Rivaroxaban for preventing adverse outcomes after acute management of acute coronary syndrome

Clinical effectiveness

4.4 The Committee considered the clinical‑effectiveness data from the ATLAS‑ACS 2‑TIMI 51 trial comparing rivaroxaban in combination with aspirin plus clopidogrel or aspirin alone against aspirin plus clopidogrel or aspirin alone. It noted that this formed the basis of the clinical‑effectiveness evidence in the company's submission. The Committee considered that the ATLAS‑ACS 2‑TIMI 51 trial was of good quality but noted that a key issue highlighted by the ERG was the generalisability of the results to people diagnosed with acute coronary syndrome in England. The Committee noted that people with acute coronary syndrome in clinical practice are usually older than those patients who were recruited to ATLAS‑ACS 2‑TIMI 51. The Committee also noted that patients in the trial could be considered a relatively low‑risk population because they had little comorbidity, lower than usual use of percutaneous coronary intervention and included a relatively small proportion of people aged over 75 years or with impaired renal function. The Committee heard from the clinical experts that the average age difference between the trial population and patients seen in clinical practice was not likely to be clinically significant and that patients recruited to ATLAS‑ACS 2‑TIMI 51 were similar in terms of baseline characteristics to those recruited to other trials in acute coronary syndrome. The Committee was persuaded that the issue of generalisability was similar across all trials in this condition, and concluded that the results of ATLAS‑ACS 2‑TIMI 51 were relevant to routine clinical practice.

4.5 The Committee considered the results of the ATLAS‑ACS 2‑TIMI 51 trial. The Committee noted that the company had presented clinical‑effectiveness results for the overall trial population and also for a post hoc subgroup analysis of patients with elevated cardiac biomarkers (STEMI and NSTEMI) and no history of a stroke or TIA (80% of the total trial population). The Committee was aware from the company that this post hoc subgroup analysis was carried out at the request of the European Medicines Agency (EMA). The Committee noted that this post hoc subgroup analysis (referred to as the licensed population by the company) provided efficacy results that tended to be more favourable to rivaroxaban than the results from the overall trial population. However, it acknowledged that these differences were unlikely to be sufficiently large as to have an impact on the overall decision as to whether rivaroxaban was clinically and cost effective in its licensed indication.

4.6 The Committee discussed the numerical inconsistencies between the 2 dose groups in the trial (2.5 mg twice daily and 5 mg twice daily) for the individual components of the composite efficacy end point that had been identified by the ERG (see section 3.31). It was aware that for some individual outcomes the 2.5 mg twice‑daily dose appeared to have a greater efficacy than the 5 mg twice‑daily dose. The ERG considered that it was unlikely that the 2.5 mg twice‑daily dose would be more clinically effective than 5 mg twice daily and suggested that the results from both doses combined were more plausible than those of the individual doses. The Committee heard from the company that the summary of product characteristics for rivaroxaban specified the 2.5 mg twice‑daily dose and that the EMA had based its decision for this dose based on the balance of risk and benefits of the 2.5 mg twice‑daily dose, compared with those of the 5 mg twice‑daily dose. The Committee noted the lower bleeding risk associated with the 2.5 mg twice‑daily dose compared with the 5 mg twice‑daily dose. While it acknowledged that there were numerical inconsistencies in the efficacy results between the 2.5 mg and 5 mg twice‑daily arms, it concluded that the efficacy data from the 2.5 mg twice‑daily rivaroxaban arm were the most relevant for decision‑making because it is the licensed dose of rivaroxaban for this indication.

