CLINICAL RESEARCH

Europace (2014) 16, 652–659 doi:10.1093/europace/eut380

Ablation for atrial fibrillation

Cost-effectiveness of cryoballoon ablation for the management of paroxysmal atrial fibrillation Matthew R. Reynolds 1*, Mark Lamotte2, Derick Todd 3, Yaariv Khaykin 4, Simon Eggington 5, Stelios Tsintzos 6, and Gunnar Klein 7 1 Lahey Clinic Medical Center, 41 Mall Road, Burlington, MA 01805, USA; 2IMS HEOR, Vilvoorde, Belgium B-1800; 3Institute of Cardiovascular Medicine and Science, Liverpool Heart and Chest Hospital, Liverpool, L14 3PE, UK; 4University of Toronto, Ontario, Canada; 5Corporate Health Policy and Payment, Medtronic International Trading Sa`rl, 1131-Tolochenaz, Switzerland; 6 Medtronic Global Headquarters, Mounds View, MN 55112, USA; and 7Department of Cardiology, Hannover Medical School, Car-Neuberg-Strasse 1, 30625 Hannover, Germany

Received 19 August 2013; accepted after revision 4 November 2013; online publish-ahead-of-print 2 January 2014

Aims

----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords

Cost-effectiveness analysis † Atrial fibrillation † Quality of life † Stroke † Cryoablation

Introduction Atrial fibrillation (AF) is the most common sustained arrhythmia observed in adult populations.1 As the onset of AF is distinctly age-related, the incidence and prevalence of AF are projected to rise substantially in the coming decades.2 Individuals with AF on an average have poorer quality of life (QOL) than age-matched members of the general population.3 – 5 In addition, AF is associated with and complicates hypertension and heart failure, and is known to significantly increase the risk of stroke6 and possibly death.1 Initially described in 1998,7 catheter ablation in the vicinity of the pulmonary veins (PVs) has become an important treatment option for many AF patients. Current treatment guidelines consider AF ablation as a second-line (and potentially in selected cases, first-line) option for patients who desire and are appropriate candidates for rhythm control therapy.1 Atrial fibrillation ablation is clearly more

effective at maintaining sinus rhythm than alternative anti-arrhythmic drug (AAD) therapy in patients for whom one or more AADs have previously failed,8 and improves QOL more than AADs in this circumstance.9 When initially developed, AF ablation procedures were performed using radiofrequency (RF) ablation catheters designed for the treatment of more focal arrhythmias. Electrical isolation of all PVs—the generally accepted endpoint of AF ablation procedures—can be time-consuming, requires considerable operator skill, and is limited by relatively frequent electrical re-connection of acutely isolated areas, leading to repeat procedures in 30% or more of patients.10 More recently, a balloon catheter which isolates the PVs using cryothermal energy has been developed and commercialized. Although previous studies based on ablation outcomes with RF catheters have suggested that AF ablation is most likely cost-effective

* Corresponding author. Tel: +1 617 307 5489; fax: +1 617 307 5600, E-mail: [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2014. For permissions please email: [email protected].

Downloaded from by guest on November 14, 2014

Cryoballoon ablation is an established treatment option for the management of patients with atrial fibrillation. We sought to evaluate the cost-effectiveness of cryoablation, compared with second-line anti-arrhythmic drug (AAD) therapy in patients with paroxysmal atrial fibrillation (PAF), from a UK payer perspective. ..................................................................................................................................................................................... Methods We developed a state-transition (Markov) model to calculate the total costs and quality-adjusted life-years (QALYs) assoand results ciated with cryoablation and AAD therapy in patients with PAF. A 5-year horizon was used for the base-case. Data from a recent study of cryoballoon ablation in patients with PAF were used to model short-term health outcomes and costs, together with longer term external evidence to populate subsequent time periods. Total discounted costs were £21 162 and £17 627 for the cryoballoon ablation and AAD arms, respectively. Total QALYs of 3.565 and 3.404 therefore led to an incremental cost-effectiveness ratio of £21 957 per QALY gained. Sensitivity analysis suggested that the key drivers of the results were the model time horizon, the costs of follow-up care in patients with recurrent AF, and the costs of the ablation procedure. ..................................................................................................................................................................................... Conclusion Cryoballoon ablation provides increased quality-adjusted life expectancy compared with AAD at reasonable additional cost, representing good value for money in patients with PAF.

653

Cost-effectiveness of cryoballoon ablation for the management of PAF

What’s new? † Cryoballoon ablation is a recent alternative to radiofrequency (RF) catheter ablation and is approved in many countries for the treatment of paroxysmal atrial fibrillation (AF). † Although RF ablation has been shown to be a cost-effective intervention for AF when compared with anti-arrhythmic drug therapy, there are currently no economic evaluations which specifically relate to cryoballoon ablation in this patient population. † In this paper, we describe a cost – utility analysis of cryoballoon ablation vs. anti-arrhythmic drug therapy, based in part upon the pivotal STOP-AF study, concluding that its costeffectiveness profile is similar to that of RF ablation.

