Investigating the Value of Abatacept in the Treatment of Rheumatoid Arthritis: A Systematic Review of Cost-Effectiveness Studies

Background. Rheumatoid arthritis is a progressive inflammatory disease that affects greatly patients' quality of life and demands for aggressive management early on during the course of the disease. The discovery of biologics has equipped rheumatologists with evolutionary treatment tools but has also impacted greatly management costs. Objectives. To conduct a systematic review in order to evaluate the cost effectiveness of abatacept in the treatment of moderate to severe rheumatoid arthritis. Methods. Pubmed, the International Society for Pharmacoeconomics and Outcomes Research Outcomes Research Digest, the National Health System Economic Evaluation Database, and the Database of Abstracts of Reviews of Effects were searched. Results. In total 301 studies were identified and 42 met the inclusion criteria. Half of the selected studies evaluated abatacept in the treatment of rheumatoid arthritis, after failure of or intolerance to tumor necrosis factor alpha inhibitors. Of those, 82% were in favor of abatacept as a cost-effective or dominant strategy versus varying alternatives, whereas 18% favored other treatments. Conclusion. The majority of evidence from the published literature supports that abatacept can be a cost-effective alternative in the treatment of moderate to severe rheumatoid arthritis, especially in patients that have demonstrated inadequate response or intolerance to anti-TNF agents or conventional disease modifying antirheumatic drugs.


Introduction
Rheumatoid arthritis (RA) is a systemic inflammatory disease that presents itself in multiple joints of the musculoskeletal system. Symptoms include joint swelling, redness, and pain with gradual joint deformity in some cases. Due to its pathophysiology, RA causes not only significant morbidity and progressive loss of quality of life, but also carries a substantial economic burden, both for the individuals as well as for the society as a whole, since it is associated with high intense short-and long-term healthcare resource utilizations due to its increased prevalence and disability potential [1].
Over the last two decades, researchers have equipped rheumatologists with revolutionary therapeutic options. However, these changes have not been fully brought into effect in many European countries and other parts of the world [2,3]. Disease-modifying antirheumatic drugs (DMARDs) represent the mainstay of RA management. Corticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs) are also employed, mainly in the short-term, arthritis-related, symptomatic mitigation. Recently, several disease modifying biologic agents have been licensed for the management of RA, alone or in combination with methotrexate (MTX). These agents include abatacept (ABA), adalimumab (ADA), anakinra (ANA), certolizumab pegol (CER), etanercept (ETA), golimumab (GOL), infliximab (INF), rituximab (RTX), and tocilizumab (TOC).
Abatacept is a selective modulator that blocks T-cell activation. It has a marketing authorisation for use in combination with MTX for the treatment of moderate to severe active rheumatoid arthritis in adults whose disease has responded inadequately to previous therapy with one or more DMARDs, including MTX or a tumour necrosis factor (TNF) inhibitor [4]. Its unique mode of action results in inhibition of the inflammatory cascade by preventing activation of T-cells through binding to the natural ligands CD80 and CD86 not 2 ISRN Rheumatology allowing their interaction with CD28 on the T-lymphocyte [5].
The plethora of therapeutic options available and the variability of patient subgroups that represent target populations for each medication have led to the publication of a number of guidelines and treatment algorithms internationally. Among the agencies that have published such guidance, the National Institute of Clinical Excellence (NICE) has issued several guidelines on the use of biologics in RA [6], producing a rather complex prescribing regime [7]. According to the NICE recommendations, TNF antagonists are the only class of biologic agents that can be used as first line therapy, limiting the use of other agents with different mode of action. Biologic DMARD sequencing cannot be pinpointed accurately after intolerance to or failure of a TNF antagonist, and thus treatment regimens should be based not only to the licensed indications, but to the cost-effectiveness data available [6]. The latter has become a global objective, especially today, since scarcity of resources is more than evident.
In light of the aforementioned, the purpose of the present study was to conduct a systematic review of cost-effectiveness studies of abatacept in order to evaluate the role of this agent in the treatment of moderate to severe rheumatoid arthritis, specializing to treatment algorithms after previous inadequate response or intolerance to disease modifying antirheumatic drugs (DMARDs) and/or TNF inhibitors.

2.1.
Overview. The literature search was extended to a wide range of databases, in order to include as much and as recent information as possible. The databases that were searched included Pubmed, the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Outcomes Research Digest, the National Health System Economic Evaluation Database (NHS EED), and the Database of Abstracts of Reviews of Effects (DARE). Overlapping hits were included once. The literature search was not limited by study type or publication year; however, study selection was limited to articles written in English (inclusion and exclusion criteria are listed in detail later). For economic evaluation data, the outcomes included cost effectiveness and related types of analyses (including cost consequence, cost of treatment, and budget impact studies).

