The new European Society of Cardiology guidelines on myocardial revascularisation: an appraisal
- Correspondence to Dr J W Deckers, Thoraxcentre, Room Bd-420, Dr Molewaterplein 40, 3015 RD Rotterdam, The Netherlands;
- Accepted 13 September 2011
- Published Online First 14 October 2011
The latest European Society of Cardiology (ESC) guidelines on myocardial revascularisation are reviewed. The nearly 300 recommendations make it difficult to apply them in their totality. The authors would propose 20–30 recommendations per guideline based on sound clinical evidence. Also, the scope of the current guidelines is very wide as it includes topics already incorporated in other guidelines, such as strategies for pre-intervention diagnosis and imaging as well as on secondary prevention. Some recommendations in the new guidelines are sensible and will not be disputed. In particular, the encouragement of a balanced multidisciplinary decision process (the ‘heart team’) is welcome. Although coronary revascularisation in unstable high risk patients is well accepted, this is less the case for the low risk patient with chest pain. This issue is controversial and a balanced discussion of the pros and cons of percutaneous coronary intervention is missing. Despite convincing evidence indicating lack of benefit of percutaneous coronary intervention for chronic total occlusion, this procedure is not discouraged. Lastly, most committee members were interventional cardiologists or cardiac surgeons. Guideline committees should be representative of the whole group of professionals as the interpretation of the evidence by specialists may be biased. There may be a role for procedure oriented guidelines but, in that case, the items at issue should remain confined to matters directly related to technical aspects of the procedure.
Guidelines describe the current state of the art by formulating the current evidence base and recommendations for their application in clinical practice. On the basis of this premise, we reviewed the latest European Society of Cardiology (ESC) guidelines on myocardial revascularisation.1 The new document is a follow-up on the percutaneous coronary intervention (PCI) guidelines from 2005, which have previously been discussed in this journal.2 3 The current document is a joint effort by interventional cardiologist as well as cardiovascular surgeons, and thus considers both percutaneous and surgical revascularisation modalities. As a consequence, a ‘heart team,’ a more formal multidisciplinary team approach to coronary revascularisation, is advocated. As such, the new approach goes a long way towards establishing a sound basis for clinical decision making for patients with coronary insufficiency, and we—and others before us—have no doubt that it will fulfil that purpose.4 Nevertheless, some topics covered in the guidelines require clarification. In this review, we comment on those topics with the aim of improving future versions of such documents.
Scope of the guidelines
The new guidelines, which comprise 13 sections, provide recommendations for percutaneous as well as surgical revascularisations. In reality, however, it goes further. For instance, the various strategies for pre-intervention diagnosis and imaging are also included (chapter 5). A useful chapter without doubt, but we question the appropriateness of including this issue in this procedure oriented document. This is relevant as the worth of diagnostic tests has previously been described in the ESC guidelines on stable angina.5 Similarly, chapter 13 of the current guidelines provides recommendations for secondary prevention. That subject has been covered extensively in the guidelines on CVD prevention.6 The reader is thus left with the question of where these topics will be covered in future versions of these guidelines, and if the recommendations are different, which ones are to be followed. The scope of the current new guidelines is thus too wide.
Number and levels of recommendations
We are of the opinion that IA (‘strongly recommended’) and IIIA (‘strongly discouraged’) recommendations should be issued only in the context of relevant changes in clinical outcomes resulting from performing (or withholding) specific interventions that were rigorously scrutinised and tested, preferably in randomised clinical trials. Unfortunately, this view was not taken by the current committee, with a very large number of recommendations based on ‘expert opinion’ as a result. Leaving out such assessments would not only go a long a way towards strengthening the value of the recommendations based on sound clinical evidence but, in addition, would make it much easier to apply the guidelines in clinical practice. We also believe that some of the current recommendations are not appropriate or incorrect. For instance, the stratification scores used to estimate periprocedural intervention risks (chapter 4, table 3) have been given specific recommendations. However, we are unaware of any studies that have shown that actual patient outcome will improve as a result of the use of such scores and levels of recommendations are thus not applicable. The same is true for many other recommendations. For instance, the recommendations depicted in table 7 (chapter 5), which compare the value of different imaging techniques in (sub) groups of patients, are not based on studies in which these techniques were directly compared with each other.
However, most recommendations in the new guidelines are quite sound and will not lead to much controversy. For instance, mechanical PCI of the infarct related vessel as early as possible in ST segment elevation myocardial infarction, the preference for surgical revascularisation intervention in the case of multivessel disease, poor left ventricular function or diabetes, appropriate risk stratification and revascularisation in patients with unstable clinical conditions, as well as physiological assessment of the importance of intermediate coronary lesions prior to intervention, all seem very appropriate and sound. In particular, the encouragement of a balanced multidisciplinary decision process (‘heart team’) is a very welcome addition, which was given appropriate credit in a recent editorial by Taggart et al in this journal.4
PCI in stable angina and chronic total occlusion
Although in suitable high risk patients coronary revascularisation in unstable clinical conditions is generally well accepted, this is not true for low risk patients with stable chest pain. This is a controversial issue and, given the large number of procedures performed for this indication, not without reason. This topic is covered in chapter 6, and the reader would expect a concise but objective textual review of this subject. This section is, however, quite short and does not provide such an evaluation. However, from review of tables 8 and 9 in this chapter, it becomes clear that the authors do recommend PCI for this indication, with a (noteworthy) class IC (expert opinion) indication for one or two vessel disease not involving the proximal left anterior descending. Although the lack of references given for this recommendation is in line with the given C, the results of the randomised trials performed in these patients do not support this recommendation (see below). The table gives a IIa B recommendation for PCI when the proximal left anterior descending is involved. This recommendation is based on two meta-analyses and two other studies comparing coronary artery bypass grafting with PCI and are therefore not applicable. In fact, all but two references in this table compare the results of PCI with surgical revascularisation, and therefore do not address the issue of medical therapy versus PCI.
