Maxime Verhoeven

226 Chapter 13 In the current era of scarce resources in healthcare, also the cost effectiveness of a treatment strategy is important. 12 Overall, in treat-to-target treatment strategies aiming at remission, initiation of bDMARDs, specifically TCZ, are not cost-effective compared to initiation of a single csDMARD ( chapter 7 ). Based on our result in chapter 3 , we hypothesize that initiating as first therapy in early RA a strategy with a csDMARD combined with GC is cost-effective or even cost-saving compared to a more intensive strategy with a bDMARD, due to lower drug costs and similar effectiveness in treat- to-target treatment strategies. This is in line with a previous cost-utility analysis study showing that initial combination therapy with prednisone is preferred over initial combination therapy with a bDMARD, both in a treat-to-target strategy aiming at remission in newly diagnosed RA patients over 2 years. 13 Surprisingly, in our own cost effectiveness analyses, we observed that productivity loss costs were saved in the groups initiating TCZ, particularly in the group starting with TCZ-monotherapy ( chapter 7 ), and less so in the initial TCZ+MTX group, compared to the MTX-monotherapy group, suggesting that MTX toxicity may be partially responsible for the productivity loss. 14 However, this does not offset the cost of therapy. We expect that initiation of bDMARDs could become cost-effective as initial treatment strategy in early RA under certain circumstances. We hypothesize that initial treatment with bDMARDs combined with active tapering might enable drug free remission, an even sharper target than remission only. If this target is indeed achievable in a considerable proportion of patients over time, we expect that this increases quality adjusted life years, not only by preventing irreversible joint damage, but also adverse effects of treatment (specifically of MTX), while also saving medication costs. Further, decrease of bDMARD prices in the future driven by introduction of biosimilars will add to this; after introduction of biosimilars on the Dutch market in 2015, the mean purchase price of etanercept for Dutch hospitals decreased almost 60%. 6,15 In addition, tailoring initial bDMARD treatment strategy to specific subpopulations may be promising to further improve cost effectiveness. For example, focussing on patients who are anticipated to have an insufficient response to the anchor treatment of MTX, 1 and for those implementing prediction models in clinical practice make sense. 16 In chapter 7 we observed that, in a specific early RA subgroup (i.e., baseline DAS28 >5.1 based on prediction model as described in chapter 4 ), TCZ-based initial treatment might become more cost-effective, although the strategy did not reach generally accepted levels of cost effectiveness in the Netherlands, 17,18 probably due to the lack of active tapering in the follow-up period. Notwithstanding, analyses simulating specific stratified treatment strategies, using existing or future prediction models, should be performed more frequently and continuously to develop effective as well as cost-effective treatment strategies, specifically in light of achieving drug free remission.