Richard B. Devereux, M.D.
Dr. Richard B. Devereux is Director of the Echocardiography Laboratory at the New York Hospital - Cornell Medical Center. He also serves as Professor of Medicine at the Cornell Univeristy Medical College and Attending Physician at the New York Hospital. A Pennslvania natie, Dr. Devereux earned his M.D. Degree at the University of Pennsylvania School of Medicine in 1971. He completed his Internal Medicine training at New York and Memorial Hospitals in New York. During his postgraduate training he was appointed a Fellow in Cardiology at the Hospital of the University of Pennsylvania and was a Research Fellow for he Southeastern Pennsylvania Affiliate of the American Heart Association. He is Chairman of the Publications Committee of the American Society of Hypertension, and serves on the editorial boards of numerous journals focusing on cardiology and hypertension. He is a renowned researcher and international expert on Marfan syndrome and left ventricular hypertrophy.
Introduction of Dr. Richard Devereux
Dr. Weber: Now I would like to focus again on the myocardium, on the left ventricle and talk about the effects of angiotensin II on the structure and the function of the left ventricle and the relationship to clinical events. It is a pleasure to welcome my former colleague at Cornell, Richard Devereux. Dick is on the faculty of the Department of Medicine, one of the leaders of the cardiology division and certainly one of the leaders of our understanding of the left ventricle hypertrophy and what makes it all happen. Dick, it is a real pleasure to have you here.
AII: Left Ventricular Structure and Clinical Events
Dr. Richard Devereux
28.8 modem
14.4 modemReal Audio recording of Dr. Devereux's Presentation
Introductory remarks
Thank you very much, Michael, ladies and gentlemen. I would like to move from the last two discussions which have been examining molecular and cellular aspects of the impact of angiotensin II on the heart and circulation largely into the clinical arena with a little bit of data from intact animals and address several interrelated issues having to do with the ways in which production of angiotensin II and activation of the renin angiotensin system may affect the heart, inducing hypertrophy, affect the blood vessels, the relationships between cardiac hypertrophy and vascular morbid events and what is known and some of the ways we are going to learn more about the ability of interrupting the effects of angiotensin II to induce reversal of left ventricular hypertrophy and to identify the benefits of interrupting the effects of angiotensin II and regression of hypertrophy on clinical outcome.
This is a slightly revised version of what you've seen before, showing how one gets to the production of angiotensin II and has a series of different effects. John Reid reviewed some of the evidence with regard to vascular damage and how that can lead to myocardial infarction and stroke with both epidemiological evidence from my colleagues, Michael Alderman and John Laragh, and clinical trials' evidence showing the ability to interrupt this by ACE inhibitors. In addition, angiotensin II, either directly or acting through aldosterone causes arterial and arteriolar constriction and induces sodium reabsorption at two successive stages in the kidney leading to pressure overload with varying components of volume overload or at least prevention of the volume underload that should occur with pressure natruresis, leading to the causation of ventricular hypertrophy and arterial hypertrophy. I will show a little data dealing with the issue on both sides of the question with regard to whether in intact organisms angiotensin II has a direct effect that is totally independent of these hemodynamic effects. I will go into the issue of the relationship between cardiac hypertrophy and morbid events and where we stand with the ability to understand the benefits of angiotensin II interruption on the production of cardiac hypertrophy, the reversal of cardiac hypertrophy on morbidity and mortality, and eventually tying these three things together.
Echocardiography as a tool to assess LVH
Now the primary tool I will use in most of these studies are measurements of ventricular mass derived from echocardiograms, shown here on the left in a human, on the right in a rat where the whole left ventricle is in the space of one centimeter. And, in fact, my colleague Rebecca Hahn is doing echocardiograms on mice, adult mice and even in newborn mice, so that perhaps we will have pictures showing the in vivo geometry and function of the heart in transgenic mouse models to show at future meetings of this sort. If we use the echocardiographic measurements to calculate ventricular mass, here on the vertical axis, and compare it to the gold standard of post-mortem weight, in humans, in dogs, in rabbits, in rats, and now down off the scale here in mice, we get reasonably accurate measurements which gives us a probe we can use to examine cardiac hypertrophy in vivo.