Methods Overview We developed a Markov model to assess the cost-effectiveness of cryoablation compared with AADs in patients with PAF. The model used data from the recent STOP-AF trial14 to model 12-month outcomes, together with data sources on longer term endpoints and UK cost data to calculate the expected total costs and quality-adjusted life-years (QALYs) associated with each treatment. STOP-AF enroled patients with PAF unsuccessfully treated with ≥1 AAD. The patient population modelled reflected the characteristics of patients in STOP-AF in terms of age, baseline stroke risk, and sex. We assumed that the efficacy, safety, and QOL data from STOP-AF are transferable to the UK setting. A UK NHS perspective was taken on costs (2011 prices), and both costs and QALYs were discounted at 3.5% per year, in line with guidance from the National Institute for Health and Clinical Excellence.15 A 5-year horizon was used in the base-case analysis to reflect uncertainty in longterm outcomes in this population. The cycle length for health state transitions was set at 6 months. The model was developed using TreeAge Pro. Figure 1 shows the health states used in the model and the possible transitions between them. A microsimulation approach was used to capture the events which occurred in the model, including stroke, AF recurrence, and procedure-related complications. A sequence of AAD therapies was modelled to reflect the fact that several drug interventions could be administered in this patient population if ablation were not an option.1 The model starts at the time of the ablation procedure or commencing AAD therapy. Patients in the ablation arm thus entered the model in the ‘Sinus rhythm (post-ablation)’ state (some patients required more than one ablation procedure), transitioning through the remaining health states according to the risks of various events including AF recurrence, stroke, and death. Upon first recurrence of AF, these patients were assumed to commence their first AAD (propafenone); a subsequent

Sinus rhythm on propafenone (first AAD) Non-disabling stroke Sinus rhythm on sotalol (second AAD) Disabling stroke

Sinus rhythm on amiodarone (third AAD)

AF postrecurrence (rate control)

Dead

Figure 1 Markov model structure.

recurrence of AF then led to them being treated with sotalol, followed by amiodarone, and finally, rate control therapy alone (assumed to be metoprolol) at subsequent recurrences. Patients in the AAD arm followed the same sequence of interventions, except that they entered the model in the ‘Sinus rhythm on propafenone’ state. The model did not allow these patients to crossover to catheter ablation in the model, although this was a common occurrence in the STOP-AF study and other randomized trials of AF ablation. This is because the objective of the model was to evaluate the cost-effectiveness of catheter ablation vs. a scenario in which catheter ablation is not available. Stroke could occur from any health state, and be either ischaemic or haemorrhagic in nature; different levels of stroke severity (disabling and non-disabling) were modelled to reflect variable stroke outcomes. Mortality could occur from any other health state and was assumed to be higher for patients in the post-stroke states. We assumed that patients could not re-enter a health state which they had previously left, or move to a less severe state. To ensure convergence of the results, we simulated 100 000 patients per treatment arm for the base-case analysis. A half-cycle correction was applied, which assumes that transitions occur, on an average, half-way through the cycle.

Procedural complications Patients undergoing ablation are at risk of complications related to the procedure, of which eight were included in the model: ischaemic stroke, cardiac tamponade, phrenic nerve palsy, PV stenosis, arteriovenous fistula, bleeding requiring transfusion, femoral artery pseudoaneurysm, and subclavian vein rupture. Given the relatively small population in STOP-AF and resulting uncertainty around the incidence of these complications, we used data from a meta-analysis of cryoballoon studies comprising over 1300 patients to derive inputs for this part of the model.16

Downloaded from by guest on November 14, 2014

for patients with paroxysmal AF (PAF),11 – 13 the potential costeffectiveness of cryoballoon ablation for AF has not been assessed. We therefore sought to evaluate the cost-effectiveness of cryoballoon ablation compared with AADs over a 5-year time horizon from a UK NHS perspective, using clinical data from the Sustained Treatment of Paroxysmal Atrial Fibrillation (STOP-AF) trial14 together with the most suitable long-term data on PAF and UK unit costs.

Sinus rhythm (post-ablation)

654

Atrial fibrillation recurrence The key data on the initial recurrence of AF came from the STOP-AF trial,14 a 12-month prospective, multi-centre, randomized controlled study which compared cryoballoon ablation vs. AAD therapy in patients with PAF. At 12 months, treatment success was achieved in 69.9% of ablated patients, compared with 7.3% of AAD patients. In patients free of recurrent AF at 12 months, the longer term risk of AF following cryoballoon ablation was taken from a large, recently published series,17 while longer term recurrent rates on AAD treatment were taken from previous literature.18 Patients in both model arms were assumed to cycle through a sequence of AADs after each AF recurrence. The assumed order of drugs received was propafenone, followed by sotalol, followed by amiodarone. In the ablation arm, the same sequence of AADs was assumed after the first recurrence of AF. The probability of recurrence on propafenone and sotalol was based on the 12-month control arm data from STOP-AF and thereafter on previously published data.18 A different recurrence rate was applied to patients receiving amiodarone, based on the DIONYSOS study.19 After failure of all three AADs, patients were assumed to receive rate control therapy in the form of a beta-blocker (metoprolol); no further recurrences were modelled beyond this point.

