Journal of Human (2001) 15, 659–667  2001 Nature Publishing Group All rights reserved 0950-9240/01 $15.00 www.nature.com/jhh REVIEW ARTICLE ACE inhibition in : dangerous medicine or golden opportunity?

HC Routledge and JN Townend Department of Cardiology, Division of Medical Sciences, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK

Conventionally -converting enzyme (ACE) these agents are not only tolerated but are associated inhibitors are contraindicated in patients with aortic with acute improvements in haemodynamics and dias- stenosis. Abundant evidence is now available showing tolic function. Further studies are merited to assess the that angiotensin II has a central role in the development possible role of ACE inhibitors in aortic stenosis both of left (LVH), myocardial con- before and after valve replacement. Potential benefits tractile failure and diastolic dysfunction in response to may include prevention of LVH, improved diastolic func- pressure overload. In animal models, ACE inhibitors tion, reduction of arrhythmias and preservation of left have been shown to attenuate these pathological ventricular function. responses. In humans there is no such evidence avail- Journal of Human Hypertension (2001) 15, 659–667 able, however uncontrolled studies have shown that

Keywords: ACE inhibition; aortic stenosis; left ventricular hypertrophy

Introduction cated in patients with aortic stenosis. This is pre- sumably based upon their incorrect classification as Cardiologists are often able to practice evidence- vasodilators although they are neuro-hormonal based medicine as a result of the extensive data antagonists with no intrinsic vasodilator action. In available from large, randomised clinical trials. most patients ACE inhibitors are relatively weak However, the management of many common dis- vasodilators and it has long been recognised that orders continues to be guided by custom and belief their efficacy in chronic failure may have little based upon fundamental principles and anecdotal to do with vasodilatation.2 The activation of both clinical experience. The avoidance of vasodilator circulating and tissue -angiotensin systems has therapy in patients with aortic valve stenosis is a good example of this practice. Although no con- been shown to exert diverse deleterious results in trolled data exists, clinical experience suggests that and the use of ACE inhibitors to inhibit vasodilatation in this condition carries the risk of angiotensin II formation is now a routine manage- severe and coronary hypoperfusion as ment strategy. Both tissue and circulating renin- a result of the inability of to increase angiotensin systems are also activated in patients in the face of a fixed obstruction.1 Sadly, such ‘com- with aortic stenosis and although the pathophysiol- mon ’ medicine is often used to justify only ogy of ventricular dysfunction is distinct from heart distantly related practices and it is possible or even failure due to systolic dysfunction, the effects of likely that the avoidance of angiotensin converting angiotensin II may be equally adverse (Figure 1). enzyme (ACE) inhibitor therapy in aortic stenosis Indeed, far from being contraindicated on ‘first prin- falls in to this category. Conventional clinical teach- ciples’, current knowledge of the role of the renin- ing and the current edition of the British National angiotensin system in aortic stenosis would suggest Formulary suggest that these are contraindi- that these drugs may be highly advantageous to patients with this valve lesion. Currently, no medi- cal therapy is known to influence the natural history Correspondence: Dr HC Routledge, Department of Cardiology, of aortic stenosis, a disorder that affects up to 10% Division of Medical Sciences, Queen Elizabeth Hospital, Edgbas- of the elderly population3 and which without sur- ton, Birmingham B15 2TH, UK. E-mail: h.routledgeȰbham.ac.uk gery inevitably results in premature death. Received 13 February 2001; revised 13 May 2001; accepted 16 May 2001 ACE inhibition in aortic stenosis HC Routledge and JN Townend 660

Figure 1 Angiotensin II in the pathophysiology of ventricular dysfunction due to aortic stenosis.

