Medium-Term Determinants of Left Ventricular Mass Index After Stentless Aortic Valve Replacement

Medium-Term Determinants of Left Ventricular Mass Index After Stentless Aortic Valve Replacement Xu Y. Jin, MD, PhD, Ravi Pillai, FRCS, and Stephen Westaby, FRCS Department of Cardiac Surgery, Oxford Heart Centre, John Radcliffe Hospital, Oxford, England Background. This study aimed to investigate the risk factors for elevated left ventricular mass index 3 to 5 years after stentless aortic valve replacement, and to elucidate the underlying physiologic mechanisms. Methods. Eighty-nine patients (age, 76 6 years, 51 males) having a stentless porcine valve for aortic stenosis (n 76) or regurgitation (n 13) were prospectively studied by Doppler echocardiography 3 to 5 years after operation. Left ventricular systolic function, mass index, blood pressure, cardiac rhythm, and New York Heart Association function class were all determined. Stentless valve effective orifice area, mean pressure drop, and the presence and degree of aortic regurgitation were quantified. Results. The mean stentless aortic valve size was 24 2 mm. At follow-up time of 45 9 months, effective orifice area index was 1.2 0.35 cm 2 m 2, and mean pressure drop was 5.7 3.8 mm Hg. Left ventricular mass index was 128 47 g m 2, and ejection fraction was 63% 14%. Multivariant analysis showed a greater left ventricular mass index to be associated with nonsinus rhythm (versus sinus) (163 8 versus 131 7g m 2 ), greater pulse pressure (> 84 mm Hg) (161 7 versus 133 7g m 2 ), New York Heart Association class II or III (versus class I) (166 10 versus 128 5g m 2 ), and male sex (versus female) (160 7 versus 134 8g m 2 ), all p < 0.01. Mean pressure drop (> 8 mm Hg), effective orifice area index (< 1.0 cm 2 m 2 ), the presence of mild regurgitation of the stentless valve, or the type of previous valve disease were insignificant determinants of left ventricular mass index. Conclusions. Three to five years after the implantation, stentless aortic valve hemodynamics remain excellent. Left ventricular hypertrophy caused by previous native aortic valve disease had largely regressed. However, patient-related factors, particularly systemic blood pressure, cardiac rhythm, and function, are significant causes of late residual left ventricular hypertrophy. Thus, continued medical care and earlier surgical intervention may further improve the outlook for these patients. (Ann Thorac Surg 1999;67:411 6) 1999 by The Society of Thoracic Surgeons The type of aortic valve prosthesis has an important bearing on postoperative left ventricular performance [1 3]. The residual pressure drop and nonphysiologic flow profile across both mechanical valves and stented bioprostheses have an unfavorable influence on long-term events and outcomes [4, 5]. After stentless aortic valve replacement there is a rapid fall in left ventricular hypertrophy (mass index) brought about by a reduction in left ventricular wall thickness [3, 6, 7]. These findings are similar to those reported for an aortic homograft [3]. More recently, Westaby and associates [8] have shown a better long-term survival for stentless xenograft patients in comparison with those having a stented valve, which adds further support to the compelling argument for these valves. In our recent detailed postoperative echocardiographic study, however, we observed an unexpected upswing of mean left ventricular mass index between 2 and 3 years postoperatively [9]. Accepted for publication June 29, 1998. Address reprint requests to Dr Jin, Department of Cardiac Surgery, Oxford Heart Centre, John Radcliffe Hospital, Headington, Oxford, OX3 9DU, England. This occurred despite normal stentless valve function. In addition, we have also noted a considerable interpatient variation of left ventricular mass index at 5 years after stentless valve replacement [10]. The present study was, therefore, aimed to identify the factors that determine left ventricular structure and function late after stentless aortic valve