Imaging techniques for non-invasive assessment of coronary heart disease in hypertension: value of an integrated approach

An impressive amount of information has been obtained in recent years on the mechanisms, the clinical history and the outcome of cardiovascular diseases. This improvement in our knowledge has been made possible, among other factors, by the fast progress in methods and technologies dealing with cardiovascular imaging analysis. Routine application of these techniques in clinical cardiology has allowed us to collect observations on epidemiological scale. This has not only greatly improved our understanding of the pathophysiology of several cardiovascular diseases, but it has also substantially influenced our daily approach to their clinical management. Such an improvement also has occurred over the past 60 years within the field of arterial hypertension, where we have moved from the crude occasional measurement of arterial pressure in the physician's office and from the detection of its late effects on heart and arterial vessels [1] to the quantification of 24 h ambulatory blood pressure profiles [2], as well as developing the ability to detect early signs of target organ damage [3,4], commonly referred to as ‘preclinical’[5]. In the opinion of a substantial part of the scientific community, these early markers of preclinical disease might be even more precise indicators of the blood pressure impact on the cardiovascular system and of the treatment-related antihypertensive control than ‘casual’ arterial pressure measurement by itself [6]. Among them, electrocardiogram (ECG)-mute echocardiographic left ventricular hypertrophy has emerged as the most potent predictor of cardiovascular risk in hypertension [7] and has been definitively accepted as a key target for antihypertensive treatment [8,9]. The prognostic importance of left ventricular hypertrophy is at least in part related to the associated impairment of coronary circulation [10], as documented by a reduced coronary reserve [11]. Detection of preclinical or early clinical manifestations of coronary artery disease in the context of arterial hypertension management is therefore relevant, because it might help refining the risk stratification in a number of clinical conditions. On the basis of these considerations, two imaging methods for the non-invasive assessment of coronary heart disease also have been proposed in this field, with each of them from time to time being claimed to be of greater accuracy than the other by different investigators: stress echocardiography and stress coronary perfusion single photon emission computed tomography (SPECT). In the context of stress echocardiography, two techniques are available, based on either dypiridamol or dobutamine injection, which are probably characterized by a similar predictive ability [12–14]. A number of studies have indeed been implemented to establish which of these imaging methods, either stress echocardiography or radionuclide scintigraphy, is superior in providing reliable information on the occurrence of pathological changes in coronary circulation. Pasquet et al. [15] compared the ability of dipyridamole SPECT and echocardiography in 129 selected patients referred for vascular surgery. The sensitivity of SPECT for the prediction of early events was similar to that of stress echocardiography, and its specificity was only slightly lower. However, the specificity of SPECT in predicting more remote event occurrence over a 3-year follow-up was 58% compared to the higher value of 80% for stress echocardiography (P < 0.001). Thus, these authors concluded that stress echocardiography was superior to stress myocardial scintigraphy in this selected group of patients. However, this study, as well as others, only dealt with this issue in the frame of a ‘method-orientated’ perspective. The possibility that stress echocardiography might mainly detect gross lesions involving epicardial arteries (therefore carrying a higher risk), while stress myocardial SPECT imaging might better reach peripheral districts of coronary circulation (thus being more sensitive in detecting also lesions at lower cardiovascular risk) was not considered. In the March issue of the Journal of Hypertension, Astarita et al. [16], compared the diagnostic accuracy of Tl201 myocardial SPECT with dypiridamole–atropine echocardiography in a group of hypertensive patients with a high likelihood of coronary heart disease (chest pain + positive exercise ECG). The predictive power of the two methods was compared with the evidence of angiographic coronary lesions. In the context of high pretest probability of coronary heart disease, this comparison demonstrated only a borderline higher diagnostic accuracy for dypiridamole–atropine echocardiography. However, some additional findings of their study should also be considered, because they might stimulate some ‘method-independent’ important pathophysiological considerations, which may also provide an explanation for the subtle differences between dobutamine and dypiridamole tests in the milieu of stress echocardiography [17]. Certainly, arterial hypertension is a peculiar clinical condition in which several phenotypic features can influence the detection of pathologically relevant coronary abnormalities. These conditions include the frequent presence of left ventricular hypertrophy [18] and abnormal vascular structure [19]. Therefore, even more than in other clinical conditions, indications and results provided by any imaging test should be considered in the context of the overall