitivity analysis

AL: Effectiveness measure for PAC + CIS and high-dose chemotherapy was based on a single study

  Orr et al. [36]

(1) Platinum + CYC (600 mg/m2)

(2) Platinum + PAC (135 mg/m2). The doses administered for platinum drugs were: CIS 100 mg/m2; CAR 300 mg/m2

Total cost per patient:

(1) US$3415

(2) US$11,872

3-year survival:

(1) 74 %

(2) 66 %

(1) US$4615 per patient to achieve 3-year survival

(2) US$17,988 per patient to achieve 3-year survival

AC: The study demonstrated the feasibility of use on in vitro assay to avoid use of ineffective chemotherapy. Cost-effectiveness of assay directed chemotherapy may be increased with consideration of quality of life, avoidance of toxicity etc.

AL: No randomization was conducted and prognostic factor differences were not considered

  Dalton et al. [37]

(1) CAR (AUC 6) + PAC (180 mg/m2) q3w

(2) CAR (AUC 6) + dose-dense PAC (80 mg/m2) weekly

Total costs:

(1) US$9751

(2) US$12,152

Incremental cost:

(1) Referent

(2) US$2401

PFS:

(1) 1.37 y

(2) 1.78 y

Incremental PFS:

(1) Referent

(2) 0.41 y

(1) Referent

(2) US$5809 per PFLY gained

AC: Administration of dose-dense weekly PAC and CAR is cost-effective in patients with advanced-stage ovarian cancer

AL: All clinical scenarios were not considered as model was based on data from single trial; costs for treatment after progression was not considered; assumption of a constant hazard over time; no consideration of societal and patient-related costs which may be important in this setting

  Lairson et al. [38]

(1) No chemotherapy

(2) Other non-platinum regimen

(3) Platinum-based regimen

(4) Platinum–taxane regimen

Mean costs (early-stage, late-stage):

(1) US$83,430, US$63,804

(2) US$126,376, US$140,947

(3) US$115,364, US$120,829

(4) US$122,631, US$149,669

QALY (early-stage, late-stage):

(1) 2.3, 0.5

(2) 2.9, 1.3

(3) 3.3, 1.5

(4) 2.4, 1.3

ICER (early stage):

(1) Referent

(2) Dominated

(3) US$30,073 per QALY gained (2014 USD equivalent 34,851 per QALY gained)

(4) Dominated

ICER (late stage):

(1) Referent

(2) Dominated

(3) US$58,151 per QALY gained (2014 USD equivalent 67,390 per QALY gained)

(4) Dominated

AC: Treating elderly ovarian cancer patients with a platinum-based regimen is the most cost-effective treatment alternative

AL: Although propensity-matched samples minimized selection bias, unobservable differences between treatment groups may exist due to the observational study design; utility measures were used as a proxy for quality of life and obtained from the literature; external validity; lack of a societal perspective

 Maintenance chemotherapy (κ = 1)

  Lesnock et al. [39]

(1) CAR AUC 7.5 + 135 mg/m2 PAC over 3 h—21-day cycle for 6 cycles

(2) CAR AUC 7.5 + 135 mg/m2 PAC over 3 h—28-day cycle + maintenance PAC for 12 cycles

(3) CAR + PAC (AUC 6 + 175 mg/m2 for 1 cycle, AUC 6 + 175 mg/m2 + 15 mg/kg for 5 cycles) + 15 mg/kg BEV + maintenance BEV for 16 cycles

Incremental cost per patient:

(1) Referent

(2) US$4909

(3) US$99,012

Incremental QALYs:

(1) Referent

(2) 0.37

(3) −0.05

(1) Referent

(2) US$13,402 per QALY (2014 USD equivalent 15,531 per QALY)

(3) Dominated

AC: PAC maintenance therapy is more cost-effective in advanced ovarian cancer, while maintenance BEV is not cost-effective

AL: Model was created based on treatment arms and estimates from 3 separate clinical trials; quality of life utility values have not been validated in this population; cost estimates may have regional dependence; cost estimates did not consider outpatient issues, hospitalizations, extra laboratory testing or imaging; CAR + PAC may not be the most appropriate reference treatment arm

 Recurrent platinum-sensitive (κ = 7)

  Fisher and Gore [40]

(1) Three 0.25-mg vials and one 1-mg vial of TRA + one 50-mg vial of PLD for 6.9 mo

(2) Two 50-mg vials of PLD alone

Total costs:

(1) £41,880

(2) £23,404

Incremental cost:

(1) Referent

(2) £18,476

Total QALYs gained:

(1) 1.85

(2) 2.33

Incremental QALYs gained:

0.49

(1) Referent

(2) £38,026 per QALY (2014 USD equivalent 56,671 per QALY)

AC: Compared with the original NICE assessment that was based on interim data, this analysis based on the final data showed improvements in OS and ICER/QALY. TRA + PLD may be cost-effective given that it is a candidate for end-of-life criteria; though only NICE may make the final recommendations

AL: Lacks a comparison with platinum-based regimen which is the standard for platinum-sensitive disease; OVA-301 may be under-powered for sub-groups treatment effects; precise extrapolation of survival data across a lifetime may be difficult

  Lee et al. [41]

(1) PAC + CAR

(2) PLD + CAR

Total cost per patient:

(1) US$20,838

(2) US$21,732

Incremental cost per patient:

(1) Referent

(2) US$894

Mean overall survival:

(1) 2.59 y

(2) 2.60 y

Mean overall QALY:

(1) 1.82 QALYs

(2) 1.86 QALYs

Incremental QALYs:

0.04 QALY

(1) Referent

(2) US$21,658 per QALY (2014 USD equivalent 23,554 per QALY gained)

AC: PLD + CAR as a second-line therapy for platinum-sensitive ovarian cancer patients was cost-effective

AL: Productivity costs were not included; effectiveness data was based on proceedings

  Havrilesky et al. [42]

(1) 30 mg/m2 DOC on d 1,8 q3w followed by CAR AUC 6 q3w at first progression or after 6 cycles DOC for stable disease or a partial response

(2) 30 mg/m2 DOC on d 1,8 + CAR AUC 6 mg/mL per min on day 1 q3w

Average costs per patient:

(1) US$20,381

(2) US$25,122

Median PFS:

(1) 8.4 mo

(2) 13.7 mo

(1) Referent

(2) US$25,239 per QALY (2014 USD equivalent 28,284 per QALY gained)

AC: Despite the slightly lower quality of life, combination DOC + CAR is cost-effective for patients with recurrent platinum-sensitive ovarian cancer

AL: Cost and utility data was not prospectively collected; differences in caregiver expenses between treatment arms were not accounted for

  Montalar et al. [43]

(1) PLD alone

(2) TRA + PLD

Total cost per patient:

(1) €23,072

(2) €45,573

Incremental cost per patient:

(1) Referent

(2) €22,501

Mean overall survival:

(1) 2.13 y

(2) 2.80 y

Mean overall QALY:

(1) 1.86 QALYs

(2) 2.35 QALYs

Incremental overall survival:

(1) Referent

(2) 0.68 y

Incremental QALYs:

(1) Referent

(2) 0.49 QALYs

(1) Referent

(2) €33,335 per LYG (2014 USD equivalent 45,477 per LYG)

(1) Referent

(2) €45,592 per QALY (2014 USD equivalent 62,199 per QALY)

AC: TRA + PLD was found to be clinically effective for platinum-sensitive patients as compared with PLD alone; however, cost-effectiveness is debatable

AL: Adverse event management cost was obtained from literature, indirect costs were not included, a model was used and thus individual characteristics of patients cannot be accounted for

  Case et al. [44]

(1) BSC

(2) Second-line monotherapy (CAR)

(3) Second-line combination (CAR + PAC

(4) Third-line previous monotherapy (DOX)

(5) Third-line previous combination (DOX)

(6) Fourth-line previous monotherapy (GEM)

(7) Fourth-line previous combination (GEM)

Total cost for 10,000 patients:

(1) US$244 million

(2) US$405 million

(3) US$521 million

(4) US$625 million

(5) US$741 million

(6) US$1032 million

(7) US$1147 million

Overall survival:

(1) 6 mo

(2)14 mo

(3) 17 mo

(4)18 mo

(5) 21 mo

(6) 21 mo

(7) 24 mo

(1) Referent

(2) US$24,228 per LYG (2014 USD equivalent 34,009 per LYG)

(3) US$46,068 per LYG (2014 USD equivalent 64,666 per LYG)

(4) Dominated

(5) US$66,012 per LYG (2014 USD equivalent 92,662 per LYG)

(6) Dominated

(7) US$162,552 per LYG (2014 USD equivalent 228,177 per LYG)

AC: BSC and second-line chemotherapy is cost-effective for patients with recurrent platinum-sensitive ovarian cancer

AL: Model lacks flexibility in terms of drug choice; uncertainty in clinical estimates and non-clinical trial setting; role of secondary cytoreduction was not evaluated; quality of life, performance status or palliative interventions were not included

  Havrilesky et al. [45]

(1) CAR AUC 5 on day 1 of a 21-day cycle for a total 6 cycles

(2) 175 mg/m2 PAC + CAR AUC 5 on day 1 of a 21-day cycle for a total 6 cycles

(3) IV 1000 mg/m2 GEM on d 1,8 + CAR AUC 4 on day 1 of a 21-day cycle for total 6 cycles

Average costs:

(1) US$4018

(2) US$5982

(3) US$17,627

Mean PFS:

(1) 8.01 mo

(2) 10.05 mo

(3) 10.52 mo

Incremental PFS:

(1) Referent

(2) 2 mo

(3) 0.5 mo

(1) Referent

(2) US$15,564 per PFLY (2014 USD equivalent 20,151 per PFLY)

(3) US$278,388 per PFLY (2014 USD equivalent 360,440 per PFLY)

AC: PAC + CAR is cost-effective for patients with recurrent platinum-sensitive ovarian cancer

AL: combination vs sequential therapy was not evaluated

  Main et al.: Analysis 2 [46]

(1) Platinum-based

(2) CAP

(3) PAC

(4) PLD

(5) PAC + platinum

(6) TOP

Total cost per patient:

(1) £2876

(2) £3988

(3) £6274

(4) £7662

(5) £8841

(6) £11,276

Mean quality-adjusted overall survival:

(1) 66.3 w

(2) 69.5 w

(3) 41.2 w

(4) 58.5 w

(5) 81.2 w

(6) 41.7 w

(1) Referent

(2) £16,421 per QALY (2014 USD equivalent 24,472 per QALY)

(3) Dominated

(4) Dominated

(5) £20,950 per QALY (2014 USD equivalent 31,222 per QALY)

(6) Dominated

AC: TOP, PAC, and PLD were dominated by platinum-based chemotherapy. PAC + platinum combination was cost-effective for platinum-sensitive patients

AL: To include full range of treatment options less robust methods were used to synthesize evidence

 Recurrent platinum-resistant (κ = 1)

  Rocconi et al. [47]

(1) BSC

(2) Second-line monotherapy (DOX 40 mg/m2 for 4 mo)

(3) Second-line combination (GEM 750 mg/m2 + CIS 30 mg/m2 on day 1 and 8 every 21 d for 4 mo)

(4) Third-line previous monotherapy (TOP 1.5 mg/m2 for 5 d every 21 d for 3 cycles)

(5) Third-line previous combination (TOP 1.5 mg/m2 for 5 d every 21 d for 3 cycles)

Incremental total cost for 4000 patients:

(1) Referent

(2) US$64 million

(3) US$201 million

(4) Dominated

(5) US$203 million

Incremental overall survival:

(1) Referent

(2) 3 mo

(3) 2 mo

(4) Dominated

(5) 2 mo

(1) Referent

(2) US$64,104 per LYG (2014 USD equivalent 89,984 per LYG)

(3) US$302,316 per LYG (2014 USD equivalent 424,365 per LYG)

(4) Dominated

(5) US$303,984 per LYG (2014 USD equivalent 426,707 per LYG)

AC: BSC is cost-effective and second-line monotherapy is reasonably cost-effective. Improvements in survival may be needed for second-line combination and third-line therapy to be cost-effective

AL: Due to the clinical estimates that were used, costs for treating chemotherapy myelosuppression were not considered; quality-of-life measures were not included

 Multiple lines (κ = 1)

  Fedder et al. [48]

(1) First-line CAR followed by second-line TOP

(2) First-line CAR followed by second-line DOX

(3) First-line CAR + PAC followed by second-line DOX

(4) First-line CAR + PAC followed by second-line TOP

Total cost per patient:

(1) €20,124

(2) €22,337

(3) €29,821

(4) €31,550

Life years saved:

(1) 2.55

(2) 2.70

(3) 2.60

(4) 2.64

(1) €7892 per YOLS (2014 USD equivalent 10,144 per LYG)

(2) €8270 per YOLS (2014 USD equivalent 10,630 per LYG)

(3) €11,454 per YOLS (2014 USD equivalent 14,722 per LYG)

(4) €11,958 per YOLS (2014 USD equivalent 15,371 per LYG)

AC: At the willingness-to-pay of €45,500 per life-year saved all therapy alternatives were evaluated to be cost-effective

AL: Prognostic factors and quality of life were not considered due to limited data. Costs due to gastrointestinal disorders, nausea, vomiting etc. were not included

 Other (κ = 1)

  Main et al.: analysis 1 [46]

(1) PAC

(2) PLD

(3) TOP

Total cost per patient:

(1) £6354

(2) £7714

(3) £11,394

Mean quality adjusted overall survival:

(1) 30.9 w

(2) 40.9 w

(3) 34.2 w

(1) Referent

(2) £7033 per QALY (2014 USD equivalent 10,481 per QALY)

(3) Dominated

AC: PLD was cost-effective as compared with PAC and TOP; the results were sensitive to additional effectiveness data but PLD remained a cost-effective alternative

AL: Additional effectiveness data was based on prematurely terminated trials. PAC being used as a first-line treatment may not be considered a relevant comparator for second-line treatment

AUC area under the concentration–time curve, BEV bevacizumab, BSC best supportive care, CAP cyclophosphamide-doxorubicin-cisplatin, CAR carboplatin, CIS cisplatin, CYC cyclophosphamide, DOC docetaxel, DOX doxorubicin, GEM gemcitabine, GOG US Gynaecologic Oncology Group, IP intraperitoneal, IV intravenous, LYG life-year gained, OS overall survival, PAC paclitaxel, PFLY progression-free life-year, PFS progression-free survival, PLD pegylated liposomal doxorubicin hydrochloride, q3w every 3 weeks, QALY quality-adjusted life-year, TOP topotecan, TRA trabectedin, USD United States dollars, YOLS years of life saved

Subsequently, after GOG-111 trial findings came to the forefront, the European Organization for Research and Treatment of Cancer (EORTC) 55931 trial [50] was designed to provide confirmatory findings and also integrate an economic evaluation with prospectively collected data on costs and quality of life. Using this source, Neymark et al. conducted a cost-effectiveness analysis from the perspective of the Belgian health insurance and financing system [27]. Patients randomized to cisplatin–paclitaxel gained an additional 4 months in mean survival, coupled with an added €7017 average cost per patient when compared with the cisplatin–cyclophosphamide treatment arm. Paclitaxel cost was the major driver of the cost difference between both treatment arms; however, most patients in the cisplatin–cyclophosphamide group also received second-line paclitaxel treatment, which made overall drug acquisition costs nearly equivalent between the groups. An ICER of €20,385 per LYG was obtained. Using bootstrap simulation, the authors showed that for a threshold willingness-to-pay ranging between €12,400 and €24,800, the percentage of ICER replicates considered cost-effective would range between 20–60 %, respectively. In order to supplement decision making from trial-based evaluations, Limat et al. assessed the cost-effectiveness of paclitaxel using regular clinical practice data in a French university hospital before-and-after case study [28]. The year 1998 marked the change of standard treatment at their institution. The overall mean cost of treating a patient with cisplatin–paclitaxel was US$10,716 higher than with cisplatin–cyclophosphamide, with drug costs, hospitalization, and hematopoietic growth factors being the major cost drivers. With an incremental median benefit of 0.90 years and 0.78 quality-adjusted life-years (QALYs), the ICER for cisplatin–paclitaxel was estimated as US$11,907 per LYG and US$13,827 per QALY gained. Given the single site and small sample size (n = 59), extensive sensitivity analyses were carried out to examine uncertainties related to their cost and clinical parameters. Across the spectrum of variations, the ICER for cisplatin–paclitaxel remained less than US$20,000.