replacement. Methods We investigated 89 patients who received a stentless aortic valve replacement between 1991 and 1994 and had been followed up by prospective echocardiography for 3 years or more (mean, 45 9 months [ SD]). The study was performed with ethical committee approval as part of the clinical assessment of two new bioprostheses. Written informed consent was obtained from each participant. There were 51 men and 38 women. Age at follow-up ranged from 62 to 90 years (mean, 76 6 years). The primary valve lesion was aortic stenosis in 76 patients, aortic regurgitation in 7, and mixed disease in 6. All stentless valve replacement operations were per- 1999 by The Society of Thoracic Surgeons 0003-4975/99/$20.00 Published by Elsevier Science Inc PII S0003-4975(98)01134-5

412 JIN ET AL Ann Thorac Surg LV MASS LATE AFTER STENTLESS AVR 1999;67:411 6 formed by two surgeons (S.W. or R.P.), using the subcoronary implant method [11, 12]. Fifty-four patients had a Freestyle xenograft (Medtronic Inc, Minneapolis, MN), and the remaining patients had a Prima stentless valve (Baxter Healthcare Corp, Irvine, CA). Valve sizes ranged from 19 to 29 mm (mean, 24 2 mm). Twenty-two patients with coronary artery disease underwent concomitant coronary artery bypass grafting (2 1 grafts). At the time of echocardiographic study, 66 patients were in sinus rhythm, 15 in chronic atrial fibrillation, and 8 had a pacemaker inserted (2 with dual chamber atrialventricular sequential pacing and 6 with right ventricular pacing mode). Seventy-six patients were in New York Heart Association (NYHA) class I, 11 in class II, and 2 in class III. Echocardiographic Study Transthoracic echocardiography was performed prospectively by the same physician using a Toshiba 380A Ultrasound System (Toshiba Corp, Tochigi-Ken, Japan), with a 2.5-MHz phased-array transducer. The technical detail has previously been described by our group [7, 9, 10]. In brief, the diameter of the left ventricular outflow tract was measured from a two-dimensional image in early systole. Standard left ventricular M-mode echocardiograms [13] were recorded and stored on videotape at a speed of 50 to 100 mm/s, with simultaneous recordings of electrocardiogram and phonocardiogram. From an apical five-chamber view, flow velocities in the outflow tract (2.5-MHz pulsed Doppler) and across the stentless valve (2.5-MHz continuous-wave Doppler) were recorded for off-line analysis. Aortic regurgitation was semiquantified as absent, mild, moderate, or severe by measuring the ratio of regurgitant jet width to that of outflow tract diameter, as detected by color flow mapping from parasternal views [14]. Peak systolic, mean, and pulse pressures were determined noninvasively by the Hewlett Packard (Andover, MA) 66S hemodynamic monitoring system. Calculations Left ventricular end-diastolic dimension, septal thickness, posterior wall thickness, and end-systolic dimension were measured from M-mode echocardiograms. Left ventricular ejection fraction was thus calculated from Teichholz formulas [15, 16]. The ratio of wall thickness to cavity radius at end diastole, left ventricular muscle mass, and mean velocity of circumferential fiber shortening were calculated according to the formula of the American Society of Echocardiography [13]. Muscle mass was indexed to body surface area. In addition, left ventricular short axis and long axis at end diastole were measured from the apical four-chamber view, and the ratio of the two derived. The velocity time integral of blood flow in the left ventricular outflow tract was determined to give stroke distance. Left ventricular stroke volume was thus calculated from the product of stroke distance with outflow tract cross-sectional area [7]. The effective orifice area of the aortic valve was calculated by the continuity equation (stroke volume divided by valve flow velocity time integral as determined by continuouswave Doppler), and indexed to body surface area [7]. The mean pressure drop across the aortic valve was calculated using the