Preliminary Search.
A preliminary database search was conducted in order to challenge the systematic review strategy, focusing on the keywords to be used and the need for limiting the main literature search. Furthermore, it provided an estimate of the size of the literature available on the specific topic.

Main Search.
The main search was carried out in August 2012. Slightly different strategies were employed for each database searched due to inherent limitations pertaining to the type and classification of information. An initial search using the keywords "abatacept, cost effectiveness, and rheumatoid arthritis" was carried out in Pubmed. The Pubmed search was enriched by selecting the "related citations" option to each one of the initial results. The study that was selected as the basis for related citations was one containing all keywords in the title and was deemed the most relevant to the objective [8]. The ISPOR Outcomes Research Digest was searched by using the in-title term "abatacept" in the "rheumatoid arthritis" disease/disorder section. Additional searches employed the same strategy; however, the disease/disorder section selected was "arthritis" and "Muscular-skeletal Disorders (including Tunnel Syndrome). " No other limits were used. DARE and NHS EED databases were searched through the Centre for Reviews and Dissemination. "Abatacept" and "rheumatoid arthritis" were searched as keywords to the titles of published studies, without limiting databases or record types.

Study Selection.
Two members of the review team assessed the titles of the identified results independently. All relevant hits were then cross-checked against the inclusion and exclusion criteria. Studies that were identified by only one of the reviewers were examined in a separate meeting, in order for an inclusion decision to be reached.

Inclusion Criteria.
The criteria which all initial results were examined against are listed as follows: (i) studies comparing abatacept to other biologics or conventional therapies in rheumatoid arthritis; (ii) studies included both systematic reviews as well as stand-alone economic evaluation studies; (iii) types of costs included direct and indirect costseffectiveness measures included all life expectancy and quality of life outcome measures; (iv) studies focusing on adult patients.

Exclusion
Criteria. Exclusion criteria were as follows: (i) studies comparing abatacept to other treatments in different indications than rheumatoid arthritis; (ii) studies not reporting cost data-comparisons to other treatment options only on clinical effectiveness grounds were excluded; (iii) studies not specifying a time horizon for costs or results (for standalone economic evaluation studies); (iv) studies comparing treatments in juvenile rheumatoid arthritis; (v) studies with no research abstract and/or author details available; (vi) articles commenting on previous research studies; (vii) studies that did not meet the quality assessment criteria.

Results
In total, 301 studies were identified. The initial Pubmed search produced 19 hits and the secondary search produced another 197 results. The CRD database produced 8 hits and finally the ISPOR database revealed 50, 24, and 3 matches for "rheumatoid arthritis, " "arthritis, " and "Muscular-skeletal Disorders (including Tunnel Syndrome)" disease sections, respectively. After the inclusion and exclusion criteria were applied, 42 studies were finally selected. The majority of rejected studies were due to lack of cost data, failure to include abatacept as a comparator to other biologic agents, and failure to include RA as a treatment indication. Other reasons for exclusion included failure to meet the quality assessment criteria (1 study) and comparison of different biologic sequences, where results were presented for single agents (other than abatacept) and not the entire sequencing strategy containing abatacept (2 studies).

Abatacept after Conventional DMARD.
In total 9 studies on the cost effectiveness of abatacept in the treatment of rheumatoid arthritis after failure of or intolerance to conventional DMARDs were identified. All retrieved publications followed the cost-effectiveness analysis. Two of the previous also presented budget impact data for the agents under comparison. The results are presented in detail in Table 1.

Abatacept after TNF-Antagonist.
In total 21 studies on the cost effectiveness of abatacept in the treatment of rheumatoid arthritis after failure of or intolerance to TNF-a inhibitors were identified. Apart from the "classic" economic evaluations in the form of cost-effectiveness analysis, the results under this section included other approaches, such as cost of therapy and budget impact analysis. The results are presented in detail in Table 2.

Other Pharmacoeconomic Studies Involving Abatacept.
An additional 12 pharmacoeconomic studies relevant to the role of abatacept in the treatment of rheumatoid arthritis were identified but could not be included into one of the two major treatment pathways mentioned earlier. Those studies mainly included abatacept use after conventional DMARD therapy and anti-TNF replacement or studies that failed to specify the line or sequence of treatment and simply compared the cost effectiveness of abatacept against other modes of therapy. Those results are presented in Table 3.

Summary of Findings.
A wide range of economic evaluation and cost-analysis studies were identified with several comparators under investigation. Abatacept mainly appears to have been studied after failure of a TNF inhibitor. In regards to the economic evaluation studies, abatacept's cost effectiveness appears to range within the acceptable costeffectiveness thresholds according to the authors in the majority of studies. In general, abatacept versus comparators after conventional DMARDS appear to have a higher cost effect ratio when compared to abatacept versus comparators after TNF failure. Other cost analyses are highly dependent on the country of origin and the comparator treatments used.

