1Professional Masters in Management and Administration of Pharmaceutical Care, Faculty of Pharmacy, Fluminense Federal University (Mestrado Profissional em Gestão e Administração da Assistência Farmacêutica, Faculdade de Farmácia, Universidade Federal Fluminense), 2Department of Policy, Planning and Administration in Health, Institute of Social Medicine, Rio de Janeiro State University.(Departamento de Política, Planejamento e Administração em Saúde, Instituto de Medicina Social, Universidade do Estado do Rio de Janeiro), 3Faculty of Pharmacy, Fluminense Federal University(Faculdade de Farmácia, Universidade Federal Fluminense), 523 Mário Vianna St. 24241-000
Email: thaysgb@yahoo.com.br
Received: 24 Mar 2015 Revised and Accepted: 08 May 2015
ABSTRACT
Objective: To conduct a systematic review of cost-effectiveness analyses of trastuzumab in patients with metastatic breast cancer overexpressing human epidermal growth factor receptor 2.
Methods: A systematic review for the years 1998 to 2013 was conducted. A survey of the scientific literature was performed using six electronic databases, two search tools on the internet and a manual search of references using search strategies for each database. The selected studies were assessed for quality according to specific methodology. Data analysis was performed by converting the costs and by observations made from the incremental cost-effectiveness ratios of trastuzumab.
Results: From a detailed analysis of 521 retrieved articles, 13 articles were selected for this review. The treatment regimens adopted for cost-effectiveness were varied. Eight studies compared treatment with trastuzumab as first line for metastatic disease, and five studies compared treatment with trastuzumab as the second line. All the studies were using trastuzumab as second-line treatment was not considered a cost-effective intervention. The analyses used different thresholds to determine whether treatment with trastuzumab was cost-effective as well as to determine differences in modeling costs, outcomes and treatment patterns.
Conclusion: The use of trastuzumab, alone or combined, was cost-effective as first-line treatment. Differences were found in the quality of the included studies. Conducting new cost-effectiveness analyses of trastuzumab in metastatic breast cancer is required, in alliance with political, social and administrative factors, to help decision makers concerning its incorporation.
Keywords: Trastuzumab, Cost effectiveness, Breast cancer, Systematic review.
INTRODUCTION
According to estimates of the National Cancer Institute (Instituto Nacional do Câncer José Alencar Gomes da Silva-INCA), in 2014 and 2015, there will be approximately 576,000 new cases of cancer, of which 57,000 will be cases of female breast cancer. Breast cancer is the most frequent neoplasm in women from the Southeast, South, Midwest and Northeast regions of Brazil and the second most prevalent neoplasm in the North region, where cervical cancer is the leading neoplasm [1].
Approximately 30-50% of patients diagnosed in the initial stages will develop a metastatic disease. In this incurable phase of the disease, the therapies aim to prolong the survival and provide palliative control of symptoms [2]. The over expression of human epidermal growth factor receptor 2 (HER2) occurs in approximately 25% of all cases of breast cancer and is associated with lower disease-free interval and survival [3]. Trastuzumab, a monoclonal antibody directed at HER2, has been increasing the response, progression-free survival and/or overall survival rates in metastatic breast cancer [2, 4, 5].
Many countries have economic pressures from their healthcare systems and society to provide coverage for monoclonal antibodies due to their high prices [3]. Treatment with trastuzumab is unlikely to be affordable by most patients, unless it is covered by the public or private healthcare system [6]. The Ministry of Health has incorporated trastuzumab in the Coverage table of the Sistema Único de Saúde (SUS), the Brazilian Unified Health System, for initial and advanced breast cancer, excluding the metastatic phase [7, 8]. The main issues regarding the cost-effectiveness of trastuzumab use in metastatic breast cancer have remained unanswered in the literature, because data used to support these demands are dispersed across different studies from different organizational contexts. The studies found are often small-scale and methodologically flawed and refer to non-controlled clinical trials, making them unsuitable for formal meta-analyses [6].
It is important to consider that the successful development of a systematic review generates a reliable estimate of the effect of the intervention studied [9]. Therefore, a systematic literature review was performed to identify relevant articles concerning the cost-effectiveness of trastuzumab for metastatic breast cancer in patients with over expression of HER2 compared with other chemotherapy regimens adopted. The present systematic review secondarily aimed to add supporting information to future discussions on the cost-effectiveness ratio of trastuzumab and subsequent incorporation of trastuzumab by SUS for the treatment of metastatic breast cancer.
MATERIALS AND METHODS
Study design
A systematic review of the literature was performed with the following guiding question: “Is trastuzumab more cost-effective than other anticancer agents in patients with metastatic breast cancer with over expression of HER2?”