modified Bernoulli equation by taking the subvalvular (V 1 ) and valvular (V 2 ) flow velocities (transvalvular pressure drop 4(V 2 2 V 1 2 ), in mm Hg) [7]. In addition, global stroke volume index, stroke work index, and cardiac index were all determined [7]. Myocardial stroke work was defined as the ratio of global stroke work to its mass volume, in mj cm 3 [17]. Mean values for each measurement were derived from three heart beats in patients in sinus rhythm, and from five beats in those in atrial fibrillation or with a pacemaker. Statistics Continuous data are presented as mean SD. Statistical analysis was carried out with MINITAB statistical software [18]. Sex, previous valve disease (stenosis versus regurgitation), cardiac rhythm (sinus versus nonsinus), NYHA class (I versus II or III), age (77 years), systolic (155 mm Hg) and pulse blood pressures (84 mm Hg), ejection fraction (50%), mean transvalvular pressure drop (8 mm Hg), effective orifice area index (1.0 cm 2 m 2 ), and the presence of aortic regurgitation (none versus mild) were taken as cut-off points for dividing patients into two groups. One-way analysis of variance was performed to define the possible significance of individual factors that might affect left ventricular mass index. Those with p values less than 0.10 were then entered into a multivariant analysis of variance (general linear model) to identify significant factors that independently affected left ventricular mass index. The effects of these significant factors on left ventricular cavity size, ventricular geometry, and regional and global systolic function were further assessed by multivariant analysis of variance. A p value less than 0.05 was considered statistically significant. Results Stentless Valve Performance and Left Ventricular Function At a mean of 45 months after implantation, mean pressure drop across the stentless valve was 5.7 3.8 mm Hg. The effective valve orifice area was 2.1 0.7 cm 2, and its index to body surface area was 1.17 0.35 cm 2 m 2. Twenty-two patients had mild transvalvular aortic regurgitation (clinically insignificant). Three had moderate regurgitation. Mean values for left ventricular ejection fraction, 63% 14%, and mean velocity of circumferential fiber shortening, 1.2 0.3 s 1, were within their respective normal ranges. Left ventricular mass index was 128 47 g m 2, which was slightly but significant greater than that reported at 1 year after operation (109 36 g m 2, p 0.01) [7]. Mean values for systolic blood pressure (149 28 mm Hg) and pulse pressure (84 28 mm Hg) were also within the normal range with respect to age. However, the scatter about these mean

Ann Thorac Surg JIN ET AL 1999;67:411 6 LV MASS LATE AFTER STENTLESS AVR 413 Table 1. Results of One-Way ANOVA of Left Ventricular Mass Index a Factor Group A LV Mass Index (g m 2 ) Group B AVOVA p Value Sex 137 47.9 116 44.0 0.034 A male (n 51) B female (n 38) Age 119 36.2 137 55.8 0.086 A 77 years (n 46) B 77 years (n 43) Concomitant CABG 126 45.4 134 53.3 0.498 A without CABG (n 67) B with CABG (n 22) Valve etiology 125 45.5 146 54.9 0.128 A predominant stenosis (n 76) B predominant regurgitation (n 13) Systolic blood pressure 121 45.4 139 48.6 0.079 A 155 mm Hg (n 54) B 155 mm Hg (n 35) Blood pressure pulse 120 49.3 137 43.8 0.085 A 84 mm Hg (n 46) B 84 mm Hg (n 43) Cardiac rhythm 118 40.2 157 54.3 0.000 A sinus rhythm (n 66) B nonsinus rhythm (n 23) NYHA functional class 120 39.8 170 59.2 0.000 A class I (n 74) B class II or III (n 15) Left ventricular ejection fraction 156 45.0 121 45.5 0.005 A 50% (n 17) B 50% (n 72) Valve size 126 50.7 130 45.2 0.711 A 21 (female), 23 (male) mm (n 36) B 23 (female), 25 (male) mm (n 53) Valve: mean pressure drop 131 48.2 119 44.4 0.271 A 8mmHg(n 64) B 8mmHg(n 25) Valve: effective orifice area index 126 54.9 129 43.8 0.826 A 1.0 cm 2 m 2 (n 28) B 1.0 cm 2 m 2 (n 61) Valve regurgitation 124 44.2 139 53.7 0.170 A none regurgitation (n 64) B mild regurgitation (n 25) a Mean SD. ANOVA analysis of variance; CABG coronary artery bypass