Stroke and stroke prevention

Drug-related adverse events Serious adverse event (SAE) rates were applied to each AAD to reflect the different adverse event profiles of each drug. We selected one large study of sotalol in patients with PAF, which adequately reported SAEs,25 and used this to estimate a 6-monthly risk of SAEs for use in the sotalol health state. For the propafenone and amiodarone health states, we used the sotalol SAE risk, combined with odds ratios from a meta-analysis of AADs in the treatment of AF,26 to obtain 6-monthly

risks of SAEs on propafenone and amiodarone. For patients on rate control therapy, we modelled SAEs using data from Ku¨hlkamp et al.,27 using discontinuation rates due to SAEs as a proxy for the overall SAE rate.

Cost and resource use Various cost categories were included in the model to represent a range of care for patients with PAF. The majority of cost parameters were taken from the NHS Payment by Results tariffs,28 in addition to data from existing economic analyses, national drug price lists, personal and social care costs, and resource use estimates from large databases.29 – 34 The details of the cost and resource use data and assumptions are described in the Supplementary material online, Appendix.

Utility data Quality-of-life data were collected in STOP-AF via the SF-36, which was administered at baseline and at the 12-month follow-up visit. We used these data to generate SF-6D utility weights using a standard algorithm,35 and combined this information with existing evidence on the impact of procedural complications, drug-related adverse events, and stroke upon patients’ QOL.36,37 The utility weight associated with each health state was then multiplied by the mean time each patient spent in each health state to give overall QALYs for each treatment. The details of the utility weights used are described in the Supplementary material online, Appendix.

Mortality Mortality within the model was captured in two ways: general background mortality for patients in sinus rhythm or with AF, and death due to or following stroke. Background mortality was based upon UK life tables,38 adjusted for the age and sex distribution of patients in STOP-AF. Death due to stroke was modelled using age-specific mortality rates reported by Wardlaw et al.22 Patients surviving a stroke were assumed to be at an elevated risk of death, implemented using two relative risk parameters depending on stroke severity. If a patient suffered a non-disabling stroke, a relative risk of 1.33 was applied to the life table mortality risk for the remainder of the model; a patient who initially survived a disabling stroke had a relative risk of 3.46 applied to the life table risk. These parameters were calculated using data from a 5-year study of outcomes in stroke patients.39

Sensitivity analyses A series of one-way sensitivity analyses was undertaken by varying input parameters within plausible limits, largely based upon the reported confidence intervals for each parameter value. In instances where confidence limits were not available, plausible ranges were applied using data from other studies. A probabilistic sensitivity analysis was also undertaken to explore the effect of the joint uncertainty in all input parameters. By assigning a probability distribution to each input parameter to reflect the uncertainty in its mean value, and repeatedly taking random samples from each distribution, the uncertainty was propagated through the model to obtain a range of possible incremental cost-effectiveness ratios (ICERs). A range of distributions was used to represent the uncertainty in different parameters types: beta distributions for probabilities and utilities; gamma distributions for cost and resource use parameters; odds ratio parameters for adverse event rates; lognormal distributions for relative risk parameters; and Dirichlet distributions to sample more than two probabilities concurrently (e.g. distribution of CHADS2 scores at baseline). We ran the probabilistic analysis with 1000 sets of sampled parameter values, for each of which 100 000 patients were simulated. A full list of parameter values is given in Table 1.

Downloaded from by guest on November 14, 2014

The model included both ischaemic and haemorrhagic strokes. Ischaemic stroke risk was based on the risk factors of patients in STOP-AF, using the CHADS2 classification system and annual stroke rates for each CHADS2 score.20 These stroke risks were applied to patients in all states apart from the ‘Sinus rhythm (post-ablation)’ state. One study reported a hazard ratio of 1.6 for stroke risk in patients with AF vs. those in sinus rhythm21—this relative risk was applied in the ablation arm up until the time of first recurrence, reflecting patients’ reduced stroke risk during this period. Stroke outcome was modelled based on age-specific outcomes from the Lothian stroke registry,22 enabling us to divide stroke events into fatal, disabling, and non-disabling. Once in the stroke states, patients were assumed to stop taking AADs and begin rate control therapy. Upon the first recurrence of AF, all patients were assumed to commence oral anti-coagulation therapy (OAC) in the form of warfarin, which reduces stroke risk by 67%.23 However, the use of OACs can lead to major bleeding, which we modelled in terms of gastrointestinal bleed or haemorrhagic stroke. A major bleeding rate of 1.36% per year was applied, of which 32.1% were assumed to be haemorrhagic strokes.24 A major bleed was assumed to lead to withdrawal of warfarin and commencement of aspirin, which is associated with fewer bleeds but is less effective in stroke prevention. Minor bleeding events were also modelled but did not require a change in OAC therapy. The outcomes of haemorrhagic stroke were modelled in the same way as for ischaemic stroke, with patients dying or entering one of the two stroke states. The model allowed patients to have more than one stroke (since a stroke increases a patient’s CHADS2 score and thus their risk of subsequent stroke), with patients moving into the health state corresponding to their most severe stroke.

