J Thromb Thrombolysis DOI 10.1007/s11239-014-1104-3
Edoxaban versus warfarin for stroke prevention in non-valvular atrial fibrillation: a cost-effectiveness analysis Carla Rognoni • Monia Marchetti • Silvana Quaglini Nicola Lucio Liberato
•
Ó Springer Science+Business Media New York 2014
Abstract Edoxaban, an oral direct factor Xa inhibitor, has been found non-inferior to warfarin for preventing stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF), with a lower rate of intracranial bleeding. The aim of our investigation was to assess the cost-effectiveness of edoxaban versus warfarin from the perspective of the Italian health-care system. A Markov decision model was used to evaluate lifetime cost and quality-adjusted life expectancy of NVAF patients treated with warfarin or edoxaban. Transition probabilities were obtained from the ENGAGE AF-TIMI 48 trial, cost estimates were based on Italian prices and tariffs, utilities were obtained from the literature. One-way and second-order sensitivity analyses were performed. In the base case, lifetime costs were €18,658 for edoxaban and €14,060 for warfarin. Discounted quality-adjusted survival was 9.022 years for edoxaban and 8.425 years for warfarin, leading to an incremental cost-utility ratio of €7,713 per quality-adjusted life year (QALY) gained. Results were sensitive to time horizon, time in therapeutic range of warfarin and to the relative impact of warfarin versus
C. Rognoni Centre for Research on Health and Social Care Management (CERGAS), Bocconi University, Milan, Italy M. Marchetti Hematology Unit, Cardinal Massaia Hospital, Asti, Italy S. Quaglini Department of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, Italy N. L. Liberato (&) Department of Medicine, Azienda Ospedaliera della Provincia di Pavia, Casorate Primo, PV, Italy e-mail: [email protected]
edoxaban therapy onto quality of life. Probabilistic sensitivity analysis showed edoxaban to be cost-effective versus warfarin in 92.3 % of the simulations at a willingnessto-pay threshold of €25,000 per QALY. In conclusion, edoxaban proved to be a cost-effective alternative to warfarin in patients with moderate-to-high-risk NVAF. Keywords Pharmacoeconomic analysis Costeffectiveness analysis Atrial fibrillation Edoxaban Warfarin
Introduction Direct oral anticoagulants (DOAs) are fixed-dose drugs proved to be as effective as warfarin in preventing stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF) but also less prone to induce severe bleedings and drug–drug or drug–food interactions. Scientific societies have included DOAs into their guidelines [1, 2], drug agencies have approved their use and they are actually reimbursed in most industrialized countries for this indication. However, their cost-effectiveness versus warfarin varied across a wide range being the estimate sensitive to probabilities of ischaemic stroke (IS) and intracranial hemorrhage (ICH), patients’ CHADS2 score, time in therapeutic range (TTR) with warfarin and cost of drug [3, 4]. Most countries assigned the same price to different DOAs, however, the relative incremental costeffectiveness of DOAs one to another has been poorly assessed [5, 6] and the economic impact of DOAs onto health-care budget is still to be properly assessed in the different countries. Edoxaban is a novel oral direct factor Xa inhibitor proving fine pharmacokinetic and bioavailability. The
123
C. Rognoni et al.
Fig. 1 Simplified decision model representation. M represents a Markov process with 7 health states. The patients enter the Markov process in the NVAF-only state, where they remain unless one of six
events occurs, namely IS, ICH, minor extracranial hemorrage, major extracranial hemorrage and death. NVAF non-valvular atrial fibrillation; IS ischemic stroke; ICH intracranial hemorrhage
ENGAGE AF-TIMI 48 double-blind, double-dummy trial recently reported on 21,105 NVAF patients treated with either edoxaban or warfarin for a median of 2.8 years [7]. The present study aimed at assessing the cost-effectiveness of high dose edoxaban, as compared with warfarin, in the perspective of the Italian health-care system.
Transition probabilities were derived from annualized rates of events in the ENGAGE AF-TIMI 48 clinical trial [7] (Table 1). Patients with IS and ICH have been distributed into temporary, mild, moderate-severe and fatal according to the literature (9.1, 42.5, 40.2 and 8.2 % for IS and 7.8, 8.8, 31.8 and 51.6 % for ICH, respectively) [15, 16]. To take into account the age-related increase of cardiovascular risk, the rates of IS and ICH were assumed to increase by a factor of 1.4 [17] and 1.97 [18], respectively, per decade of life (multiplicative adjustment). Mortality rates were further adjusted for age according to Italian mortality tables. After an IS or ICH, the risk of further events was assumed to increase 2.6-fold [17, 19]. Quality-adjusted life expectancy is estimated by multiplying survival in health states by weight factors (utility coefficients) ranging from 0 (death) to 1 (perfect quality of life). In our model, the utility coefficients were obtained from a sub-analysis of the RE-LY study, assuming an impact of edoxaban on patients’ quality of life comparable to the dabigatran one [9] (Table 1). Temporary decrements in quality of life due to minor extracranial hemorrhage, minor IS/ICH and major extracranial hemorrhage have been accounted as disutilities. We considered major IS and ICH either fatal or with a permanent decrement in utility due to neurologic sequelae.