graft operation; LV left ventricle; NYHA New York Heart Association. values was wide, suggesting substantial patient-topatient variation. Determinants of Left Ventricular Mass Index Univariant analysis identified that male sex, nonsinus cardiac rhythm (atrial fibrillation or pacing), NYHA functional class II or III, and a left ventricular ejection fraction less than 50% associated with a greater left ventricular mass index ( p 0.034 to 0.001). Advanced age ( 77 years) and increased systolic blood pressure ( 155 mm Hg) or pressure pulse ( 84 mm Hg) were of borderline significance ( p 0.079 to 0.086). Concomitant coronary artery bypass grafting, native valve disease (stenosis or regurgitation), stentless valve size ( 21 mm in women, 23 mm in men), mean stentless valve pressure gradient ( 8 mm Hg), valve effective orifice area index ( 1.0 cm 2 m 2 ), and the presence of mild aortic regurgitation showed no significant effects on left ventricular mass index (Table 1). By taking both significant and borderline significant factors into a multivariant analysis, we found only male sex ( p 0.004), nonsinus cardiac rhythm ( p 0.001), impaired NYHA function class (II or III) ( p 0.002), and pulse pressure more than 84 mm Hg ( p 0.004) to be significant independent

414 JIN ET AL Ann Thorac Surg LV MASS LATE AFTER STENTLESS AVR 1999;67:411 6 Table 2. Results of Multivariant Analysis of Left Ventricular Cavity Size, Mass Index, Geometry, and Systolic Function a Variable Sex (male vs female) Cardiac Rhythm (sinus vs non-sinus) Pulse Pressure ( 84 vs 84 mm Hg) NYHA Functional Class (I vs II II) LV mass index (g m 2 ) 160 7vs134 8 131 7vs163 8 133 7vs161 7 128 5vs166 10 p 0.004 p 0.001 p 0.001 p 0.002 LV end diastolic dimension 55 1vs49 1 50 1vs54 1 NS 47 1vs56 2 (mm) p 0.001 p 0.050 p 0.001 LV posterior wall thickness 11.5 0.4 vs 9.8 0.4 NS 9.8 0.4 vs 11.5 0.4 NS (mm) p 0.001 p 0.001 LV septal thickness (mm) NS NS 12.9 0.7 vs 13.7 0.9 NS p 0.025 Ratio of LV wall thickness to NS NS 0.46 0.02 vs 0.52 0.03 NS cavity radius p 0.034 Ratio of LV short axis to NS 0.61 0.02 vs 0.69 0.02 NS NS long axis p 0.001 LV ejection fraction (%) 51 2vs58 2 NS NS 57 2vs52 2 p 0.002 p 0.001 LV stroke work index NS NS 510 41 vs 658 42 NS (mj m 2 ) p 0.004 Myocardial stroke work (mj cm 3 ) 3.9 0.5 vs 5.3 0.5 5.5 0.5 vs 3.8 0.6 NS NS p 0.027 p 0.015 a Mean SD. LV left ventricle; NS insignificant ( p 0.05); NYHA New York Heart Association. determinants of a greater left ventricular mass index (Table 2). Relationship Between Determinants of Mass Index and Left Ventricular Cavity Size, Geometry, and Systolic Function A greater left ventricular mass index was associated with a larger cavity size in patients with nonsinus rhythm or NYHA class II or III. Increased ventricular wall thickness was found in those with an increased pulse pressure. Male patients had greater values for both cavity size and wall thickness in comparison with females. In addition, a wider aortic pulse pressure resulted in greater global stroke work, whereas nonsinus rhythm was associated with a greater ratio of minor to major axis of left ventricle, and a lower myocardial stroke work. Female sex was associated with a higher ejection fraction and greater myocardial stroke work (Table 2). Comment The favorable hemodynamic performance of stentless bioprostheses results in early rapid fall of left ventricular mass to normal levels in the majority of aortic stenosis patients by 12 months postoperatively [6, 7]. Indeed, the function of a stentless valve appears equivalent to normal elderly aortic valve, and this is translated into a very low incidence of thromboembolism, endocarditis, and dysrhythmia [8]. The present study now suggests that 3 to 5 years after stentless valve replacement factors other than aortic valve performance determine left ventricular mass index. This underlines the fact that a stentless valve replacement has essentially eliminated the influence of previous aortic valve disease on ventricular hypertrophy, though medical care remains