Rheumatoid arthritis (RA) is a chronic, progressive auto immune disease that causes severe articular damage and functional impotence in the affected joints.1 The prevalence of RA is reported to be 0.5%–1% in developed countries, with a higher prevalence among females (ratio 2:1).1,2 Its prevalence is 0.5% in Spain.3 The therapy of RA aims at early disease control and induction of sustained remission; successful treatment is reflected by sustained quality of live and ability to work.4

Treatment with nonsteroidal anti-inflammatory drugs, conventional disease-modifying antirheumatic drugs and biologic treatment have been assessed in individuals with RA.5 Biologics drugs (BD) approved for use in RA include TNF inhibitors (TNFi), Tocilizumab (Tcz), Rituximab (Rtx), Abatacept (Aba) and Janus kinase inhibitors tofacitinib and baricitinib (Jak).6–8 The TNFi registered for the indication of RA are adalimumab (Ada), certolizumab pegol (Ctz), etanercept (Etn), golimumab (Goli), and infliximab (Ifx)5,7,9; TNFi have improved outcomes for patients who are refractory or intolerant to conventional treatments, inducing long-term remission in some cases.10,11 If TNFi fails, switching to another TNFi or an agent with another mode of action should be considered.6–9 The cost of BD for treating rheumatic diseases has dramatically increased in Spanish hospitals.12 Due to the high cost of BD, it is important to evaluate real costs of use of these agents.13

The main objective was to calculate the annual cost per patient and the cost of each biological treatment of patients with RA in real practice in a tertiary hospital in Spain for eighteen years (2009–2017). Other secondary objectives were to analyze factors related to treatment costs (the prescription of biosimilars instead of original drugs, discounts and negotiated rebates or biologic regimes optimization according to drug and anti-drug antibodies serum levels.14–16

We conducted a retrospective observational study between 2009 and 2017 approved by Ethics Committee of La Paz University hospital in April 2017.

Patients diagnosed of RA who were dispensed BD by the pharmacy department in the study period were included. These dispensations were recorded in a CPOE program (FarmaTools 2.5 Dominion). This software allows pharmacists to register regimes, drugs and unit-drugs used by patient and related them with costs.

In 2009, 2013, and 2017 were treated with BD 320, 270, and 389 patients respectively. Patient's characteristics are shown in Table 1. No statistically significant difference was found between study groups except in DAS28.

Fig. 1 shows the evolution of the results for RA per average-dispensed-patient, annual cost, and annual cost per average-dispensed-patient. “We observed an average-dispensed-patient decrease from 2009 to 2013, and an upward trend from 2014 to 2017 (Fig. 1a)”. The same tendency in terms of annual cost of RA was observed, with a minimum data in 2013 (Fig. 1b)”. However, the annual cost per patient decreased (P<.001 from 2009 to 2013) (Fig. 1c).

Fig. 1.

Data evolution in rheumatoid arthritis from 2009 to 2017. (a) Average dispended patient from 2009 to 2017. (b) Annual cost for rheumatoid arthritis 2009–2017. (c) Annual cost 2017 per average dispensed patient 2009–2017.

(0.27MB).

When evaluating costs according to each drug used a similar trend was observed (Table 2). Annual cost per patient from 2009 to 2017 decreased significantly for Ifx and Eta (P<.001), Ctz and Aba IV (P<.05), and from 2009 to 2013 for Ada (<.001) and Toci IV (P<.05).

Biologics acquired by our center (units) as well as the data for BD dispensed (units) to patients with RA are shown in Table 3. We detected that there was an increase in the number of marketed BD and total savings per drug. In order of appearance, official discounts and negotiated rebates in 2017 were 11.5% for Ada, 15.5% for Eta, 30.9% for Ifx, 17.9% for Goli, 17.8 for Ctz and 17.4% for Rtx, and 19.8% for Aba SC, 15.9% for Tcz, 10.3% for Bari and 7.5% for Tofa, the latest released biological. We checked that the release of a biosimilar infliximab increased the rebates up to 43.1% in 2017, with a gradual increase in bonus units over time while original Ifx rebate reached a 24.6%.

