Regression of left ventricular hypertrophy; what are appropriate therapeutic objectives?

Introduction

Cardiac hypertrophy is a morphological adaptive increase in myocardial mass in response to chronic work overload and is a common clinical finding affecting 23% of men and 33% of women over the age of 59 years [1]. Pressure or volume overload on the myocardium results in an increase in myocardial wall stress and hypertrophy may be seen as an attempt to normalise wall stress and oxygen demand. Although initially protective, the increased myocardial mass requires an increase in coronary blood flow to maintain function; indeed, ventricular hypertrophy may be associated with myocardial ischaemia even with angiographically normal coronary arteries [2–4]. Left ventricular hypertrophy (LVH) significantly increases the risk of myocardial infarction, congestive heart failure and sudden cardiac death [5–7]. It is also associated with a greater prevalence of cardiac arrhythmias [8] and is an important risk factor for cardiac morbidity and mortality [9–10]. Understanding the potential mechanisms for these adverse effects has been the focus of much interest in recent years.

Pathophysiology of LVH

Coronary reserve, defined as the difference between basal flow and maximal flow during infusion of agents such as adenosine is reduced in left ventricular hypertrophy due to hypertension [3, 11], aortic stenosis [2, 4], aortic regurgitation [12], supravalvar aortic stenosis and hypertrophic cardiomyopathy [13, 14], and in experimental models of LVH [15–18]. This impairment is marked in subendocardial layers, with some evidence of an alteration in the normal subendocardial-subepicardial distribution of perfusion at rest [19]. Available evidence suggests that several factors contribute to this impaired coronary reserve. Coronary vascular resistance measured during maximal vasodilatation (minimal coronary vascular resistance) is increased in cardiac hypertrophy [17]. Since coronary resistance is minimal during maximal vasodilatation, increased resistance in hypertrophy is likely to reflect a reduction in the available luminal cross sectional area per unit volume of myocardium. Several studies have suggested a reduction in the number of resistance vessels in hypertrophied myocardium [20, 21] while others suggest a lag in vascularisation and that arteriolar density normalises later in hypertrophy [22, 23]. Systolic impedance to coronary flow as a result of perivascular compression during systole is well known and experimental studies indicate that this is increased in hypertrophied hearts [15]. The reduced coronary reserve limits the ability of hypertrophied hearts to meet blood flow and metabolic requirements when demand is increased. Thus despite normal myocardial oxygen consumption [24] and myocardial perfusion per unit mass at rest, the hypertrophied heart is more vulnerable to ischaemia [18, 25, 26].

LVH is associated with an increased risk of sudden cardiac death as well as total cardiovascular mortality [7, 8]. Electrocardiographic or echocardiographic evidence of LVH is associated with an increased prevalence of ventricular arrhythmias, independent of treatment (diuretic) induced hypokalaemia suggesting an independent arrhythmogenic effect of LVH [27–29]. The increased risk of sudden cardiac death and arrhythmias associated with LVH has focused renewed interest on the underlying electrophysiological disturbances in hypertrophy. Numerous studies have reported prolongation of ventricular action potential duration (APD) in hypertrophy. Possible mechanisms for this include changes in ]i[ L-type calcium channels, ]ii[ sodium-calcium exchange, ]iii[ the delayed rectifier current (I_K) and ]iv[ the transient outward current [30]. Delayed conduction also occurs in hypertrophied myocardium and this is associated with an increased propensity to arrhythmias [31]. This reduced conduction velocity appears to be due at least in part to increased intercellular electrical impedance [32], possibly related to a reduction in intracellular pH [33] and alterations in gap junction morphology, which may also reduce gap junction conductance [34]. Thus delayed conduction and repolarisation as well as increased dispersion of repolarisation [35, 36] are important features of hypertrophy and are known to favour conditions suitable for the development of re-entry arrhythmias.