important for these elderly patients. Late Determinants of Left Ventricular Mass Index After aortic replacement with mechanical valves or stented bioprostheses, there is a 20% to 30% fall of left ventricular mass index in the first year, largely caused by regression of myocyte hypertrophy [19]. In the succeeding 3 to 5 years, there is a further 10% reduction through resolution of interstitial fibrosis, providing the operation has been performed before irreversible myocardial damage has occurred [20]. However, these findings were recorded predominantly in younger patients (mean age, 60 years), and not in the elderly. Furthermore, even in these relatively younger patients, regression of left ventricular hypertrophy is incomplete partly owing to a residual pressure drop across the prosthesis. The effects of persistent elevated afterload on ventricular mechanics are well recognized [20, 21] and have been shown to have an adverse influence on long-term outcome [5]. In contrast, the benefit of stentless aortic valve replacement is the particularly low residual pressure drop, which falls further with time, together with a corresponding increase in prosthetic effective orifice area [6, 7]. In our elderly aortic stenosis patients, the time course for resolution of left ventricular hypertrophy to normal levels was approximately 1 year [7]. Previous studies have shown that the stenotic or regurgitant nature of the original valve disease still affects left ventricular cavity size and mass index within the first 2 years after stentless valve replacement [3]. By 3 years in the present study, the same functional and structural end points appeared to have been reached by all patients regardless of whether the origional lesion was stenosis or regurgitation. Furthermore, mild prosthetic

Ann Thorac Surg JIN ET AL 1999;67:411 6 LV MASS LATE AFTER STENTLESS AVR 415 regurgitation (clinicaly insignificant) or a low transvalvular pressure drop (mean systolic pressure drop of 8 to 10 mm Hg) proved to have no effect on left ventricular mass index in the medium term. Nevertheless, our main finding, that patient-related rather than valve-related factors are the determinants of left ventricular mass index, has significant implications in the clinical management of these elderly patients. The underlying pathophysiology should thus be further investigated. Mechanisms of Elevation of Left Ventricular Mass A greater left ventricular mass index occurs with increased left ventricular wall thickness alone or in combination with enlarged left ventricular cavity size. Incomplete recovery of left ventricular function after delayed aortic valve replacement is frequently associated with residual hypertrophy, through a larger cavity size with normal wall thickness [21, 22]. A greater left ventricular mass index is known to be associated with male sex through both a larger cavity size and greater wall thickness [23]. The effects of systemic hypertension on left ventricular hypertrophy are well recognized and constitute a major source of morbidity in the elderly [24, 25]. Our finding of an association between increased pulse pressure, rather than systolic pressure, and greater mass index is consistent with the increasingly recognized importance of pulse pressure measurement for prediction of long-term outcome [25]. Age-related cardiovascular problems, such as atheroma or systemic hypertension, have dominant effects on ventricular wall thickness rather than on cavity size [24, 26, 27]. The highly significant association between chronic atrial fibrillation or pacemaker rhythm and greater left ventricular mass index after aortic valve replacement has not been clearly defined in the past. However, not only is the cavity size larger, but left ventricular geometry is altered to a more spherical shape by nonsinus rhythm, indicating complex physiologic changes, which certainly adds adverse effects on left ventricular mechanics [28]. Limitations of the Study Measurements of left ventricular mass were based on M-mode echocardiographic