M.R. Reynolds et al.

655

Cost-effectiveness of cryoballoon ablation for the management of PAF

Table 1 Table of parameter inputs Input parameter

Table 1 Continued Value

................................................................................ Procedure-related parameters Number of cryoballoons per procedure Initial procedure Repeat procedure Number of Freezor catheters per procedure Initial procedure Repeat procedure Probability of repeat ablation procedure within 3 months Procedural complication probabilities Cardiac tamponade

Disabling stroke Other adverse event parameters

14

1.61

1.1314 14

0.52

0.6514 14

0.19

0.009116

Stroke/TIA Phrenic nerve palsy

0.003216

Persisting beyond 1 year—as % of above rate Arteriovenous fistula

0.0473

0.07516 0.004116 16

Bleeding

0.0041

Pseudoaneurysm Subclavian vein rupture

0.003216 0.000816 14

CHADS2 ¼ 0 CHADS2 ¼ 1

50.6% 37.9%14

CHADS2 ¼ 2

14

CHADS2 ¼ 3 AF recurrence probabilities

10.6%

14

0.8%

Per cycle post-12 months Propafenone and sotalol 0– 6 months 6– 12 months

Probability of minor bleed on warfarin (per year) Relative risk of bleeding on aspirin (vs. warfarin)

0.15840

Minor bleed Probability of drug-related SAE on sotalol (per cycle) Sotalol Odds ratios for risk of SAE (vs. sotalol) Amiodarone Propafenone

0.227 0.06314 17

0.0501

0.86614 0.45414 18

0.117 0.23819

Discontinuation on amiodarone (per cycle)

0.06919

Annual stroke probabilities by CHADS2 score (with AF) CHADS2 ¼ 0 CHADS2 ¼ 1

0.008320 0.015420

CHADS2 ¼ 2

0.023520

CHADS2 ¼ 3 CHADS2 ¼ 4

0.042920 0.090620

CHADS2 ¼ 5

0.110220 0.13420

Hazard ratio of stroke in AF vs. sinus rhythm

1.621

Relative risk of stroke on warfarin (vs. placebo) Relative risk of stroke on warfarin (vs. aspirin)

0.3323 0.5923

Stroke outcome (disabling/non-disabled/dead)

Age-dependent22

0.5823 0.4523

0.20725 0.03927 2.1426 1.3526

Resource use parameters Hospital admissions per cycle Sinus rhythm states Rate control and stroke states Outpatient visits per cycle Sinus rhythm states Rate control and stroke states GP visits per cycle Rate control and stroke states Utility parameters

14

Per-cycle post-12 months Amiodarone (per cycle)

CHADS2 ¼ 6 Other stroke-related parameters

0.320624

Sinus rhythm states

Ablation arm 0– 6 months 6– 12 months

Probability that a major bleed is a haemorrhagic stroke

0.3434 0.7134 3.434 4.7434 9.2234 11.2534

Baseline utility

0.7314

Increment in sinus rhythm Increment in first cycle immediately after first recurrence post-ablation Increment in first cycle immediately after recurrence on AADs

0.0814 0.0614

Decrement with phrenic nerve palsy Decrement for procedural complication or SAE (1 month decrement) Utility multiplier for

0.0314 0.136

Non-disabling stroke Disabling stroke Daily drug dose parameters (mg)

Downloaded from by guest on November 14, 2014

Baseline CHADS2 score distribution

3.4639 0.013624

Rate control 16

1.3339

Probability of major bleed on warfarin (per year)

Major bleed 0.005716

Value

................................................................................ Non-disabling stroke

Pulmonary vein stenosis

Persisting post-procedure

Input parameter

0.0114

0.8737 0.5237

Warfarin

512

Aspirin Sotalol

7531 24014

Propafenone

45014

Amiodarone Metoprolol

20012 15032

Costs—ablation procedure

Relative risk of mortality following

Continued

Procedure Cryoballoon (Arctic Front—23/28 mm)

£314928 £188841

Freezor catheter

£80741

Continued

656

M.R. Reynolds et al.

Table 1 Continued

Table 2 Deterministic cost-effectiveness results Treatment arm

Total costs

Total QALYs

................................................................................

................................................................................

Costs—drugs (per mg)

Anti-arrhythmic drugs

£17 627

3.404

Cryoballoon ablation Incremental

£21 162 £3535

3.565 0.161

Input parameter

Value

Warfarin Aspirin

£0.0006632 £0.0001332

Sotalol

£0.0004332

Propafenone Amiodarone

£0.0005232 £0.00033932

Metoprolol

£0.00033332

QALYs, quality-adjusted life-years.