Factors that impact on genesis of LVH
We've done many different studies that have revealed aspects that influence the impact of various physiological factors on cardiac hypertrophy in vivo. One of our studies was an examination of three major different independent hemodynamic axes that impact cardiac hypertrophy. One is the level of blood pressure, the second is the level of stroke volume, and the third is the contractile efficiency of the heart. Most of the studies that have been done and reviewed a little bit here today have looked at blood pressure as a control variable, but not at the other hemodynamic factors.
Relation between plasma AII levels and LVH: Clinical Studies
To summarize the evidence directly bearing on the angiotensin II -- left ventricular hypertrophy relationship, this is a predominance of positive evidence from both experiments in animals or tissue cultures and in humans studies, but the evidence is not 100 percent concordant. John Reid showed one of the early negative studies using ECG evidence of LVH from Hans Brunner and John Laragh, and we have had a similar negative study using echocardiographic data when we classified patients into high, normal, and low renin groups where the levels of renin and presumably of angiotensin II show partial overlap. Several studies now have directly related the circulating levels of angiotensin II, on the x axis here, to wall thicknesses and ventricular dimensions. This is from a paper that Roland Schmieder published some years ago showing a moderate correlation that was statistically independent of the level of blood pressure [1]. Two months ago, two additional papers were published. This is one from a group in Denmark which had shown an independent relation of ventricular mass of comparable magnitude to ambulatory systolic pressure and to the level of angiotensin II and showed that one got an overall model...this is the systolic pressure, an r value of about 0.6, with equal contributions from angiotensin II and from the systolic blood pressure [2].
This is from a paper published in March in Circulation by Harrap's group [3] showing that after the level of blood pressure was taken into account, the adjusted level of left ventricular mass was related to the level of angiotensin II, again adjusted against the mean for the population. And there is a positive correlation with an r value of about 0.4. So there are now a number of human studies, each of which shows modest but independent relationships between angiotensin II levels and left ventricular mass that are independent of blood pressure. But the studies have not yet examined the concomitant circulatory volume load and contractile efficiency of the myocardium. So this is intriguing and highly suggestive evidence but not yet totally complete.
Relation between plasma AII levels and LVH: Animal data
Experimental studies also tend to favor positive relationships between angiotensin II levels, either measured in the circulation or actually here in a study from Japan in the tissues in relation to ventricular weight, showing a moderate positive correlation. There have been some studies that suggest that this is thoroughly independent of the level of hemodynamic workload; beautiful studies from James Scheuer's group here in New York showing that the heterotopically transplanted heart weighs more in animals that received angiotensin II [4].
We have done a study which suggests that the effect of angiotensin II on cardiac weight and on molecular transformations in the heart is not totally independent of hemodynamic load. This is a study that Peter Buttrick at Montefiore was the first author on in which we tried to raise activity of the renin angiotensin system by two totally independent interventions [5]. We had rats that we put on a severely salt-deprived diet-- close to zero but not quite zero for sodium chloride, normal diet, and high- salt diet and then either induced the two-kidney-one-clip classical Goldblatt model or left the animals in the sham condition. This is the measurement of plasma renin activity, and we induced similar elevations of plasma renin activity either by salt deprivation, in which case blood pressure was perfectly normal, or by renal artery clipping, in which case blood pressure was high. We reasoned that if we could make the heart grow in intact adult animals purely by stimulating angiotensin II that we should get a high body weight to heart rate ratio or at least a shift from alpha to beta myosin in the adult animals in the group that was salt deprived. If one needed at least a modest degree of hemodynamic overload in addition to the angiotensin II stimulus, we wouldn't see this effect. This is what we have in terms of the heart rate to body weight ratio--really no difference in the salt-deprived animals from the normal animals here on normal salt diet. But when there is an increase in pressure in addition to stimulation of the renin angiotensin system, there is quite a bit of hypertrophy. In terms of the molecular transformations in normal or high-salt conditions, one gets more production of beta myosin, the slow myosin, the alpha to beta myosin ratio falls. But with salt deprivation, whether or not the renal artery is clipped, one gets the alternative change towards more of the fast myosin. So this would suggest that we need to have concomitant hemodynamic overload and can't purely drive the heart to hypertrophy in adult animals.