3.2.2 Intraperitoneal Compared with Intravenous Route of Administration (κ = 2)

Administering adjuvant platinum–taxane chemotherapy via the intraperitoneal (IP) route over the traditional intravenous (IV) route was evaluated by two studies [29, 30]. Treatment arms that were compared in both studies were based on GOG-158 [51] and GOG-172 [52] protocols. While Bristow et al. [29] compared IV carboplatin–paclitaxel (from GOG-158) with IP cisplatin–paclitaxel (from GOG-172), Havrilesky et al. [30] additionally incorporated IV cisplatin–paclitaxel (from GOG-158) and IV cisplatin–paclitaxel (from GOG-172) in their analysis (Table 1). Using a societal perspective, the ICERs for IP cisplatin–paclitaxel were estimated at US$60,976 and US$180,022 per QALY gained compared with IV carboplatin–paclitaxel in the studies by Bristow et al. and Havrilesky et al., respectively (see Table 2 for 2014 USD equivalent ICERs). The considerable differences in ICERs could be explained by differences in time horizon, survival time, source of cost data and number of completed intraperitoneal treatment cycles. Costs related to hospitalization for the IP administration and its related toxicities were the prime components of higher total costs in both studies. IV cisplatin–paclitaxel was dominated (more costly and less effective) compared with IV carboplatin–paclitaxel.

3.2.3 Targeted Biologics (κ = 4)

Administration of a humanized anti-vascular endothelial growth factor antibody, such as bevacizumab, has gained popularity in the ovarian cancer treatment regimen because of its superior progression-free survival advantage observed in the GOG-218 and ICON-7 (Gynecologic Cancer Intergroup International Collaboration on Ovarian Neoplasms 7) trials [53, 54]. Using the earliest abstract findings from the GOG-218 trial, Cohn et al. [31] evaluated the cost-effectiveness of adding bevacizumab to standard platinum–taxane adjuvant chemotherapy in a simplified decision analysis model for a hypothetical cohort of 600 patients (Table 1). The total cost of treatment increased across the three groups, ranging from carboplatin–paclitaxel (US$2.5 million) to carboplatin–paclitaxel–bevacizumab (US$21.4 million), and carboplatin–paclitaxel–bevacizumab + maintenance bevacizumab (US$78.3 million). Compared with standard carboplatin–paclitaxel, treatment with combination bevacizumab resulted in an ICER estimate of US$479,712 per PFLY gained; and compared with combination treatment with bevacizumab, treatment with maintenance bevacizumab resulted in an ICER of US$401,088 per PFLY gained (Table 2). Thus, adding bevacizumab to standard adjuvant chemotherapy was not a cost-effective alternative. When post-hoc sub-group analysis for the ICON-7 trial demonstrated that the high-risk stage IIIC/IV sub-group experienced an additional 8-month overall survival benefit with bevacizumab, Chan and colleagues conducted a cost-effectiveness evaluation to compare the standard combination carboplatin–paclitaxel treatment arm with the same combination with bevacizumab plus 12 cycles of maintenance bevacizumab [33]. The dose and duration of bevacizumab in this trial was shorter than GOG-218, impacting overall costs in their analyses. Using Medicare payments as their source of costs, the ICER for adding bevacizumab was estimated as US$167,771 per life-year saved (LYS) from a healthcare system perspective. At a willingness-to-pay threshold of US$200,000, only 37 % of their simulations were cost-effective. To complement the earlier economic evaluations, Mehta and Hay assessed the cost-effectiveness of adding bevacizumab to first-line treatment from the perspective of US society [34]. Separate analyses were performed for an overall population based on the GOG-218 trial and a high-risk population based on the ICON-7 trial. The ICER was estimated as US$2,420,691 per QALY (GOG-218) and US$225,515 per QALY (ICON-7) for adding bevacizumab to combination carboplatin–paclitaxel plus maintenance bevacizumab. Use of biosimilar bevacizumab was additionally investigated, assuming a 30 % discount on the Federal Supply Schedule drug price. The resulting ICER was US$1,702,968 per QALY (GOG-218) and US$161,603 per QALY (ICON-7). At a societal willingness-to-pay threshold of US$150,000 per QALY, adding bevacizumab to front-line therapy was not considered cost-effective. Bevacizumab drives treatment costs, which cannot be ignored, despite its role in prolonging survival. Barnett et al. designed a cost-effectiveness study to determine the best strategy for incorporating bevacizumab in first-line therapy based on the ICON-7 trial, including use of a simulated predictive biomarker test to identify patients that will benefit most with bevacizumab treatment [32]. For the base-case scenario, investigators compared carboplatin–paclitaxel for all patients, carboplatin–paclitaxel–bevacizumab plus maintenance bevacizumab for high-risk patients, and for all patients. Administration of bevacizumab to the high-risk patients yielded an ICER of US$168,610 per QALY compared with standard carboplatin–paclitaxel for all patients. Use of bevacizumab in all patients was more expensive and less effective compared with use in high-risk patients. For the alternative scenario, the investigators additionally evaluated use of a biomarker predictive test wherein only patients that tested positive received bevacizumab. The resulting ICER was US$128,928 per QALY compared with standard carboplatin–paclitaxel for all patients (Table 2). A strategy using concurrent bevacizumab plus maintenance bevacizumab for all patients and high-risk patients was dominated by the biomarker predictive test. Although administration of bevacizumab was more cost-effective in the high-risk population, it failed to fall within the traditional willingness-to-pay threshold.

3.2.4 Miscellaneous (κ = 4)

Around the same period when most studies evaluated the economics of various platinum–taxane treatment strategies, Orr et al. assessed the cost-effectiveness of using an in-vitro drug resistance assay to guide individual chemotherapy for patients with stage III ovarian cancer after cytoreductive surgery [36]. Using average costs per patient and 3-year survival estimates from their primary phase II trial, cost-effectiveness for platinum–cyclophosphamide, platinum–paclitaxel, and assay-directed regimen was obtained by dividing total costs by 3-year survival. The costs per patient to obtain a 3-year survival were estimated as US$4615, US$17,988, and US$9768, respectively. The authors did not report on their source of cost data, comparison group, or any ICER estimates. Alternatively, Messori and colleagues evaluated the cost-effectiveness ratios for a high-dose treatment with hematopoietic rescue (using peripheral blood stem cell reinfusion or autologous bone marrow transplantation), conventional dose cisplatin, and cisplatin–paclitaxel [55]. With no survival difference found in their analyses between cisplatin and cisplatin–paclitaxel, the investigators did not report any cost-effectiveness ratio for these treatment alternatives [55]. However, the high-dose treatment with hematopoietic rescue yielded an ICER of US$25,641 per LYG compared with the conventional cisplatin regimen. Dalton et al. evaluated the cost-effectiveness of administering dose-dense weekly paclitaxel plus carboplatin compared with paclitaxel plus carboplatin every 3 weeks in patients with advanced ovarian cancer [37]. The treatment regimen with dose-dense paclitaxel yielded an incremental cost and effect of US$2401 and 0.41 PFY, respectively, with an ICER of US$5809 per PFLY gained, well within the traditional willingness-to-pay thresholds. The cost–utility of platinum-based chemotherapy using community level data was evaluated in one study [38]. Lairson et al. used the Surveillance Epidemiology and End Results (SEER) Medicare-linked database to assess the cost-effectiveness of using only a platinum-based regimen versus a platinum–taxane combination and other non-platinum regimens in a large US cohort of elderly patients. The investigators found that compared with the no-chemotherapy group, the ICERs for the only platinum-based regimen were US$30,073 per QALY and US$58,151 per QALY for early- and late-stage disease, respectively. Use of a platinum–taxane combination and other non-platinum chemotherapy was dominated. The investigators used a propensity-score matched cohort to minimize selection bias and confounding resulting from their observational study design. They addressed methodological uncertainty by incorporating alternative scenarios for utility and a net-benefit approach.