Role of the renin angiotensin system in of the renin-angiotensin system are produced and aortic stenosis function locally in the heart and other tissues.7 Increased expression of both cardiac renin mRNA Left ventricular hypertrophy and cardiac angiotensinogen mRNA are seen under conditions of experimental chronic pressure overload Left ventricular hypertrophy (LVH) is an inevitable 8,9 consequence of aortic valve stenosis.4 Hitherto, this in rats. Expression of the ACE has also been dem- has been regarded as an important adaptation for onstrated to be upregulated within the myocardium 8–11 preservation of pump function but it is now known of rats with pressure overload induced LVH. that increased left ventricular (LV) mass of any These findings suggest that increased expression of cause is associated with substantially increased car- ACE in the presence of increased local production diovascular and all-cause morbidity and mortality.5 of renin and angiotensinogen produce increases in The mechanisms by which sustained pressure over- angiotensin II, as a response to pressure overload or 12,13 load leads to compensatory cardiac hypertrophy in damage, in the left ventricle. This activation of aortic stenosis are not clearly defined. There is how- the cardiac renin-angiotensin system occurs inde- ever accumulating evidence for the role of neuro- pendently of systemic angiotensin II generation. humoral mechanisms and in particular angiotensin The cardiac receptors upon which angiotensin II II as an effector in this adaptive process.6 If acts, as a paracrine modulator, have been defined it were possible to demonstrate that angiotensin II and divided into two functionally distinct types— 14 is produced in increasing concentrations as a AT1 and AT2—encoded by different genes. AT1a+b response to haemodynamic load and then acts on receptors, expressed on cardiomyocytes and fibro- specific receptors to stimulate myocardial growth blasts, mediate a cellular hypertrophic response.15,16 then we would have a target for ACE inhibition in AT2 receptors are found on coronary endothelial reducing the LV hypertrophic response to aortic cells and fibroblasts7 and it has been proposed that stenosis. To some extent this evidence is now avail- they exert antiproliferative effects counteracting able. growth-promoting effects of AT1 receptor stimu- Recent work has clearly demonstrated that rather lation.15 Distinct patterns of receptor upregulation than being solely a circulating system, components have been demonstrated in animal models of LV

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 661 pressure overload. Left ventricular AT1 and AT2 itself results in a cardiomyocyte hypertrophic receptor mRNA levels are both increased in hyper- response.6 Once again this is reproducible in adult 28 tensive rats, whilst AT2 receptors dominate in rats isolated beating heart models, using a LV balloon. after aortic banding.17,18 In with established An increase in angiotensinogen mRNA, angiotensin LVH AT2 receptor expression is increased and at the II release and an upregulation of the number of angi- same time the new protein synthesis response to otensin receptors is seen during this response sug- angiotensin II seems to be blunted. This may rep- gesting that the local renin-angiotensin system does 6 resent adaptive AT2 receptor upregulation in order play a role in the response to mechanical stimuli. to prevent further inappropriate cardiac Consistent with this hypothesis angiotensin AT1 growth.15,16,19 To date it remains unclear exactly receptor-blockers (whose influence in vivo is what effect alterations in receptor subtype density described below), partially blunt the myocyte hyper- in humans have on response to angiotensin II in the trophic response to mechanical stretch when added context of LV pressure overload, although the devel- to the neonatal cell cultures.28 opment of specificAT1 receptor-blockers is helping to elucidate this.15 Once cardiac angiotensin II has bound to local Diastolic dysfunction receptors a number of biological effects have been Filling curves in patients with aortic stenosis, LVH demonstrated using a cell culture model of ventricu- and early heart failure demonstrate impaired LV lar cardiac myocytes from neonatal rats. Direct diastolic distensibility.29 Few animal studies have stimulation of myocyte culture systems with angio- specifically examined this phenomenon, but again tensin II is seen to induce the transcription of proto- there is some evidence linking the pathophysiology oncogenes (characteristic of foetal cardiac cells)20 directly to the cardiac renin-angiotensin system. In and expression of growth factors.21 These increases isolated hypertrophied myocytes taken from hyper- in growth factor and foetal gene expression can also tensive rats and to a lesser extent myocytes taken be demonstrated in isolated perfused adult rat hearts from control animals, angiotensin II was reported to in response to angiotensin II.22 Changes of this nat- delay and impair relaxation. These results suggest ure, such as increased expression of the beta-myosin that the high levels of angiotensin II found in aortic heavy chain mRNA and insulin-like growth factor stenosis may not only mediate a hypertrophic production, are similar to those seen in human car- response but may actively potentiate impairment diomyocytes with pressure overload induced of relaxation.30 hypertrophy in vivo.23,24 The pathological remodelling of the ventricle occurring in pressure overload is not however simply caused by increases Transition to heart failure in myocyte size, but also involves a progressive interstitial fibrosis. Angiotensin II binding to the In both human patients and experimental models of AT1 receptor has also been shown to be responsible aortic stenosis, following the adaptive hypertrophic for a significant increase in collagen synthesis by phase, persisting pressure overload eventually leads cardiac fibroblasts. Fibroblast growth and increased to progressive ventricular dilatation and dysfunc- collagen synthesis can also be stimulated by the tion. The pathophysiology of this transition remains addition of to cell cultures25 and in vivo unclear but is likely to be related to altered exci- local cardiac aldosterone synthesis is increased as a tation contraction coupling at the level of the myo- result of angiotensin II binding to its cardiac recep- cyte. These changes may result from altered gene tors.26 In the rat heart aldosterone acts on receptors expression in hypertrophied cells leading to a down- found locally, both on fibroblasts and myocytes, pro- regulation of Ca2+-ATPase mRNA.20,31 Such a change moting collagen mRNA expression and protein syn- has been demonstrated in aortic-banded guinea pigs, thesis and accumulation thus contributing to the as they develop pressure-overloaded heart failure, hypertrophic effect.27 Interstitial collagen is and the decrease in Ca2+-ATPase expression can be degraded by the enzyme metalloproteinase-1 and attenuated significantly by 8 weeks of treatment the activity of this enzyme is inhibited by the pres- with an ACE inhibitor.32 Neuroendocrine activity ence of angiotensin II.25 Alterations in the myocar- has also been implicated, in particular local pro- dium induced by angiotensin II binding to cardiac duction of angiotensin II, in the apoptosis hypoth- AT1 receptors thus closely mimic those seen in esis for explaining progressive alterations in response to pressure overload caused by aortic sten- chamber contractility. Myocyte apoptosis has been osis, and include an increase in myocyte size, demonstrated to contribute in rats and mice to the change in myocyte contractile protein expression transition from hypertrophy to early heart fail- and an increase in interstitial fibrotic tissue. ure.33,34 The demonstrations that angiotensin II There is further evidence that angiotensin II applied directly, or released in response to mechan- mediates the effects of mechanical stress upon the ical stretch, induced apoptosis in adult rat cardi- ventricular myocardium. In neonatal cell culture omyocytes in vitro35,36 suggest a further pathophysi- systems, in the absence of serum containing growth ological role for the hormone in the progression of factors or hormones, passive mechanical stretch myocardial in aortic stenosis.