Methods A previously described [8] Markov model has been adapted to compare costs and quality-adjusted life expectancy associated with adjusted-dose warfarin and high dose (60 mg daily) edoxaban strategies in patients with moderate-to-high-risk (CHADS2 score higher than 1) NVAF. The health states considered in the Markov model are: NVAF with no other morbidities (NVAF-only); temporary, mild, moderate-severe or fatal IS; temporary, mild, moderatesevere or fatal ICH; minor and major extracranial bleedings; and death (Fig. 1). We assumed that a mild/moderatesevere ICH or major extracranial bleeding from the NVAFonly state on anticoagulant therapy resulted in 1-month discontinuation of anticoagulation. A 3-month Markov cycle length and a lifetime horizon were chosen for baseline analysis.
123
Edoxaban vs warfarin for stroke prevention in NVAF Table 1 Model parameters and range variations for sensitivity analyses Base-case value
Range
Table 1 continued
Reference Major extracranial bleeding
Transition probabilities between health states (quarterly) NVAF-only to IS or systemic embolism Warfarin Edoxaban
0.0048 RR = 0.88
0.75–1.02
[7] [7]
NVAF-only to ICH Warfarin
0.0018
Edoxaban
RR = 0.47
[7] 0.34–0.63
0.0054
Edoxaban
RR = 0.80
[7] 0.70–0.90
[7]
NVAF-only to minor extracranial bleedings Warfarin
0.0286
Edoxaban
RR = 0.86
[7] 0.82–0.91
Range
Reference
4,032
3,277–4,854
DRG 174
Vascular death
5,500
4,466–6,626
[14]
Warfarin therapyc
3.89
3.1–4.7
NHS price list
Edoxaban therapyc
198.74
160–239
NHS price list
[7]
NVAF-only to major extracranial bleedings Warfarin
Base-case value
[7]
Each if not otherwise specified RR relative risk versus warfarin a
1 month duration
b
5 days duration
c
every 3 months
NVAF-only to myocardial infarction Warfarin
0.0018
Edoxaban
RR = 0.94
0.75–1.19
[7] [7]
0.00982 RR = 0.92
0.82–1.01
[7] [7]
NVAF on warfarin
0.777
0.747–0.807
[9]
NVAF on edoxaban
0.783
0.750–0.816
[9]
Permanent moderate/ severe disability due to IS or ICH
0.51
0.32–0.7
[10]
Permanent mild disability due to IS or ICH
0.85
0.51–0.99
[10]
Temporary disability due to minor IS or ICHa
-0.1385
0.118–0.160
[11]
Major extracranial hemorrhagea
-0.1385
0.118–0.160
[11]
Minor extracranial hemorrhageb
-0.06
0.04–0.08
[12]
Follow-up of every NVAF patientsc
7.38
6–9
NHS price list
Follow-up of warfarin patientsc
87.22
70–106
NHS price list
The study was conducted from the perspective of the Italian National Health System (NHS), and only direct costs were considered (Table 1). All costs were expressed in Euros (2014 value). Being edoxaban still not reimbursed for NVAF indication in Italy, the daily cost of dabigatran (€2.18) was used as a proxy. The cost of edoxaban therapy included, besides the drug cost, one visit with renal function determination (creatinine) every year. Monitoring of patients with NVAF without complications was assumed to include one cardiologic visit and one electrocardiogram every year. Future costs and life-years were discounted at 3.5 % per year [20]. One-way sensitivity analysis of all variables included in the decision model over their reasonable ranges was performed (Table 1). Confidence intervals for hazard ratios from the ENGAGE AF-TIMI 48 trial were used to derive the ranges for clinical events. The impact of TTR was assessed by varying the rate of stroke and systemic embolism and of major bleeding on warfarin along with the hazard ratios of edoxaban versus warfarin for the same endpoints according to the trial. Similarly, subgroup analysis according to CHADS2 was performed. In addition, second-order Monte-Carlo analysis was performed by running 10,000 simulations in order to estimate the probabilities of edoxaban of being cost effective at different thresholds in different clinical scenarios.