There are a lot of hemodynamic effects of angiotensin II. You get vasoconstriction causing increased resistance and arterial stiffness, volume expansion, and coronary vasoconstriction can impaired coronary reserve and adversely affect cardiac contractility, which can then in turn induce hypertrophy.
LVH and mortality risk
We've talked about the roles of angiotensin directly and hemodynamics in inducing hypertensive LVH. I would like to address briefly why it is important and what we can do about it. It is important because cardiac hypertrophy predicts dead bodies, basically. This is from one of our studies in which we separated a group of hypertensive adults into those with normal ventricular mass, here, and those with high ventricular mass, here [6]. And using the simple and usually uncontroversial end point of "all-cause mortality", we see substantial divergence between the lines with those with cardiac hypertrophy having approximately a four-fold risk of being dead by the end of the period of study. Follow up, despite the fact that a number of the individuals without cardiac hypertrophy died of cancer. So we didn't exclude competing causes of mortality. When we further examined the data in this long-term follow-up study using either cardiovascular death, all-cause mortality, or non-fatal plus fatal cardiovascular events, we had an opportunity to look at both the univariate prediction of adverse outcome and a multivariate prediction of adverse outcome. In univariate analyses age, systolic pressure, and ventricular mass were consistent independent predictors of all morbid events with additional effects from cholesterol when we added some non-fatal events and had more end points. But in multivariate analyses, only age and ventricular mass achieved statistical significance in any of these analyses.
Somewhat analogous data were presented by the Framingham investigators [7]. Groups with relatively low, more average, and high ventricular mass, both men and women, in the solid and stippled bars, showing that the age-adjusted mortality rate was on average at least twice as high in those with high ventricular mass. Dan Levy and his colleagues concluded that only age and ventricular mass were strong and consistent predictors of either death or non- fatal cardiovascular events in their study. This has been confirmed by additional studies from a number of other investigative groups.
Regression of LVH and mortality risk
There is additional evidence suggesting that LVH regression has an influence on prognosis from four different studies. In one of ours we looked either at whether LV mass increased or decreased or whether people had hypertrophy using specific partition values and had approximately a two- or three-fold difference in the proportion of patients who had morbid events [8]. In each case higher mass or LVH was worse. A study from Eastern Europe--three- and four-fold differences [9]. An ECG-based study from Dan Levy from Framingham, a two-fold difference [10]. And an echocardiographic study, as was the Muiesan study from Europe, about five- and six-fold differences [11]. So the data are quite internally consistent. However, the Koren, Levy, and Muiesan studies are all observational studies without control for treatment. The study by Yurenev et al was a controlled clinical trial that was prospective, but echoes were read by many investigators and the analyses were retrospective. So we do not as yet have data on the relationship between LVH change and prognosis in the setting of a controlled clinical trial in which it is properly known what people were taking, what the intervening blood pressures were, and with other confounders being available for analysis.