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 662 Animal studies of ACE inhibitors in obtained by lowering LV noradrenaline content by aortic stenosis sympathectomy,38 suggest that this mechanism is less likely. Prevention of LVH by ACE inhibition As cardiac ACE was elevated in untreated aortic The effects of ACE inhibition on the postulated stenosis rats whilst circulating serum ACE and mechanisms for myocardial hypertrophy have been plasma renin activity were unchanged, it seems investigated extensively using a live rat model of unlikely that ACE inhibitor mediated LVH ascending aortic stenosis.37 Although the supraval- regression is due to an effect on the circulating vular location of the obstructive lesion, above the renin-angiotensin system. Ventricular systolic press- origins of the coronary arteries, in the rat model ures were minimally affected by (ෂ6% means direct comparisons with human aortic sten- reduction) or in this aortic stenosis model, osis must be made with care, the progression and no benefit was observed with treat- through compensated LVH to dilation, systolic dys- ment. Thus it is also unlikely that reduction of after- function and cardiac failure is well demonstrated.31 load is a major factor responsible for the regression In addition, using this model, changes occurring of LV hypertrophy by inhibition of ACE.37 within the heart in response to manipulation of the renin-angiotensin axis can be examined in the con- ACE inhibitors and improvement in LV diastolic text of a fixed increased load on the ventricle. function In these rats, echocardiographically determined LV mass index was significantly elevated 4 weeks Echocardiographic studies of aortic-banded rats after banding of the ascending aorta and further pro- demonstrated a restrictive filling pattern, and this gression of LVH was seen in untreated rats. However indicator of diastolic dysfunction has been exam- 6 weeks of treatment with the ACE inhibitor, rami- ined after treatment with the ACE inhibitor fosino- pril, or with an angiotensin II receptor antagonist pril. Measurements of LV stiffness were significantly (losartan) caused a reduction in LV mass index com- lower in treated compared with untreated rats, pared to control. To demonstrate that this effect was though not completely normalised when compared unlikely to be due to the (ෂ18%) reduction in sys- with rats with no valvular obstruction.31 Such temic pressure, a similar chronic drop in improvement in diastolic function may have been was achieved by hydralazine without due to the regression of myocyte hypertrophy and any attenuation of the LV hypertrophic response.37 interstitial fibrosis over the ACE inhibitor treatment Confirmation of this result was obtained by excision period. In the untreated rats however, between 6 and of the hearts after 12 weeks of ACE inhibition at 18 weeks after aortic banding, LV wall thickness did which time it was shown that the LV–body weight not increase any further whilst the abnormal dias- ratio increase was blunted in the ramipril and losar- tolic filling pattern did become more marked. The tan-treated rats 1.5-fold compared to hydralazine investigators therefore suggested then that ACE inhi- treated rats. In addition, individual myocyte widths bition was directly enhancing relaxation in the which were higher in the untreated hypertrophied hypertrophied hearts as a result of a reduction in hearts than those in control rats, were no different cardiac angiotensin II levels. from controls in the ACE inhibitor treated group. In these same rats with aortic stenosis, cardiac (but not ACE inhibitors and prevention of LV dilatation serum) ACE activity was markedly increased with a and remodelling close correlation between cardiac ACE and LV–body weight ratio. Ramipril treatment resulted in approxi- Whether ACE inhibition can attenuate the pro- mately 70% inhibition of cardiac ACE activity, gression of LV dysfunction following pressure over- returning levels in aortic stenosis rats to near nor- load, or whether in fact the reduction of myocyte mal.37 hypertrophy achieved, as described above, rep- The conclusions to these studies were that long- resents the loss of an adaptive mechanism for pre- term blockade of the renin-angiotensin system serving systolic function, has been examined in the arrested cardiac growth and even resulted in rat model. In order to dissociate the effects of ACE regression of established LVH in aortic stenosis. As inhibition on systemic from local cardiac renin-angi- a similar effect was seen with AT1-receptor blockade otensin system, the same method of aortic clipping the effect seems likely to be mediated by angiotensin was used to produce a rat with fixed increased LV II rather than other neurohormonal mechanisms afterload. Rats demonstrated a stage of compensated influenced by ACE inhibitors such as . concentric LV hypertrophy, 6–8 weeks after banding The possibility that ACE inhibition and AT1-recep- with elevated intracardiac ACE mRNA and angio- tor blockade were both blunting a trophic effect of tensin II generation.39 This was followed by cardiac angiotensin II mediated sympathetic stimulation decompensation and premature death after about 20 cannot be excluded. The findings of an earlier study weeks. Investigators hypothesised that if the acti- however, in which ACE inhibition with quinalapril vation of the cardiac tissue renin-angiotensin system caused an 80% regression of LVH in a rat model of contributed to the transition to cardiac failure in rats aortic stenosis, but similar benefits were not with established hypertrophy then chronic adminis-