Although antihypertensive therapy in clinical practice has been effective in reducing mortality from stroke and renal failure its effect on cardiac mortality appears to be less impressive [37, 38]. A number of factors are likely to account for this; a) the slope of the association between stroke and blood pressure is greater than that for cardiac morbidity, thereby offering a potentially greater opportunity for stroke prevention with anti-hypertensive treatment. Anti-hypertensive studies to date have had greater statistical power for assessing effects on stroke prevention than on cardiac mortality [39]; b) a review of available anti-hypertensive studies suggests that benefits of lowering blood pressure in terms of coronary heart disease prevention are less than expected from epidemiological observations [39]; c) studies of cardiac morbidity have tended to focus on coronary heart disease, ignoring other important cardiac end points, notably hypertensive heart failure; d) improving cardiac morbidity is partly dependent on reversing or modifying chronic processes such as atheroma and left ventricular hypertrophy and this has focused attention on the ability of different classes of antihypertensive agents to do so. In this context regression of left ventricular mass has received most attention to date.

Regression of LVH

i) LV mass

Echocardiography has provided an excellent tool for the study of LV mass and regression of left ventricular hypertrophy has been widely reported to occur following treatment of hypertension with β-adrenoceptor antagonists [40, 41], α-adrenoceptor antagonists [42], calcium antagonists [43, 44], angiotensin converting enzyme (ACE) inhibitors [45, 46] and diuretics [47], as well as by lifestyle modification that included weight loss and reduction in salt and alcohol intake [48]. Normalisation of LV mass has not been demonstrated after any medical treatments. Regression of LV hypertrophy does occur following aortic valve replacement [49] but may remain incomplete in 50% of such patients [50]. A number of meta-analyses have compared the relative efficacy of various classes of anti-hypertensive agent in regressing LVH [51–53]. These have suggested that ACE inhibitors may be most effective but the results need to be considered with caution as some uncontrolled studies were included [51, 52] and even when trials were carefully selected for good design features, they were often of small size and involved short periods of treatment [53]. These meta-analyses have usefully highlighted potentially important baseline characteristics which may influence response to treatment, particularly gender, ethnic mix and pre-treatment LV mass indices. For example the Treatment of Mild Hypertension Study (TOMHS) randomized patients to receive one of five different anti-hypertensive agents in addition to lifestyle modification. Patients who received diuretic treatment had the greatest regression of LV mass but also had the highest pre-treatment values [48]. A more recent trial [54] compared monotherapy using six agents from different anti-hypertensive classes in regressing LVH in 587 hypertensive men of whom 45% had LVH at entry. After 1 year groups signed to ACE inhibitor or diuretic treatment had the greatest regression. This study was limited by the lack of females and low follow-up rate (only 21% of randomized patients had echocardiographic measurements at 1 year).

Ideal design features for future trials have been discussed previously [55]. Double-blind, randomized, comparative trials should include both males and females with appropriate ethnic mix and treatment should be maintained for at least 1 year. The striking increase in LVH with age underscores the importance of studying elderly patients in sufficient numbers. Anatomically validated measurements of LV mass should be made in a central laboratory by experienced staff blind to treatment and trial size should be appropriate for the variance of the methods used, approximately 150 to 200 per treatment arm in the case of echocardiography. The LIVE study is a randomized controlled international trial to compare the efficacy of 12 months’ treatment with an ACE inhibitor, enalapril, or a diuretic, indapamide, in regressing LVH in hypertensive patients with LVH using a central laboratory for blind measurement of LV mass [56]. The PRESERVE trial will compare the efficacy of 12 months’ treatment with enalapril or nifedipine in reducing LV mass and patients will be followed for a further 3 years to assess longterm morbidity and mortality [57]. The LIFE study will compare the efficacy of atenolol vs angiotensin II receptor antagonism with losartan in reducing LVH and cardiovascular mortality [58]. These large prospective trials should help to clarify optimal methods for regressing LVH and its effects. Beyond documenting the extent of LVH regression future studies will also need to determine the impact this has on longterm outcome. Epidemiological and retrospective studies suggest that LVH regression does improve prognosis [59, 60]. Documenting that pharmacologically induced LVH regression reduces the prevalence of heart failure and cardiac death in prospectively controlled clinical trials would be of great clinical importance.