Disease activity decreased annually in patients with optimized regimes when compared with patients without optimized regimes (P<.001) (Table 4a).

As Fig. 2 shows, active patients and percentage using optimized regimes from 2009 to 2017, reached 51.5% and 35.2% of patients with optimization by 2013 and 2017 respectively. The optimized therapies per drug and annually was analyzed (Table 4b).

Fig. 2.

Proportion of active patients with optimized regimes.

(0.14MB).

Costs evolution according to the factors studied (results in 2017 are shown in Table 5). Thus, costs savings related to therapy optimization (830,000€), costs savings by monitoring drug and anti-drug antibody (ADA) serum levels in 2017 represented a 73.87% (613,101€).

Moreover, costs savings by drugs monitoring were 88.08% (322,882€) in 2011, 75.38% (797,906€) in 2013 and a 79.19% (730,810€) in 2015.

Moreover, we found that from 2009 to 2017 the total savings increased (Table 6). The greatest contribution to economic savings was therapy optimization (24.93%). Savings associated with official discounts and negotiated rebates (13.77%) in 2017 (Table 6).

The results obtained are in line with an article that we recently published in patient with Spondyloarthritis.16 Over the study period there was a marked decrease in annual cost per-average-patient diagnosed with RA (incremental difference: −34.9%), however average-dispensed-patient trend increased. Also annual cost per drug decreased during 2009–2017.

In Spain, the Royal Decree Law 4/2010 implementation in June 2010 lead to decreased the prices of all medications by 7.5%19 this fact was associated with cost reduction from 2010 to 2011. Therapy optimization, use of biosimilar TNFi, and official discounts or negotiated rebates that lowered prices in some biologics were other factors associated to the cost reduction for 2011–2017.16

Different published studies have analyzed the economic impact of biological therapies in RA. Gómez-DeRueda et al.,20 in a study conducted from 2013 to 2015, concluded that Ifx (€10,717) had the lowest cost per patient per year under the established practice, followed by Etn (€11,015) and Ada (€11,977). Our study differs in that the costs of Ifx, Etn, and Ada were lower (41.3%, 28.5%, and 41.7%, respectively), compared with the aforementioned study in 2015. Mariatena et al.,21 in a study conducted in 2013, concluded that Ifx (€10,073) had the lowest cost per patient per year under the established practice, followed by Toci (€10,798), Eta (€11,056), and Ada (€11,512). Our study differs in that the costs of Ifx, Toci, Etn, and Ada were lower (46.1%, 22.3%, 33.1% and 37.0%, respectively), compared with the aforementioned study in 2013. Toci, Eta and Ada doses in the first study were optimized empirically and they were reduced a 13.3%, 6.9% and 10.7% for Toci, Eta and Ada, respectively. Ramírez-Herraiz et al.22 concluded that mean doses used were significantly lower with Eta than with Ada and Ifx and they used 81.0%, 93.02% and 135.73% of recommended dose for Eta, Ada and Ifx, respectively. In this study, BD were optimized empirically, controlling for disease activity. Thus patient-year cost in 2011 were €9594, €11,962 and €10,094 for Eta, Ada and Ifx, respectively. Our study differs in that the costs of Ada and Ifx were lower (15.4%, 25.6% respectively), and costs of Eta was higher (3.4%). Finally, Ivorra et al.23 published annual costs per patient and per drug referred to 2013 and our data for the same period showed that these therapies were cheaper 54.3% for Ifx, 43.6% in Ada, 40.1% in Tcz, 37.6% in Eta, 36.0% in Aba IV, 32.1% in Ctz and 19.8% in Goli, that reported in the aforementioned study.

Although in most of these studies the treatments were empirically optimized, our results showed marked differences in RA. This could be explained by the fact that the monitoring of drugs (Etn, Ifx, Ada, Toci) helps the clinician to optimize treatments earlier, with greater safety, and lower doses and wider dosing intervals regarding empirical optimization.