3.3 Maintenance Therapy

Cost-effectiveness of maintenance therapy was evaluated in one study [39]. Although randomized controlled trials for maintenance therapy have shown promising improvements in progression-free survival, toxicity concerns and lack of an overall survival advantage have limited its wide acceptance. The study evaluated the cost-effectiveness of consolidation treatment with paclitaxel compared with bevacizumab in patients with advanced stage ovarian cancer that had undergone standard cytoreductive surgery followed by six cycles of adjuvant carboplatin–paclitaxel chemotherapy. Survival estimates were derived from three different trials and utility values from inputs by gynaecological oncology experts. With an ICER of US$13,402 per QALY, maintenance therapy with paclitaxel was cost-effective at the willingness-to-pay threshold of US$100,000 per QALY, from the US healthcare system perspective as compared with no maintenance therapy. In addition, it dominated bevacizumab maintenance therapy; that is, maintenance therapy with bevacizumab was more costly and less effective compared with paclitaxel maintenance therapy.

3.4 Second-Line Chemotherapy

3.4.1 Recurrent Platinum-Sensitive Disease

Patients that relapse after 6 months of completing a first-line platinum-based regimen are assumed to be platinum-sensitive and usually offered subsequent platinum therapy. We identified five studies with varying second-line treatment regimens that had a platinum agent in one of its treatment arms. The first study compared a regimen containing carboplatin only, carboplatin–paclitaxel, and carboplatin–gemcitabine [45]. The ICER for carboplatin–paclitaxel was estimated as US$15,564 per PFLY gained compared with carboplatin alone, making it a cost-effective second-line regimen. In contrast, the ICER for carboplatin–gemcitabine was substantially higher at US$278,388 per PFLY gained, compared with carboplatin–paclitaxel. In another study, Havrilesky et al. evaluated the cost-utility of administering concurrent versus sequential platinum–taxane chemotherapy [42]. The ICER for concurrent carboplatin–docetaxel was US$25,239 per QALY compared with sequential carboplatin–docetaxel, suggesting it was a cost-effective alternative from a third-party payer perspective. In the third study, Main et al. compared a range of treatment alternatives for use in patients with platinum-sensitive disease [46]. The investigators derived clinical estimates from multiple trials, and their model comprised pegylated liposomal doxorubicin (PLD) hydrochloride, topotecan, paclitaxel monotherapy, cyclophosphamide–doxorubicin–cisplatin combination therapy (CAP), paclitaxel–platinum combination therapy and platinum-based monotherapy. Since topotecan, paclitaxel monotherapy, and PLD hydrochloride were all dominated by platinum-based monotherapy, these regimens were excluded in their analyses. Compared with platinum-based monotherapy, the ICER for CAP was £16,421 per QALY, and compared with CAP, the ICER for paclitaxel–platinum combination therapy was £20,950 per QALY. Lee et al. evaluated the cost-utility of carboplatin–PLD compared with carboplatin–paclitaxel as a second-line treatment strategy [41]. Using a Korean societal perspective, treatment with carboplatin–PLD incurred an incremental US$894 cost and 0.04 extra QALYs per patient. The resulting ICER was US$21,658 per QALY, and at a willingness-to-pay of US$20,000 (Korean GDP per capita in 2010), there was 49.4 % probability that carboplatin–PLD was cost-effective. The fifth study by Case et al. evaluated various chemotherapeutic strategies administered as second-line or subsequent treatment for patients with recurrent platinum-sensitive ovarian cancer [44]. Best supportive care, second-line mono-chemotherapy (carboplatin), second-line combination chemotherapy (carboplatin–paclitaxel), two third-line chemotherapy regimens (doxorubicin) and two fourth-line regimens (gemcitabine) were compared in a decision model for a hypothetical cohort of 10,000 patients. Overall survival estimates were used from clinical experience and multiple phase II/III trials, while costs were incorporated for drugs and administration. There was a significant increase in total treatment costs as the line of treatment advanced (incremental costs were 161 million, 116 million, 220 million, and 406 million for second-line monotherapy, second-line combination, third-line previous combination, and fourth-line previous combination therapy, respectively); though the corresponding improvement in overall survival was poor (incremental overall survival rates were 8 months, 3 months, 4 months and 3 months, respectively). With ICERs of US$24,228 per LYG and US$46,068 per LYG, second-line mono- and combination therapies emerged as cost-effective strategies compared with best supportive care.

Patients that experience severe toxicities from platinum-based treatment or those considered to be partially platinum-sensitive (relapse within 6–12 months) may not benefit from second-line platinum-based treatment. The OVA-301 trial [56] was designed to compare use of PLD plus trabectedin with PLD alone in this population. Clinical parameters from this trial were used in two cost-effectiveness studies [40, 43], one from the UK National Health Service/Personal Social Services perspective and the other from the Spanish National Health Service perspective. The incremental costs for treating patients with PLD plus trabectedin were £18,476 and €22,501, respectively, for a total 0.49 QALYs gained in both studies. The corresponding ICERs were estimated as £38,026 per QALY and €45,592 per QALY. Both studies concluded that the cost-effectiveness of PLD plus trabectedin was debatable based on these ICER estimates, and dependent on the healthcare payers’ willingness-to-pay.

3.4.2 Recurrent Platinum-Resistant Disease

Treating patients that relapse within 6 months of completing standard therapy (i.e., cytoreductive surgery + adjuvant platinum–taxane chemotherapy) is an on-going challenge, given that overall survival in this cohort of platinum-resistant advanced ovarian cancer patients is <12 months. Rocconi et al. compared best supportive care, second-line mono-chemotherapy (doxorubicin), second-line combination chemotherapy (gemcitabine–cisplatin), and two third-line chemotherapy regimens (topotecan) [47]. Survival estimates were ascertained from several clinical trials and cost-estimates were derived for the drug and its administration. The investigators did not consider costs for treating chemotherapy-related toxicities or complications. For their hypothetical cohort of 4000 patients, administration of best supportive care was the most cost-effective salvage therapy. Increasing the subsequent line of treatment translated into substantial cost with minimal improvement in overall survival. Second-line monotherapy with doxorubicin presented an ICER of US$64,104 per LYG, making it the next most cost-effective strategy contingent on the payer’s willingness-to-pay threshold. All other higher-line treatments were either dominated or resulted in non-cost-effective ICERs.

3.5 Multiple-Line Chemotherapy Regimen

With evidence from clinical trials showing that first-line platinum–taxane combination chemotherapy was as efficacious as first-line platinum monotherapy, and limited considerations for subsequent second-line regimens, Fedders et al. designed a study to assess the cost-effectiveness of multiple-line chemotherapy regimens [48]. The authors evaluated carboplatin followed by topotecan, carboplatin followed by liposomal doxorubicin, carboplatin–paclitaxel followed by topotecan, and carboplatin–paclitaxel followed by liposomal doxorubicin as the alternative arms of their Markov model for a cohort of epithelial ovarian cancer patients staged from I–IV. They used clinical estimates from seven published trials and cost estimates from a University hospital in Germany. Although the study did not report the ICERs, the cost per LYS for the different treatment alternatives was estimated as €7892, €8270, €11,454, and €11,958, respectively. Considering a social willingness-to-pay threshold of €45,000 per LYG, all treatment regimens were deemed cost-effective. Notably, this study also lacked an explicit perspective as well as any sensitivity analyses.

3.6 Other Treatment Regimens

Significant limitations identified in a review evaluating economic evidence for PLD hydrochloride, topotecan, and paclitaxel led to the design of a cost-effectiveness study by Main et al. [46]. The investigators compared PLD hydrochloride, topotecan, and paclitaxel administered as a monotherapy for second-line or subsequent treatment of patients with platinum-sensitive, -refractory and -resistant advanced ovarian cancer. Compared with paclitaxel, the ICER for PLD hydrochloride was estimated as £7033 per QALY in the overall patient population, £5777 per additional QALY in the platinum-sensitive population, and £9555 per additional QALY in the platinum-resistant and -refractory population; while topotecan use was dominated by PLD hydrochloride overall and in the sub-group analysis. The probability of PLD hydrochloride being cost-effective for the overall population was 69–92 %, when the willingness-to-pay threshold ranged between £10,000 and £50,000 per additional QALY, respectively.