Follow-up of edoxaban patientsc
6.05
4.9–7.3
NHS price list
Results
IS (acute and postacute phases) ICH (acute and postacute phases)
10,802
8,800–13,017
[13]
17,383
14,166–20,870
[13]
NVAF-only to death Warfarin Edoxaban QALY weights
Disutilities
Costs (€)
The number of events related to IS, ICH, major extracranial bleedings, cardiovascular deaths and total deaths of a simulation at 2.8 years (median follow-up of the ENGAGE
123
C. Rognoni et al. Table 2 Baseline results (95 % confidence intervals)
Total cost (€)
Life-years
QALYs
ICUR (€/QALY)
Undiscounted Warfarin
15,345
12.295
9.047
–
Edoxaban
20,337
13.036
9.722
–
Incremental
4,992
0.741
0.675
7,396
(€2,280–7,444)
(0.255–1.139)
(0.044–1.267)
(€2,169–€37,192)
Discounted Warfarin
14,060
11.424
8.425
–
Edoxaban
18,658
12.070
9.022
–
Incremental
4,598
0.646
0.596
7,713
(€2,140–6,866)
(0.219–0.998)
(0.020–1.140)
(€1,755–€40,642)
Fig. 2 Tornado diagram showing univariate sensitivity analyses on ICUR value over plausible ranges for variables. Upper and lower limits of variables’ values referring to the ICUR extremes are indicated next to the bars
AF-TIMI 48 trial) matched the results of the trial, no value deviating by more than 0.5 %. In the base case, the use of warfarin and edoxaban for a lifetime horizon projected an average quality-adjusted survival of 8.425 and 9.022 QALYs, respectively. Being the average global costs for the two strategies €14,060 and €18,658, respectively, the incremental cost-utility ratio (ICUR) was €7,713/QALY gained (Table 2). One-way sensitivity (Fig. 2) analysis showed the baseline result being significantly influenced by the horizon of the analysis, with an ICUR above €25,000/QALY for an horizon below 51 months. We analyzed the impact of quality of life’s change by considering the difference between the utility coefficients of edoxaban and warfarin NVAF-only health states, which was 0.006 in the base case: ICUR increased above €25,000/QALY when such difference decreased below -0.030, i.e. if edoxaban had a significant detrimental effect onto quality of life with respect to warfarin. When considering the thrombotic risk, ICUR was €9,438/QALY for patients with CHADS2 score lower or equal to 3 and €5,363/QALY for CHADS2 score higher than 3. Finally, ICUR was €10,040/QALY for
123
patients managed by centers showing median TTR values above 66.4 %, while it was €6,479/QALY for patients managed in centers with lower TTR. In the probabilistic sensitivity analysis, edoxaban resulted cost-effective versus warfarin in 92.3 % of the simulations using a willingness-to-pay threshold of €25,000 per QALY and in 96 % of the simulations using a willingness-to-pay threshold of €50,000 per QALY.
Discussion In this study we adapted a previously published decision model to the clinical data reported by the ENGAGE AFTIMI 48 trial, demonstrating a favorable incremental costutility of edoxaban versus warfarin. One can wonder whether the cost-effectiveness profile of all DOAs is similar in the NVAF clinical setting, provided that similar market prices are assigned to the different drugs. Indeed, sensitivity analysis showed that quality of life of patients receiving DOAs is a key parameter influencing the results of the present and prior published cost-utility analyses;
Edoxaban vs warfarin for stroke prevention in NVAF
similarly, the results are also sensitive to some efficacy endpoints, which can vary among different DOAs [3, 4]. The present study specifically addresses the patient subgroup enrolled into the ENGAGE AF-TIMI 48 trial, that is NVAF patients with a CHADS2 score of 2 or higher, and, unfortunately, no comparison could be made versus other DOAs, since CHADS2-based subgroup analyses are not available from the ENGAGE AF-TIMI 48 trial, yet. Despite no direct comparison could be held among different DOAs, based on available data, edoxaban costeffectiveness compares positively with the other DOAs in the perspective of the Italian Health Care System. We are aware that applicability of cost-effectiveness results outside the original setting can be misleading. However, based on previously published economic evaluations of DOAs, differences among models may be even larger than differences among countries. Therefore, we deem that the results reported for the Italian health-care setting can be probably extended to other European countries. Economic data are of great interest for regulatory decisions such as approval of new drugs, formulary inclusion, insurance coverage, reimbursement rate [21]. Health economic evaluations are basic tools for stakeholders: constrained budgets impose them an increasing reliance on formal cost-effectiveness and cost-utility analyses, which have become the key issue for perception of the drug’s value by the regulator and among most important factors affecting drug prices. Value-based pricing has been recently introduced by some European countries: it is a method assessing whether the additional benefits of an innovative technology are worth the additional proposed costs [22, 23]. Therefore, economic models and simulations of budget impact analyses are recommended to be performed early in the drug-development process [24]. Based on the data reported by the present study, important health-care policy implications can be expected. First of all, market price of DOAs is expected to decline due to drug competition. Second, cost-containment policies can be proposed to companies, i.e. risk sharing, and to Italian regions, i.e. volume ceiling, in order to prevent massive shifting of patients taking warfarin with a good compliance and an acceptable TTR, while international guidelines do not recommend DOAs preferably to warfarin, yet. Third, cost-effectiveness of DOAs should be reassessed by health-care authorities as real-world safety data are gathered. Fourth, health-care authorities might impose master decision models with similar structure and assumptions in order to preserve transparency and comparability among new drugs belonging to the same class, as DOAs. Finally, we deem that a specific international effort should be devoted to quality of life of NVAF patients receiving oral anticoagulants in order to make cost-effectiveness estimations real-life tools for clinical governance.
Acknowledgments No sources of funding were used to assist in the conduct of this study Conflict of interest of interests.