Costs—ischaemic stroke Non-disabling stroke Per cycle in first year Per cycle after first year Disabling stroke Per cycle in first year Per cycle after first year Immediate death

Per cycle after first year

£145

Anti-coagulation

£13 68029 – 31

Per cycle after first year

£186429 – 31 £93.2728

AADs

£237529 – 31 £10429 – 31

£15 000

£14 22429 – 31 £133729 – 31 £183128 £39128

Warfarin monitoring (per year)

£253.1342

PV stenosis

Follow-up/ hospitalizations Adverse events

£20 000

Major gastrointestinal bleed Minor bleed Costs—miscellaneous Cardiac tamponade

£25 000

£10 000

£5 000

£378328 £410928 28

CT scan (applies to 50% of patients with PV stenosis)

£118

GP visit Inpatient hospital admission

£2533 £123928

Cardiology outpatient visit

£10828

Cardiac surgery outpatient visit Arteriovenous fistula

£20028 £205528

Access site bleeding

£12543

Femoral artery pseudoaneurysm Subclavian vein rupture (blood vessel injury admission, plus 2 cardiac surgery outpatient visits)

£205528 £147628

SAE on AADs

£123928

Initiation of amiodarone Amiodarone monitoring (per cycle)

£10828 £22928

Results The deterministic results of the cost-effectiveness analysis are shown in Table 2, based on a 5-year time horizon and with both costs and QALYs discounted at 3.5% per year. Cryoballoon ablation, compared with drug therapy, results in greater QALYs (3.565 vs. 3.404), and greater costs (£21 162 vs. £17 627). These results yield an ICER of £21 957 per QALY gained.

£0 AADs

Cryoballoon

Figure 2 Cost breakdown by treatment arm.

Figure 2 shows a breakdown of the costs into several categories. A large proportion of the overall costs were related to on-going follow-up and AF-related hospitalizations, which is consistent with previous studies.19,44 – 46 The ablation procedure also formed a significant proportion of the cost in the ablation arm, which was partially offset by reduced OAC use and drug-related SAEs. One-way sensitivity analyses (Figure 3) showed that the model result was the most sensitive to the time horizon used, the costs of follow-up care in patients with recurrent AF, and the total cost of the ablation procedure. Assuming the same risk of stroke for patients in sinus rhythm and those with recurrent AF had little impact upon the results, with the ICER increasing slightly to £24 265 per QALY gained. The results of the probabilistic analysis are shown on the scatterplot in Figure 4. In all replications, cryoballoon ablation was projected to provide additional QALYs but also additional costs. The majority of model replications predicted a gain of between 0.1 and 0.2 QALYs, at an incremental cost of £2000 –6000 per patient. A cost-effectiveness acceptability curve is shown in Figure 5. The model results suggest that cryoballoon ablation is unlikely to be

Downloaded from by guest on November 14, 2014

Disabling haemorrhagic stroke Per cycle in first year

Ablation procedure

29 – 31

Costs—bleeding Non-disabling haemorrhagic stroke Per cycle in first year

Rate control therapy

£228429 – 31

657

Cost-effectiveness of cryoballoon ablation for the management of PAF

2

Time horizon (years) 10 Costs of FU in AF per 6 months £2 500

£8 250

£5 500

Total procedure cost (initial) Cost of drug-related SAE

£1 086

£500

£3 858

Prob of 1-year recurrence (Cryo)

0.181

Utility gain in sinus rhythm

0.099

Prob of 1-year recurrence (AADs)

0.927

Repeat ablation rate

10% £0

0.366 0.061 0.825 30%

£10 000 £20 000 £30 000 £40 000 £50 000 £60 000 £70 000 £80 000 £90 000 £100 000

Incremental cost-effectiveness ratio

Incremental costs

£7 000 £6 000 £5 000 £4 000 £3 000 £2 000 £1 000 £0 0

0.05

0.2 0.1 0.15 Incremental QALYs

0.25

0.3

Figure 4 Cost-effectiveness scatter plot.

preferred at a willingness-to-pay threshold below £20 000 per QALY. However, beyond a threshold of £22 000 per QALY gained, ablation becomes the more cost-effective intervention, with probabilities of 86% and 97.2% of being cost-effective at thresholds of £30 000 and £40 000 per QALY gained, respectively.

Discussion Cryoballoon ablation is a recently introduced alternative to RF catheter ablation and is currently approved in many countries for the treatment of PAF. In this health economic analysis performed from the perspective of the UK NHS, we found that cryoballoon ablation increased both QALYs and aggregate medical care costs, with a point estimate of £22 000 per QALY gained. This result was sensitive to

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Cryoablation AADS

£0

£20 000 £40 000 £60 000 £80 000 £100 000 Cost-effectiveness threshold

Figure 5 Cost-effectiveness acceptability curve.

assumptions about the time horizon, the on-going cost of AF followup care, and the ablation procedure cost. Several previous studies have evaluated the potential costeffectiveness of AF ablation in different countries; however, this is the first cost-effectiveness analysis specifically examining cryoballoon ablation. Khaykin et al.47 published a cost analysis of AF ablation compared with AADs based on an assessment of ablation costs and AF care patterns in Canada. This analysis suggested that by better controlling the rhythm, ablation would result in reduced downstream healthcare resource utilization, and hence that the upfront expenditures on ablation would be recouped somewhere between 3 and 8 years after the procedure, depending on the assumptions. Two additional cost-effectiveness models of AF ablation have been published since then. Reynolds et al.13 compared, from a US perspective, AF ablation with AADs as second-line therapy for PAF—a

Downloaded from by guest on November 14, 2014

£8 000

Probability of cost-effectiveness

Figure 3 Tornado diagram (one-way sensitivity analysis).