Discussion
The disease and economic burden of RA remains substantial for the patients and health systems of developed economies worldwide. The continuously increasing therapeutic armamentarium provides clinicians with numerous options for the treatment of RA. Even when adopting the health economics point of view, it is safe to emphasize on the need for mutual patient-clinician agreement on the treatment options and strategy to be followed throughout the course of the disease, as recommended by an international task force on RA [3]. However, the context of "scarce resources versus infinite needs" in which health systems are obliged to operate nowadays (more than ever) necessitates that all options should be evaluated both in terms of their cost as well as in terms of their effectiveness, in order to rationalize the allocation of resources and achieve economic efficiency, that is, more and better "health" for the given resources. Placed in the previous line of thought, the present study aimed to review the existing economic evaluation evidence for abatacept, a recently approved agent for the treatment of RA [9]. For that purpose, a systematic literature review was performed in order to depict a thorough overview of the clinicoeconomic value of abatacept in the management of rheumatoid arthritis, compared to existing alternatives.

Cost Effectiveness of the Use of Abatacept following Conventional DMARDs.
The criteria to initiate a biologic agent after failure of conventional DMARDs have been analyzed by several rheumatology bodies (ACR, EULAR, BSR) [6]. Usually, a biologic is started after failure to achieve a DAS-28 of <3.2 with conventional therapies [51]. The results of the systematic review on the use of abatacept after failure of DMARDs included studies mostly performed in American countries and in lesser cases in the European setting, evaluating the use of abatacept or other agents, after failure of methotrexate and taking into account a third party payer (social insurance) perspective.
In all 9 studies that were retrieved, abatacept was a cost-effective strategy against comparators after failure of methotrexate, although in varying degrees. This ranged from  ABA therapy would be economically attractive from a Canadian payer perspective.  The results suggest that in patients with an IR to at least one anti-TNF agent, biologic sequences including ABA appear more efficacious and cost-effective than similar sequences including RTX or cycled anti TNF agents. Compared to MTX, ABA is a cost-effective treatment. Canada 2012 [7] Research article Pubmed CEA ABA versus sequential use of anti-TNF agents As first biologic agent, ABA appears significantly more cost effective compared to the sequential use of anti-TNF agents ( < 0.001). ABA, as second biologic agent, appears significantly more cost effective compared to the sequential use of anti-TNF agents ( < 0.001).
ABA after failure of DMARDs and after failure of one anti-TNF agent.
US 2008 [43] Research article Pubmed Cost analysis relative to QoL ABA versus placebo ABA led to greater reduction in medical expenditure over time in MTX failure ($152 lower) and anti-TNF failure patients ($122 lower) compared with placebo. Likewise, significantly more reduction in likelihood for current and future job loss was achieved with ABA.
Two RCTs. One with patients who have not responded to MTX and the other with patients unresponsive to anti-TNF agents.
US 2008 [44] Research article Pubmed Cost analysis relative to lost productivity ABA versus no treatment The lost productivity cost of RA for a firm of 10,000 was $1.69 million. In the base case analysis 37% of the acquisition cost of abatacept was offset by reductions in the cost of RA-related productivity losses. In some industry groups (Utilities and Finance) and in models that included presenteeism, reductions in lost productivity costs exceeded the abatacept cost.
Productivity losses for a private firm of 10,000 due to employee absences and reduced effectiveness at work because of rheumatoid arthritis.
Canada 2008 [45] Poster  abatacept being a dominant strategy (i.e., more effective and less costly), versus etanercept, adalimumab, infliximab, and tocilizumab ( [14], for the health system of Peru) and tocilizumab alone ( [15], for the health system of Colombia), to borderline cost effective, the latter depending on the acceptability threshold that is used for local reimbursement decisions.
In some cases of the referenced studies, abatacept led to reductions in the total costs of treatment accompanied (in all cases) by gains in Quality Adjusted Life Expectancy, thus leading to favorable economic evaluation results.
Specifically, when compared to infliximab, abatacept proved to be associated with higher treatment costs, especially during the initiation phase [12,14,16,17]. However, all of the previous studies concluded on the cost effectiveness of abatacept compared to infliximab, as these costs were offset by the increased number of QALYs gained. One study [17] estimated the offsetting of costs to take place at a maximum of around 16 months after initiation, claiming that abatacept appears to be an economically attractive, long-term treatment option in RA. Another study comparing the combination of abatacept and methotrexate with conventional DMARD treatment after failure of methotrexate alone also demonstrated that the abatacept strategy was able to offset the higher treatment costs, producing an ICER per QALY gained of £27,657. The indirect comparison to other biologic therapies, such as etanercept and adalimumab [13,16], demonstrated the cost effectiveness of abatacept in active rheumatoid arthritis after inadequate response to conventional DMARDs. One study compared rituximab with abatacept [13]. Regarding RTX, ABA has an incremental cost-effectiveness ratio of S/75 493 per QALY gained.