clinical evaluation of a given individual patient. The probability that a hypertensive patient with typical chest pain and positive ECG during low-workload exercise does not have coronary heart disease is low and the diagnosis can thus be issued with a high degree of accuracy. Such patients are likely to have both positive stress echocardiography and coronary angiography [20]. However, coronary lesions may not necessarily involve epicardial arteries, but may also affect smaller vessels which cannot be precisely identified by coronary angiography. In the study by Astarita et al. [16], dipiridamole–atropine echocardiography was negative in all 30 patients with normal epicardial coronary arteries, but more than 50% of patients with normal coronary angiograms were positive to Tl201 SPECT. This finding can of course yield speculations on the sensitivity and specificity of the different methods compared to the arbitrary gold standard represented by coronary angiography. However, a more pathophysiologically orientated inference might also lead to another conclusion, i.e. that coronary angiography may not necessarily always represent the gold standard for the diagnosis of coronary heart disease, at least when left ventricular anatomic abnormalities are present. A positive coronary angiogram documents ‘only’ haemodynamic abnormalities at the level of epicardial arteries (i.e. a coronary artery disease), while a negative coronary artery angiogram does not allow exclusion of lesions in the intramyocardial circulation (i.e. a coronary heart disease). This obvious consideration is also clearly supported by a number of previous studies [21–23]. In patients with arterial hypertension and a positive exercise test (thus with high pretest probability of coronary heart disease), exercise SPECT is probably more sensitive than dypiridamole–atropine echocardiography for detecting microvascular abnormalities, in addition to macrovascular lesions, whereas dypiridamole-atropine stress echocardiography is more accurate in the selection of patients that might require coronary angiography (i.e. those with high probability of epicardial lesions). Another concept emerging from these studies is that wall-motion abnormalities (the visual evidence of ischemia) more likely result from gross epicardial lesions than from microvascular abnormalities [24]. Finally, the evidence that Tl201 SPECT, particularly when considering delayed images, reflects the distribution of the potassium pool and hence myocardial viability [25], emphasizes the ability of this imaging technique to provide additional important pathopysiological information, besides reflecting regional myocardial blood flow. Under this comprehensive view, dypiridamole–atropine stress echocardiography might be used to exclude epicardial coronary lesions (thus providing information on short- and mid-term cardiovascular risk), while stress myocardial SPECT imaging might be used to identify a more generic increase in cardiovascular risk due to coronary heart disease in the long term (i.e also in the absence of a high probability of epicardial lesions) (Table 1). Taking together the information provided by both imaging methods with the information provided by a simple exercise ECG test might therefore help in identifying hypertensive patients with a high probability of coronary microvascular lesions and a low probability of epicardial lesions, as well as patients for whom a coronary angiography might be indicated. In this scenario, a hypertensive patient with negative stress echocardiography but positive SPECT has a low probability of significant epicardial coronary lesions and a high probability of microvascular alterations. Similarly, a patient with low-workload exercise ECG test is most probably affected by major coronary artery alterations. Future studies should address decision making in relation to this proposed integration, and attempt to optimize its cost : effectiveness ratio.Table 1: Possible interpretation of the results obtained by different tests for detection of coronary heart disease when assessing coronary circulation status in hypertensive patients with high pretest probability of coronary heart diseaseUnfortunately, the high number of follow-up studies available in the literature may not help in clarifying whether this view is fully correct because of their short observation periods. In the short or medium term, patients with epicardial lesions are at a higher risk than patients with isolated microvascular abnormalities and are indeed expected to suffer more cardiovascular adverse events. This is probably the pathophysiological explanation for the apparently better performance of stress echocardiography in most of the available longitudinal investigations. With this background, a more general consideration to be made is that evaluation of diagnostic tests for coronary heart disease merely in terms of sensitivity, specificity and diagnostic accuracy might sometimes lead to misinterpretation of some important pathophysiological messages carried by clinical studies, especially when arbitrarily defined gold standard reference procedures are adopted and the overall clinical presentation of a patient is sacrificed in favour of strictly ‘method-orientated’ observations. The practical implications of these concepts is that the selection of which diagnostic test should be used in a given individual hypertensive patient with suspected coronary heart disease should always be made based not only on theoretical considerations, but also on the background of a thorough clinical evaluation.