The authors declare that they have no conflict
References 1. Anderson JL, Halperin JL, Albert NM, Bozkurt B, Brindis RG, Curtis LH, DeMets D, Guyton RA, Hochman JS, Kovacs RJ, Ohman EM, Pressler SJ, Sellke FW, Shen W-K (2013) Management of patients with atrial fibrillation (compilation of 2006 ACCF/AHA/ESC and 2011 ACCF/AHA/HRS guideline recommendations): a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines. Circulation 127(18):1916–1926 2. Camm AJ, Lip GY, De Caterina R, Savelieva I, Atar D, Hohnloser SH, Hindricks G, Kirchhof P, Kirchhof P (2012) 2012 focused update of the ESC guidelines for the management of atrial fibrillation: an update of the 2010 ESC guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association. Eur Heart J 33(21):2719–2747 3. Kansal AR, Zheng Y, Pokora T, Sorensen SV (2013) Costeffectiveness of new oral anticoagulants in the prevention of stroke in patients with atrial fibrillation. Best Pract Res Clin Haematol 26(2):225–237 4. Limone BL, Baker WL, Kluger J, Coleman CI (2013) Novel anticoagulants for stroke prevention in atrial fibrillation: a systematic review of cost-effectiveness models. PLoS One 8(4):1–15 5. Kansal AR, Sharma M, Bradley-Kennedy C, Clemens A, Monz BU, Peng S, Roskell N, Sorensen SV (2012) Dabigatran versus rivaroxaban for the prevention of stroke and systemic embolism in atrial fibrillation in Canada. Comparative efficacy and costeffectiveness. Thromb Haemost 108(4):672–682 6. Lip GY, Kongnakorn T, Phatak H, Kuznik A, Lanitis T, Liu LZ, Iloeje U, Hernandez L, Dorian P (2014) Cost-effectiveness of apixaban versus other new oral anticoagulants for stroke prevention in atrial fibrillation. Clin Ther 36(2):192-210.e20 7. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, Waldo AL, Ezekowitz MD, Weitz JI, Sˇpinar J, Ruzyllo W, Ruda M, Koretsune Y, Betcher J, Shi M, Grip LT, Patel SP, Patel I, Hanyok JJ, Mercuri M, Antman EM, ENGAGE AF-TIMI 48 Investigators (2013) Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med 369(22):2093–2104 8. Rognoni C, Marchetti M, Quaglini S, Liberato NL (2014) Apixaban, dabigatran, and rivaroxaban versus warfarin for stroke prevention in non-valvular atrial fibrillation: a cost-effectiveness analysis. Clin Drug Investig 34(1):9–17 9. Monz BU, Connolly SJ, Korhonen M, Noack H, Pooley J (2013) Assessing the impact of dabigatran and warfarin on health-related quality of life: results from an RE-LY sub-study. Int J Cardiol 168(3):2540–2547 10. Gage BF, Cardinalli AB, Owens DK (1998) Cost-effectiveness of preference-based antithrombotic therapy for patients with nonvalvular atrial fibrillation. Stroke 29(6):1083–1091 11. Sullivan PW, Arant TW, Ellis SL, Ulrich H (2006) The cost effectiveness of anticoagulation management services for patients with atrial fibrillation and at high risk of stroke in the US. Pharmacoeconomics 24(10):1021–1033 12. Meckley LM, Gudgeon JM, Anderson JL, Williams MS, Veenstra DL (2010) A policy model to evaluate the benefits, risks and costs of warfarin pharmacogenomic testing. Pharmacoeconomics 28(1):61–74
123
C. Rognoni et al. 13. Fattore G, Torbica A, Susi A, Giovanni A, Benelli G, Gozzo M, Toso V (2012) The social and economic burden of stroke survivors in Italy: a prospective, incident-based, multi-centre cost of illness study. BMC Neurol 12:137–147 14. Fedele F, D’Ambrosi A, Bruno N, Caira C, Brasolin B, Mancone M (2011) Cost-effectiveness of levosimendan in patients with acute heart failure. J Cardiovasc Pharmacol 58(4):363–366 15. O’Brien CL, Gage BF (2005) Costs and effectiveness of ximelagatran for stroke prophylaxis in chronic atrial fibrillation. JAMA 293:699–706 16. Rosand J, Eckman MH, Knudsen KA, Singer DE, Greenberg SM (2004) The effect of warfarin and intensity of anticoagulation on outcome of intracerebral hemorrhage. Arch Intern Med 164(8):880–884 17. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med 154(13):1449–1457 18. Ariesen MJ, Claus SP, Rinkel GJ, Algra A (2003) Risk factors for intracerebral hemorrhage in the general population: a systematic review. Stroke 34(8):2060–2065 19. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ (2001) Validation of clinical classification schemes
123
20.
21.
22.
23.
24.