3.7 Sensitivity Analyses

One-way sensitivity analyses were performed in a majority of the studies and some more recent studies have also conducted probabilistic sensitivity analyses. Overall survival, progression-free survival and costs were varied across a range of values, specifically to determine the value at which cost-effectiveness for a treatment alternative may be reached. Alternatively, some studies also varied model probabilities, discount rate, utilities/quality-of-life indexes, infusion time for paclitaxel, and toxicity rates. Survival estimates were mostly varied in the ranges of ±25–75 %, while cost estimates were varied by ±20–30 %. Rarely were these ranges justified by prior evidence. The results of the studies that evaluated first-line platinum–taxane treatment were robust to the sensitivity analyses that were performed, and conclusions were not affected by the range of values examined. The ICER estimates for studies that evaluated bevacizumab were sensitive to cost and survival estimates [31, 33, 34]. In the first-line setting, bevacizumab would become cost-effective if its cost was to decline to approximately 50 % of the current cost [32, 34], while in the maintenance therapy setting, its cost would have to decline to <12 % of its current cost for it to be cost-effective [39]. The findings for three studies that evaluated second-line treatment for platinum-sensitive recurrent disease were sensitive to changes in clinical parameters [40, 41, 45]. There was only one study that did not report sensitivity analyses [48].

3.8 Quality Assessment

Quality assessment using the QHES instrument for all studies is shown in Table 3. Of the 28 studies reviewed, 14 with a time horizon of >1 year discounted both costs and effectiveness. Quality-of-life adjustment due to treatments administered and associated adverse events was carried out for 13 studies. Three studies used utility values estimated from Functional Assessment of Cancer Therapy–General/Ovarian quality of life questionnaire [29, 30, 42], while Barnett et al. assessed utilities from the EORTC quality of life questionnaire [32]. Two studies evaluating treatments among platinum-sensitive ovarian cancer patients applied utilities obtained from the OVA-301 study which utilized the EuroQol EQ-5D questionnaire [40, 43]. Five studies obtained the utilities from the literature, three of which used a study by Havrilesky et al. that assessed health state utilities for ovarian cancer patients using the visual analog score and time trade-off method [34, 38, 39, 41, 46]. Ortega et al. estimated the health state utilities by conducting interviews of 40 subjects using the time trade-off technique [25], while Limat et al. computed QALYs by adapting the quality-adjusted time without symptoms or toxicity (Q-TWiST) methodology and assumed a utility of 0.5 for the phase with toxicity and from progression to death [28].
Table 3

Proportion of the included studies reporting each of the items of the Quality of Health Economic Studies (QHES) instrument

QHES questions

Additional clarification/comments

N (%)

Q1. Was the study objective presented in a clear, specific and measurable manner?

 

26 (92.9)

Q2. Were the perspective of the analysis and reasons for its selection stated?

Specification of perspective was given a score of 1

25 (89.3)

Q3. Were variable estimates used in the analysis from the best available source?

Systematic reviews, meta-analyses, previous RCTs would be given a score of 1

25 (89.3)

Q4. If estimates came from a subgroup analysis, were the groups pre-specified at the beginning of the study?

 

2a

Q5. Was uncertainty handled by [1] statistical analysis to address random events, [2] sensitivity analysis to cover a range of assumptions?

Any kind of sensitivity analysis would be given a score of 1

27 (96.4)

Q6. Was incremental analysis performed between alternatives for resources and costs?

 

26 (92.9)

Q7. Was the methodology for data abstraction stated?

 

26 (92.9)

Q8. Did the analytic horizon allow time for all relevant and important outcomes? Were benefits and costs that went beyond 1 year discounted and justification given for the discount rate?

For time horizons beyond 1 year, if both costs and benefits were discounted then a score of 1 was given

14 (50.0)

Q9. Was the measurement of costs appropriate and the methodology for the estimation of quantities and unit costs clearly described?

If studies included costs for drug acquisition, drug administration, adverse events and associated hospitalization costs then a score of 1 was given

19 (67.9)

Q10. Were the primary outcome measure(s) for the economic evaluation clearly stated and did they include the major short-term, long-term, and negative outcomes?

 

28 (100)

Q11. Were the health outcome measures/scales valid and reliable? If previously tested valid and reliable measures were not available, was justification given for the measures/scales used?

Quality-adjusted outcomes (base-case or sensitivity analysis) were considered valid and a score of 1 was given

14 (50.0)

Q12. Were the economic model (including structure), study methods and analysis, and the components of the numerator and denominator displayed in a clear, transparent manner?

 

26 (92.9)

Q13. Were the choice of economic model, main assumptions, and limitations of the study stated and justified?

 

28 (100.0)

Q14. Did the author(s) explicitly discuss direction and magnitude of potential biases?

 

9 (32.1)

Q15. Were the conclusions/recommendations of the study justified and based on the study results?

 

27 (96.4)

Q16. Was there a statement disclosing the source of funding for the study?

 

19 (67.9)

aApplicable to only 2 of the articles reviewed

Clinical effectiveness (i.e., progression-free life-years, life-years, or quality-adjusted life-years) was estimated from clinical trial data for 25 studies, limiting the generalizability of the cost-effectiveness evaluations. Three studies used observational data to evaluate cost-effectiveness of primary treatment in the US [38], France [28], and Canada [22]. Lairson et al. used longitudinal population-based registry data from various geographic areas in the US to estimate overall survival [38]. Limat et al. estimated overall survival for patients treated in a French university hospital using a before-after study design, while Covens et al. obtained survival information from patient charts at a cancer center in Canada [22, 28]. Almost all studies stated (89 %) a perspective, of which four studies considered a societal perspective. Two studies considered wages lost along with caregiver costs [29, 34], one study only included caregiver cost [30], and the remaining study considered only transportation costs [41]. Direct costs most commonly included were related to treatment (i.e., drug cost), treatment administration, hospitalization, and adverse events. However, only 19 studies included all of the above-mentioned costs in their analyses. A mix of micro-costing and gross-costing approaches were used with cost data estimated from fee schedules [22, 23, 24, 26, 27, 34, 40, 41, 45, 46], hospitals [22, 23, 24, 25, 28, 44, 47, 48], Medicare reimbursement data [30, 31, 32, 38, 39, 42, 45], secondary databases (e.g., Agency for Healthcare Research and Quality—Healthcare Cost and Utilization Project) [29, 30, 32, 39, 43], and literature [21, 34, 37, 46, 55]. Similarly, drug costs were obtained from various sources like hospitals [22, 23, 25, 29, 37, 48], formularies or fee schedules [24, 26, 34, 40, 41, 43, 46], Medicare reimbursement data [30, 31, 33, 37, 38, 39, 42, 45], or average wholesale prices [28, 39, 44, 47].

4 Discussion

4.1 Summary of Evidence

The evidence from eighteen CEAs for first-line treatment can be broadly categorized as fourteen studies incorporating platinum–taxane chemotherapy only and four studies that also include targeted biologics such as bevacizumab. Eight of the fourteen studies specifically assessed the cost-effectiveness of combination cisplatin–paclitaxel compared with cisplatin–cyclophosphamide [21, 22, 23, 24, 25, 26, 27, 28]. Paclitaxel was universally associated with an increase in costs and effectiveness, with a majority of these studies explicitly recommending cisplatin–paclitaxel as a cost-effective treatment option for advanced ovarian cancer. Two studies evaluated administration of IP cisplatin–paclitaxel compared with IV carboplatin–paclitaxel. While the approach showed good value in one study [29], contradictory findings were obtained for the other [30]. In addition, alternative strategies such as use of in-vitro drug resistance assay to guide chemotherapy [36], high-dose treatment with hematopoietic rescue [55], and dose-dense weekly paclitaxel administration [37] were also considered cost-effective compared with standard platinum–taxane regimen, cisplatin regimen, and 3-weekly paclitaxel, respectively. In one study that focussed solely on the elderly population, the platinum-based regimen was cost-effective compared with no chemotherapy for both early- and late-stage disease [38]. Studies unanimously agreed that it is not cost-effective to include bevacizumab as a front-line treatment compared with the existing platinum–taxane standard for the overall ovarian cancer population [31, 32, 33, 34]. Nevertheless, two studies showed that in the high-risk sub-groups, use of bevacizumab may yield better value than treating all patients [32, 33]. Also, bevacizumab will become off-patent in 2019 in the U.S., which may alter its cost-effectiveness [57]. For maintenance therapy, paclitaxel administration was cost-effective compared with no maintenance therapy [39].