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 663 tration of would favourably modify this in functioning cardiac myocytes due to angiotensin outcome. At the start of fosinopril treatment all II mediated apoptosis. aortic-banded rats already had established concen- tric LVH but had normal systolic function. At the end of the 12-week treatment period LV mass by echo was again returned towards the normal range Increased survival by fosinopril treatment.31 In fosinopril-treated rats there was a modest reduction in arterial systolic The eventual outcomes of these seemingly beneficial pressure but LV systolic pressure was not different alterations in pathophysiology are reflected in the between treated and untreated groups. After 15 animals’ survival data. Rats in the fosinopril study weeks, the untreated aortic stenosis rats showed a showed a striking survival benefit over untreated marked increase in LV systolic dimension and a cor- rats with aortic stenosis: 3% vs 31% after 15 weeks responding reduction in function (via fractional of treatment (P Ͻ 0.05).39 Initially premature deaths shortening), which was not present in the fosinopril- in the untreated aortic-banded animals were attri- treated group.31 Left ventricular contractile function buted to development of severe LV dysfunction and in the isolated perfused hearts after removal was dilatation. Inspection of the hearts from animals that also better in the fosinopril-treated rats.39 died prematurely however, did not support this The benefit of ACE inhibition in ischaemic heart hypothesis. Another possibility is that ACE inhi- failure with a dilated ventricle,40 was originally bition prevented premature deaths from cardiac thought of as simply a result of modification of the arrhythmia possibly triggered by myocardial ischae- haemodynamic responses to systolic dysfunction. mia or by adverse changes in cardiac autonomic By blocking systemic angiotensin II production, control.39 Both sympathetic tone and cardiac vagal reductions in LV systolic load, systolic pressure and activity are known to be favourably modified by the cavity dilation (by increased venous capacitance) reduction of circulating angiotensin II.43,44 In the occur. A fall in heart rate and contractility are also ramipril-treated aortic banded rats, regression of seen as a result of withdrawal of angiotensin II’s compensatory hypertrophy was also paralleled by actions on sympathetic and vagal activity.41 In con- improved survival: 31% died in the placebo treated trast in aortic banded rats, the beneficial effects of group during 12 weeks follow-up compared with ACE inhibition in preventing serial changes in LV 11% in the ramipril-treated group.37 end systolic or diastolic dimension or preserving ejection fraction did not depend on a reduction in LV systolic pressure or heart rate, which remained equivalent to those in the untreated rats. If the bene- Angiotensin AT1 receptor inhibition fits in this model are not due to improved haemody- namics and the adaptive mechanism of myocyte The beneficial effects of ACE inhibition on hyper- hypertrophy has been lost, then alternative mech- trophic remodelling have thus been attributed to the anisms by which ACE inhibition preserves contrac- inhibition of cardiac ACE activity and reduced acti- tility must be sought. vation of AT1 receptors. Although this hypothesis The fosinopril study examined developed press- has been supported by the results of Bruckshelgel ure–calcium relation and found this to be signifi- and colleagues using the AT1-receptor blocker losar- cantly improved with a shift upwards and to the left tan,37 it has been challenged by the results of in isolated perfused hearts from ACE inhibitor another important investigation. Using identical treated aortic stenosis rats.39 Thus, an improved con- methods to their fosinopril study Weinberg and col- tractile response to extracellular calcium appears to leagues39 treated aortic-banded rats with the specific be compensating for the regression of hypertrophy AT1-receptor blocker . They found no or loss of force bearing units. Confirmatory evidence regression of LVH, no improvement in survival, and comes from isolated myocyte models in which the no change in contractile function compared with calcium-shortening relation was significantly untreated rats. A further unexpected observation enhanced in fosinopril-treated rats compared to was that elevated LV ACE mRNA levels in aortic- untreated rats with aortic stenosis.39 Similar results banded rats were normalised by treatment with both have been seen with improvement of contractile the AT1-receptor blocker and by and response in chronic treatment of post this correlated with an improvement in LV end dias- infarction heart failure rats.42 tolic pressure. The investigators concluded that an There is also evidence of a role for ACE inhibition increased cardiac ACE expression may not be criti- in the reduction of apoptosis in pressure overload cal for pressure-overload LV hypertrophy. Differ- ventricular hypertrophy. AT1-receptor blockade ences in pharmacological effects may also be (losartan) but not AT2 blockade, was shown to pre- explained by the ability of ACE inhibitors, in con- vent the five-fold increase in apoptosis caused by trast to the AT1-receptor blockers, to decrease acti- 35,36 angiotensin II in rat cardiomyocytes. Thus, vation of the AT2 receptor whose role in the regu- another possible target for the use of ACE inhibitors lation of myocardial growth and function in cardiac in aortic stenosis may be prevention of a reduction disease states remains poorly defined.