for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA 285(22):2864–2870 National Institute for Health and Clinical Excellence (NICE). Guide to the methods of technology appraisal. 2008. Available from: http://www.nice.org.uk/media/b52/a7/tamethodsguideupda tedjune2008.pdf Accessed January 10, 2014 Frick K, Neissen L, Bridges J, Walker D, Wilson RF, Bass EB (2012) Usefulness of economic evaluation data in systematic reviews of evidence. Agency for healthcare research and quality (US) Report no. 12(13) Danzon P, Towse A, Mestre-Ferrandiz J (2013) Value based differential pricing: efficient prices for drugs in a global context. Health Econ. doi:10.1002/hec.3021 Hall PS, McCabe C, Brown JM, Cameron DA (2010) Health economics in drug development: efficient research to inform healthcare funding decisions. Eur J Cancer 46(15):2674–2680 van Nooten F, Holmstrom S, Green J, Wiklund I, Odeyemi IA, Wilcox TK (2012) Health economics and outcomes research within drug development: challenges and opportunities for reimbursement and market access within biopharma research. Drug Discov Today 17(11–12):615–622
Edoxaban versus warfarin for stroke prevention in non-valvular atrial fibrillation: a cost-effectiveness analysis Carla Rognoni • Monia Marchetti • Silvana Quaglini Nicola Lucio Liberato
•
Ó Springer Science+Business Media New York 2014
Abstract Edoxaban, an oral direct factor Xa inhibitor, has been found non-inferior to warfarin for preventing stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF), with a lower rate of intracranial bleeding. The aim of our investigation was to assess the cost-effectiveness of edoxaban versus warfarin from the perspective of the Italian health-care system. A Markov decision model was used to evaluate lifetime cost and quality-adjusted life expectancy of NVAF patients treated with warfarin or edoxaban. Transition probabilities were obtained from the ENGAGE AF-TIMI 48 trial, cost estimates were based on Italian prices and tariffs, utilities were obtained from the literature. One-way and second-order sensitivity analyses were performed. In the base case, lifetime costs were €18,658 for edoxaban and €14,060 for warfarin. Discounted quality-adjusted survival was 9.022 years for edoxaban and 8.425 years for warfarin, leading to an incremental cost-utility ratio of €7,713 per quality-adjusted life year (QALY) gained. Results were sensitive to time horizon, time in therapeutic range of warfarin and to the relative impact of warfarin versus
C. Rognoni Centre for Research on Health and Social Care Management (CERGAS), Bocconi University, Milan, Italy M. Marchetti Hematology Unit, Cardinal Massaia Hospital, Asti, Italy S. Quaglini Department of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, Italy N. L. Liberato (&) Department of Medicine, Azienda Ospedaliera della Provincia di Pavia, Casorate Primo, PV, Italy e-mail: [email protected]
edoxaban therapy onto quality of life. Probabilistic sensitivity analysis showed edoxaban to be cost-effective versus warfarin in 92.3 % of the simulations at a willingnessto-pay threshold of €25,000 per QALY. In conclusion, edoxaban proved to be a cost-effective alternative to warfarin in patients with moderate-to-high-risk NVAF. Keywords Pharmacoeconomic analysis Costeffectiveness analysis Atrial fibrillation Edoxaban Warfarin
Introduction Direct oral anticoagulants (DOAs) are fixed-dose drugs proved to be as effective as warfarin in preventing stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF) but also less prone to induce severe bleedings and drug–drug or drug–food interactions. Scientific societies have included DOAs into their guidelines [1, 2], drug agencies have approved their use and they are actually reimbursed in most industrialized countries for this indication. However, their cost-effectiveness versus warfarin varied across a wide range being the estimate sensitive to probabilities of ischaemic stroke (IS) and intracranial hemorrhage (ICH), patients’ CHADS2 score, time in therapeutic range (TTR) with warfarin and cost of drug [3, 4]. Most countries assigned the same price to different DOAs, however, the relative incremental costeffectiveness of DOAs one to another has been poorly assessed [5, 6] and the economic impact of DOAs onto health-care budget is still to be properly assessed in the different countries. Edoxaban is a novel oral direct factor Xa inhibitor proving fine pharmacokinetic and bioavailability. The
123
C. Rognoni et al.
Fig. 1 Simplified decision model representation. M represents a Markov process with 7 health states. The patients enter the Markov process in the NVAF-only state, where they remain unless one of six
events occurs, namely IS, ICH, minor extracranial hemorrage, major extracranial hemorrage and death. NVAF non-valvular atrial fibrillation; IS ischemic stroke; ICH intracranial hemorrhage
ENGAGE AF-TIMI 48 double-blind, double-dummy trial recently reported on 21,105 NVAF patients treated with either edoxaban or warfarin for a median of 2.8 years [7]. The present study aimed at assessing the cost-effectiveness of high dose edoxaban, as compared with warfarin, in the perspective of the Italian health-care system.
Transition probabilities were derived from annualized rates of events in the ENGAGE AF-TIMI 48 clinical trial [7] (Table 1). Patients with IS and ICH have been distributed into temporary, mild, moderate-severe and fatal according to the literature (9.1, 42.5, 40.2 and 8.2 % for IS and 7.8, 8.8, 31.8 and 51.6 % for ICH, respectively) [15, 16]. To take into account the age-related increase of cardiovascular risk, the rates of IS and ICH were assumed to increase by a factor of 1.4 [17] and 1.97 [18], respectively, per decade of life (multiplicative adjustment). Mortality rates were further adjusted for age according to Italian mortality tables. After an IS or ICH, the risk of further events was assumed to increase 2.6-fold [17, 19]. Quality-adjusted life expectancy is estimated by multiplying survival in health states by weight factors (utility coefficients) ranging from 0 (death) to 1 (perfect quality of life). In our model, the utility coefficients were obtained from a sub-analysis of the RE-LY study, assuming an impact of edoxaban on patients’ quality of life comparable to the dabigatran one [9] (Table 1). Temporary decrements in quality of life due to minor extracranial hemorrhage, minor IS/ICH and major extracranial hemorrhage have been accounted as disutilities. We considered major IS and ICH either fatal or with a permanent decrement in utility due to neurologic sequelae.