Data sources
The following electronic databases were used to identify published studies on the cost-effectiveness of trastuzumab for metastatic breast cancer: MEDLINE (via PubMed), EMBASE (via Portal Saúde Baseadaem Evidências), LILACS (via BIREME), Web of Science and Science Direct. Registers from the Cochrane Collaboration (via BIREME), the National Health System of the United Kingdom through the NHSEED database and other international technological evaluation agencies of the International Network of Agencies for Health Technology Assessment (INAHTA) were also analyzed. To retrieve additional studies, the search was expanded using the Google scholar search engine. As a complement, a manual search for references cited in the relevant publications obtained was performed in the mentioned databases. The search was limited to the period from 1998 to 2013, and there were no language restrictions in this step.
The search strategy was designed based on the following descriptors and their synonyms of each database: MeSH for MEDLINE (trastuzumab AND breast neoplasms AND cost-benefit analysis AND secondary), EmTree for EMBASE (trastuzumab AND “cost effectiveness analysis” AND “breast metastasis”)and Decs for LILACS [(trastuzumabe OR trastuzumab OR herceptin) AND ("custo-efetividade" OR "cost-benefit analysis" OR "cost effectiveness" OR "análisis costo-beneficio" OR "análise custo-benefício") AND ("neoplasias da mama" OR "câncer de mama" OR "breast neoplasms" OR "neoplasias de la mama") AND (secundário OR metastático OR secondary OR metastatic OR secundario)]. For the remaining electronic databases, keywords relating to the disease (“metastatic breast cancer”), intervention (“trastuzumab”) and type of economic analysis (“cost effectiveness”) were used with the necessary adaptations for each bibliographic database.
Studies from each database were placed in Microsoft® Excel (version 2010) spreadsheets to eliminate duplicates and create a reference database.
Selection of studies
Studies from the bibliographic search were independently analyzed by two researchers (T. S. A. and T. F. G. S.) to manually eliminate duplicate studies. The authors, titles and journal names were compared, along with the volume, number and year of the publication. The remaining studies were examined based on their title and abstract. Original studies that mentioned the technology, health problem studied and the type of economic evaluation (trastuzumab, metastatic breast cancer and cost-effectiveness analysis) in Portuguese, Spanish and English were selected.
Comments, editorials, letters, case studies, review articles, conference posters, systematic reviews and meta-analyses were excluded, as were duplicate studies performed by the same research group addressing the same object of the initial study. Studies that did not include information on costs and outcomes of the intervention, that addressed the impact of the metastatic breast cancer management costs on the health budget or that addressed the costs of therapeutic interventions used to correct adverse reactions caused by trastuzumab were also excluded from the review.
Articles selected as potentially relevant continued to the next step, in which they were thoroughly evaluated based on a full reading of the text, and only those that met al. l eligibility criteria, mentioned below, were included in the systematic review.
Disagreements in both steps regarding the inclusion or exclusion of studies in the review were resolved by a third reviewer (G. B. G. M.).
Data extraction and quality assessment of studies
The selected articles were fully read by the two reviewers, and the items listed below were independently extracted to a spreadsheet:
Differences in the extraction of information were resolved by consensus with a third reviewer upon the reassessment of the original article.
The quality assessment was independently performed by two reviewers adopting the format used by Teerawattananon et al. [10] with two different approaches. First, the practices reported for economic evaluation studies were assessed, namely, (a) the expression of the perspective used; (b) the description of the interventions to be compared; (c) the relationship between time horizon and discount; (d) the ICER data; (e) sensitivity analysis; and (f) the financial support statement.
The studies were assessed in a second stage according to the quality of the evidence used [11, 12]. The items comprising this assessment were: (a) the size of the clinical effects; (b) adverse events and complications; (c) basal clinical data; (d) the use of resources; (e) costs; and (f) utility (only applicable to cost-utility analyses). Coyle & Lee [11] asserted that it is necessary to identify specific factors related to the data sources for the main parameters that determine their suitability regarding the economic analysis in question. The data sources of each component were ranked from 1 to 6 in descending order. Studies are class 1 if the parameters are derived from the most reliable data sources. This quality assessment was summed using a table, in which the results found in each study were organized.
Synthesis and analysis of results
Data were extracted and interpreted regarding the ICER of trastuzumab (combined or alone) in metastatic breast cancer. The costs were converted to Brazilian Real using the conversion tool available on the website of the Central Bank of Brazil (Banco Central do Brasil) [13], and the results were compared with the budget ceiling recommended by the World Health Organization (WHO), but in Brazilian currency. According to the WHO, developing countries should adopt the threshold of up to three times the value of their gross domestic product (GDP) per capita in cost-effectiveness analyses of new technologies [14, 15]. Information on the GDP and Brazilian population was obtained from the website of the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística) [16].
