Dabigatran etexilate for the prevention of stroke and systemic embolism in atrial fibrillation

4.1 The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of dabigatran, having considered evidence on the nature of atrial fibrillation and the value placed on the benefits of dabigatran by people with the condition, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources.

4.2 The Committee heard from the clinical specialists and patient experts that the current standard treatment for the prevention of stroke and systemic embolism in people with atrial fibrillation is warfarin, and that because of its lower efficacy, aspirin is used only in people for whom warfarin is unsuitable. The Committee also heard that warfarin, although an effective treatment, is associated with a number of problems. The main concerns for people with atrial fibrillation were fear of having a stroke and anxiety about the difficulty of keeping the INR within the satisfactory therapeutic range. The Committee heard from the patient experts that stroke is a major concern for people with atrial fibrillation and that stroke severity is usually greater in this group than in people who have strokes from other causes. The patient experts also highlighted that many people taking warfarin are outside their target therapeutic INR range at any one time and that warfarin, unlike dabigatran, is associated with a number of inconveniences that make adherence difficult. These include numerous food and drug interactions that can have an impact on people's work, social and family life, and regular monitoring and dose adjustments that can cause disruption and inconvenience. The Committee accepted the limitations of warfarin therapy, and the considerable effect that it may have on the lives of the people who take it, and recognised the potential benefits of dabigatran for people with atrial fibrillation.

4.3 The Committee considered the clinical-effectiveness data from the RE-LY trial comparing dabigatran with warfarin. It noted that this formed most of the clinical-effectiveness evidence in the manufacturer's submission and was the largest published trial in people with atrial fibrillation. The Committee considered that the RE-LY trial was of good quality but noted that a key uncertainty highlighted by the ERG was the generalisability of the results to people diagnosed with atrial fibrillation in the NHS. The Committee noted that the definition of moderate to high risk of stroke in the RE-LY trial was different from the definition used in NICE clinical guideline 36 and did not include people aged 65 years and over with no additional risk factors for stroke, resulting in a higher risk profile in the trial than in the general population eligible for anticoagulation prophylaxis. However, the Committee was persuaded by the clinical specialists that the RE-LY trial included a broad range of people that reflected those seen in UK clinical practice and that the results were applicable to a wide range of people with atrial fibrillation. The Committee concluded that the population included in the trial was appropriate and broadly relevant to UK clinical practice.

4.4 The Committee considered the results of the RE-LY trial. It noted that dabigatran 150 mg twice daily was associated with a statistically significantly lower incidence of stroke or systemic embolism, ischaemic stroke and vascular mortality compared with warfarin, but that there were no statistically significant differences in these outcomes between dabigatran 110 mg twice daily and warfarin. It also noted that both doses of dabigatran were associated with an increased risk of acute myocardial infarction compared with warfarin but that this was not statistically significant. The Committee heard from the clinical specialists that this reflected a small absolute difference in the incidence of acute myocardial infarction between the treatment groups, but it was unclear whether this was because of a protective effect of warfarin or a negative effect of dabigatran treatment, and that the effects did not appear to translate into an increased vascular mortality risk. The Committee concluded that dabigatran 150 mg twice daily was more clinically effective than warfarin in reducing the risk of stroke or systemic embolism, ischaemic stroke and vascular mortality and that this represented an important development for people with atrial fibrillation. It also concluded that the lower 110 mg dabigatran twice-daily dose had shown non-inferiority to warfarin.

4.5 The Committee considered the results of the manufacturer's subgroup analyses. It was aware, however, that the manufacturer's analyses by age had been defined post hoc and it therefore considered that the results should be interpreted with caution. The Committee also considered the results of the manufacturer's pre-planned analyses of people naive to vitamin K antagonists and people who have previously used vitamin K antagonists. It noted that dabigatran 150 mg twice daily was associated with a statistically significant reduction in the incidence of stroke or systemic embolism compared with warfarin in both vitamin K antagonist-naive and vitamin K antagonist-experienced subgroups, but dabigatran 110 mg twice daily did not show a statistically significant reduction in either group. The Committee concluded that dabigatran 150 mg twice daily showed increased efficacy compared with warfarin in people with atrial fibrillation irrespective of their previous exposure to vitamin K antagonists.

4.6 The Committee discussed the effectiveness of dabigatran compared with warfarin according to INR control. It noted the evidence presented by the ERG that people with good INR control with warfarin may not gain additional clinical benefit by taking dabigatran. However, the clinical specialists emphasised the importance of the significantly lower rates of intracranial haemorrhage and haemorrhagic stroke associated with both doses of dabigatran compared with warfarin in the RE-LY trial, and that this effect is maintained in people with good INR control. The Committee heard that haemorrhagic stroke and intracranial haemorrhage have devastating and life-threatening consequences and concluded that the lower rates associated with dabigatran represent an important advance in the treatment of atrial fibrillation alongside reduction in ischaemic stroke. It concluded that this applied to all patients with atrial fibrillation, including those with good INR control, and that there were also benefits of taking a treatment that didn't need INR monitoring or dietary restriction.

4.7 The Committee considered the additional adverse events reported in the RE-LY trial. It noted that both doses of dabigatran were associated with statistically significant reductions in the incidence of life-threatening bleeds compared with warfarin. However, it also noted that the incidence of gastrointestinal bleeding, in contrast to cerebral haemorrhage, was statistically significantly higher for both doses of dabigatran, and the comment from the manufacturer that this may be the result of a local effect of the orally administered drug on the gastrointestinal mucosa. Dabigatran 150 mg twice daily was associated with a statistically significantly higher incidence of major and life-threatening gastrointestinal bleeding. The Committee noted that even small changes in total gastrointestinal bleeding rates might have a substantial impact on the provision of services and that major gastrointestinal bleeding is associated with a significant mortality risk. The Committee concluded that treatment with dabigatran resulted in more gastrointestinal bleeding than warfarin, but also recognised the particular importance of the effects of dabigatran on reducing the risk of haemorrhagic stroke and intracranial haemorrhage for people with atrial fibrillation when compared with warfarin.

4.8 The Committee was aware that health-related quality of life data were collected in a sub-study of the RE-LY trial. It noted that baseline utility values for people with atrial fibrillation were derived from the sub-study. The Committee agreed that because the sub-study was reasonably representative of the overall RE-LY population, this approach was appropriate.

4.9 The Committee considered the manufacturer's economic model and the critique and exploratory analyses performed by the ERG. The Committee considered the utility values used in the model and noted that it was unclear how the utility values relating to the effect of stroke were derived. However, the Committee agreed with the ERG that the general approach taken by the manufacturer to estimate the lifetime cost effectiveness of dabigatran was appropriate.

4.10 The Committee noted that the manufacturer presented a single-dose model, and a sequential regimen model in which people younger than 80 years began treatment with dabigatran 150 mg twice daily, and at the age of 80 years were switched to dabigatran 110 mg twice daily. As the summary of product characteristics for dabigatran excludes people older than 80 years from treatment with dabigatran 150 mg twice daily because of additional risks in this group, the Committee concluded that the sequence of dabigatran 150 mg twice daily followed by dabigatran 110 mg twice daily once people reach 80 years would be the only regimen appropriate for the assessment of the cost effectiveness of dabigatran relative to warfarin in the whole eligible UK population.

4.11 The Committee heard from the ERG that the relative risks used to inform the manufacturer's original sequential regimen model were derived from people in the younger than 80 years and older than 80 years subgroups of the RE-LY trial that were defined post hoc. It also heard that using relative risks from the whole RE-LY trial population would be more appropriate to determine reliable effectiveness estimates for the dabigatran sequence. Therefore the Committee asked the manufacturer to submit a re-analysis of the data for discussion at the second Appraisal Committee meeting using the relative risks from the whole RE-LY trial population.

4.12 At the first Appraisal Committee meeting, the Committee noted that the ERG had highlighted a number of uncertainties relating to assumptions used in the manufacturer's economic model. First, the Committee noted the ERG's view that an analysis based on an older patient cohort with a lower risk of stroke using data reported by Gallagher et al. (2008) would be more representative of people with atrial fibrillation in the UK than the cohort from the RE-LY trial used by the manufacturer. The Committee accepted that there was uncertainty around which cohort most realistically reflected the population of people with atrial fibrillation in the UK.

4.13 Second, the Committee noted that the ERG questioned whether disability and mortality were independent of the treatment received. The Committee heard from the clinical specialists that the manufacturer's assumption that a stroke would be less severe after treatment with dabigatran than warfarin was plausible and that there is evidence that both the incidence and the severity of stroke may vary according to the treatment received. The Committee also noted the ERG's views about disutility of dabigatran and the inclusion of dyspepsia management costs throughout treatment (see sections 3.30 and 3.31). The Committee agreed that including all of these assumptions would be a more conservative approach.

4.14 Third, the Committee noted the ERG's view that the cost of INR monitoring had been overestimated in the manufacturer's model. The Committee heard from the clinical specialists that the introduction of dabigatran would not result in complete closure of anticoagulation services with release of all the funding, and the manufacturer's estimate (£414.90) was likely to be too high. It also heard that INR monitoring costs varied in different settings and could not be quantified precisely. The Committee agreed that exploring the effect of assuming the alternative INR monitoring costs put forward by the ERG (£115.14, £241.54, £279.36), in addition to the cost assumed in the manufacturer's submission, would enable it to make a more accurate judgement about the cost effectiveness of dabigatran.

4.15 Finally, the Committee noted the ERG's comments that the cost effectiveness of dabigatran compared with warfarin varied substantially according to level of INR control in those already being treated with warfarin. In the appraisal consultation document, the manufacturer of dabigatran was therefore asked to provide further analyses addressing the uncertainties outlined in sections 4.11 to 4.15.

4.16 The Committee discussed the manufacturer's revised analyses and the critique and the exploratory analyses performed by the ERG. The Committee noted that the manufacturer's revised analysis included the relative risks from the whole RE-LY trial population rather than from the post hoc subgroup analysis and had explored the effect of varying the cost of INR monitoring as requested. It also noted that the manufacturer's revised analysis incorporated an INR monitoring cost of £241.54 in its base case as opposed to £414.90 in the original submission. The Committee was aware that comments received during the consultation largely agreed that INR monitoring costs are likely to be higher than the ERG's lower estimate of £115.14 and possibly higher than £414.90 in some cases. The Committee accepted the manufacturer's approach, acknowledging that although INR costs may vary widely, this assumption was reasonable.

4.17 The Committee discussed the manufacturer's approach to including the ERG's other preferred assumptions in the revised analysis (see section 3.33). The Committee noted that the Gallagher et al. (2008) data on atrial fibrillation had not been incorporated. However, the Committee accepted the manufacturer's rationale and the supporting views of the ERG for using General Practice Research Database data instead (see section 3.39). The Committee noted that, in its revised analyses, the manufacturer had incorporated the ERG's preferred assumptions about dyspepsia management costs throughout treatment, disability and mortality risks being treatment independent, and disutility associated with dabigatran. It further noted that combining all of these assumptions together with an INR monitoring cost of £241.54 resulted in an ICER for dabigatran of £17,700 per QALY gained for the full sequential regimen in people starting treatment when younger than 80 years and £18,400 per QALY gained in people starting treatment at 80 years and older, compared with warfarin. Finally, the Committee noted that the ERG's analysis, which included all of the requested assumptions, an INR monitoring cost of £241.54, and the corrected values for ischaemic stroke and disability rates (see section 3.39) resulted in an ICER of £18,900 per QALY gained for the sequential regimen in people starting treatment younger than 80 years, compared with warfarin. The Committee concluded that this was broadly in line with the manufacturer's estimate and that the ICERs presented by the manufacturer were robust to the changes requested. The Committee therefore accepted the manufacturer's approach and concluded that the most plausible ICERs for the whole population eligible for dabigatran were within the range normally considered a cost-effective use of NHS resources, being less than £20,000 per QALY gained.

4.18 The Committee discussed comments from consultees that suggested it may be appropriate to recommend dabigatran for use only in people with atrial fibrillation whose INR is not well controlled on warfarin. The Committee was satisfied that the technology was a cost-effective treatment for the whole patient group. It noted that robust evidence of differential clinical effectiveness and cost effectiveness, with clear justification of the threshold level chosen, would be needed to select out a subgroup, based on INR control, for whom dabigatran would not be recommended.

4.19 The Committee was aware of the need for guidance to apply equally to those already on warfarin and to those newly diagnosed with atrial fibrillation. The Committee noted that, for people newly diagnosed but not already taking an anticoagulant, any stratification of the population according to INR control would mean that all patients would have to try warfarin for at least a few months to assess whether the INR was well controlled and to estimate the time in therapeutic range. The Committee heard from clinical specialists that many of the significant complications of warfarin therapy are experienced in the first months of treatment before the dose is established and stabilised. The Committee accepted therefore that a large number of people having a trial of warfarin at initial diagnosis could be expected to switch to dabigatran. It also accepted that it was not reasonable to expect all patients to try warfarin first, with the associated risks, for the purpose of selecting out a subgroup for whom dabigatran was less cost effective.

4.20 The Committee was also aware of the estimates of the time that the INR in people already taking warfarin would need to be in the target range for the ICERs for dabigatran compared with warfarin to be above £30,000 per QALY gained. Assuming an INR monitoring cost of £241.54 per annum, the manufacturer and ERG estimated an average of 83% to 85% and 75% to 76% of the time respectively. The Committee noted that this would apply to only a proportion of the whole population. The Committee was aware that the average time spent in therapeutic range for the UK centres in the RE-LY trial was 72%, and in the UK-based study by Jones et al. (2005) there was an average time in therapeutic range of 67.9%. It noted the ERG's analysis that explored the effects of time in therapeutic range on the cost effectiveness of dabigatran compared with warfarin. This calculated the ICER for the people with the best-controlled INR (that is, within range 83.7% of the time) at £47,000 per QALY gained. However, this figure incorporated INR monitoring costs of £241.54 (per annum) and the ICER reduced considerably if higher INR monitoring costs of £414.90 per annum were used. The Committee concluded that evidence for stratifying by INR control was insufficient to exclude the minority of people with very good control from the recommendation of dabigatran as a potential treatment option, and that the ICER for the whole population should be the basis of the recommendation.

4.21 The Committee was mindful of the higher gastrointestinal bleeding rates associated with dabigatran and of the relatively short-term safety data compared with the established standard of care, warfarin. It was also mindful that for those with very well-controlled INR on warfarin, the clinical benefits are likely to be less than for those with poorly controlled INR. The Committee therefore concluded that the decision about whether to start treatment with dabigatran in people with atrial fibrillation should be made after an informed discussion between the responsible clinician and the person about the safety risks and benefits of dabigatran compared with warfarin. It also concluded that, for people currently receiving warfarin, the potential risks and benefits of switching to dabigatran should be considered in light of their level of INR control.

4.22 The Committee considered whether there were any equalities considerations affecting population groups protected by equality legislation and concluded that there were no equality issues relating to this appraisal that needed addressing in the guidance.