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 664 Clinical studies findings raise the possibility that treatment with an ACE inhibitor for patients with aortic stenosis may Potential benefits be of benefit both preoperatively to avoid irrevers- Because of the persisting opinion that ACE inhibi- ible structural remodelling of the myocardium, and tors are contraindicated in patients with aortic sten- postoperatively to aid regression of myocardial osis there are very few clinical studies and no ran- hypertrophy. In addition, a beneficial effect on auto- domised-controlled trials in this group of patients. nomic function may reduce the risk of sudden If we can use the evidence now available from these arrhythmic death. animal models, to design the appropriate clinical studies, we may reveal a number of possible roles Safety, tolerability and haemodynamics for this group of drugs in the treatment of patients with aortic stenosis. Whilst there is little doubt that Before recommending a randomised controlled trial the only ‘cure’ for aortic stenosis is to relieve the to evaluate the effects of these agents on this patient obstruction by valve replacement, there are situ- group, the possible adverse effects must be con- ations in which medical therapy could confer a sidered. When examining the animal data it must benefit. The most common scenario in which they also be remembered that the different position of the are currently considered (and then usually rejected) supravalvular obstruction in the rat, relative to the is in patients with co-existing hypertension. ACE coronary ostia, may be of importance. The clinical inhibitors are well established antihypertensives evidence for adverse effects of ACE inhibitors is and within this role they decrease cardiovascular however lacking. morbidity and mortality.45 In patients with a bicus- There are no case reports published of acute hypo- pid aortic valve (ෂ45–50% cases) evidence suggests tensive reactions attributable to the use of an ACE that co-existing systemic hypertension increases the inhibitor in a patient with aortic stenosis. This lack rate of progression of valvular stenosis.46 In addition of confirmed cases is found despite the observation it is clear that in patients with aortic valve disease, that as many as 48% of cases of aortic stenosis go a second stimulus to progressive LVH, a well estab- undiagnosed—only presenting at post mortem, and lished risk factor for all major cardiovascular events, many more wait a considerable period of time from is disadvantageous.47 Hence effective treatments for presentation with cardiac symptoms before the cor- hypertension in this patient group are a necessity rect diagnosis is made.46 Bearing in mind that and an agent, which by its effects on the local car- approximately 15% of patients are being treated for diac renin-angiotensin system, can also arrest car- co-existing hypertension and many more for pain diac growth and even result in regression of estab- attributed to ischaemic heart disease,46 one might lished myocyte hypertrophy and interstitial assume that a number of patients with aortic sten- fibrosis48 should be of further benefit. osis do receive an ACE inhibitor at some time. If this The second scenario, in which prescription of an is the case then the incidence of adverse reactions ACE inhibitor might now be considered, is in treat- is either negligible or simply goes unreported. The ing the adverse effects of aortic stenosis on the LV only published case reports describing chronic use myocardium. In the large number of asymptomatic of ACE inhibitors in patients with aortic stenosis patients under follow-up, there are currently no both involve single patients, neither of whom treatment options which have been shown to reduce experienced adverse effects on normal doses of an LVH or to alter morbidity and mortality. In sympto- ACE inhibitor.62,63 matic aortic stenosis valve replacement is urgently Cox et al64 recently reviewed the only studies required, but for those who refuse or are refused for specifically addressing the question of ACE inhibitor surgery on the grounds of extreme age or co- prescription in aortic stenosis. The larger of the two morbidity there is no effective medical therapy.49 published series concerns 22 patients studied in Currently operative mortality stands at 2–4%, but Mexico with critical aortic stenosis (mean gradient: rises to between 9 and 20% in the presence of 93 mm Hg ± 38) who were given oral captopril.65 extreme age or poor LV function.50–55 Whilst there Patients were monitored with a Swan Ganz catheter is significant functional improvement in almost all over 48 h and six doses (12.5 mg/8 mg). Seven of survivors of aortic valve replacement, irrespective of these patients fell into the higher risk group for age or ventricular function,52,53,56,57 long-term sur- hypotension having signs and haemodynamic meas- vival rates post aortic valve replacement depend on ures consistent with heart failure. Results were a the absence of increased heart size or low ejection predictable fall in systemic vascular resistance, with fraction preoperatively.58–60 In autopsies performed a small fall in systolic blood pressure. Cardiac out- in 51 patients who died 0 days to 14 years after put (by thermodilution) was increased significantly; aortic valve replacement all hearts remained rather than remaining ‘fixed’ enlarged by heart weight index and measures of LV increased from a mean of 47 to 64 ml even in the wall thickness, indicating that regression of patients with heart failure. Mean pulmonary capil- hypertrophy is incomplete after surgery. The cardiac lary wedge pressure fell from 19 to 16 (P = 0.04). pathology was predictable by preoperative scores of All patients demonstrated improved haemodynamic ECG hypertrophy and cardiothoracic index.61 These parameters following acute dosing with no deleteri-