The selected studies presented the use of trastuzumab in two different ways: as a first-or second-line therapy for metastatic breast cancer. The use of trastuzumab was performed alone or combined with other drugs. Given these characteristics, the conclusions of the cost-effectiveness analyses regarding the introduction of trastuzumab were grouped according to the existence of previous treatment of the metastatic diseases of patients.
Results
Search and selection of studies
In total, 521 studies were identified. Of these, 161 were excluded due to being duplicates, leaving 360 studies. A total of 294 studies were discarded after reading their titles because they did not meet the criteria specified for this step (studies related to the cost-effectiveness analysis of trastuzumab in metastatic breast cancer). Studies on the use of trastuzumab as an adjuvant and neoadjuvant therapy for breast cancer were excluded. Studies whose titles were not clear regarding the inclusion criteria were added to the subsequent step of reading of the abstracts, along with the other studies identified. A total of 66 studies were selected for reading of their abstracts. Forty-two studies were excluded by the same criteria mentioned previously, resulting in 24 studies to be fully read. No studies were excluded for being in a different language from those established for the present review. No additional studies were included by searching gray literature or by manually searching the references cited in the relevant publications obtained.
In the second step of the selection, 11 studies were excluded because 10 were conference posters, and 01 was an abstract of a technology evaluation report, whose objective was not to evaluate trastuzumab but to evaluate another drug. Finally, 13 studies comprised the present systematic review.
Data extraction and quality assessment of studies
Regarding the publication countries, 12 studies involved countries from North America and Europe. Only one study was developed in Brazil.
The studies were conducted within the last 10 years. The therapeutic regimens adopted in the cost-effectiveness analyses differed: eight studies compared the treatment with trastuzumab as a first-line therapy for metastatic disease, and five studies compared it as a second-line therapy; three of these latter studies used trastuzumab as a comparator in the evaluation of the cost-effectiveness ratio of other therapies, not as an intervention.
Table 1 shows the main characteristics of the studies. Table 2 shows the characteristics related to the cost-effectiveness analyses. All the studies were grouped according to the use of trastuzumab as a first-and second-line treatment.
Table 1: Main characteristics of the studies included
| Authors | Year | Country | Intervention | Comparator | HER2 test |
| Trastuzumab as 1st line therapy for metastatic disease | |||||
| [17] | 2012 | Greece | trast+dxt | docetaxel | not reported |
| [18] | 2004 | USA | IHC test, trast+txt for+3 | paclitaxel | IHC |
| IHC test confirmed by FISH for IHC+2 and+3, trast+txt | IHC confirmed by FISH for IHC+2 and+3 | ||||
| IHC test confirmed by FISH for IHC+2, trast+txt | IHC confirmed by FISH for IHC+2 | ||||
| IHC test, trast+txt for+2 and+3 | IHC | ||||
| FISH test, trast+txt for positive results | FISH | ||||
| trast+txt | none | ||||
| [19] | 2011 | United Kingdom | trast+anastrozole | anastrozole | not reported |
| [20] | 2009 | USA | trast+chemotherapy | chemotherapy | IHC and/or FISH (without details) |
| [21] | 2008 | Sweden | IHC test, trast+dxt for+3 | docetaxel | IHC |
| IHC test, trast+dxt for+2 and+3 | IHC | ||||
| IHC test confirmed by FISH for IHC+2 and+3, trast+dxt for FISH-positive results | IHC confirmed by FISH for IHC+2 and+3 | ||||
| FISH test, trast+dxt for FISH-positive results | FISH | ||||
| [22] | 2005 | Norway | trastuzumab+docetaxel | docetaxel | IHC |
| trastuzumab+anthracycline+cyclophosphamide | anthracycline+cyclophosphamide | ||||
| trastuzumab+paclitaxel | paclitaxel | ||||
| [23] | 2009 | France | trastuzumab (with or without taxanes) | Standard chemotherapy (taxanes and/or anthracycline-based chemotherapy) | IHC+3 and/or FISH |
| [24] | 2008 | France | trastuzumab+paclitaxel | conventional chemotherapy | IHC+3 |
| Trastuzumab as 2nd line therapy for metastatic disease | |||||
| [25] | 2005 | Italy | Cyclophosphamide+low-dose methotrexate | Trastuzumab+paclitaxel | not reported |
| [26] | 2012 | United Kingdom | Lapatinib+capecitabine | Trastuzumab+capecitabine | not reported |
| [27] | 2012 | Brazil | Lapatinib+capecitabine | Trastuzumab+capecitabine | not reported |
| [28] | 2010 | Sweden | Trastuzumab+capecitabine | Capecitabine | IHC or FISH |
| [29] | 2005 | Belgium | Trastuzumab | no trastuzumab (details not reported) | FISH |
Legend: dxt: docetaxel; USA: United States of America; FISH: fluorescence in situ hybridization; IHC: immunohistochemical; trast: trastuzumab; txt: paclitaxel.
