Rationale and impact
- Dual antiplatelet therapy for acute STEMI intended for primary PCI
- Antithrombin therapy during primary PCI for acute STEMI
- Complete revascularisation with PCI or culprit vessel only PCI for acute STEMI
- Drug-eluting stents
- Antiplatelet therapy for STEMI not treated with PCI
- Initial antithrombin therapy for unstable angina and NSTEMI
- Early invasive versus conservative management for unstable angina and NSTEMI
- Antiplatelet therapy for unstable angina and NSTEMI
- Antiplatelet therapy for people with an indication for anticoagulation
- Duration of beta-blocker treatment after an MI
These sections briefly explain why the committee made the recommendations and how they might affect practice.
Evidence was reviewed comparing the clinical effectiveness of clopidogrel, prasugrel and ticagrelor, each in combination with aspirin, at time-points of 30 days and 1 year. Prasugrel and ticagrelor were more effective than clopidogrel at both time-points. In a network meta-analysis of the 30‑day data, prasugrel was more effective than ticagrelor, although with some uncertainty around this conclusion. Prasugrel was more effective than ticagrelor at 1 year with noteworthy differences in all-cause mortality and re‑infarction. A detailed cost-effectiveness analysis was also performed incorporating data at both time-points with either prasugrel or ticagrelor being most cost effective in different scenarios using different clinical data. The results favouring prasugrel were driven by the ISAR‑REACT 5 trial that directly compared ticagrelor and prasugrel and these were considered the most relevant. The committee agreed that clinical evidence and cost-effectiveness results are directly applicable to the treatment of ST‑segment elevation myocardial infarction (STEMI) in the NHS, and recommended prasugrel for people with STEMI undergoing percutaneous coronary intervention (PCI).
The final wording of the recommendation reflects the wording of the summary of product characteristics for prasugrel. If there is particular concern about bleeding risk from prasugrel in a person aged 75 or over, either ticagrelor or clopidogrel might be used instead; although ticagrelor is the more cost effective of the two, it also carries a higher bleeding risk.
An exception was added for people needing anticoagulation for a separate reason (for example, ongoing atrial fibrillation). For these people, clopidogrel is preferred because of the high bleeding risk of full anticoagulation plus prasugrel.
In the UK, prasugrel is currently used less than ticagrelor or clopidogrel. The recommendation will therefore require a change in prescribing for most centres, but should be easily achievable. Prasugrel costs less than ticagrelor, but considerably more than clopidogrel, and although some areas will see a cost saving from switching to prasugrel from ticagrelor, the overall effect of this recommendation will be an increase in cost to the NHS.
When considering the evidence on the effectiveness of bivalirudin for people with acute STEMI undergoing primary PCI, the committee gave more weight to studies that were closest to current UK practice. These included studies that used bailout or selective, rather than routine, glycoprotein inhibitors (GPIs) and radial artery rather than femoral artery access. The committee concluded that there was no convincing difference between bivalirudin and the main alternative, heparin, in terms of mortality, and that bivalirudin is inferior to heparin in reducing the need for subsequent unplanned revascularisation. The committee discussed data on bleeding risk and agreed that there is no clinically significant difference between bivalirudin and heparin when radial access is used, but bivalirudin probably lowers the bleeding risk when access via the femoral artery is needed. The committee noted that heparin is cheaper than bivalirudin and easier to administer.
The committee agreed that the recommendations generally reflect current practice and are not expected to result in a substantial resource impact to the NHS in England.
Evidence showed that complete revascularisation with multivessel PCI reduced cardiovascular mortality, myocardial infarction (MI) and repeat revascularisation at 1 year, compared with culprit vessel only PCI for people with acute STEMI without cardiogenic shock. It was also associated with lower overall costs.
Although the evidence clearly favoured complete revascularisation, there was less certainty about the timing of the non-culprit procedure. There are a number of different possible approaches to multivessel PCI: undertaking multivessel revascularisation at the time of primary PCI; treating the culprit vessel during the primary procedure and then bringing the person back to the catheter laboratory for revascularisation of other vessels later in the index admission; or treating the culprit vessel during primary PCI, discharging the person and then electively readmitting them for further revascularisation. The committee agreed that multivessel PCI during the index admission should be considered, either at the time of primary PCI or later during the same admission. They were concerned that the clinical benefits may be lower and costs may be higher when people are discharged and readmitted, and noted that delaying treatment of the non-culprit lesions is worrying for patients. However, they agreed that the optimal timing within the index admission will depend on a number of variables and is best left to the discretion of the clinical team.
People with cardiogenic shock were excluded from these studies of multivessel PCI and the committee agreed that, in view of the results from the separate CULPRIT-SHOCK trial, it was not appropriate to recommend multivessel PCI for this group during the index admission.
Current practice is variable across centres and also within centres. Some offer multivessel PCI during the first procedure for acute STEMI but others may postpone this (either to later within the index admission or to a later readmission). Some operate on the culprit vessel only. The recommendations are therefore likely to result in a change in practice, but not for all centres or all professionals performing PCI. Because the recommendations allow for multivessel PCI to be either at the time of primary PCI or later within the index admission, they offer flexibility to accommodate situations in which there are a number of other people waiting for primary PCI. Healthcare professionals can move on to treat the next person after completing revascularisation of the culprit vessel, minimising the overall impact on primary PCI services.
There will be a resource impact for centres not currently undertaking multivessel PCI, because multivessel PCI has higher costs than culprit vessel only PCI. Audit data reported by MINAP (the Myocardial Ischaemia National Audit Project) between April 2016 and March 2017 show there were 33,797 cases of STEMI reported in England, Wales, Northern Ireland and the Isle of Man. It is estimated that around 30% will present with multivessel disease, which would be around 10,000 people. However, it is unclear for how many of these people multivessel PCI would be suitable. The change from current practice is likely to be cost saving overall because of the reduction in later revascularisation procedures.
Evidence from angiography studies showed that drug-eluting stents are less likely to fail than bare metal stents in terms of both recurrence of obstruction to the target vessel and the need for further revascularisation. The evidence also shows that drug-eluting stents may be beneficial in reducing deaths (all-cause and cardiac) and there is a reduced incidence of MI in the 3 years after revascularisation when drug-eluting stents are used. Costs of drug-eluting stents are higher than bare metal stents, but analyses using current cost and benefit data suggest that they are a cost-effective use of resources.
The use of drug-eluting stents has been slowly increasing over recent years in the UK and the most recent national audit data show that 91% of PCIs for acute coronary syndromes used stents and 97% of these used drug-eluting stents. The recommendation will therefore involve little change from current practice and will not have a substantial resource impact for the NHS in England.
The UK licence for prasugrel is for people with acute coronary syndrome who are proceeding to coronary angiography with a view to PCI. Although this is usual practice for most people with STEMI, for some people, either medical management without coronary revascularisation or coronary artery surgery are better options. Direct evidence in these patient groups was lacking; the evidence comparing the clinical effectiveness of clopidogrel and ticagrelor was largely for people receiving PCI. This showed convincing superiority of ticagrelor in reducing mortality (cardiac and all-cause) and in preventing re‑infarction and the need for future revascularisation procedures, although there was some evidence of an increased risk of bleeding complications. The committee agreed to recommend ticagrelor for people with STEMI having medical management unless they are at high risk of bleeding when clopidogrel or no second antiplatelet may be the safer option. However, the committee were aware that some of the excess bleeding risk comes from complications of a PCI procedure, which is not relevant to this particular group of people with acute coronary syndrome. They therefore made a recommendation to offer ticagrelor in most cases but to consider clopidogrel as an alternative when the bleeding risk is high.
In the UK, both ticagrelor and clopidogrel are currently used for STEMI that is managed without PCI. The recommendations require a change in practice for most, but not all, people who would otherwise receive clopidogrel. Ticagrelor costs considerably more than clopidogrel, and although the recommendations apply to a minority of people with STEMI, the effect will be an increase in cost to the NHS.