Second-line treatment studies were categorized into recurrent platinum-sensitive (seven studies) and one platinum-resistant disease study. Varying regimens in the platinum-sensitive disease scenario were evaluated, with a majority of studies including a platinum–taxane agent in one treatment arm. Three studies concluded that second-line platinum–paclitaxel was a cost-effective treatment option (carboplatin was the platinum agent for two studies [44, 45] and unknown in the third study [46]). Comparison groups for these studies were carboplatin alone [45], best-supportive care [44], and cisplatin alone [46], respectively. The use of carboplatin monotherapy was also cost-effective compared with best supportive care [44]. Similarly, carboplatin–PLD and concurrent carboplatin–docetaxel emerged as more cost-effective alternatives than carboplatin–paclitaxel [41] and sequential carboplatin–docetaxel [42], respectively. For patients with partially platinum-sensitive disease that may not fully benefit from platinum-based second-line treatment, the cost-utility of PLD plus trabectedin was compared with PLD alone in two studies [40, 43]. ICERs from both studies were high and cost-effectiveness was dependent on the payer’s willingness-to-pay above commonly cited benchmarks. In the setting of recurrent platinum-resistant disease, evidence showed that best supportive care was the most cost-effective strategy, followed by second-line monotherapy with doxorubicin [47].

Only one study evaluating multiple-line chemotherapy regimens [48] concluded that first-line carboplatin followed by second-line topotecan, first-line carboplatin followed by second-line liposomal doxorubicin, first-line carboplatin–paclitaxel followed by second-line topotecan, and first-line carboplatin–paclitaxel followed by second-line liposomal doxorubicin were cost-effective treatment alternatives. In another study, PLD monotherapy emerged as cost-effective compared with paclitaxel or topotecan monotherapy for second-line or subsequent treatment of patients with platinum-sensitive, -refractory and -resistant advanced ovarian cancer [46].

4.2 Quality of Evidence

The studies included in this review were published over a period of 18 years (1996–2014) and were conducted from various perspectives in several countries with different payment systems. Hence, direct comparison between studies was difficult. Nevertheless, general attributes such as health state utility sources, estimation of effectiveness and type of costs were evaluated. Only about half of the studies included in the review accounted for quality of life and thereby overall effectiveness of treatments may be overestimated. Moreover, effectiveness was primarily estimated using clinical trial data and thus more studies using real-world data may be needed to validate trial-based economic evaluations. While the societal perspective is the preferred perspective for economic evaluation, a majority of the studies excluded indirect costs which may underestimate total costs incurred by the patient and their family. A number of US-based studies estimated costs using Medicare reimbursement data, which may satisfactorily estimate costs for the elderly population (≥65 years) but not for the younger, commercially insured population. Thus, cost-effectiveness results vary primarily due to the various approaches and multiplicity of sources from which costs were estimated.

4.3 Strengths and Limitations

To our knowledge, this is the first study to systematically review the cost-effectiveness literature on available chemotherapeutic alternatives and targeted biologics for ovarian cancer treatment. Contrary to previous systematic and narrative reviews that were outdated or restricted to a specific drug (e.g., paclitaxel) [58, 59], this is the most comprehensive review incorporating economic evaluations over an extended period of time with quality assessed using a validated instrument. Additional strengths of this systematic review included the development of search strategies by a public health librarian with experience in systematic reviews (HVV). She worked in conjunction with the authors (IBP, RCP) to ensure that the appropriate terminology was employed. Both Medline (Ovid) and PubMed were searched even though they overlap significantly (Medline is a subset of PubMed); Embase (Ovid) was included to ensure that articles in journals with a non-North American focus were included. These databases include the vast majority of health economics publications. Great care was taken while screening to ensure each screener was blinded to the other’s screening decisions. Additionally, each screener was blinded to article authors and publication journals to reduce the potential for bias. However, the results should be interpreted in light of certain limitations. Our study selection criteria did not include any population criteria in terms of age, ethnicity, residence, or tumor stage. However, a majority of the included studies were based on modeling, and estimates for the economic evaluations were derived from clinical trials. The average age of the population varied across the clinical trials; though a median age of approximately 56–60 years was reported for the commonly used GOG-111, GOG-218, ICON-7, and OVA-301 trials. Additionally, over 75 % of the studies comprised patients with an advanced stage (stage III or higher). Some relevant studies may have been overlooked in our review, especially those that were not published in the English language. Similarly, we did not formally assess potential publication bias that may have occurred due to the lack of inclusion of unpublished studies (e.g., industry-sponsored evaluations), which may have had unfavorable findings. Methodological quality of studies as well as reporting procedures have immensely improved over the last 2 decades, and it may not be appropriate to directly compare some of the older studies with those conducted more recently. While most studies mentioned their cost year, for some studies that did not, inflation adjustment or conversion using purchasing power parity was not conducted. In addition, to achieve more comparability across studies we excluded those that were partial economic evaluations, cost-minimization analyses, and cost-benefit analyses.

5 Conclusion

Based on the available evidence, we conclude that administration of cisplatin–paclitaxel combination chemotherapy for first-line treatment appears to be the most cost-effective alternative. Notably though, most of the first-line platinum–taxane economic evaluations were based on a cisplatin–paclitaxel treatment arm, despite carboplatin being the preferred platinum choice in clinical settings for almost a decade [5, 60, 61]. Given that carboplatin has a favorable side-effect profile and tolerability compared with cisplatin, it may be important to determine its position in the cost-effectiveness literature for ovarian cancer. Use of targeted biologics (e.g., bevacizumab) did not demonstrate similar value. With drug price changes and bio-similar products becoming available over the next decade, cost-effectiveness of biologics should be revisited. There was reasonable agreement that in patients with recurrent platinum-sensitive disease, administration of a second-line platinum–paclitaxel combination or platinum monotherapy was cost-effective compared with platinum monotherapy or best supportive care, respectively. In contrast, there was uncertainty if PLD plus trabectedin was cost-effective compared with PLD alone in patients with recurrent partially platinum-sensitive disease. Given the limited evidence, it was difficult to draw strong conclusions about treatment alternatives for patients with recurrent platinum-resistant disease. Nevertheless, it appears that best supportive care may be most cost-effective, followed by second-line monotherapy with doxorubicin.

Notes

Acknowledgments

This study was supported in part by a grant from the Agency for Healthcare Research and Quality (R01-HS018956). None of the authors have any potential conflict of interest to report.

Author contributions

Dr. Poonawalla, Mr. Parikh and Ms. VonVille determined the search strategies. Dr. Poonawalla and Mr. Parikh identified the articles for inclusion in the review and drafted the manuscript. Drs. Lairson and Du supervised the overall progress and provided a final review and revisions. Dr. Lairson will serve as the overall guarantor of this review.

Supplementary material

40273_2015_304_MOESM1_ESM.docx (29 kb)
Supplementary material 1 (DOCX 29 kb)