Table 2: Main characteristics related to cost-effectiveness analyses
| Authors | Economic model | Time horizon | Perspective | Discount rate | Currency, year | Costs of drugs | Other costs | Outcomes evaluated |
| Trastuzumab as 1st line therapy for metastatic disease | ||||||||
| [17] | Area under the curve | 12 years | Third-party payer | 3.5% | Euro, 2011 | Trastuzumab: €3280/g Docetaxel: €5950/g |
Infusion, consultations and other drugs | Life years, QALY |
| [18] | Markov | Lifetime | Society | 3% | American dollar, 2002 | Trastuzumab: US$ 2301/cycle (no waste) Paclitaxel: US$ 1848/cycle |
Diagnostic tests, outpatient infusion, medical visit, pre-CT drugs, adverse reactions (treatment and monitoring), tickets, post-treatment progression and productivity of patients | Life years, QALY |
| [19] | Markov | 20 years | National Health System of the United Kingdom | 3.5% | Pound sterling, 2008-2009 | Trastuzumab: £1657.86/loading dose and £1292.88/regular dose Anastrozole: £68.56/box with 28 tablets |
Pharmaceutical services, heart rate monitoring, outpatient care, supportive care pre and post-progression, exemestane, drug waste, terminal treatment | Life years, QALY |
| [20] | Markov | 18 years | Not reported | 3% | American dollar, 1998 | Not reported | Costs of the diagnostic test, monitoring and treatment of adverse reactions | QALY |
| [21] | Markov | Lifetime | Society | Not reported | Swedish Krona (SEK), 2005 | Docetaxel: SEK 15406/cycle Trastuzumab: SEK 5653/week |
Diagnostic tests, outpatient care, adverse reactions (monitoring and treatment), palliative care, outpatient support after 1st line therapy | QALY |
| [22] | Comparison | 40 weeks | Third-party payer | 5% | Euro, 2003 | Trastuzumab: €39454/patient treated Other drugs: not reported |
Preparation and administration of chemotherapy, diagnostic test, daily outpatient rate, treatment and monitoring of adverse reactions | Life years |
| [23] | Design before and after | 56 months | Hospital | Not reported | Euro, 2002 | Not reported | Hospitalization, chemotherapy drugs and adjuvants, imaging and laboratory tests | Life years |
| [24] | Clinical trial | Cost: 24 weeks; Effectiveness: 14 months | Hospital | Not reported | Euro, 2002 | Trastuzumab: €4200/g Paclitaxel: €4600/g Docetaxel: €8000/g Epirubicin: €1900/g |
Diagnostic tests, pre-treatment assessment, monitoring, daily hospital rate | Life years |
| Trastuzumab as 2nd line therapy for metastatic disease | ||||||||
| [25] | Comparison of clinical trials | Not reported | National Italian health system | Not reported | Euro, 2003 | CTX+MTX: €38/patient Trast+paclitaxel: €42423/patient |
Administration of chemotherapy, hospitalization, medical visits, clinical tests, supportive care, administration of support treatment | Progression-free life years and overall tumor response |
| [26] | Area under the curve | 5 years | National health system of the United Kingdom | 3.5% | Pound, 2008 | lap+cap: £13835/patient trast+cap: £12780/patient |
Preparation, administration, medical visits, hospitalization, laboratory and imaging tests, radiotherapy | Life years, progression-free and post-progression life years, QALY |
| [27] | Area under the curve | 5 years | National Brazilian health system | 5% | Brazilian Real, 2010 | lap+cap: R$66775/patient trast+cap: R$88833/patient |
Medical care, hospital care, laboratory and imaging tests | Years, progression-free survival years, QALY |
| [28] | Markov | Lifetime | Swiss health system | 0% | Euro, 2009 | Capecitabine: €1.29/tablet at0,15 g and €3.85/tablet at0,5 g Trastuzumab: €5250/g |
Preparation and administration of chemotherapy, laboratory tests, treatment and monitoring of adverse reactions, disease progression treatment | QALY |
| [29] | Developed by the authors | Lifetime | Hospital or care provider | Not reported | Euro, 2002-2003 | Not separately reported from the total cost of the disease | Professionals involved, drugs, materials, equipment, hospitalization, preparation of drugs, sterilization of materials, equipment maintenance, laboratory tests, surgical procedures | Life months |
Legend: cap: capecitabine; tab: tablet; CTX: cyclophosphamide; lap: lapatinib; MTX: methotrexate; QALY: quality-adjusted life years; CT: chemotherapy; trast: Trastuzumab.