The 2010 NICE guideline on unstable angina and NSTEMI recommended fondaparinux rather than low molecular weight heparin for initial management. The recommendation was based mainly on evidence from a single large study (the OASIS‑5 study). This study showed a small risk of catheter thrombosis when fondaparinux was the only antithrombin used before angiography, and therefore the 2010 guideline recommended not to use fondaparinux when angiography is planned within 24 hours. The thrombosis risk was noted by the OASIS‑5 investigators before the study ended, and in the later phase of the study, people were given intravenous unfractionated heparin with fondaparinux during angiography; this appeared to remove the excess risk of catheter thrombosis. Two further small studies published after 2010 have confirmed that giving unfractionated heparin during angiography to people already receiving fondaparinux removed the excess risk of catheter thrombosis. The committee considered that unfractionated heparin is already used in this way in many centres, agreed with the 2010 guideline that fondaparinux is the most cost-effective option, and were able to remove the caveat about avoiding fondaparinux if catheterisation is planned within 24 hours. They recommended that fondaparinux should be given to people who are not at high risk of bleeding unless they are having immediate angiography. People receiving fondaparinux should be given additional systemic unfractionated heparin in the catheter laboratory.
Fondaparinux is already used before angiography in many centres in the UK, with additional unfractionated heparin given during the procedure. The recommendations will affect those centres currently withholding fondaparinux from people having angiography in the next 24 hours. Fondaparinux is a cheaper option than low molecular weight heparin so the recommendation could be cost saving in these centres.
The 2010 guideline on unstable angina and non-ST-segment elevation MI (NSTEMI) recommended a comprehensive assessment of baseline risk of adverse events. The committee agreed that this should influence the choice between early invasive intervention (coronary angiography, with PCI if indicated) and conservative management (initial medical management, proceeding to coronary angiography and PCI if there is evidence of recurrent ischaemia). Studies comparing these options show a short-term harm with an invasive strategy, but this is offset by the clinical benefits in the months following the procedure. A cost-effectiveness analysis found that routine early invasive intervention was cost effective in people at higher risk of adverse events, but conservative management was the most cost-effective option for people at lower risk. This was because overall health gains were greater in those at higher baseline risk.
Most of the evidence was already available at the time of the 2010 guideline, and the committee recognised that the data may be less applicable to modern practice than had been the case in 2010. Nonetheless, they agreed that early angiography should be the default recommendation for most people at intermediate or higher baseline risk of adverse outcomes. They accepted the previous committee's interpretation of the appropriate risk cut‑offs based on their detailed work mapping of the evidence to real-world UK risk data. However, the committee also recognised that the risk prediction models might be less applicable to the youngest people with unstable angina and NSTEMI who are relatively under-represented in the dataset, and therefore added a cautionary recommendation to this effect.
The committee noted that the 2010 guideline had recommended that angiography should be done within 96 hours of admission for those who are likely to benefit from an early invasive strategy. However, they considered this a conservative target and knew that angiography within 72 hours is now common practice. This allows time for a correct diagnosis, immediate stabilisation and treatment of symptoms, and transfer to a centre with PCI facilities if necessary. The available evidence does not permit a definitive statement about the optimal timing, but a recommendation to consider angiography within 72 hours was agreed.
The recommendations largely reflect current NHS practice. Although the timeframe for early invasive management has been reduced from 96 hours, 72 hours has been specified in the NICE quality standard for a number of years and a best practice tariff on the same basis was introduced in 2017. Audit data are only currently available from the same year as the introduction of the tariff and report that, of people who are admitted to a hospital that can perform angiography, 56% received angiography within 72 hours and 69% within 96 hours. The proportion receiving angiography within 72 hours is likely to be higher since the introduction of the best practice tariff. Performing angiography earlier is likely to result in a shorter hospital stay. The recommendations are unlikely to result in a substantial resource impact for the NHS.
The committee agreed that dual antiplatelet therapy should not be offered to people with chest pain before a diagnosis of unstable angina or NSTEMI is made. In their experience, dual antiplatelet therapy had caused harm (bleeding) in some people presenting with chest pain not caused by an acute coronary syndrome.