658

similar in magnitude to that previously estimated for RF ablation in a UK technology assessment. The similarity of cost-effectiveness results between RF and cryoballoon ablation vs. AADs suggests that both RF and cryoballoon ablation should be treated similarly from a health policy perspective.

Supplementary material Supplementary material is available at Europace online. Conflict of interest: M.R.R., M.L., D.T., Y.K., and G.K. all received honoraria from Medtronic. S.E. and S.T. are employees of Medtronic.

Funding This work was supported by Medtronic International, Tolochenaz, Switzerland.

References 1. Camm AJ, Kirchof P, Lip GY, Schotten U, Savelieva I, Ernst S et al. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Europace 2010;12:1360 –420. 2. Miyasaka Y, Barnes ME, Gersh BJ, Cha SS, Bailey KR, Abhayratna WP et al. Secular trends in incidence of atrial fibrillation in Olmsted County, Minnesota, 1980 to 2000, and implications on the projections for future prevalence. Circulation 2006; 114:119 – 25. 3. Dorian P, Jung W, Newman D, Paquette M, Wood K, Avers GM et al. The impairment of health-related quality of life in patients with intermittent atrial fibrillation: implications for the assessment of investigational therapy. J Am Coll Cardiol 2000;36:1303 –9. 4. Reynolds MR, Lavelle T, Essebag V, Cohen DJ, Zimetbaum P. Influence of age, sex, and atrial fibrillation recurrence on quality of life outcomes in a population of patients with new-onset atrial fibrillation: the Fibrillation Registry Assessing Costs, Therapies, Adverse events and Lifestyle (FRACTAL) study. Am Heart J 2006;152:1097 –103. 5. Jung W, Herwig S, Newman D, Akhtar M, Wood K, Ayers G et al. Impact of atrial fibrillation on quality of life: a prospective, multicenter study. J Am Coll Cardiol 1999;33(Suppl A):104A. 6. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation: a major contributor to stroke in the elderly. The Framingham Study. Arch Int Med 1987;147:1561 –4. 7. Haı¨ssaguerre M, Jais P, Shah DC, Takahashi A, Hocini M, Quiniou G et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. New Engl J Med 1998;339:659 – 66. 8. Terasawa T, Balk EM, Chung M, Garlitski AC, Alsheikh-Ali AA, Lau J et al. Systematic review: comparative effectiveness of radiofrequency catheter ablation for atrial fibrillation. Ann Int Med 2009;151:191 –202. 9. Reynolds MR, Walczac J, White SA, Cohen DJ, Wilber DJ. Improvements in symptoms and quality of life in patients with paroxysmal atrial fibrillation treated with radiofrequency catheter ablation versus antiarrhythmic drugs. Circ Cardiovasc Qual Outcomes 2010;3:315 –23. 10. Calkins H, Reynolds MR, Spector P, Sondhi M, Xu Y, Martin A et al. Treatment of atrial fibrillation with antiarrhythmic drugs or radiofrequency ablation: two systematic literature reviews and meta-analyses. Circ Arrhythm Electrophysiol 2009;2:349 –61. 11. Khaykin Y, Wang X, Natale A, Wazni OM, Skanes AC, Humphries KH et al. Cost comparison of ablation versus antiarrhythmic drugs as first-line therapy for atrial fibrillation: an economic evaluation of the RAAFT pilot study. J Cardiovasc Electrophysiol 2009;20:7 –12. 12. McKenna C, Palmer S, Rodgers M, Chambers D, Hawkins N, Golder S et al. Costeffectiveness of radiofrequency catheter ablation for the treatment of atrial fibrillation in the United Kingdom. Heart 2009;95:542 – 9. 13. Reynolds MR, Zimetbaum P, Josephson ME, Ellis E, Danilov T, Cohen DJ. Costeffectiveness of radiofrequency catheter ablation compared with antiarrhythmic drug therapy for paroxysmal atrial fibrillation. Circ Arrhythm Electrophysiol 2009;2: 362 –9. 14. Packer DL, Kowal RC, Wheelan KR, Irwin JM, Champagne J, Guerra PG et al. Cryoballoon ablation of pulmonary veins for paroxysmal atrial fibrillation. First results of the North American Arctic Front STOP-AF pivotal trial. J Am Coll Cardiol 2013;61: 1713 –23. 15. National Institute for Health and Clinical Excellence. Guide to the methods of technology appraisal. June 2008. 16. Andrade JG, Khairy P, Guerra PG, Deyell MW, Rivard L, Macle L et al. Efficacy and safety of cryoballoon ablation for atrial fibrillation: a systematic review of published studies. Heart Rhythm 2011;8:1444 –51.