We realise that an objective assessment of the appropriateness of PCI in stable angina is not easy and is confounded by specific issues. Observational data showing the superiority of revascularisation over medical treatment are not helpful because of significant treatment bias8 and—unfortunately—the actual evidence base is quite small.
A summary of all randomised controlled clinical trials comparing PCI with medical therapy for stable angina is given in table 2 and figure 1. This analysis differs from previous analyses because we have excluded trials performed in the early 1990s.18–22 In that era, medical treatment was more or less placebo: for instance, statins and ACE inhibitors—currently considered standard medical treatment—were not available or not used. In addition, we excluded trials that investigated early post-myocardial infarction patients.23–27 Overall, the combined results of these trials provide limited evidence for the superiority of PCI over medical therapy in stable chest pain syndromes. Of course, we acknowledge that some of these trials have come under criticism, and that patients switched from medical treatment to PCI (or coronary artery bypass graft) (typically in about 25%). On the other hand, all subjects considered in these trials qualified for intervention on the basis of their specific coronary anatomy and were therefore quite selected. This is different from the practical situation in which the patient with chest pain but with unknown coronary anatomy is contemplated. In the latter situation, eligibility for revascularisation (and PCI) is uncertain, and a considerable number of such patients will not be eligible for such a procedure because of anatomical or other circumstances. For instance, in the ICTUS trial, which investigated the possible benefit of revascularisation in subjects with unstable angina, less than 60% of the patients scheduled to undergo intervention actually qualified for subsequent revascularisation.8 Thus, in the depicted trials, the true effectiveness of PCI in clinical practice may well have been exaggerated, and one could argue that, given their specific eligibility criteria, the randomised trials overestimated the true worth of PCI. Against this background, a balanced discussion of the pro and cons of PCI in this setting would have been helpful in the guidelines. It is unreasonable to base a positive recommendation for PCI on the basis of results obtained in a subgroup of a subgroup of the (overall negative) COURAGE trial: an ‘evidence base’ of 100 patients does not meet the criteria to justify millions of costly invasive procedures of uncertain clinical benefit.8 28 A strategy of initial medical treatment is certainly the preferable option for the low risk patient with stable angina or in someone in whom the presence of obstructive coronary artery disease has just been established during diagnostic coronary angiography.
It is also noteworthy that, despite two negative randomised trials that specifically addressed this topic, the performance of percutaneous intervention for chronic total occlusion is not discouraged. Instead, an ‘experienced’ intervention team is called to order, perhaps suggesting that the care was substandard in the trials that investigated this very issue.
Format of the guidelines and composition of the writing committee
The current format of the guidelines differs significantly from the previous version, and this makes it difficult to assess why, and on the basis of what information, the current recommendations have changed from those issued in the earlier document. We recommend that future versions of the guidelines be based on and, where necessary, expand on previous versions.
Specific to these guidelines, we have noted that the majority of the current committee members (and reviewers) were interventional cardiologists or cardiac surgeons. By contrast, other guideline development groups (such as NICE) also include other stakeholders, such as primary care physicians, nurses, health economists, epidemiologists, cardiac radiologists, patient members and pharmacists (Timmis A. et al Stable angina: full guideline. 2011. http://www.nice.org.uk). Guideline committees should be representative of the whole group of professionals and stakeholders as it is likely that the interpretation of the evidence by specialists will be biased. There may be some role or place for procedure oriented guidelines but, in that case, the items at issue should remain confined to matters directly related to technical aspects of the actual procedure.
Regarding the number of recommendations, we believe that 20–30 sound recommendations would be a reasonable number per guideline, in view of our earlier comments and given the low number of recommendations based on high quality clinical data; this should not be too difficult. A summary of our recommendations is given in box 1, and we hope that these will be helpful in developing future versions of the current as well as upcoming guidelines.
Summary of our comments and suggestions
Medical treatment should be the initial management of the patient with stable chest pain. Strong recommendation, high quality data (class IA).
Percutaneous opening of a chronic total coronary occlusion is not recommended. High quality data (class IIIA).
The composition of the guideline committee must be representative of the profession at large.
The number of recommendations per guideline should be limited (eg, between 20 and 30), and be based on sound (high quality) clinical evidence.
As a rule, recommendations based on opinion are not very useful and should be avoided. This pertains to most class II recommendations.
The scope of the guidelines must not be too wide.
New versions of guidelines must build on and expand on previous versions, and relevant changes in recommendations must be addressed specifically.
Competing interests JWD was a member of the ESC Committee on Practice Guidelines from 2002 to 2006, and was the review coordinator of the (previous) PCI guideline.
Provenance and peer review Not commissioned; internally peer reviewed.