Regarding effectiveness, most economic evaluations based their efficacy estimates on the results from four controlled randomized clinical trials: three economic evaluations based on Marty et al. [30], three based on Slamon et al. [31], three based on Von Minckwitz et al. [32] and one evaluation based on Kaufman et al. [33]. Four economic evaluations used results from non-randomized studies, including one comparison of the mean survival between two single-arm clinical studies [25], one phase II clinical trial [29], one before-and-after design study of a single treatment center [23] and one comparison of the results between four treatment centers that treated patients with trastuzumabe, with six centers that did not treat patients with trastuzumab [24]. Similar to the effectiveness, comparators and treatment regimens were also selected based on randomized controlled clinical trials. The trastuzumab dose was very similar among the studies, ranging from a loading dose of 0.004 g/kg, followed by 0.002 g/kg weekly, to 0.008 g/kg as the loading dose and 0.006 g/kg every 3 weeks. There was a wide variety of other drugs used as comparators that were already used in the treatment of the metastatic disease (anastrozole, anthracyclines, capecitabine, cyclophosphamide, lapatinib, methotrexate, docetaxel, and paclitaxel) [34]. Most studies reported using one or two biochemical tests to determine the HER2 status in the tumor cell membrane. However, only Elkin et al. [18] and Lidgren et al. [21] listed the cost-effectiveness results for each HER2 test strategy used.
The Markov model was the most used economic method among the evaluations. Two (progression-free and post-progression) to three (stable disease, response, progressive disease) stages were used, followed by death. There was great variation between the time horizons adopted by the studies, from 24 weeks to the lifetime of the patient.
Most of the evaluations used direct costs, from the perspective of the national health system, whereas two evaluations used the society’s perspective. The resources used varied drastically between these evaluations. For example, one of them used only medical visits and the administration of drugs [17], whereas another evaluation also used palliative care [19]. In addition, five evaluations did not report the use of the discount rate in the costs or outcomes [21, 23-25, 29], and there was even one evaluation [28] that reported that their costs and outcomes were not discounted (0%).
Similar to the effectiveness data, all economic evaluations also obtained the values of outcomes from studies published in the literature.
All the studies calculated the ICER, shown in table 3. Because two studies [25, 27] reported the ICER regarding two outcomes different from the remaining studies, this ratio was removed from the table, leaving only calculations related to life years and QALY gained.
Table 3: Results of the incremental cost-effectiveness ratio
| Authors | Additional cost | Additional life years | Additional QALY | ICER (cost per life years) | ICER (cost per QALY) | Conclusion of the authors |
| Trastuzumab as 1st line therapy for metastatic disease | ||||||
| [17] | €27323.98 | 0.729 | 0.449 | 37759.97 | €61323.33 | Cost-effective |
[18] IHC confirmed by FISH for IHC+2 and+3, trast+txt vs no test, txt |
US$10388 | Not applicable | 0.08 | Not applicable | US$125100 | Cost-effective |
[18] FISH, trast+txt for positives vs IHC confirmed by FISH for IHC+2 and+3, trast+txt |
US$1036 | Not applicable | 0.01 | Not applicable | US$145400 | Cost-effective |
| [19] | £37899 | 0.669 | 0.545 | Not reported | £69514 | Not cost-effective |
| [20] | US$47728 | Not applicable | 0.56 | Not applicable | US$85676 | Not reported |
[21] IHC confirmed by FISH for IHC+2 and+3, trast+dxt vs no test, dxt |
SEK85064 | Not reported | 0.176 | SEK332252 | SEK485039 | Cost-effective |
[21] FISH, trast+dxt vs IHC confirmed by FISH for IHC+2 and+3, trast+dxt |
SEK8442 | Not reported | 0.015 | SEK384427 | SEK561207 | Cost-effective |
| [22] | €44196 | 0.3 to 0.7 | Not applicable | €69212 to €162417 | Not applicable | Not cost-effective |
| [23] | €26812 | 1.5 | Not applicable | €17874.66 | Not applicable | Cost-effective |
| [24] | €21980 | 1.43 | Not applicable | €15370 | Not applicable | Cost-effective |
| Trastuzumab as 2st line therapy for metastatic disease | ||||||
| [25] | €64164 | 0.477 progression-free life years | Not applicable | €134516/progression-free life years | Not applicable | Not cost-effective |
| [26] | €107 | 0.019 | 0.031 | Dominated | Dominated | Not cost-effective |
| [27] | R$18430 | 0.23 | 0.131 | Dominated | Dominated | Not cost-effective |
| [28] | €33980 | Not applicable | 0.35 | Not applicable | €98329 | Not cost-effective |
| [29] | €3258.42 | 0.82 life months | Not applicable | €3981.44/life months gained | Not applicable | High cost-effectiveness ratio |
Legend: dxt: docetaxel; FISH: fluorescence in situ hybridization; ICER: incremental cost-effectiveness ratio; IHC: immunohistochemical; QALY: quality-adjusted life years; trast: trastuzumabe; txt: paclitaxel.
Of the eight evaluations that used trastuzumab as a first-line therapy, five concluded that this drug could be considered cost-effective. Of these, three evaluations [17, 21, 23] specified the threshold against which the ICER was compared. One evaluation related a minimum value that the society was willing to pay (“willingness to pay”) for health improvements [18], and another reported that the ICER was “considered as acceptable in a developed country where the health system is willing to offer the population full access to innovative treatments.” [24]. All the studies that used trastuzumab as a second-line therapy did not consider the intervention as cost-effective.
Regarding the first approach of the quality assessment (table 4), and following the methodology for economic evaluations [10], the least met criteria were the use of the discount rate (53.8%) and disclosure of funding sources (61.5%).
The ICER was calculated in all the studies. Two studies [24, 29] did not perform the sensitivity analysis, whereas four did not report the perspective used in the evaluations [19-24]. Four other studies [18, 20, 22, 29] did not adequately describe the comparators used (dosage and duration of treatment).
Table 4: Results of the first stage of the quality assessment
| Recommendations | Number of studies fulfilling recommendation | Percentage (%) |
| Perspective specified | 09 | 69.2 |
| Descriptionof comparators | 09 | 69.2 |
| Used discounting for costs and/oroutcomes iftime horizonwas>1 year | 07 | 53.8 |
| Calculated and reported ICER | 13 | 100 |
| Performed uncertainty analysis | 11 | 84.6 |
| Disclosed funding sources | 08 | 61.54 |
Table 5 shows the second approach adapted by Cooper et al. [12]. The utilities and base clinical data criteria were not mandatory for every article, and thus were the least met criteria. Eight studies used randomized controlled clinical trials, measuring end results, to insert clinical effects data into economic evaluations, except for five studies [20, 23-25, 29]. Only three studies collected their own administrative (hospital) data to define the types and amount of resources used during the treatment of metastatic breast cancer [23, 24, 29].
Table 5: Representation of results of the second stage of the quality assessment
| Evidence rank | Size of clinical effects/adverse events and complications [n (%)] |
Base clinical data (if applicable) [n (%)] | Use of resources [n (%)] | Costs [n (%)] |
Utilities (if applicable) [n (%)] |
| 1+ | |||||
| 1 | 8 (61.5) | 3 (60) | 3 (25) | 12 (92.3) | |
| 2+ | |||||
| 2 | 7 (58.3) | 1 (7.7) | |||
| 3+ | |||||
| 3 | 8 (100) | ||||
| 4 | 4 (30.8) | 2 (40) | |||
| 5 | |||||
| 6 | 2 (16.7) | ||||
| 9 | 1 (7.7) |
Summary and analysis of results
The ICERs were converted to Brazilian Real (R$) to be compared with the threshold usually recommended by technology evaluation agencies and adopted by the Department of Science and Technology of the Ministry of Health (table 6). According to WHO, a technology is considered to be very cost-effective if the ICER is less than the GDP per capita, cost-effective if the ICER is between 1 and 3 times the GDP per capita, and not cost-effective if the ICER is greater than 3 times the GDP per capita [15].
Table 6: Results of the incremental cost-effectiveness ratio using costs transformed into Brazilian Real
| Authors | Additional cost (R$) | Additional life years | Additional QALY | ICER (cost per life years) (R$) | ICER (cost per QALY) (R$) |
| Trastuzumab as 1st line therapy for metastatic disease | |||||
| [17] | 87521.44 | 0.729 | 0.449 | 120056.84 | 194925.25 |
[18] IHC confirmed by FISH for IHC+2 and+3, trast+txt vs no test, txt |
24170.80 | Not applicable | 0.08 | Not applicable | 302135.00 |
[18] FISH, trast+txt for positives vs IHC confirmed by FISH for IHC+2 and+3, trast+txt |
2410.56 | Not applicable | 0.01 | Not applicable | 241056.00 |
| [19] | 144524.05 | 0.669 | 0.545 | 216029.97 | 265181.74 |
| [20] | 111053.51 | Not applicable | 0.56 | Not applicable | 198309.83 |
[21] IHC confirmed by FISH for IHC+2 and+3, trast+dxt vs no test, dxt |
30342.33 | Not reported | 0.176 | 118514.28 | 172399.60 |
[21] FISH, trast+dxt vs IHC confirmed by FISH for IHC+2 and+3, trast+dxt |
3011.26 | Not reported | 0.015 | 137125.11 | 200750.66 |
| [22] | 141564.21 | 0.3 to 0.7 | Not applicable | 202234.58 to 471880.70 | Not applicable |
| [23] | 85881.52 | 1.5 | Not applicable | 57254.35 | Not applicable |
| [24] | 70404.14 | 1.43 | Not applicable | 49233.66 | Not applicable |
| Trastuzumab as 2st line therapy for metastatic disease | |||||
| [25] | 205523.71 | 0.477 progression-free life years | Not applicable | 430867.31/progression-free life years | Not applicable |
| [26] | 342.73 | 0.019 | 0.031 | Dominated | Dominated |
| [27] | 18430 | 0.23 | 0.131 | Dominated | Dominated |
| [28] | 108841.34 | Not applicable | 0.35 | Not applicable | 310,975.25 |
| [29] | 10437.05 | 0.82 life months or 0.0683 life years | Not applicable | 12728.11/life months gained or 152811.85/life years |
Not applicable |
Based on the R$ 67,206.00 Brazilian threshold (3 × GDP per capita of R$ 22402.00), none of the evaluations reported an ICER that allows trastuzumab to be considered cost-effective.
DISCUSSION
The present review indicated that the use of trastuzumab, alone or in combination, was cost-effective as a first-line therapy for metastatic disease, in contrast to when it is used as a second-line therapy. In the five studies that used trastuzumab as a second-line therapy, four authors [26-29] stated that the differences in the drug prices were the main responsible factors for the differences between the treatments. Bocci et al. [25] stated that their intervention (cyclophosphamide+methotrexate) represented a “good value for money and efficient use of health care resources”.
Important limitations were observed in some economic evaluations, particularly in those that used non-randomized approaches to estimate efficacy data [23,24]. Comparing the results of different treatment centers, as performed by Poncet et al. [24], using the time (“before-and-after”), as Perez-Ellis et al. [23] did, or making indirect comparisons between different clinical trials, as performed by Bocci et al. [25], Delea et al. [26] and Machado & Einarson [27], may be biased if there are differences between the trials regarding the methods (for example, treatment and measurement of results) or characteristics of the patients that modify the effects of the compared treatments. There is a strong recommendation to use the highest evidence levels when assessing the efficacy of interventions. Randomized clinical trials, syntheses of systematic reviews and meta-analyses are the most robust studies because their design allows controlling for potential biases and confounding factors [9, 14, 35].
Currently, there are two biochemical assays on the market to determine the HER2 status in tumor cell membranes. The immunohistochemical (IHC) assay tests the overexpression of the HER2 protein, whereas the fluorescence in situ hybridization (FISH) detects the amplification of the HER2 gene. This latter test is considered as a gold standard, but it is also the most expensive one [36].
The studies of Elkin et al. [18] and Lidgren et al. [21] combined the use of the HER2 test with the treatment strategy, and this test had an important influence on the cost-effectiveness ratio of the therapy. When the “test+treatment” strategies were compared, the differences between the tests (with the same treatments) could already make the strategy cost-effective. These two authors found that the most cost-effective strategy is to test all patients using IHC, followed by confirmation of IHC2+and IHC3+with FISH, a strategy that is consonant with the systematic review conducted by Dendukuri et al. [36], and trastuzumab plus chemotherapy for FISH-positive results. And also testing all patients with FISH and treatment of positives, if the threshold values are higher than those commonly practiced.
Another important point raised in the studies analyzed that may have generated different conclusions was the value of the drugs and resources used in different countries. The prices and expenses may vary according to the type of health system and the study’s perspective. Any decrease in the costs would result in a subsequent decrease in the cost-effectiveness ratio [37]. Matter-Walstra et al. [28], in their sensitivity analysis, showed that, depending on the body surface, a decrease of 30-60% in the cost of trastuzumab would result in an ICER below the threshold used in the study, indicating that the use of trastuzumab would be cost-effective as a second-line therapy.
Some evaluations excluded treatment beyond disease progression, whereas others considered trastuzumab waste when calculating its cost, a situation that does not always occur in clinical practice. There were other differences regarding the use of resources, such as considering the monitoring and treatment of adverse reactions, palliative care, and pharmaceutical services. All these factors contributed to the calculation of the total cost of the treatment, thus influencing the cost-effectiveness ratio. For Parkinson et al. [6], the rate of adverse reactions and their monitoring and treatment, the extension of treatment beyond progression, the level of trastuzumab waste and the use of chemotherapy after disease progression most likely can only be precisely determined after the widespread use of the intervention, instead of using data from a controlled randomized clinical trial.