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 665 ous reduction in blood pressure. A further human The concerns regarding safety and tolerability of study suggests that these acute improvements are these agents in patients with aortic stenosis are due to changes in diastolic function. In 20 patients based on the misclassification of ACE inhibitors as with LVH due to aortic stenosis, Friedrich et al66 vasodilators rather than being founded on the demonstrated improved ventricular diastolic dis- results of clinical investigation. In those 30 patients tensibility when infusing into the left who have been closely observed, the acute haemo- coronary artery. Left ventricular end-diastolic press- dynamic response to ACE inhibition has been ben- ure dropped by 20%, and simultaneous changes in eficial in 28 and no adverse effects have been LV pressure and dimensions demonstrating a down- recorded. ward shift in LV pressure–dimension relations. Further tolerability and safety studies are needed These improvements in relaxation were confined to but if initial results are confirmed, a randomised the anterior segments perfused with enalaprilat and controlled will be required to further systemic neurohumoral changes did not occur. investigate the effects of ACE inhibitors on symp- There were no changes in heart rate, cardiac output toms, disease progression, LVH and survival. The or right atrial pressure with ACE inhibitor adminis- choice of ACE inhibitor for such a trial cannot be tration to explain the regional increases in peak fill- based on clinical experience but logic would suggest ing rate. A group of patients with dilated cardiomy- that the maximal benefit and lowest risk might be opathy failed to demonstrate these changes. The associated with an agent with high tissue binding authors concluded that activation of the cardiac and minimal hypotensive effect. renin-angiotensin system occurs in patients with pressure overload LVH, that this contributes to References impaired diastolic function and can be prevented acutely by ACE inhibition. 1 Cohn JN, Franciosa JA. Vasodilator therapy of cardiac The effect of chronic ACE inhibitor therapy has failure (second of two parts). N Engl J Med 1977; 297: been studied in eight patients with severe aortic 254–258. stenosis and fluid retention despite . A pul- 2 Cleland JG, Oakley CM. Vascular tone in heart failure: the neuroendocrine-therapeutic interface. Br Heart J monary artery catheter was again used to observe the 1991; 66: 264–267. acute haemodynamic response to a first dose, but the 3 Iivanainen AM et al. Natural history of aortic valve 67 treatment was continued. After 6.25 mg oral capto- stenosis of varying severity in the elderly. Am J Cardiol pril, six patients demonstrated a fall in LV end-dias- 1996; 78:97–101. tolic pressure and an increase in cardiac output of 4 Seiler C, Jenni R. Severe aortic stenosis without left 21%. Blood pressure did not change significantly. ventricular hypertrophy: prevalence, predictors, and These patients were said to demonstrate an short-term follow up after aortic valve replacement. improved exercise tolerance and decreased fluid Heart 1996; 76: 250–255. retention with chronic dosing although these uncon- 5 Levy D, Garrison RJ, Savage DD. Prognostic impli- trolled results must be interpreted with caution. In cations of echocardiographically determined left ven- tricular mass in the Framingham Heart Study. N Engl the remaining two patients there was no significant JMed1990; 322: 1561–1565. acute haemodynamic response and no change in 6 Sadoshima J et al. Autocrine release of angiotensin II symptoms with chronic dosing. mediates stretch-induced hypertrophy of cardiac myo- It seems then that although the reluctance to pre- cytes in vitro. Cell 1993; 75: 977–984. scribe ACE inhibitors in aortic stenosis is based on 7 Wollert KC, Drexler H. The renin-angiotensin system some logical concerns there are no substantive clini- and experimental heart failure. Cardiovasc Res 1999; cal data to support them. In addition the only clini- 43: 838–849. cal data we do have, although numbers are small 8 Studer R et al. Increased angiotensin-I converting and studies uncontrolled, show not only tolerability enzyme gene expression in the failing human heart. but acute haemodynamic improvement in keeping Quantification by competitive RNA polymerase chain reaction. J Clin Invest 1994; 94: 301–310. with the results of animal studies. 