Regression of LVH with different classes of antihypertensive drugs
Now what about what we can do about LVH regression? This figure is from one of our analyses about ten years ago of the initial trials establishes one point-- that if you lower blood pressure, ventricular mass tends to go down [12]. These are data points for groups in trials, not individual patients, and this relationship with an r value of about .5 has been consistently confirmed as analyses have been updated. However, when one takes blood pressure into account and its change and other important factors, such as the duration of treatment and the baseline levels of ventricular mass, there are, in fact, suggestions that the type of agent that is used makes a difference. These data are taken from a paper published in today's issue of the Journal of the American Medical Association [13], with which I have written an accompanying editorial [14]. It shows results from randomized, double-blind studies using echocardiographic calculation of ventricular mass and several other features of at least moderate study quality. As one can see from the relationship between the number of studies and patients, the studies were on average small, but pooled data provide reasonable amounts of information. One can see that there are not very large decreases in ventricular mass after adjustment for covariates with use of diuretics or beta blockers. Most prominent decreases occur with ACE inhibitors and somewhat intermediate levels of decrease appear to occur with the calcium blockers. But these studies clearly have some limitations. This simply summarizes the evidence so far that a renin system interruption may be best for reversing hypertrophy. Here are the better meta analyses, including the Schmieder one which was finally published. Supporting experimental studies have been done, which I haven't shown. And definitive human tests are on the way. Now to do a definitive study to assess the relationship of LVH regression and prognosis, there are certain criteria [15]. The study should be large; both genders; ethnically diverse; prospective; serial measurements of LVH with prediction of morbid events, LVH needs to be measured by an anatomically validated method--either echo, MRI or CT, or by ECG indices that can estimate it accurately; and have various analytical characteristics.
The PRESERVE and LIFE trials
There are two studies I would like to mention, both sponsored by Merck. The PRESERVE trial, which is described in the American Journal of Cardiology [16] is a study of a moderate number, nearly 500, women and men from three continents and diverse ethnicity comparing enalapril as an interrupter of the renin angiotensin system to Nifedipine in individuals who have LVH at baseline, follow up echoes for LVH regression, and at least a three-year subsequent follow-up period for morbid events that will give insight into both the issues of the impact of LVH regression on morbidity and mortality and while it is somewhat underpowered may give some information about drug treatment contrasts and the methodology as appropriate. The really major study in this area is the LIFE study [17]; I'll ease up to it in terms of numbers. The LIFE study is a large treatment trial which I will describe in full on the next slide. We are going to do echocardiograms in a substantial number of individuals in this group. It is a randomized, double-blind comparison of losartan and atenolol with echocardiograms at baseline and serially through four years of treatment with centralized reading and appropriate analyses. The full LIFE trial is the 800-pound gorilla of studies in this area. The target population is 8,300 women and men ages 55 to 80, an age range that should assure plenty of morbid events, with electrocardiographic LVH. The Cornell voltage duration product and very high values of the Sokolow-Lyon voltage combination will allow people to get into the study. The Cornell voltage, I should point out, was the ECG measure used by Dan Levy at Framingham to identify a relationship in their data between change in LVH and subsequent prognosis [10] giving this already, despite the fact that it has been developed relatively recently by our group, prognostic validation. This study with 8,300 high-risk patients and a plan to use as endpoints the occurrence of cardiovascular death or non-fatal clinically apparent myocardial infarction or definite stroke is getting at the major endpoints that drive our desire to treat hypertensive patients. We will be able to use ECG changes in the entire population to get at the issue of whether LVH regression beneficially influences morbidity and mortality. The comparison between losartan and atenolol will compare agents with different degrees of interruption of the renin angiotensin system, to borrow John Reid's phrase, unsurmountable at the final step, with losartan, and partial but still significant at an early step with atenolol. This study is well into the recruitment phase. The better part of 3,000 individuals have already been enrolled and randomized. Somewhere between 4 1/2 and 5 years from now by the time the analyses are done, we would expect to have the results.
Other clinical studies in progress
There are in addition to PRESERVE and LIFE other studies that will help to address this issue. A similarly sized study of 480 men and women from France comparing enalapril to indapamide with serial echocardiograms and some degree of subsequent follow up. And a larger study from New Zealand, also double-blind and randomized in which echocardiographic left ventricular mass will be measured in individuals with manifest atherosclerotic disease somewhere or other in the circulation who are randomized to remapril versus placebo on top of their pre-existing medication. They will have serial evaluations which will provide a minimum of two years of follow up and in many individuals, 3 to 3.5 years of follow-up after the in-study echocardiogram. So a series of studies are being undertaken using agents that interrupt the renin angiotensin system that will provide insight into the ability of this type of intervention to induce regression of hypertrophy, and into the relationship between hypertrophy regression and prevention of morbid events, which looks very promising, and into the ability of interruption of the renin angiotensin system to prevent morbid events, which also looks very promising. The LIFE study is by far the largest of these studies. It should prove the most definitive. To my somewhat biased eye as vice-chairman of the steering committee, it has the unusual property of being designed to test two major hypotheses related to the effect of renin system inhibitio on morbidity and mortality and the effect of LVH regression on morbidity and mortality with adequate power to examine both of those issues.