measurements. Despite potential limitations of M-mode echocardiography in calculating left ventricular volume and mass, it remains the most widely used method. A recent study confirmed that no practical difference was demonstrated when compared with two-dimensional images [16]. The use of a 75% lower limit, rather than the median value, of ventricular ejection fraction, and valve gradient or effective area, was to perform a clinally more significant comparison. It is thus unlikely that the conclusions would be altered had we used the median value. The enrollment of patients between the two valves used in the present study was not randomized, so we did not attempt any comparison between the two valves. Clinical Implications The excellent hemodynamic performance of stentless aortic valve 3 to 5 years after implantation leads to native aortic valve disease-related left ventricular hypertrophy to regress completely. Furthermore, we found no evidence to suggest any prosthetic valve-related ventricular dysfunction or hypertrophy. In these circumstances, patient-related factors, such as blood pressure, cardiac rhythm and function, become primary determinants of left ventricular mass index, and thus reflect the elderly population (mean age, 76 years) in whom valve replacement is now mainly undertaken. It should, therefore, be reinforced that continuous medical care should be given to these elderly patients even when the best available stentless prosthesis is used. With all the hemodynamic advantages of a stentless valve, one should still realize that the long-term outcome for these elderly patients now depends on the treatment of hypertension, atrial fibrillation, and preexisting ventricular disease. Furthermore, use of the criterion of regression of left ventricular hypertrophy as the sole index of the long-term effectiveness of aortic valve replacement should be treated with caution. In conclusion, earlier surgical intervention and continued medical care may yet improve further the outlook for patients undergoing stentless aortic valve replacement. We are grateful for invaluable comments on the manuscript from Derek Gibson, FRCP, of Royal Brompton Hospital. This study was supported in part by Medtronic Inc (Minneapolis, MN) and Baxter Inc (Irvine, CA). References 1. Jin XY, Gibson DG, Pepper JR. Left ventricular function after aortic valve replacement. In: Yacoub M, Pepper JR, eds. Annual of cardiac surgery, 8th ed., London: Current Science, 1995:125 34. 2. Jin XY, Gibson DG, Yacoub MH, Pepper JR. Perioperative assessment of aortic homograft, Toronto stentless valve, and stented valve in the aortic position. Ann Thorac Surg 1995; 60:S395 401. 3. Jin XY, Zhang ZM, Gibson DG, Yacoub MH, Pepper JR. Effects of valve substitute on changes in left ventricular function and hypertrophy after aortic valve replacement. Ann Thorac Surg 1996;62:683 90. 4. Morris JJ, Schaff HV, Mullany CJ, et al. Determinants of survival and recovery of left ventricular function after aortic valve replacement. Ann Thorac Surg 1993;56:22 30. 5. Lund O. Valve replacement for aortic stenosis: the curative potential of early operation. Scand J Thorac Cardiovasc Surg 1993;40(Suppl):1 137. 6. Del Rizzo DF, Goldman BS, Christakis GT, David TE. Hemodynamic benefits of the Toronto stentless valve. J Thorac Cardiovasc Surg 1996;112:1431 46. 7. Jin XY, Westaby S, Gibson DG, Pillai R, Taggart D. Left ventricular remodelling and improvement in Freestyle stentless valve haemodynamics. Eur J Cardiothorac Surg 1997;12: 63 9. 8. Westaby S, Huysmans HA, David TE. Stentless aortic bioprosthesis: compelling data from the Second International Symposium. Ann Thorac Surg 1998;65:235 40. 9. Westaby S, Jin XY, Katsumata T, Arifi A, Braidley P. Stentless aortic valve replacement: versatility of the porcine root. J Thorac Cardiovasc Surg 1998;116:477 84. 10. Jin XY, Dhital K, Bhattacharya K, Pieris R, Amarasena N, Pillai R. Fifth-year hemodynamic performance of the Prima stentless aortic valve. Ann Thorac Surg 1998;66:805 9. 11. Pillai R, