Department of Cardiology, Rikshospitalet University Hospital, Oslo, Norway
In clinical trials, the treatment effect of an intervention is usually measured as the impact on a prespecified end point. All-cause mortality is an ultimate end point and often regarded as the "gold standard" for cardiovascular trials. However, since major therapeutic breakthroughs like acute revascularisation with percutaneous coronary intervention, use of antithrombotic agents, β blockers, ACE inhibitors and statins have improved prognosis after acute myocardial infarction (AMI), large-scale studies are necessary to demonstrate improved survival by a new treatment. Another option is to use an intermediate end point like quality of life or functional capacity, which may be more closely linked to disease progression. From a patient perspective, improvement of such intermediate end points may be more important than the sole prolongation of life.1
An alternative to an ultimate or intermediate end point is to use a surrogate end point. The FDA defines a surrogate end point as a laboratory measurement or physical sign which is used in therapeutic trials as a substitute for a clinically meaningful end point that is a direct measure of how a patient feels, functions or survives and is expected to predict the effect of the treatment. Changes of the surrogate should reflect changes in disease progression and occurrence of clinical events, and there should be a biological explanation for the connection.
There are several successful examples of surrogate end points that translated to clinically meaningful end points. Statins were first shown to improve the surrogate lipid profile, and later to improve survival in patients with ischaemic heart disease.2 The effect of vasodilator therapy on the surrogate left ventricular ejection fraction (LVEF) was shown to predict survival in patients with congestive heart failure.3 The positive impact of the surrogate may, however, not necessarily translate into a clinical benefit. Torcetrapib was known to improve the lipid profile, but increased mortality and the rate of cardiovascular events in high-risk patients,4 and the inotropic drug milrinone increased mortality in patients with congestive heart failure,5 even though it was known to improve the LVEF. Therefore, there are sound reasons to question the clinical relevance when an intervention is found to change even a seemingly meaningful surrogate end point.
Most clinical trials assessing the effects of cell-based treatments in patients with cardiac disease have relied on the change in LVEF.6–9 The choice of LVEF was reasonable, given the expectations of improved cardiac pump function through myocardial regeneration.10 However, LVEF is a global measure of myocardial function and often well preserved despite large infarctions because of hypercontractility of non-infarcted segments.11 Therefore, measurement of regional function may be more informative.12 Indeed, the wall motion score index, a measure of regional myocardial systolic function, was recently found to be a better prognostic indicator than LVEF after myocardial infarction.13 Newer ultrasound techniques such as tissue Doppler imaging and speckle tracking may be more accurate than the semiquantitative obtainment of the wall motion score index,14 and are candidates for measurement of treatment efficacy in clinical trials.
In this issue of Heart, Chang et al studied the effects of intracoronary administration of mononuclear cells obtained from peripheral blood after mobilisation with granulocyte-colony stimulating factor (G-CSF) in patients with AMI (see page 995).15 Of the 40 patients in the study, 18 were recruited from their previously published MAGIC-Cell-3-DES trial,16 and as in that trial, cell infusion was found to improve LVEF compared with the control group, where no improvement was found. Furthermore, they used tissue Doppler imaging to identify a positive effect of cell therapy on myocardial synchronicity. The standard deviation of the time to peak systolic velocity obtained from 12 basal left ventricular segments (Ts-SD or the Yu index) was chosen as their end point. The Yu index has been found to provide independent prognostic information in patients with congestive heart disease,17 and to predict responders to cardiac resynchronisation therapy.18 However, the Yu index and several other markers of mechanical dyssynchrony recently failed to add substantial information beyond QRS duration to predict response to cardiac resynchronisation therapy in the multicentre PROSPECT study.19 Furthermore, in patients with myocardial infarction, an increased Yu index seems to be caused by regional differences in myocardial contractility (dyssynergy) rather than abnormal excitation—contraction (dyssynchrony),20 and the Yu index correlates closely with infarct size.21 Thus, the delay in systolic motion is probably secondary to the scar, and not the primary reason for impaired cardiac function.20 While Chang et al obtained data in all patients with excellent repeatability, the Yu index was only obtained in 50% of patients in the PROSPECT study with high interlaboratory variation.19 Although standard drug treatment for congestive heart failure has been shown to improve measures of dyssynchrony,20 little is known about the clinical relevance of improving the Yu index itself. Thus, it remains to be shown that the Yu index is a reasonable surrogate end point in patients with AMI.
作者:admin@医学,生命科学 2011-08-28 17:17