Cost Effectiveness of the Use of Abatacept following TNF-a
Inhibitors. The majority of retrieved results investigated the cost effectiveness of abatacept use after inadequate response or intolerance to one or more TNF-a antagonists ( = 21). Most studies focus on the comparison between abatacept and rituximab, either "head to head" or with the inclusion of other alternatives. This comparison is in favor of abatacept, since 10 studies report that abatacept is more cost effective or dominant versus rituximab and 4 state the opposite. However, it needs to be noted that 2 of the 4 studies in favor of rituximab are based on the (rather bold) hypothesis of equal efficacy across all agents under evaluation, thus selecting the cheapest one as the dominant option ( [28] for the Brazilian healthcare setting and [29] for the Italian health care setting).
Other findings of interest include the improved cost effectiveness of abatacept compared to methotrexate in Hungary [31] and the UK [38], the association of abatacept with lower overall costs of treatment from a managed care perspective in the US [32], and the favorable cost-effectiveness results of abatacept in comparison to DMARDs in general [23,25].

Abatacept and Its Value as Presented in Other Types of Pharmacoeconomic Analyses.
A series of other economic evaluations of abatacept could not be classified to the previously mentioned major treatment pathways. Nevertheless, they provide some useful insights of the value of abatacept in the treatment of RA patients. Indicatively (a) the cost utility analysis by Wong et al. [41] (comparing methotrexate and abatacept regimens versus methotrexate and rituximab or methotrexate alone), which demonstrated that the combination of methotrexate and abatacept was well within the US cost-effectiveness threshold and (b) the analysis of abatacept versus sequential use of anti-TNF agents for the Canadian health care setting [7] which demonstrated that abatacept was highly cost effective.
Regarding the macroeconomics of the healthcare system, cost-of-illness analyses can be of great use to decision makers. In this field, Wong et al. have demonstrated lower acquisition and administration costs for abatacept [41] although a recent review [40] did not include the agent among those with the lowest cost per responder. Of particular interest is the study of Cole et al. [43] who showed that the use of abatacept can fully offset productivity losses for RA patients, a finding similar to that by Yuan et al. [42].

Limitations.
Being performed in an environment with inherent uncertainty, economic evaluations of health interventions, especially in the rheumatoid arthritis, are characterized by a number of limitations points of discussion that should be addressed when interpreting the outcomes of any analysis. A limitation that applies for most of the economic evaluations presented here lies in the fact that the majority of them adopts a third party payer perspective, thus including only those costs that are relevant for a social security organization and not incorporating the so called "societal costs. " Third party payer perspective analyses are really useful when pricing or, especially, when reimbursement decisions must be supported; this is why they are currently acceptable according to most national and international guidelines for economic evaluations and performed more often than evaluations with a societal perspective. However, this way entails exclusion of a large fraction of indirect (nonhealth) costs, mainly productivity losses and costs of informal care, which in the case of RA is important, given the disabling nature of the disease.
Another issue that should be taken into account when interpreting the results of the present analysis lies within an inherent problem of economic evaluations, that is, the transferability of cost-effectiveness data among varying healthcare settings. Although guidelines on the best practices for economic evaluation and the uniformity of calculations do exist, variability in the organization, administration of care, and, thus, costs, between health systems, renders the caution on the interpretation of the results a necessity.
Other important limitations that specifically apply to this study are the exclusion of research presented in languages other than English and the lack of full paper articles for some of the research presented. Specifically, for the majority of the ISPOR studies, only abstracts were retrievable. So, there remains a (small) degree of uncertainty as to whether all the information available internationally regarding the value of abatacept has been in fact included in the present review and therefore, the findings have to be treated with caution. Future similar reviews will be able to identify fulltext availability of the included research in abstract format. Finally, it has to be noted that one of the included studies [20] is a health technology assessment for the British National Institute for Health and Clinical Excellence (NICE), for which a small number of included studies may have been used in the analysis. The authors decided to include this HTA review, irrespective of this minor duplication, due to the significance of such a study. The review by Liu et al. [40] does not include any studies that have been identified and included in this analysis.
Rheumatoid arthritis is a disease with high prevalence and a significant social burden. Interventions with proven efficacy, such as abatacept, are an important addition to the clinician's armament. However, fiscal realities and scarce resources make cost-effectiveness data essential for decisions about treatment in the micro-(patient) as well as the macro-(healthcare system) level. The majority of data from the international research shows that abatacept is both clinically efficient and highly cost effective in the treatment of moderate to severe rheumatoid arthritis, especially in patients that have demonstrated inadequate response or intolerance to anti-TNF agents or DMARDs.