In the present review, it was also found that the studies analyzed used different thresholds and decision analyses when evaluating the ICER of trastuzumab. The choice of a cost-effectiveness threshold depends on who is making the decision, what is the purpose of the analysis, how much the society is willing to pay for health improvements, what are the resources available, and how the decision maker values health, money and risk [38]. This explains the different cost-effectiveness thresholds found in the present review because the evaluation of authors from each country (even within the same country) was based on each one of the previously described factors. The two evaluations [23,24] that reported higher estimates for life years gained (approximately 1.5 life years gained) provided the lowest ICER, and these estimates may have influenced the ICER. This result occurred because these two evaluations used results from non-randomized studies, in which selection or allocation (patients with a worse prognosis in the comparator arm, for example) biases may have occurred that would overestimate the efficacy of trastuzumab.
The findings of the present systematic review are consistent with those of other published reviews on the subject. A literature search showed five systematic reviews on the cost-effectiveness ratio of trastuzumab in metastatic breast cancer: Lewis et al. [39]; Ferrusi et al. [40]; Blank et al. [37]; Fleeman et al. [19]; and Parkinson et al. [6]. The review by Ferrusi et al. [40] covered breast cancer treatment as a whole-initial and metastatic-but the other four reviews evaluated only the metastatic phase of the disease. The study by Lewis et al. [39] included only two economic evaluations, extracted from the submission of Roche to the National Health System (NHS) to obtain the trastuzumab registration. The study by Fleeman et al. [19] did not include any economic evaluation in their review; therefore, the authors decided to make their own cost-effectiveness analysis. The systematic review by Blank et al. [37] covered not only trastuzumab but also other anticancer drugs and molecular targeted agents. However, the authors did not perform a quality assessment. Parkinson et al. [6] did not use a formal methodology to assess the quality of the studies and only discussed issues that could affect the quality of an economic evaluation based on the checklist of Drummond et al. [41].
Regarding the quality assessment of the economic evaluations, the methodology used by Lewis et al. [39] and Ferrusi et al. [40] in their systematic reviews was the one developed by Drummond et al. [41]. This methodology results in a thorough interpretation of economic evaluations because it covers all the important items for its development and is presented as open questions. To make this methodology more practical, Lewis et al. [39] adapted it to a checklist with closed questions that could be answered by binary categorical variables (yes/no).
In the present review, the methodology used by Teerawattananon et al. [10] was adopted. The advantage of this method is that it provides a more thorough analysis of the evidence that was used as a basis to develop the economic studies. It was concluded that the included studies were of good methodological quality, both in the first and in the second approach, because there was little heterogeneity among the criteria evaluated.
The limitation of the present study was that the second approach of the quality assessment was not conducted by a third reviewer, a strategy that was different from the other steps of the systematic review. In addition, conference abstracts were excluded for not meeting the eligibility criteria, but they could contain relevant information on the cost-effectiveness ratio of trastuzumab for metastatic breast cancer.
CONCLUSION
The present study aimed to address the cost-effectiveness ratio of interventions with trastuzumab in the treatment of metastatic breast cancer, investigating whether the existing literature would allow drawing any conclusions about this issue. Secondarily, the study also allowed exploration of the main drivers of the conflicting conclusions among the authors, leading to the thorough examination of the cost-effectiveness ratio of trastuzumab for metastatic breast cancer. This was possible because the included economic evaluations showed differences in the modeling of costs, outcomes, treatment patterns between the countries (such as the method for HER2 tests) and thresholds considered for a technology to be cost-effective.
After converting the values used in the studies to Brazilian Real (R$), trastuzumab would not be considered cost-effective in any assessment using the Brazilian threshold. However, the differences in the use of resources and funding between the various health care systems may limit the transferability of foreign results into the Brazilian context. To consider trastuzumab as cost-effective for metastatic breast cancer, some topics must be considered: first, the treatment protocol (combination with other drugs or patients with or without recurrence). Second, the decision must also consider the use of the HER2 test and which HER2 status (2+and/or 3+, confirmed or not by FISH) will receive treatment because this test affects the cost-effectiveness results.
New cost-effectiveness analyses of trastuzumab for metastatic breast cancer are required, preferably combined with future randomized clinical trials in which data on the cost and effectiveness are collected simultaneously with the comparison of groups, providing more accurate data and allowing the comparison of the adverse effects profile of both therapies. It is also important that these economic evaluations cover new drugs already available for metastatic breast cancer (e. g., lapatinib and pertuzumab) to generate information useful particularly for treatment beyond progression (second-line or subsequent treatment).
Finally, any decisions on the incorporation of trastuzumab for the treatment of metastatic breast cancer in Brazil should not be based only on economic studies. The decision of introducing a new public health intervention should consider multiple factors, including, in addition to the effectiveness of the intervention, its safety, the financial costs of initiating and maintaining the treatment, the monitoring of the clinical results of its use and the necessary infrastructure to successfully provide the intervention, as well as the political will and equity promotion factors.
CONFLICT OF INTERESTS
Declared None
REFERENCES