|Home | About | Journals | Submit | Contact Us | Français|
See article on page 703
Ever since the inception of percutaneous coronary intervention (PCI), it has been apparent that some myocardial injury was often associated with the procedure. Myocardial injury was obvious when associated with procedural complications such as occlusion of a side branch, occlusive dissection or no‐reflow, but could also be detected quite often in asymptomatic patients in whom the procedure had been uneventful and successful by all angiographic criteria. In the latter case, it was often attributed to distal embolisation of atheromatous material during the procedure or occlusion of minute side branches.1
As methods and markers of myocardial injury became more and more sensitive, the frequency and significance of such periprocedural myocardial injury has increased and become intensely debated. Specifically, as troponins gradually replaced creatine kinase‐MB as sensitive and specific markers for myocardial injury, the frequency of reporting of periprocedural myocardial injury increased precipitously, invoving up to 40% of patients undergoing PCI.
Obviously, the frequency of sampling, the type of clinical setting and patient mix, the type of procedure performed, and the criteria used for defining thresholds, all impact on the rate of such events. Recent studies have highlighted that interpretation of troponin increase is largely confounded when pre‐procedural troponin levels are increased.2 There has been considerable controversy with respect to the prognostic impact of periprocedural myocardial injury. Some have argued that many, if not most, of these events were simply “troponin leaks”, with no impact on clinical or angiographic outcomes. However, over time, evidence has accumulated to suggest that there is a quantitative, although probably curvilinear, relation between the extent of biomarker release and clinical outcomes.3,4,5 At the higher end of the spectrum, it is easy to discern the prognostic impact; at the lower end, the impact is minimal and may require very large studies to be demonstrated.6 Because troponins have the ability to detect much smaller amounts of myocardial injury than creatine kinase, most (but not all) increases in troponin levels will have minimal prognostic impact.7 However, even when completely asymptomatic, many such “troponin leaks” are associated with clearly delineated areas of irreversible myocardial injury on MRI, and the size of these lesions is well correlated to the degree of increase of biomarkers.8 Because of these observations, defining a binary threshold above which periprocedural injury would be prognostic or should be defined as myocardial infarction is largely arbitrary. Yet, it is an important tool for clinical research, as most studies involving PCI will use composite clinical end points of death, myocardial infarction and revascularisation. Standardisation of definitions is important for comparison between studies, and is also essential in understanding the magnitude of the clinical benefit associated with avoidance of such “periprocedural myocardial infarctions”. There is ongoing effort by a unified task force of the European Society of Cardiology, American College of Cardiology, American Heart Association and World Heart Federation to update the definition of myocardial infarction published in 2000,9 and include a new definition of periprocedural infarction. This task force will help define biomarker thresholds to be used in the identification of infarction associated with instrumentation of the heart, although such thresholds are often largely arbitrary. Clearly, the definition will be different from that of spontaneous infarction.
Factors which predict periprocedural myocardial injury are numerous and related to patients' clinical characteristics (older age, left ventricular dysfunction, baseline inflammation), angiographic characteristics (multivessel disease, native lesion or saphenous graft, angiographic complexity of the lesion, and, specifically, presence of thrombus within the target lesion), procedural characteristics (number of target vessels, use of atherectomy devices, duration of inflations)5 and the type and intensity of antithrombotic therapy used. The use of protection devices, such as distal filters, retrieval or aspiration devices, has proved disappointing, with the exception of treatment of saphenous vein graft lesions. Antiplatelet treatment has been the focus of many studies, most of which demonstrate better protection against periprocedural myocardial injury by use of more intense or more effective antiplatelet treatment. It was shown that patients with poor response to aspirin experienced more frequent myonecrosis during PCI.10 Likewise, use of higher loading doses of thienopyridines has been associated with reduction in myonecrosis.11 Finally, in the “pre‐clopidogrel era”, use of glycoprotein IIb/IIIa receptor blockers in high‐risk patients undergoing PCI was associated with reduction in myonecrosis and also with a mortality benefit.12 Other interventions to reduce myocardial injury in the context of PCI, such as statins, have been studied.13
The study by Bonello et al14, published in this issue of Heart, addresses prevention of myonecrosis using trimetazidine, an anti‐ischaemic agent. In that study, 266 patients were randomly assigned to pretreatment with an oral loading dose of 60 mg trimetazidine or placebo. The group treated with trimetazidine had troponin values consistently lower than controls, and the total amount of post‐PCI troponin release was reduced. This result was achieved despite a slight imbalance in baseline risk for periprocedural events against the trimetazidine group (in which patients were older, more frequently had diabetes mellitus and had a high body mass index and a high angina class). Because trimetazidine is devoid of haemodynamic effects and has overall excellent clinical tolerance, it is an attractive candidate as adjunctive treatment to PCI to decrease periprocedural myocardial injury. There are, however, limitations to this study: it is a single‐centre, open study; in addition, although troponin release was reduced by trimetazidine, the frequency of troponin rise was not. Clearly, larger clinical studies powered on genuine clinical outcomes are warranted to confirm the findings of Bonello et al before such treatment can be adopted in clinical practice.
Competing interests: None declared.