Methods A previously described [8] Markov model has been adapted to compare costs and quality-adjusted life expectancy associated with adjusted-dose warfarin and high dose (60 mg daily) edoxaban strategies in patients with moderate-to-high-risk (CHADS2 score higher than 1) NVAF. The health states considered in the Markov model are: NVAF with no other morbidities (NVAF-only); temporary, mild, moderate-severe or fatal IS; temporary, mild, moderatesevere or fatal ICH; minor and major extracranial bleedings; and death (Fig. 1). We assumed that a mild/moderatesevere ICH or major extracranial bleeding from the NVAFonly state on anticoagulant therapy resulted in 1-month discontinuation of anticoagulation. A 3-month Markov cycle length and a lifetime horizon were chosen for baseline analysis.
123
Edoxaban vs warfarin for stroke prevention in NVAF Table 1 Model parameters and range variations for sensitivity analyses Base-case value
Range
Table 1 continued
Reference Major extracranial bleeding
Transition probabilities between health states (quarterly) NVAF-only to IS or systemic embolism Warfarin Edoxaban
0.0048 RR = 0.88
0.75–1.02
[7] [7]
NVAF-only to ICH Warfarin
0.0018
Edoxaban
RR = 0.47
[7] 0.34–0.63
0.0054
Edoxaban
RR = 0.80
[7] 0.70–0.90
[7]
NVAF-only to minor extracranial bleedings Warfarin
0.0286
Edoxaban
RR = 0.86
[7] 0.82–0.91
Range
Reference
4,032
3,277–4,854
DRG 174
Vascular death
5,500
4,466–6,626
[14]
Warfarin therapyc
3.89
3.1–4.7
NHS price list
Edoxaban therapyc
198.74
160–239
NHS price list
[7]
NVAF-only to major extracranial bleedings Warfarin
Base-case value
[7]
Each if not otherwise specified RR relative risk versus warfarin a
1 month duration
b
5 days duration
c
every 3 months
NVAF-only to myocardial infarction Warfarin
0.0018
Edoxaban
RR = 0.94
0.75–1.19
[7] [7]
0.00982 RR = 0.92
0.82–1.01
[7] [7]
NVAF on warfarin
0.777
0.747–0.807
[9]
NVAF on edoxaban
0.783
0.750–0.816
[9]
Permanent moderate/ severe disability due to IS or ICH
0.51
0.32–0.7
[10]
Permanent mild disability due to IS or ICH
0.85
0.51–0.99
[10]
Temporary disability due to minor IS or ICHa
-0.1385
0.118–0.160
[11]
Major extracranial hemorrhagea
-0.1385
0.118–0.160
[11]
Minor extracranial hemorrhageb
-0.06
0.04–0.08
[12]
Follow-up of every NVAF patientsc
7.38
6–9
NHS price list
Follow-up of warfarin patientsc
87.22
70–106
NHS price list
The study was conducted from the perspective of the Italian National Health System (NHS), and only direct costs were considered (Table 1). All costs were expressed in Euros (2014 value). Being edoxaban still not reimbursed for NVAF indication in Italy, the daily cost of dabigatran (€2.18) was used as a proxy. The cost of edoxaban therapy included, besides the drug cost, one visit with renal function determination (creatinine) every year. Monitoring of patients with NVAF without complications was assumed to include one cardiologic visit and one electrocardiogram every year. Future costs and life-years were discounted at 3.5 % per year [20]. One-way sensitivity analysis of all variables included in the decision model over their reasonable ranges was performed (Table 1). Confidence intervals for hazard ratios from the ENGAGE AF-TIMI 48 trial were used to derive the ranges for clinical events. The impact of TTR was assessed by varying the rate of stroke and systemic embolism and of major bleeding on warfarin along with the hazard ratios of edoxaban versus warfarin for the same endpoints according to the trial. Similarly, subgroup analysis according to CHADS2 was performed. In addition, second-order Monte-Carlo analysis was performed by running 10,000 simulations in order to estimate the probabilities of edoxaban of being cost effective at different thresholds in different clinical scenarios.