ii) LVH pathophysiology

In contrast to the intense interest in regression of LV mass, few studies have examined whether this is accompanied by reversal of the pathophysiological effects of LVH and available studies are often small and results conflicting. Reduction in the size of epicardial coronary arteries has been demonstrated following aortic valve replacement [61]. One study compared coronary sinus blood flow measurements in patients with aortic stenosis with those in a separate group following aortic valve replacement and observed lower resting flow measurements in the latter [62]. No correlation between coronary flow reserve and LV mass could be demonstrated in that study. Improvement in coronary flow reserve has also been reported following treatment of hypertensive patients with verapamil [63], however in that study this effect was independent of changes in LV mass and the same study found no improvement in flow reserve following treatment with an ACE inhibitor, suggesting a direct effect of verapamil [63]. Disturbances in LV diastolic function associated with LVH have been shown to improve with regression of LV mass following antihypertensive treatment in some [64] but not all studies [65]. One study has also demonstrated improvement in QT dispersion with antihypertensive treatment [36].

Experimental studies have reported that antihypertensive treatment of spontaneously hypertensive rats reduces blood pressure and LVH [66–68], with improvements in LV compliance [67] and reduced vulnerability to ischaemia [69] although coronary reserve remained impaired [70]. In contrast other studies have demonstrated some recovery in coronary vascular morphology [71, 72] and coronary reserve [73] with regression of experimental hypertrophy. There is also some evidence that regression of experimental hypertrophy may reduce arrhythmogenicity and allow some electrophysiological recovery.

A better understanding of the factors which regulate myocardial growth is emerging from recent studies. The existence and functional significance of angiotensin-II (Ang-II) receptors and synthesis of RAS components in the heart have been demonstrated [74] and there is evidence that components of the cardiac RAS are upregulated in hypertrophy [75]. ACE inhibitors are effective in regressing LV mass in clinical, and experimental hypertrophy. While this effect on LVH may be partly attributed to interference with the RAS resulting in reduced systemic blood pressure and afterload, recent evidence suggests that doses of ACE inhibitors with no effect on pressure responses to angiotensin 1 can regress LV mass in rats in the absence of reductions in afterload or plasma renin activity [76]. There is also evidence that fibrosis/connective tissue deposition [77] and pathological growth signals in the heart can be mediated by locally produced Ang-II acting on cardiac fibroblasts and myocytes. Mechanical stretch induces Ang II secretion from cultured myocytes suggesting an important role for the RAS in stretch induced hypertrophy [78]. In addition vascular hypertrophy has been induced by direct infusion of Ang-II [79] or by increased local expression of ACE using gene transfer [80]. These findings suggest that the paracrine/autocrine action of Ang-II is an important growth factor in the development and maintenance of hypertrophy. Ang-II interacts with two pharmacologically and molecularly distinct receptor subtypes, AT₁ and AT₂. The AT₁ receptor mediates most of the biological actions of Ang-II [81] including myocyte hypertrophy [82]. AT₂ receptors may have a developmental role [83] and while Ang-II stimulates fibroblast collagen synthesis by both AT₁ and AT₂ receptors, inhibition of collagenase activity is specifically mediated by the AT₂ subtype [84].

Conclusions

(i) There is ample evidence that at least partial regression of LVH can be achieved with antihypertensive treatment. Recent work has begun to shed light on the mechanisms involved in the development of hypertrophy and therefore provides potential targets for more focused treatments. Clinical trials with appropriate power and design are needed to clarify whether antihypertensive treatments which also target these mechanisms are more effective in achieving regression.

(ii) At present the extent to which reversal of the pathophysiological effects of LVH accompanies reduction in LV mass remains unclear. The ability to measure LV mass simply and non-invasively using echocardiography was an important milestone in the study of LVH and its regression. However LV mass alone does not provide an adequate indication of the severity or nature of LVH. For example LV mass may be markedly increased in athletes with values well within the range of what would be regarded as pathological for hypertensive patients, yet available evidence suggests that their hearts function normally. Work is therefore needed to establish whether the recognised abnormalities in cardiac electrophysiology, coronary haemodynamics and contractile function associated with LVH can be reversed and to identify optimal treatments for doing so.

(iii) The ultimate question which needs to be answered is whether regression of LVH and its pathophysiology improves long term prognosis. Epidemiological studies suggest that this is so: long term clinical trials are needed to examine the extent to which pharmacological induced regression may do so.