References

  1. 1.
    American Cancer Society. Cancer Facts & Figures 2014. Atlanta: American Cancer Society; 2014 [Internet]. Available from: http://www.cancer.org/research/cancerfactsstatistics/cancerfactsfigures2014/.
  2. 2.
    Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JW, Comber H, et al. Cancer incidence and mortality patterns in Europe: estimates for 40 countries in 2012. Eur J Cancer. 2013;49(6):1374–403.CrossRefPubMedGoogle Scholar
  3. 3.
    American Cancer Society. Ovarian Cancer. Detailed Guide. [Internet]. Available from: http://www.cancer.org/cancer/ovariancancer/detailedguide/index.
  4. 4.
    Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics (2014). CA Cancer J Clin. 2014;64(1):9–29.CrossRefPubMedGoogle Scholar
  5. 5.
    Morgan RJ, Armstrong DK, Alvarez RD, Chen L, Copeland L, Crispens MA, et al. NCCN Clinical Practice Guidelines in Ovarian Cancer. v.3.2014.Google Scholar
  6. 6.
    Ovarian Cancer National Alliance and SEER Cancer Statistics Review, 1975–2005, National Cancer Institute. Bethesda, Md. http://seer.cancer.gov/csr/1975_2005/.
  7. 7.
    Szucs TD, Wyss P, Dedes KJ. Cost-effectiveness studies in ovarian cancer. Int J Gynecol Cancer. 2003;13(Suppl 2):212–9.CrossRefPubMedGoogle Scholar
  8. 8.
    Johnston SR. Ovarian cancer: review of the National Institute for Clinical Excellence (NICE) guidance recommendations. Cancer Invest. 2004;22(5):730–42.CrossRefPubMedGoogle Scholar
  9. 9.
    Sfakianos GP, Havrilesky LJ. A review of cost-effectiveness studies in ovarian cancer. Cancer Control. 2011;18(1):59–64.PubMedGoogle Scholar
  10. 10.
    Havrilesky LJ, Fountain C. Can we maximize both value and quality in gynecologic cancer care? A work in progress. Am Soc Clin Oncol Educ Book. 2014;34:e268–75.CrossRefGoogle Scholar
  11. 11.
    Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151(4):264,9, W64.Google Scholar
  12. 12.
    VonVille HM. Excel Workbook for Tracking Systematic Review Searches & Strategies and Screening Titles & Abstracts. 2014.Google Scholar
  13. 13.
    McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb). 2012;22(3):276–82.PubMedCentralCrossRefPubMedGoogle Scholar
  14. 14.
    Sim J, Wright CC. The kappa statistic in reliability studies: use, interpretation, and sample size requirements. Phys Ther. 2005;85(3):257–68.PubMedGoogle Scholar
  15. 15.
    VonVille HM. Excel workbook to calculate Cohen’s kappa. 2014.Google Scholar
  16. 16.
    Smieliauskas F, Chien CR, Shen C, Geynisman DM, Shih YC. Cost-effectiveness analyses of targeted oral anti-cancer drugs: a systematic review. Pharmacoeconomics. 2014;32(7):651–80.CrossRefPubMedGoogle Scholar
  17. 17.
    International Monetary Fund. World Economic Outlook Database. [Internet]. Available from: http://www.imf.org/external/pubs/ft/weo/2009/02/weodata/index.aspx.
  18. 18.
    Chiou CF, Hay JW, Wallace JF, Bloom BS, Neumann PJ, Sullivan SD, et al. Development and validation of a grading system for the quality of cost-effectiveness studies. Med Care. 2003;41(1):32–44.CrossRefPubMedGoogle Scholar
  19. 19.
    Zhang W, Islam N, Ma C, Anis AH. Systematic review of cost-effectiveness analyses of treatments for psoriasis. Pharmacoeconomics. 2015;33(4):327–40.CrossRefPubMedGoogle Scholar
  20. 20.
    Krysinski J, Placzek J, Fijalkowska K, Tujakowski J, Sadowska A, Windorbska W. Treatment of advanced ovarian cancer: cost-effectiveness analysis. Curr Gynecol Oncol. 2011;9(3):147–57.Google Scholar
  21. 21.
    Messori A, Trippoli S, Becagli P, Tendi E. Pharmacoeconomic profile of paclitaxel as a first-line treatment for patients with advanced ovarian carcinoma. A lifetime cost-effectiveness analysis. Cancer. 1996;78(11):2366–73.CrossRefPubMedGoogle Scholar
  22. 22.
    Covens A, Boucher S, Roche K, Macdonald M, Pettitt D, Jolain B, et al. Is paclitaxel and cisplatin a cost-effective first-line therapy for advance ovarian carcinoma? Cancer. 1996;77(10):2086–91.CrossRefPubMedGoogle Scholar
  23. 23.
    Elit LM, Gafni A, Levine MN. Economic and policy implications of adopting paclitaxel as first-line therapy for advanced ovarian cancer: an Ontario perspective. J Clin Oncol. 1997;15(2):632–9.PubMedGoogle Scholar
  24. 24.
    McGuire W, Neugut AI, Arikian S, Doyle J, Dezii CM. Analysis of the cost-effectiveness of paclitaxel as alternative combination therapy for advanced ovarian cancer. J Clin Oncol. 1997;15(2):640–5.PubMedGoogle Scholar
  25. 25.
    Ortega A, Dranitsaris G, Sturgeon J, Sutherland H, Oza A. Cost-utility analysis of paclitaxel in combination with cisplatin for patients with advanced ovarian cancer. Gynecol Oncol. 1997;66(3):454–63.CrossRefPubMedGoogle Scholar
  26. 26.
    Berger K, Fischer T, Szucs TD. Cost-effectiveness analysis of paclitaxel and cisplatin versus cyclophosphamide and cisplatin as first-line therapy in advanced ovarian cancer. A European perspective. Eur J Cancer. 1998;34(12):1894–901.CrossRefPubMedGoogle Scholar
  27. 27.
    Neymark N, Gorlia T, Adriaenssen I, Baron B, Piccart M. Cost effectiveness of paclitaxel/cisplatin compared with cyclophosphamide/cisplatin in the treatment of advanced ovarian cancer in Belgium. Pharmacoeconomics. 2002;20(7):485–97.CrossRefPubMedGoogle Scholar
  28. 28.
    Limat S, Woronoff-Lemsi MC, Menat C, Madroszyk-Flandin A, Merrouche Y. From randomised clinical trials to clinical practice : a pragmatic cost-effectiveness analysis of Paclitaxel in first-line therapy for advanced ovarian cancer. Pharmacoeconomics. 2004;22(10):633–41.CrossRefPubMedGoogle Scholar
  29. 29.
    Bristow RE, Santillan A, Salani R, Diaz-Montes TP, Giuntoli RL 2nd, Meisner BC, et al. Intraperitoneal cisplatin and paclitaxel versus intravenous carboplatin and paclitaxel chemotherapy for Stage III ovarian cancer: a cost-effectiveness analysis. Gynecol Oncol. 2007;106(3):476–81.CrossRefPubMedGoogle Scholar
  30. 30.
    Havrilesky LJ, Secord AA, Darcy KM, Armstrong DK, Kulasingam S, Gynecologic Oncology G. Cost effectiveness of intraperitoneal compared with intravenous chemotherapy for women with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol. 2008;26(25):4144–50.PubMedCentralCrossRefPubMedGoogle Scholar
  31. 31.
    Cohn DE, Kim KH, Resnick KE, O’Malley DM, Straughn JM Jr. At what cost does a potential survival advantage of bevacizumab make sense for the primary treatment of ovarian cancer? A cost-effectiveness analysis. J Clin Oncol. 2011;29(10):1247–51.CrossRefPubMedGoogle Scholar
  32. 32.
    Barnett JC, Alvarez Secord A, Cohn DE, Leath CA 3rd, Myers ER, Havrilesky LJ. Cost effectiveness of alternative strategies for incorporating bevacizumab into the primary treatment of ovarian cancer. Cancer. 2013;119(20):3653–61.PubMedGoogle Scholar
  33. 33.
    Chan JK, Herzog TJ, Hu L, Monk BJ, Kiet T, Blansit K, et al. Bevacizumab in treatment of high-risk ovarian cancer–a cost-effectiveness analysis. Oncologist. 2014;19(5):523–7.PubMedCentralCrossRefPubMedGoogle Scholar
  34. 34.
    Mehta DA, Hay JW. Cost-effectiveness of adding bevacizumab to first line therapy for patients with advanced ovarian cancer. Gynecol Oncol. 2014;132(3):677–83.CrossRefPubMedGoogle Scholar
  35. 35.
    Messori A, Trippoli S. Pharmacoeconomic profile of taxanes in advanced ovarian cancer. Anticancer Drugs. 1998;9(10):909–16.CrossRefPubMedGoogle Scholar