Downloaded from by guest on November 14, 2014

patient population similar to that studied in STOP-AF—using a Markov model. That analysis did not assume any benefit from ablation on the risk of stroke, but did model gains in QOL after successful treatment based on several empirical datasets, and additionally projected downstream cost offsets related to improved arrhythmia control with ablation. That model reported a base-case ICER of $51 000 (US) per QALY gained. In a similar timeframe, McKenna et al.12 published a health technology assessment, including a cost-effectiveness analysis of RF AF ablation, from a UK perspective. Based on the available literature at the time, and assuming both QOL benefits and potentially small improvements in the risk of stroke and death from sinus rhythm maintenance, the study estimated an ICER of £25 000 per QALY gained over a 5-year time horizon, and an ICER of ,£10 000 per QALY gained if the QALY gains of the intervention are assumed to persist over the patient’s lifetime.12 Our study has structural similarities to previous models of AF ablation,12,13 and hence, our base-case results are convergent with those previously reported. A markedly different result would have been surprising, given that the observed efficacy rates in the pivotal randomized trial of cryoballoon ablation were similar—both for ablation and for drug therapy—to those from similar randomized studies using the older RF ablation technology.48,49 In addition, the changes in SF-6D utility scores following cryoballoon ablation in the STOP-AF trial were very similar to those reported from previous series.9,13 This latter point is important, as the patient population of STOP-AF and most other studies of AF ablation is relatively young and healthy. Therefore, the projected rate of stroke and death in this population is low in the short-to-medium term, and assumptions about the potential benefit of ablation on stroke and death have only a minor impact on our model results. The difference in projected QALYs between ablation and drug-treated patients is therefore determined to a greater degree by the changes in QOL. As observed previously,13 an important implication of this finding is that catheter ablation is most likely to be cost-effective for the patients with the highest symptom burden at baseline, assuming a good result. Like all health economic evaluations, our study has important limitations to be considered. First, as a new technology, the observed follow-up of patients treated with cryoballoon ablation is limited. We therefore incorporated data from a series with median follow-up of 30 months,17 representing the longest data yet reported. Also, treatment pathways in PAF vary considerably and long-term outcomes are not well-studied. Consequently, we limited the time horizon of our model to 5 years, which is a conservative assumption from the perspective of the technology. Cryoballoon ablation is specifically designed for isolation of the PVs, which is accepted as the cornerstone of AF ablation procedures. However, it is accepted that non-paroxysmal forms of AF may require additional ablation beyond PV isolation, and data suggest that overall success rates are lower with ablation in non-paroxysmal patients. As there is no presently randomized data comparing the success of cryoballoon ablation with drugs for non-PAF patients, the results of our model cannot be extrapolated to that patient population. In conclusion, based on the best currently available data, we project that cryoballoon ablation over a 5-year period improves QALYs compared with AADs, with a cost-effectiveness ratio

M.R. Reynolds et al.

Cost-effectiveness of cryoballoon ablation for the management of PAF

31. National Institute for Health and Clinical Excellence. Stroke: National clinical guideline for diagnosis and initial management of acute stroke and transient ischaemic attack (TIA). 2008. 32. British National Formulary, Number 62. Pharmaceutical Press, 2011. 33. Curtis L. Unit Costs of Health and Social Care. Kent, UK: Personal Social Services Research Unit, 2011. 34. Ladapo JA, David G, Gunnarsson CL, Hao SC, White SA, March JL et al. Healthcare utilisation and expenditures in patients with atrial fibrillation treated with catheter ablation. J Cardiovasc Electrophysiol 2012;23:1 –8. 35. Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Econ 2002;21:271 –92. 36. A˚kerborg O, Nilsson J, Bascle S, Lindgren P, Reynolds M. Cost-effectiveness of dronedarone in atrial fibrillation: results for Canada, Italy, Sweden and Switzerland. Clin Ther 2012;34:1788 –802. 37. Tengs TO, Lin TH. A meta-analysis of quality-of-life estimates for stroke. Pharmacoeconomics 2003;21:191 –200. 38. Government Actuary’s Department. Interim life tables, 2007–09. http://www.gad. gov.uk/Demography%20Data/Life%20Tables/ (3 October 2012, date last accessed) 39. Slot KN, Berge E, Dorman P, Lewis S, Dennis M, Sandercock P. Impact of functional status at six months on long-term survival in patients with ischaemic stroke: prospective cohort studies. BMJ 2008;336:376 –9. 40. National Institute for Health and Clinical Excellence. Atrial Fibrillation: the management of atrial fibrillation. Clinical Guideline Number 36. 2006. 41. Eucomed. European Sales Data, 2011. 42. National Institute for Health and Clinical Excellence. Dabigatran etexilate for the prevention of stroke and systemic embolism in atrial fibrillation. Technology Appraisal Guidance number 249. 2012. 43. Ker K, Edwards P, Perel P, Shakur H, Roberts I. Effect of tranexamic acid on surgical bleeding: systematic review and meta-analysis. BMJ 2012;344:1 –13. 44. McBride D, Mattenklotz AM, Willich SN, Bru¨ggenju¨rgen B. The costs of care in atrial fibrillation and the effect of treatment modalities in Germany. Value Health 2009;12: 293 –301. 45. Reynolds MR, Essebag V, Zimetbaum P, Cohen DJ. Healthcare resource utilization and costs associated with recurrent episodes of atrial fibrillation: the FRACTAL registry. J Cardiovasc Electrophysiol 2007;18:628 –33. 46. Stewart S, Murphy N, Walker A, McGuire A, McMurray JJV. Cost of an emerging epidemic: an economic analysis of atrial fibrillation in the UK. Heart 2004;90:286 –92. 47. Khaykin Y, Morillo CA, Skanes AC, McCracken A, Humphries K, Kerr CR. Cost comparison of catheter ablation and medical therapy in atrial fibrillation. J Cardiovasc Electrophysiol 2007;18:907 –13. 48. Wilber DJ, Pappone C, Neuzil P, De Paola A, Marchlinski F, Natale A et al. Comparison of antiarrhythmic drug therapy and radiofrequency catheter ablation in patients with paroxysmal atrial fibrillation: a randomized controlled trial. JAMA 2010;303:333–40. 49. Jaı¨s P, Cauchemez B, Macle L, Daoud E, Khairy P, Subbiah R et al. Catheter ablation versus antiarrhythmic drugs for atrial fibrillation: the A4 Study. Circulation 2008; 118:2498 –505.