9 Baker KM et al. Renin-angiotensin system involvement in pressure-overload cardiac hypertrophy in rats. Am Conclusion J Physiol 1990; 259: H324–H332. 10 Schunkert H, Jackson B, Tang SS. Distribution and There is accumulating evidence that cardiac angio- functional significance of cardiac angiotensin con- tensin II production may, at least in part, be respon- verting enzyme in hypertrophied rat hearts. Circu- sible for the development of myocardial hypertro- lation 1993; 87: 1328–1339. phy and diastolic dysfunction in hearts subjected to 11 Weinberg EO et al. Angiotensin AT1 receptor inhi- pressure loading. Prevention of angiotensin II bition. Effects on hypertrophic remodeling and ACE mediated myocyte growth, interstitial fibrosis, expression in rats with pressure-overload hypertrophy due to ascending aortic stenosis. Circulation 1997; 95: impairment of relaxation and induction of apoptosis 1592–1600. are some of the mechanisms by which blockade of 12 Passier RC et al. Activation of angiotensin-converting the cardiac renin-angiotensin system (or its specific enzyme expression in infarct zone following myocar- receptors) might benefit patients prior to surgical dial infarction. Am J Physiol 1995; 269: H1268–H1276. intervention. 13 Hokimoto S et al. Expression of angiotensin-con-

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 666 verting enzyme in remaining viable myocytes of in rats with pressure-overload hypertrophy. Chronic human ventricles after . Circu- angiotensin-converting enzyme inhibition attenuates lation 1996; 94: 1513–1518. the transition to heart failure. Circulation 1995; 91: 14 Nio Y et al. Regulation of gene transcription of angio- 2642–2654. tensin II receptor subtypes in myocardial infarction. J 32 Takeishi Y, Bhagwat A, Ball NA, Kirkpatrick DL. Clin Invest 1995; 95:46–54. Effects of angiotensin-converting enzyme inhibition on 15 Bartunek J et al. Angiotensin II type 2 receptor block- protein kinase C and SR proteins in heart failure. Am ade amplifies the early signals of cardiac growth J Physiol 1999; 276: H53–H62. response to angiotensin II in hypertrophied hearts. 33 Condorelli G et al. Increased cardiomyocyte apoptosis Circulation 1999; 99:22–25. and changes in proapoptotic and antiapoptotic genes 16 Goodfriend TL, Elliott ME, Catt KJ. Angiotensin recep- bax and bcl-2 during left ventricular adaptations to tors and their antagonists. N Engl J Med 1996; 334: chronic pressure overload in the rat. Circulation 1999; 1649–1654. 99: 3071–3078. 17 Lopez JJ et al. Distribution and function of cardiac 34 Ding B et al. Left ventricular hypertrophy in ascending angiotensin AT1- and AT2-receptor subtypes in aortic stenosis mice: anoikis and the progression to hypertrophied rat hearts. Am J Physiol 1994; 267: early failure. Circulation 2000; 101: 2854–2862. H844–H852. 35 Leri A et al. Up-regulation of AT(1) and AT(2) recep- 18 Meggs LG et al. Regulation of angiotensin II receptors tors in postinfarcted hypertrophied myocytes and on ventricular myocytes after myocardial infarction in stretch-mediated apoptotic cell death. Am J Pathol rats. Circ Res 1993; 72: 1149–1152. 2000; 156: 1663–1672. 19 Schunkert H et al. Alteration of growth responses in 36 Kajstura J et al. Angiotensin II induces apoptosis of established cardiac pressure overload hypertrophy in adult ventricular myocytes in vitro. J Molec Cell Car- rats with aortic banding. J Clin Invest 1995; 96: diol 1997; 29: 859–870. 2768–2774. 37 Bruckschlegel G et al. Blockade of the renin-angioten- 20 Feldman AM et al. Selective changes in cardiac gene sin system in cardiac pressure-overload hypertrophy expression during compensated hypertrophy and the in rats. Hypertension 1995; 25: 250–259. transition to cardiac decompensation in rats with 38 Kromer EP, Elsner D, Riegger GA. Role of neuro- chronic aortic banding. Circ Res 1993; 73: 184–192. humoral systems for pressure induced left ventricular 21 Sadoshima J, Izumo S. Molecular characterization of hypertrophy in experimental supravalvular aortic angiotensin II-induced hypertrophy of cardiac myo- stenosis in rats. Am J Hypertens 1991; 4: 521–524. cytes and hyperplasia of cardiac fibroblasts. Critical 39 Weinberg EO et al. Angiotensin-converting enzyme role of the AT1 receptor subtype. Circ Res 1993; 73: inhibition prolongs survival and modifies the tran- 413–423. sition to heart failure in rats with pressure overload 22 Schunkert H et al. Angiotensin II-induced growth hypertrophy due to ascending aortic stenosis. Circu- responses in isolated adult rat hearts. Evidence for lation 1994; 90: 1410–1422. load-independent induction of cardiac protein syn- 40 Investigators TS. Effect of on survival in thesis by angiotensin II. Circ Res 1995; 76: 489–497. patients with reduced left ventricular ejection frac- 23 Calderone A et al. Pressure- and volume-induced left tions and congestive heart failure. N Engl J Med 1991; ventricular hypertrophies are associated with distinct myocyte phenotypes and differential induction of pep- 325: 293–302. tide growth factor mRNAs. Circulation 1995; 92: 41 Dzau VJ. Autocrine and paracrine mechanisms in the 2385–2390. pathophysiology of heart failure. Am J Cardiol 1992; 24 Donohue TJ et al. Induction of myocardial insulin-like 70:4C–11C. growth factor-I gene expression in left ventricular 42 Litwin SE, Morgan JP. Captopril enhances intracellular hypertrophy. Circulation 1994; 89: 799–809. calcium handling and beta-adrenergic responsiveness 25 Brilla CG, Maisch B, Zhou G, Weber KT. Hormonal of myocardium from rats with postinfarction failure. regulation of cardiac fibroblast function. Eur Heart J Circ Res 1992; 71: 797–807. 