Follow-up of edoxaban patientsc
6.05
4.9–7.3
NHS price list
Results
IS (acute and postacute phases) ICH (acute and postacute phases)
10,802
8,800–13,017
[13]
17,383
14,166–20,870
[13]
NVAF-only to death Warfarin Edoxaban QALY weights
Disutilities
Costs (€)
The number of events related to IS, ICH, major extracranial bleedings, cardiovascular deaths and total deaths of a simulation at 2.8 years (median follow-up of the ENGAGE
123
C. Rognoni et al. Table 2 Baseline results (95 % confidence intervals)
Total cost (€)
Life-years
QALYs
ICUR (€/QALY)
Undiscounted Warfarin
15,345
12.295
9.047
–
Edoxaban
20,337
13.036
9.722
–
Incremental
4,992
0.741
0.675
7,396
(€2,280–7,444)
(0.255–1.139)
(0.044–1.267)
(€2,169–€37,192)
Discounted Warfarin
14,060
11.424
8.425
–
Edoxaban
18,658
12.070
9.022
–
Incremental
4,598
0.646
0.596
7,713
(€2,140–6,866)
(0.219–0.998)
(0.020–1.140)
(€1,755–€40,642)
Fig. 2 Tornado diagram showing univariate sensitivity analyses on ICUR value over plausible ranges for variables. Upper and lower limits of variables’ values referring to the ICUR extremes are indicated next to the bars
AF-TIMI 48 trial) matched the results of the trial, no value deviating by more than 0.5 %. In the base case, the use of warfarin and edoxaban for a lifetime horizon projected an average quality-adjusted survival of 8.425 and 9.022 QALYs, respectively. Being the average global costs for the two strategies €14,060 and €18,658, respectively, the incremental cost-utility ratio (ICUR) was €7,713/QALY gained (Table 2). One-way sensitivity (Fig. 2) analysis showed the baseline result being significantly influenced by the horizon of the analysis, with an ICUR above €25,000/QALY for an horizon below 51 months. We analyzed the impact of quality of life’s change by considering the difference between the utility coefficients of edoxaban and warfarin NVAF-only health states, which was 0.006 in the base case: ICUR increased above €25,000/QALY when such difference decreased below -0.030, i.e. if edoxaban had a significant detrimental effect onto quality of life with respect to warfarin. When considering the thrombotic risk, ICUR was €9,438/QALY for patients with CHADS2 score lower or equal to 3 and €5,363/QALY for CHADS2 score higher than 3. Finally, ICUR was €10,040/QALY for
123
patients managed by centers showing median TTR values above 66.4 %, while it was €6,479/QALY for patients managed in centers with lower TTR. In the probabilistic sensitivity analysis, edoxaban resulted cost-effective versus warfarin in 92.3 % of the simulations using a willingness-to-pay threshold of €25,000 per QALY and in 96 % of the simulations using a willingness-to-pay threshold of €50,000 per QALY.
Discussion In this study we adapted a previously published decision model to the clinical data reported by the ENGAGE AFTIMI 48 trial, demonstrating a favorable incremental costutility of edoxaban versus warfarin. One can wonder whether the cost-effectiveness profile of all DOAs is similar in the NVAF clinical setting, provided that similar market prices are assigned to the different drugs. Indeed, sensitivity analysis showed that quality of life of patients receiving DOAs is a key parameter influencing the results of the present and prior published cost-utility analyses;
Edoxaban vs warfarin for stroke prevention in NVAF
similarly, the results are also sensitive to some efficacy endpoints, which can vary among different DOAs [3, 4]. The present study specifically addresses the patient subgroup enrolled into the ENGAGE AF-TIMI 48 trial, that is NVAF patients with a CHADS2 score of 2 or higher, and, unfortunately, no comparison could be made versus other DOAs, since CHADS2-based subgroup analyses are not available from the ENGAGE AF-TIMI 48 trial, yet. Despite no direct comparison could be held among different DOAs, based on available data, edoxaban costeffectiveness compares positively with the other DOAs in the perspective of the Italian Health Care System. We are aware that applicability of cost-effectiveness results outside the original setting can be misleading. However, based on previously published economic evaluations of DOAs, differences among models may be even larger than differences among countries. Therefore, we deem that the results reported for the Italian health-care setting can be probably extended to other European countries. Economic data are of great interest for regulatory decisions such as approval of new drugs, formulary inclusion, insurance coverage, reimbursement rate [21]. Health economic evaluations are basic tools for stakeholders: constrained budgets impose them an increasing reliance on formal cost-effectiveness and cost-utility analyses, which have become the key issue for perception of the drug’s value by the regulator and among most important factors affecting drug prices. Value-based pricing has been recently introduced by some European countries: it is a method assessing whether the additional benefits of an innovative technology are worth the additional proposed costs [22, 23]. Therefore, economic models and simulations of budget impact analyses are recommended to be performed early in the drug-development process [24]. Based on the data reported by the present study, important health-care policy implications can be expected. First of all, market price of DOAs is expected to decline due to drug competition. Second, cost-containment policies can be proposed to companies, i.e. risk sharing, and to Italian regions, i.e. volume ceiling, in order to prevent massive shifting of patients taking warfarin with a good compliance and an acceptable TTR, while international guidelines do not recommend DOAs preferably to warfarin, yet. Third, cost-effectiveness of DOAs should be reassessed by health-care authorities as real-world safety data are gathered. Fourth, health-care authorities might impose master decision models with similar structure and assumptions in order to preserve transparency and comparability among new drugs belonging to the same class, as DOAs. Finally, we deem that a specific international effort should be devoted to quality of life of NVAF patients receiving oral anticoagulants in order to make cost-effectiveness estimations real-life tools for clinical governance.
Acknowledgments No sources of funding were used to assist in the conduct of this study Conflict of interest of interests.