Downloaded from by guest on November 14, 2014

17. Vogt J, Heintze J, Gutleben KJ, Muntean B, Horskotte D, No¨lker G. Long term outcomes after cryoballoon pulmonary vein isolation: results from a prospective study in 605 patients. J Am Coll Cardiol 2013;61:1707 –12. 18. Pappone C, Rosanio S, Augello G, Gallus G, Vicedomini G, Mazzoni P et al. Mortality, morbidity and quality of life after circumferential pulmonary vein ablation for atrial fibrillation. J Am Coll Cardiol 2003;42:185 –97. 19. Le Heuzey JY, de Ferrari GM, Radzik D, Santini M, Zhu J, Davy J-M. A short-term, double-blind, parallel-group study to evaluate the efficacy and safety of dronedarone versus amiodarone in patients with persistent atrial fibrillation: the DIONYSIS Study. J Cardiovasc Electrophysiol 2010;21:597 –605. 20. Rietbrock S, Heeley E, Plumb J, van Staa T. Chronic atrial fibrillation: incidence, prevalence, and prediction of stroke using the Congestive heart failure, Hypertension, Age.75, Diabetes mellitus, and prior Stroke or transient ischaemic attack (CHADS2) risk stratification scheme. Am Heart J 2008;156:57–64. 21. Sherman DG, Kim SG, Boop BS, Corley SD, DiMarco JP, Hart RG et al. Occurrence and characteristics of stroke events in the Atrial Fibrillation Follow-Up Investigation of Sinus Rhythm Management (AFFIRM) study. Arch Int Med 2005;165:1185 –91. 22. Wardlaw JM, Chappell FM, Stevenson M, de Nigris E, Thomas S, Gillard J et al. Accurate, practical and cost-effective assessment of carotid stenosis in the UK. Health Technol Assess 2006;10:1 –128. 23. Lip GYH, Edwards SJ. Stroke prevention with aspirin, warfarin and ximelagatran in patients with non-valvular atrial fibrillation: a systematic review and meta-analysis. Thromb Res 2006;118:321 – 33. 24. Miller CS, Grandi SM, Shimony A, Filion KB, Eisenberg MJ. Meta-analysis of efficacy and safety of new oral anticoagulants (dabigatran, rivaroxaban, apixaban) versus warfarin in patients with atrial fibrillation. Am J Cardiol 2012;110:453 –60. 25. Patten M, Maas R, Bauer P, Lu¨deritz B, Sonntag F, Dluzniewski M et al. Suppression of paroxysmal atrial tachyarrhythmias—results of the SOPAT trial. Eur Heart J 2004;25: 1395– 404. 26. Sullivan SD, Orme ME, Morais E, Mitchell SA. Interventions for the treatment of atrial fibrillation: a systematic literature review and meta-analysis. Int J Cardiol 2013;165: 229 –36. 27. Ku¨hlkamp V, Schirdewan A, Stangl K, Homberg M, Ploch M, Beck OA. Use of metoprolol CR/XL to maintain sinus rhythm after conversion from persistent atrial fibrillation. A randomised, double-blind, placebo-controlled study. J Am Coll Cardiol 2000; 36:139 –46. 28. Department of Health. 2011 –12 Payment by Results tariffs: http://www.dh.gov.uk/ prod_consum_dh/groups/dh_digitalassets/documents/digitalasset/dh_125398.xls (21 September 2012, date last accessed). 29. McGuire AJ, Raikou M, Whittle I, Christensen MC. Long-term mortality, morbidity and hospital care following intracerebral haemorrhage: an 11-year cohort study. Cerebrovasc Dis 2007;23:221 –8. 30. Jones L, Griffin S, Palmer S, Main C, Orton V, Sculpher M et al. Clinical effectiveness and cost-effectiveness of clopidogrel and modified-release dipyridamole in the secondary prevention of occlusive vascular events: a systematic review and economic evaluation. Health Technol Assess 2004;8:1 –196.

659