According to the EULAR recommendations,7 tapering of a biological drug can be considered in patients that achieve persistent remission. REDOSER project established criteria for reducing doses of biological therapies for RA, both extending the dosing interval and/or reducing the dose. In addition, serum drug levels and ADAs in serum, when available, can help to clinicians to optimize biological therapy and the clinical monitoring.24

We observed that patients with optimized regimes increased from 12 (4.6%) to 116 (35.2%) patients (2009–2017). Monitoring of Ifx, Etn and Ada using serum levels is used by clinicians in clinical practice in our center from 201114–16; serum levels for monitoring Toci began in 2014 and Goli and Rtx began in 2015, and were available in usual practice in 2017, and for their optimization, rheumatologists have stablished clinical protocols.

Our results show that optimization of biological therapies leads to a marked costs reduction. Moreover, other authors proved that dosing regimen optimization of biologicals does not mean an increase in disease activity parameters, no differences with patients under full dose regimens were found.16,25

In parallel with the beginning of the optimization of treatments, costs decreased. Ada and Ifx annual costs decreased mainly in 2011 and 2012 and Etn in 2013.

When analyzing savings related to therapy optimizations, we detected that the main factor contributing to these savings was optimization by drug serum levels monitorization that were around 80%, respect saving related to empirically optimizations, for analyzed years as we described in result section.

The majority of drugs that contributed to cost savings by optimization were Ada, Etn, Ifx and Tcz group over Goli and Ctz, coinciding with the percentage of optimized regimes for these drugs, in which Goli and Ctz were optimized in a lower percentage than first group probably they joined later.

The presence in the market of many drugs for a pathology produces an economic competition.26 However, bonus units and discounts can then reduce the expenditure on medicines. In our hospital, we have observed that introduction of Goli, Ctz, and Aba or Tcz SC was accompanied by significant invoice discounts of between 15.9% and 37% in different years; and bonus units gradually rose during the study period.

It is known that when a biosimilar is released there is an increased access and a lower health cost burden. According to the law in Spain, when a biosimilar is marketed the original have to decrease its price to the same level of the biosimilar.27,28 Over the study period the European Medicines Agency approved biosimilars of Ifx and Etn in 2013 and 2015 respectively, which led to an increase in discounts for Ifx and Eta.

Original Ifx rebate in our study (24.6%) are in line with the reduction in the price of infliximab published articles,29 however biosimilar Ifx retabe obtained (43.1%) exceed published data.30

Taking together all factors influencing annual RA cost per patient we observed that when the annual cost decreased slightly, increased the number of treated patients and the total saved costs. Our results show that the greatest saving contributions were biological therapy were optimizations, followed by official discounts and negotiated rebates.

Taking into account our results future strategies leading toward the implementation of therapeutic drug monitoring based on scientific evidence31 should be promoted in order to reduce costs and maintaining disease control at the same time.

There were several study limitations. Farmatools does not provide definitive reports of economic and clinical results in order to make a posterior statistical analysis. We have to do a data treatment before use them. Moreover, the annual theoretical cost of Ifx could be overestimated because we considered an estimated average weight of 70kg for all patients treated. Moreover, the saved by optimized regimes could be overestimated because units not dispensed by the possible lack of adherence to treatment are not included. Finally, costs from 2009 to 2017 were not adjusted.

The most important strength of our study is the very long analysis period and the large sample size, which allowed us to analyze and to quantify influential factors in decreasing cost per patient and to prove that optimization was the strategy that most influenced this decline.

Also, annual cost per average-dispensed-patient allows us to compare our data with other hospitals in Spain.

Our study proves that the greatest contribution to economic savings in biological therapy in rheumatoid arthritis was biological therapy optimization by monitoring drug and ADA serum levels when comparing with official discounts, negotiated rebates.

No financial support.

No conflicts of interest have been declared.