The authors declare that they have no conflict
References 1. Anderson JL, Halperin JL, Albert NM, Bozkurt B, Brindis RG, Curtis LH, DeMets D, Guyton RA, Hochman JS, Kovacs RJ, Ohman EM, Pressler SJ, Sellke FW, Shen W-K (2013) Management of patients with atrial fibrillation (compilation of 2006 ACCF/AHA/ESC and 2011 ACCF/AHA/HRS guideline recommendations): a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines. Circulation 127(18):1916–1926 2. Camm AJ, Lip GY, De Caterina R, Savelieva I, Atar D, Hohnloser SH, Hindricks G, Kirchhof P, Kirchhof P (2012) 2012 focused update of the ESC guidelines for the management of atrial fibrillation: an update of the 2010 ESC guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association. Eur Heart J 33(21):2719–2747 3. Kansal AR, Zheng Y, Pokora T, Sorensen SV (2013) Costeffectiveness of new oral anticoagulants in the prevention of stroke in patients with atrial fibrillation. Best Pract Res Clin Haematol 26(2):225–237 4. Limone BL, Baker WL, Kluger J, Coleman CI (2013) Novel anticoagulants for stroke prevention in atrial fibrillation: a systematic review of cost-effectiveness models. PLoS One 8(4):1–15 5. Kansal AR, Sharma M, Bradley-Kennedy C, Clemens A, Monz BU, Peng S, Roskell N, Sorensen SV (2012) Dabigatran versus rivaroxaban for the prevention of stroke and systemic embolism in atrial fibrillation in Canada. Comparative efficacy and costeffectiveness. Thromb Haemost 108(4):672–682 6. Lip GY, Kongnakorn T, Phatak H, Kuznik A, Lanitis T, Liu LZ, Iloeje U, Hernandez L, Dorian P (2014) Cost-effectiveness of apixaban versus other new oral anticoagulants for stroke prevention in atrial fibrillation. Clin Ther 36(2):192-210.e20 7. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, Waldo AL, Ezekowitz MD, Weitz JI, Sˇpinar J, Ruzyllo W, Ruda M, Koretsune Y, Betcher J, Shi M, Grip LT, Patel SP, Patel I, Hanyok JJ, Mercuri M, Antman EM, ENGAGE AF-TIMI 48 Investigators (2013) Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med 369(22):2093–2104 8. Rognoni C, Marchetti M, Quaglini S, Liberato NL (2014) Apixaban, dabigatran, and rivaroxaban versus warfarin for stroke prevention in non-valvular atrial fibrillation: a cost-effectiveness analysis. Clin Drug Investig 34(1):9–17 9. Monz BU, Connolly SJ, Korhonen M, Noack H, Pooley J (2013) Assessing the impact of dabigatran and warfarin on health-related quality of life: results from an RE-LY sub-study. Int J Cardiol 168(3):2540–2547 10. Gage BF, Cardinalli AB, Owens DK (1998) Cost-effectiveness of preference-based antithrombotic therapy for patients with nonvalvular atrial fibrillation. Stroke 29(6):1083–1091 11. Sullivan PW, Arant TW, Ellis SL, Ulrich H (2006) The cost effectiveness of anticoagulation management services for patients with atrial fibrillation and at high risk of stroke in the US. Pharmacoeconomics 24(10):1021–1033 12. Meckley LM, Gudgeon JM, Anderson JL, Williams MS, Veenstra DL (2010) A policy model to evaluate the benefits, risks and costs of warfarin pharmacogenomic testing. Pharmacoeconomics 28(1):61–74
123
C. Rognoni et al. 13. Fattore G, Torbica A, Susi A, Giovanni A, Benelli G, Gozzo M, Toso V (2012) The social and economic burden of stroke survivors in Italy: a prospective, incident-based, multi-centre cost of illness study. BMC Neurol 12:137–147 14. Fedele F, D’Ambrosi A, Bruno N, Caira C, Brasolin B, Mancone M (2011) Cost-effectiveness of levosimendan in patients with acute heart failure. J Cardiovasc Pharmacol 58(4):363–366 15. O’Brien CL, Gage BF (2005) Costs and effectiveness of ximelagatran for stroke prophylaxis in chronic atrial fibrillation. JAMA 293:699–706 16. Rosand J, Eckman MH, Knudsen KA, Singer DE, Greenberg SM (2004) The effect of warfarin and intensity of anticoagulation on outcome of intracerebral hemorrhage. Arch Intern Med 164(8):880–884 17. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med 154(13):1449–1457 18. Ariesen MJ, Claus SP, Rinkel GJ, Algra A (2003) Risk factors for intracerebral hemorrhage in the general population: a systematic review. Stroke 34(8):2060–2065 19. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ (2001) Validation of clinical classification schemes
123
20.
21.
22.
23.
24.
for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA 285(22):2864–2870 National Institute for Health and Clinical Excellence (NICE). Guide to the methods of technology appraisal. 2008. Available from: http://www.nice.org.uk/media/b52/a7/tamethodsguideupda tedjune2008.pdf Accessed January 10, 2014 Frick K, Neissen L, Bridges J, Walker D, Wilson RF, Bass EB (2012) Usefulness of economic evaluation data in systematic reviews of evidence. Agency for healthcare research and quality (US) Report no. 12(13) Danzon P, Towse A, Mestre-Ferrandiz J (2013) Value based differential pricing: efficient prices for drugs in a global context. Health Econ. doi:10.1002/hec.3021 Hall PS, McCabe C, Brown JM, Cameron DA (2010) Health economics in drug development: efficient research to inform healthcare funding decisions. Eur J Cancer 46(15):2674–2680 van Nooten F, Holmstrom S, Green J, Wiklund I, Odeyemi IA, Wilcox TK (2012) Health economics and outcomes research within drug development: challenges and opportunities for reimbursement and market access within biopharma research. Drug Discov Today 17(11–12):615–622