  36. 36.
    Orr JW Jr, Orr P, Kern DH. Cost-effective treatment of women with advanced ovarian cancer by cytoreductive surgery and chemotherapy directed by an in vitro assay for drug resistance. Cancer J Sci Am. 1999;5(3):174–8.PubMedGoogle Scholar
  37. 37.
    Dalton HJ, Yu X, Hu L, Kapp DS, Benjamin I, Monk BJ, et al. An economic analysis of dose dense weekly paclitaxel plus carboplatin versus every-3-week paclitaxel plus carboplatin in the treatment of advanced ovarian cancer. Gynecol Oncol. 2012;124(2):199–204.CrossRefPubMedGoogle Scholar
  38. 38.
    Lairson DR, Parikh RC, Cormier JN, Du XL. Cost-utility analysis of platinum-based chemotherapy versus taxane and other regimens for ovarian cancer. Value Health. 2014;17(1):34–42.CrossRefPubMedGoogle Scholar
  39. 39.
    Lesnock JL, Farris C, Krivak TC, Smith KJ, Markman M. Consolidation paclitaxel is more cost-effective than bevacizumab following upfront treatment of advanced epithelial ovarian cancer. Gynecol Oncol. 2011;122(3):473–8.PubMedCentralCrossRefPubMedGoogle Scholar
  40. 40.
    Fisher M, Gore M. Cost-effectiveness of trabectedin plus pegylated liposomal doxorubicin for the treatment of women with relapsed platinum-sensitive ovarian cancer in the UK: analysis based on the final survival data of the OVA-301 trial. Value Health. 2013;16(4):507–16.CrossRefPubMedGoogle Scholar
  41. 41.
    Lee HY, Yang BM, Hong JM, Lee TJ, Kim BG, Kim JW, et al. Cost-utility analysis for platinum-sensitive recurrent ovarian cancer therapy in South Korea: results of the polyethylene glycolated liposomal doxorubicin/carboplatin sequencing model. Clinicoecon Outcomes Res. 2013;5:297–307.PubMedCentralCrossRefPubMedGoogle Scholar
  42. 42.
    Havrilesky LJ, Pokrzywinski R, Revicki D, Higgins RV, Nycum LR, Kohler MF, et al. Cost-effectiveness of combination versus sequential docetaxel and carboplatin for the treatment of platinum-sensitive, recurrent ovarian cancer. Cancer. 2012;118(2):386–91.CrossRefPubMedGoogle Scholar
  43. 43.
    Montalar J, Cajaraville G, Carreras M-, Rubio MJ, Andres BG-, Oyaguez I, et al. Trabectedin plus PLD versus PLD monotherapy in patients with platinum-sensitive relapsed ovarian cancer: a cost-effectiveness analysis in Spain. Eur J Hosp Pharm Sci Pract. 2012;19(4):364–9.CrossRefGoogle Scholar
  44. 44.
    Case AS, Rocconi RP, Partridge EE, Straughn JM Jr. A cost-effectiveness analysis of chemotherapy for patients with recurrent platinum-sensitive epithelial ovarian cancer. Gynecol Oncol. 2007;105(1):223–7.CrossRefPubMedGoogle Scholar
  45. 45.
    Havrilesky LJ, Secord AA, Kulasingam S, Myers E. Management of platinum-sensitive recurrent ovarian cancer: a cost-effectiveness analysis. Gynecol Oncol. 2007;107(2):211–8.CrossRefPubMedGoogle Scholar
  46. 46.
    Main C, Bojke L, Griffin S, Norman G, Barbieri M, Mather L, et al. Topotecan, pegylated liposomal doxorubicin hydrochloride and paclitaxel for second-line or subsequent treatment of advanced ovarian cancer: a systematic review and economic evaluation. Health Technol Assess. 2006;10(9):1–132.CrossRefGoogle Scholar
  47. 47.
    Rocconi RP, Case AS, Straughn JM Jr, Estes JM, Partridge EE. Role of chemotherapy for patients with recurrent platinum-resistant advanced epithelial ovarian cancer: a cost-effectiveness analysis. Cancer. 2006;107(3):536–43.CrossRefPubMedGoogle Scholar
  48. 48.
    Fedders M, Hartmann M, Schneider A, Kath R, Camara O, Oelschlager H. Markov-modeling for the administration of platinum analogues and paclitaxel as first-line chemotherapy as well as topotecan and liposomal doxorubicin as second-line chemotherapy with epithelial ovarian carcinoma. J Cancer Res Clin Oncol. 2007;133(9):619–25.CrossRefPubMedGoogle Scholar
  49. 49.
    McGuire WP, Hoskins WJ, Brady MF, Kucera PR, Partridge EE, Look KY, et al. Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med. 1996;334(1):1–6.CrossRefPubMedGoogle Scholar
  50. 50.
    Piccart MJ, Bertelsen K, James K, Cassidy J, Mangioni C, Simonsen E, et al. Randomized intergroup trial of cisplatin-paclitaxel versus cisplatin-cyclophosphamide in women with advanced epithelial ovarian cancer: three-year results. J Natl Cancer Inst. 2000;92(9):699–708.CrossRefPubMedGoogle Scholar
  51. 51.
    Ozols RF, Bundy BN, Greer BE, Fowler JM, Clarke-Pearson D, Burger RA, et al. Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study. J Clin Oncol. 2003;21(17):3194–200.CrossRefPubMedGoogle Scholar
  52. 52.
    Armstrong DK, Bundy B, Wenzel L, Huang HQ, Baergen R, Lele S, et al. Intraperitoneal cisplatin and paclitaxel in ovarian cancer. N Engl J Med. 2006;354(1):34–43.CrossRefPubMedGoogle Scholar
  53. 53.
    Burger RA, Brady MF, Bookman MA, Fleming GF, Monk BJ, Huang H, et al. Incorporation of bevacizumab in the primary treatment of ovarian cancer. N Engl J Med. 2011;365(26):2473–83.CrossRefPubMedGoogle Scholar
  54. 54.
    Perren TJ, Swart AM, Pfisterer J, Ledermann JA, Pujade-Lauraine E, Kristensen G, et al. A phase 3 trial of bevacizumab in ovarian cancer. N Engl J Med. 2011;365(26):2484–96.CrossRefPubMedGoogle Scholar
  55. 55.
    Messori A, Trippoli S, Becagli P, Tendi E. Treatments for newly diagnosed advanced ovarian cancer: analysis of survival data and cost-effectiveness evaluation. Anticancer Drugs. 1998;9(6):491–502.PubMedGoogle Scholar
  56. 56.
    Monk BJ, Herzog TJ, Kaye SB, Krasner CN, Vermorken JB, Muggia FM, et al. Trabectedin plus pegylated liposomal Doxorubicin in recurrent ovarian cancer. J Clin Oncol. 2010;28(19):3107–14.CrossRefPubMedGoogle Scholar
  57. 57.
    Mullard A. Can next-generation antibodies offset biosimilar competition? Nat Rev Drug Discov. 2012;11(6):426–8.CrossRefPubMedGoogle Scholar
  58. 58.
    Young M, Plosker GL. Paclitaxel: a pharmacoeconomic review of its use in the treatment of ovarian cancer. Pharmacoeconomics. 2001;19(12):1227–59.CrossRefPubMedGoogle Scholar
  59. 59.
    Dedes KJ, Bramkamp M, Szucs TD. Paclitaxel: cost-effectiveness in ovarian cancer. Expert Rev Pharmacoecon Outcomes Res. 2005;5(3):235–43.CrossRefPubMedGoogle Scholar
  60. 60.
    du Bois A, Luck HJ, Meier W, Adams HP, Mobus V, Costa S, et al. A randomized clinical trial of cisplatin/paclitaxel versus carboplatin/paclitaxel as first-line treatment of ovarian cancer. J Natl Cancer Inst. 2003;95(17):1320–9.CrossRefPubMedGoogle Scholar
  61. 61.
    Neijt JP, Engelholm SA, Tuxen MK, Sorensen PG, Hansen M, Sessa C, et al. Exploratory phase III study of paclitaxel and cisplatin versus paclitaxel and carboplatin in advanced ovarian cancer. J Clin Oncol. 2000;18(17):3084–92.PubMedGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Insiya B. Poonawalla
    • 1
  • Rohan C. Parikh
    • 2
  • Xianglin L. Du
    • 1
    • 2
  • Helena M. VonVille
    • 3
  • David R. Lairson
    • 2
    • 4
  1. 1.Department of Epidemiology, Human Genetics and Environmental Sciences, School of Public HealthUniversity of Texas Health Science Center at HoustonHoustonUSA
  2. 2.Department of Management, Policy and Community Health, School of Public HealthUniversity of Texas Health Science Center at HoustonHoustonUSA
  3. 3.Library, School of Public HealthUniversity of Texas Health Science Center at HoustonHoustonUSA
  4. 4.HoustonUSA

Personalised recommendations