Thank you very much for your attention.
References
1. Schmieder RE, Messerli FH, Garavaglia GE, Nunez B, MacPhee AA, Re RN. Does the renin- angiotensin-aldosterone system modify cardiac structure and function in esential hypertension? Am J Med 1988;84(suppl 3A):136-139.
2. Schroeder AP, Sihm I, Morn B, Thygesen K, Pedersen EB, Lederballe O: Influence of humoral and neurohormonal factors on cardiovascular hypertrophy in untreated essential hypertensives. Am J Hypertens 1996;9:207-215.
3. Harrap SB, Dominiczak AF, Fraser R, Lever AF, Morton JJ, Foy CJ, Watt GCM: Plasma angiotensin II: predisposition to hypertension and left ventricular size in healthy young adults. Circulation 1996;93:1148-1154.
4. Geenen DL, Malhotra A, Scheuer J. Angiotensin II increases cardiac protein synthesis in adult rat heart. Am J Physiol 1993;265(Heart Circ Physiol 34):H238-H243.
5. Buttrick P, Kaplan M, Camargo MJF, de Simone G, Laragh JH, Wallerson D, Devereux RB. The influence of dietary salt and plasma renin activity on myosin heavy chain gene expression in rat hearts. Am J Hypertens 1993;6:579-585.
6. Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med 1991; 114:345-352.
7. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostin implications of echocardiographically determined left ventricular mass in the Framinghan Heart Study. N Engl J Med 1990;322:1561-1566.
8. Koren MJ, Ulin RJ, Laragh JH, Devereux RB. Reduction of left ventricular mass during treatment of essential hypertension is associated with improved prognosis. Am J Hypertens 1991;4:1A.
9. Yurenev AP, Dyakonova HG, Novikov ID, Vitols A, Pahl L, Haynemann G, et al. Management of essential hypertension in patients with different degrees of left ventricular hypertrophy. Multicenter trial. Am J Hypertens 1992;5:182s-190s.
10. Levy D, Salomon M, D'Agostino RB, Belanger AJ, Kannel WB. Prognostic implications of baseline electrocardiographic features and their serial changes in subjects with left ventricular hypertrophy. Circulation 1994:90:1786-1793.
11. Muiesan ML, Salvetti M, Rizzoni D, Castellano M, Donato F, Agabiti-Rosei E. Association of change in left ventricular mass with prognosis during long-term anithypertensive treatment. J Hypertens 1995:13:1091-1105.
12. Devereux RB, Pickering TG, Cody RJ, Laragh JH. Relation of renin-angiotensin system actiity to left ventricular hypertrophy and function in experimental and humans hypertension. J Clin Hypertens 19878;3(1):87-103.
13. Schmieder RE, Martus P, Klingbeil A. Reversal of left ventricular hypetrophy in essential hypertension: Meta-analysis of randomized double-blind studies. JAMA 1996;275:1507-1518.
14. Devereux RB. Regression of left ventricular hypertrophy: How and why? JAMA 196;275:1517- 1518.
15. Devereux RB, Agabiti-Rosei E, Dahlof B, Gosse P, Han RT, Okin PM, Roman MJ. Regression of left ventricular hypertrophy as a surrogate end-point for morbid events in hypertesnion treatment trials. J Hypertens (in press).
16. Devereux RB, Dahlof B, Levy D, Pfeffer MA. Comparison of enalapril vs. nifedipine to decrease left ventricular hypertrophy in systemic hypertension (The PRESERVE Trial). Am J Cardiol 1996 (in press).