Evidence was reviewed comparing the clinical effectiveness of clopidogrel, prasugrel and ticagrelor, each in combination with aspirin, at time-points of 30 days and 1 year. A detailed cost-effectiveness analysis for people with unstable angina or NSTEMI undergoing PCI was performed incorporating these data. Although the overall conclusion was that prasugrel is a more effective agent than ticagrelor, which in turn is more effective than clopidogrel, there was considerable uncertainty around the cost-effectiveness results, with either prasugrel or ticagrelor being most cost effective in different scenarios using different clinical data. The results favouring prasugrel were driven by the ISAR‑REACT 5 trial, in which time to angiography was much shorter than is currently achieved in the UK for people with unstable angina or NSTEMI. This could cause practical difficulty in using prasugrel because its licence effectively prevents its use before angiography, and this could leave people with unstable angina or NSTEMI without dual antiplatelet therapy for several days. The committee therefore recommended either prasugrel or ticagrelor for people with unstable angina or NSTEMI intended for PCI, depending on individual circumstances. The final wording of the recommendation reflects the wording of the summary of product characteristics for prasugrel.
Although many people with unstable angina or NSTEMI proceed to PCI, for some medical management without coronary revascularisation or coronary artery surgery are better options. Prasugrel is not licensed in these circumstances. The evidence available for medical management shows better outcomes with ticagrelor than clopidogrel. This is consistent with results from the larger datasets for people having PCI. The committee therefore recommended ticagrelor for people with unstable angina or NSTEMI having either medical management without coronary revascularisation or coronary artery surgery. However, the committee also noted that clopidogrel may be the safer agent for people who are at high risk of bleeding but still need dual antiplatelet therapy, although this is based on evidence of higher bleeding risk with ticagrelor in people with an acute coronary syndrome generally rather than specific evidence from those at higher risk. They therefore made a recommendation to consider using clopidogrel in this situation.
In the UK, prasugrel is currently used less than ticagrelor or clopidogrel. The recommendations may therefore involve a change in practice for some centres. Prasugrel costs less than ticagrelor, but considerably more than clopidogrel, and although some areas will see a cost saving from switching to prasugrel from ticagrelor, others will see an increase where either prasugrel or ticagrelor is used instead of clopidogrel. The overall effect of these recommendations will be an increase in cost to the NHS.
The committee noted that current practice is to use dual antiplatelet therapy at the time of PCI, and found no evidence to recommend changing this practice for people who are on an anticoagulant at the time of admission. In practice, the anticoagulant will often be suspended for a short period (perhaps using heparin cover). The evidence available showed that continuing dual antiplatelet therapy plus an anticoagulant after the acute PCI phase increases the risk of bleeding complications, and so a recommendation raising awareness of this issue was included. The committee therefore agreed that either aspirin or the second antiplatelet agent should be stopped, but unfortunately there was no evidence to show at what point this should happen.
In the absence of any conclusive data, recommendations for treatment after the initial phase were based on the knowledge and experience of the committee. For people who have had PCI and stent insertion, they agreed that it would be safest to combine an anticoagulant with a potent antiplatelet agent (clopidogrel), whereas for those who have had medical management or had angioplasty without stenting, the anticoagulant should be combined with aspirin. There was not enough evidence for the committee to recommend a particular anticoagulant.
Current practice is variable, with people taking different combinations of antiplatelets and anticoagulants. The number of people affected is small. It is estimated that between 5% and 15% of people with an acute coronary syndrome will have an indication for oral anticoagulation. The recommendations are mostly unchanged from the 2013 guideline and the minor changes that have been made are unlikely to result in a substantial resource impact for the NHS in England.
There was no direct evidence on the optimal duration of beta-blocker treatment for people who have had an MI but do not have reduced left ventricular ejection fraction. The 2013 guideline recommended beta-blocker treatment for at least 12 months. In the absence of any conclusive evidence, the committee agreed that they could not recommend a definite time for stopping treatment. However, they agreed that healthcare professionals should discuss the absence of clear evidence for benefit of continuing beyond 12 months with people taking beta-blockers after an MI who have normal left ventricular function. This should prompt a personalised approach to stopping or continuing beta-blockers based on the person's attitude to risk and experience of side effects.
Beta-blockers are currently offered for at least 12 months after an MI to people without reduced left ventricular ejection fraction. Audit data show that around 97% of people with MI are discharged on beta-blockers. A discussion of the absence of clear evidence for benefit of continuing treatment beyond 12 months is likely to lead to more people deciding to stop treatment at this point. Any reduction in prescriptions for beta-blockers will be cost saving.