1995; 16 (Suppl C): 45–50. 43 Townend JN et al. Modulation of cardiac autonomic 26 Silvestre JS et al. Myocardial production of aldos- control in humans by angiotensin II. Hypertension terone and corticosterone in the rat. Physiological 1995; 25: 1270–1275. regulation. J Biolog Chem 1998; 273: 4883–4891. 44 Zimmerman BG. Adrenergic facilitation by angioten- 27 Zhou G et al. Effects of angiotensin II and aldosterone sin: does it serve a physiological function? Clin Sci on collagen gene expression and protein turnover in 1981; 60: 343–348. cardiac fibroblasts. Molec Cell Biochem 1996; 154: 45 Hannson L. STOP – Hypertension-2 and best practice 171–178. for the future. Heart 2000; 84 (Suppl 1): i2–i4. 28 Thienelt CD et al. Load-induced growth responses in 46 Wren C, Petch MC. Calcific aortic stenosis. J Royal isolated adult rat hearts. Role of the AT1 receptor. College Physic London 1983; 17: 192–195. Circulation 1997; 95: 2677–2683. 47 Levy D. Clinical significance of left ventricular 29 Villari B et al. Effect of aortic valve stenosis (pressure hypertrophy: insights from the Framingham Study. J overload) and regurgitation (volume overload) on left Cardiovasc Pharmacol 1991; 17 (Suppl 2): S1–S6. ventricular systolic and diastolic function. Am J Car- 48 Brilla CG, Funck RC, Rupp H. -mediated diol 1992; 69: 927–934. regression of myocardial fibrosis in patients with 30 Neyses L, Vetter H. Impaired relaxation of the hypertensive heart disease. Circulation 2000; 102: hypertrophied myocardium is potentiated by angioten- 1388–1393. sin II. J Hypertens Suppl 1989; 7: S104–S105. 49 Carabello BA, Crawford FA Jr. . 31 Litwin SE et al. Serial echocardiographic-Doppler N Engl J Med 1997; 337:32–41. assessment of left ventricular geometry and function 50 Lindblom D, Lindblom U, Ivert T. Heart valve replace-

Journal of Human Hypertension ACE inhibition in aortic stenosis HC Routledge and JN Townend 667 ment in septuagenarians. Scand J Thoracic & Cardio- 60 O’Toole JD et al. Effect of preoperative ejection frac- vasc Surg 1989; 23:29–32. tion on survival and hemodynamic improvement fol- 51 Lindblom D et al. Long-term relative survival rates lowing aortic valve replacement. Circulation 1978; 58: after heart valve replacement. J Am Coll Cardiol 1990; 1175–1184. 15: 566–573. 61 Lund O, Larsen KE. Cardiac pathology after isolated 52 Smith N, McAnulty JH, Rahimtoola SH. Severe aortic valve replacement for aortic stenosis in relation to pre- stenosis with impaired left ventricular function and operative patient status. Early and late autopsy find- clinical heart failure: results of valve replacement. Cir- ings. Scand J Thoracic & Cardiovasc Surg 1989; 23: culation 1978; 58: 255–264. 263–270. 53 Connolly HM et al. Aortic valve replacement for aortic 62 McGrath BP, Denham IM, Johnston CI. Clinical stenosis with severe left ventricular dysfunction. Prog- improvement and hormonal changes in severe cardiac nostic indicators. Circulation 1997; 95: 2395–2400. failure after captopril treatment. Austr N Z J Med 1981; 54 Logeais Y et al. Aortic valve replacement for aortic 11: 639–644. stenosis in 200 consecutive octogenarians. J Heart 63 Wong SS, Long CC, Holt PJ. Lichenoid eruption Valve Dis 1995; 4 (Suppl 1): S64–S71. induced by low dose captopril. Acta Dermato-Venere- 55 Copeland JG et al. Long-term follow-up after isolated ologica 1992; 72: 358–359. aortic valve replacement. J Thoracic & Cardiovasc 64 Cox NL, Abdul-Hamid AR, Mulley GP. Why deny ACE Surg 1977; 74: 875–889. inhibitors to patients with aortic stenosis? [letter]. 56 Murphy ES et al. Severe aortic stenosis in patients 60 Lancet 1998; 352: 111–112. years of age or older: left ventricular function and 10- 65 Martinez SC et al. Hemodynamic effects of oral capto- year survival after valve replacement. Circulation pril in patients with critical aortic stenosis. Archivos 1981; 64: II184–II188. del Instituto de Cardiologia de Mexico 1996; 66: 57 Olsson M et al. Quality of life in octogenarians after 322–330. valve replacement due to aortic stenosis. A prospec- 66 Friedrich SP et al. Intracardiac angiotensin-converting tive comparison with younger patients. Eur Heart J enzyme inhibition improves diastolic function in 1996; 17: 583–589. patients with left ventricular hypertrophy due to aortic 58 Lund O. Preoperative risk evaluation and stratification stenosis. Circulation 1994; 90: 2761–2771. of long-term survival after valve replacement for aortic 67 Grace AA, Brooks NH, Schofield PM. Beneficial clini- stenosis. Reasons for earlier operative intervention. cal and haemodynamic effects of captopril in severe Circulation 1990; 82: 124–139. symptomatic aortic stenosis. Eur Heart J 1991; 12 59 Schwarz F et al. The effect of aortic valve replacement (Suppl): 740. on survival. Circulation 1982; 66: 1105–1110.

Journal of Human Hypertension