17. Dahlof B, Devereux RB, de Faire U, Fyhrquist F, Hedner T, et al. for the LIFE Study Group. Losartan intervention for end-point reduction in hypertension (The LIFE Study). Am J Hypertens 1996;9:26A.
QUESTIONS AND DISCUSSION Speaker:
On the issue of the long-term clinical studies that you are initiating, the question arises as to the different imperatives of pharmaceutical-company sponsored research and investigator-initiated research. My suspicion is that, from a cardiologist's point of view, the comparison they want to see is a side-by-side comparison of an ACE inhibitor with an AT1 antagonist. Do you agree?
Dr. Devereux:
No, I don't. I think that in terms of morbid event trials we have to follow a stepwise, sequential logic that since it is no longer ethical, certainly in high-risk hypertensives or any but the lowest risk hypertensives, to do a placebo-controlled morbid-events trial. We have to select as comparison agents to know where we are standing, a type of agent that has been shown already to have cardio-protective and life-prolonging effects in hypertension and cardiovascular disease. There are enough placebo-controlled trials based either on beta blockers with diuretics added or diuretics with beta blockers added plus evidence in other conditions that either of those classes of agent could be used as a comparison agent and then know if one has further benefits, that one has gone even further beyond the initial benefit compared to placebo. If we compare an ACE inhibitor, which has not yet been shown in hypertension to prevent heart attacks, strokes, or prolong life, to losartan, whatever the contrast is, we wouldn't know how the comparison related to placebo treatment, so it wouldn't be the right design.
Audience:
Dick, that was a nice presentation, as usual. Some hypertensives have inactivation of the renin system and others didn't. I would like to ask you... I didn't see in your paper the renin activity in those longevity studies. Does it fall out that the low renin is still protective or not? So is there any difference in function of those hearts with the high renin and the low renins?
Dr. Devereux:
Two very good questions. I took out some slides that I had had in from Mickey Alderman and John Laragh's data showing stepwise increases in the risk of myocardial infarction associated with higher renin activities, associated at least in small part with low dietary salt intake that is obviously a stimulus to it. I find those data to be pretty persuasive. Another thing I didn't include in the interest of time--what we see is in several different contrasts, either among the renin subgroups of essential hypertension or comparing renovascular hypertension to essential hypertension, that the high renin forms of hypertension have lower left ventricular systolic function. The simple hypothesis of that is that there is coronary vasoconstriction and probably vascular disease that is clinically inapparent. So we see important functional correlates. There many be many necroses scattered through the myocardium, as was initially shown by Harry Gavras and may in fact play a role in some of Victor Dzau's observations.
Dr. Weber:
Dick, let me ask you just one last question. Let's think a moment again about the big LIFE study. Let me put a hypothetical to you. Let's assume that the patients treated with losartan do better than the ones treated with the beta blocker. They have fewer major events, and they have a greater regression of left ventricular muscle mass. How could you be sure that the reduction in events is linked to the regression of the hypertrophy and not to one of the other attributes of the losartan, for example some antithrombotic activity or some other effect on the vascular wall that may reduce heart attacks or major clinical events and is not necessarily a reflection of regression of the hypertrophy.
Dr. Devereux:
I'm not sure there is a perfect answer to that. But the analytical strategy in terms of addressing the LVH issue would be to use the assigned treatment. Analyses will be intention to treat. They use the assigned treatment as a control variable with regard to the attempt to identify the benefit or lack of benefit of LVH regression, which would then be at least a reasonable surrogate for the whole spectrum of effects of losartan, similarly with regard to the main contrast between losartan and atenolol, at some point in the analyses taking into account the change in ECG LVH, which will be available in everybody, will be able to provide some subanalyses insight into the extent to which one might think that LVH regression was an important pathway for that... and then similarly in the echo substudy and other computerized ECG substudies will be able to look at that.
Dr. Weber:
Okay, thank you, Dick. That's a very challenging project, and thank you for telling us about it.
Use back arrow on your browser to go back to HDCN home page (faster) or click on the arrow.