4.7 The Committee discussed the high discontinuation rates from the trial (see section 3.10). The Committee was aware that 15.5% of the total randomised population prematurely discontinued from the trial and that the discontinuation rate was higher in ATLAS‑ACS 2‑TIMI 51 than in other similar randomised trials in patients with acute coronary syndrome. The Committee heard from the clinical experts that high discontinuation rates were common in trials in patients with acute coronary syndrome and that this is replicated in the adherence rates seen in clinical practice, because current treatment protocols mean that people who had acute coronary syndrome are already taking 5 separate medications. The clinical experts highlighted that clinicians are mindful of the effect on patient adherence of adding any additional treatments to those already prescribed. The Committee acknowledged that the discontinuation rates in ATLAS‑ACS 2‑TIMI 51 were high but that this was a concern for other trials carried out in people with acute coronary syndrome, and that it was also an issue that is seen in clinical practice.

4.8 The Committee discussed the missing data from people who withdrew or were lost from the trial (see section 3.10). The Committee was aware of the ERG's concerns that missing data may result in informative censoring (that is, patients who drop out, and whose data are therefore censored, have different outcomes to those who remain in the trial) leading to bias. The Committee was aware from the company that extensive efforts had been made to trace trial participants, to clarify reasons for withdrawal and to find out if they had died. The company stated that this had reduced the proportion of patients for whom vital status was unknown to 3.2% of people who had been recruited to the trial. The company explained that it had been unable to obtain the data for the remaining participants for whom vital status was unknown, because of restrictions imposed by the countries in which the people lived. The clinical experts acknowledged that this was a problem for many multinational trials in people with acute coronary syndrome but expressed their concern about the level of missing data, given the bleeding risks associated with rivaroxaban added to antiplatelet therapy. The Committee understood that the EMA's assessment report had not discussed this issue in detail but acknowledged the clinical experts' concerns regarding the missing data. The Committee concluded that the missing data from those who withdrew or were lost from the trial remained of concern, but the magnitude of any bias introduced by informative censoring was unknown.

4.9 The Committee considered the effectiveness of rivaroxaban in combination with aspirin plus clopidogrel or with aspirin alone, compared with aspirin plus clopidogrel or aspirin alone in the licensed population (that is, people with elevated cardiac biomarkers and without a history of stroke or TIA). The Committee noted that rivaroxaban 2.5 mg twice daily in combination with aspirin plus clopidogrel or with aspirin alone reduced the composite risk of myocardial infarction, stroke and death from cardiovascular causes by 20% compared with aspirin plus clopidogrel or with aspirin alone (see table 2: ALL strata, 2.5 mg twice‑daily rivaroxaban dose). The Committee understood that this composite reduction in risk was driven by reductions in cardiovascular death and myocardial infarction. The Committee concluded that rivaroxaban 2.5 mg twice daily in combination with aspirin plus clopidogrel or with aspirin alone was more effective than aspirin plus clopidogrel or aspirin alone for preventing myocardial infarction and death from cardiovascular causes in people with acute coronary syndrome and elevated cardiac biomarkers.

4.10 The Committee discussed the concerns about safety and adverse effects associated with rivaroxaban. The Committee was aware that the results for the licensed population showed that there was a dose‑dependent increase in the rate of non‑CABG TIMI major bleeds (major bleeding assessed using 'Thrombolysis in Myocardial Infarction' criteria not related to coronary‑artery bypass grafting) for rivaroxaban added to antiplatelet therapy compared with antiplatelet therapy alone. It noted that in the 2.5 mg rivaroxaban twice‑daily arm of the ATLAS‑ACS 2‑TIMI 51 study, there was a 3 times greater risk of non‑CABG TIMI major bleeding with rivaroxaban in combination with aspirin plus clopidogrel or aspirin compared with these antiplatelet therapies alone. The Committee acknowledged that all antiplatelet and anticoagulant treatments have an associated risk of bleeding but noted the comments it had heard from the clinical and patient experts that the risk of bleeding was a key consideration when deciding on a particular treatment. The Committee was aware that no data had been presented by the company or other stakeholders comparing the risk of bleeding with rivaroxaban in combination with antiplatelet agents compared with other treatment regimens with antiplatelet agents such as ticagrelor and prasugrel, because these were not included in the final scope. The Committee was therefore unable to compare the effectiveness and safety profile of a treatment strategy in which rivaroxaban is added to clopidogrel and aspirin at least 24 hours after admission to hospital, with strategies in which ticagrelor and prasugrel are added to aspirin from the start of treatment. The Committee concluded that treatment with rivaroxaban resulted in more non‑CABG‑related major bleeding than aspirin plus clopidogrel or aspirin alone, but also recognised the particular importance of the effects of rivaroxaban in reducing the risk of myocardial infarction and death from cardiovascular causes. The Committee also concluded that clinicians should undertake a careful assessment of whether the bleeding risk is outweighed by the benefits of rivaroxaban in preventing further ischaemic events for individual patients when deciding whether to start or continue treatment. The Committee noted that the summary of product characteristics states that treatment should be regularly evaluated and, in particular, that careful consideration should be given to whether treatment is continued beyond 12 months because experience of treatment with rivaroxaban up to 24 months is limited.

Cost effectiveness

4.11 The Committee considered the company's economic model and the review and exploratory sensitivity analyses performed by the ERG. The Committee noted that the ICERs presented by the company in its base‑case analysis and in the ERG's exploratory sensitivity analyses were all lower than £10,000 per QALY gained. The Committee was aware of the ERG's concerns about the structure of the company's economic model and, in particular, that the model is relatively inflexible. This meant that the ERG could not carry out all the exploratory analyses that it deemed potentially relevant. These included amendments to the hazard ratio for fatal bleeds and adjusting for the possibility of informative censoring. The Committee noted that, to explore its concerns about informative censoring, the ERG had undertaken a 'crude' exploratory analysis to explore the effects on the ICER of increasing the number of patients who experienced a fatal bleed with rivaroxaban. The Committee was aware that the analysis showed that even if rivaroxaban 2.5 mg twice daily caused as many as 20 more fatal bleeds than observed in the trial, the estimated ICER remained below £10,000 per QALY gained. The Committee concluded that despite the inflexibility of the company's economic model and the resulting constraints on the ERG's ability to undertake further exploratory analyses, the ICERs presented were a suitable basis for decision‑making on the cost effectiveness of rivaroxaban in addition to clopidogrel plus aspirin or with aspirin alone.

4.12 The Committee considered the ICER for the 2.5 mg dose of rivaroxaban in combination with aspirin plus clopidogrel or with aspirin alone, compared with aspirin plus clopidogrel or aspirin alone in patients with acute coronary syndrome with elevated cardiac biomarkers (STEMI or NSTEMI) and no history of stroke or TIA. The Committee noted that the company's base‑case ICER was £6203 per QALY gained, and the ERG's preferred base‑case estimate was £5622 per QALY gained. It accepted that there is uncertainty about the validity of the results based on ATLAS‑ACS 2‑TIMI 51 because of the risk of bias resulting from missing data and informative censoring. However, the Committee considered that the ICERs presented were all within the range that could be considered cost effective and that the results of the ERG's exploratory sensitivity and scenario analyses suggested that the ICER was unlikely to increase to the extent that it would become unacceptable. It concluded that rivaroxaban can be considered a cost‑effective use of NHS resources.

4.13 The Committee was aware that there is an increased risk of bleeding when rivaroxaban is added to aspirin or aspirin plus clopidogrel and it would be important for clinicians to carefully assess a person's individual bleeding risk and for patients to have an informed discussion with their clinician about the potential risks and benefits before starting treatment with rivaroxaban. The Committee noted that the summary of product characteristics states that after initiation, there should be regular assessment of the risks and benefits of continuing treatment with rivaroxaban and extension of treatment beyond 12 months should be done on an individual patient basis as experience up to 24 months is limited. The Committee concluded that it was appropriate that a formal assessment of whether to continue treatment should be made no later than 12 months after starting rivaroxaban.

Summary of Appraisal Committee's key conclusions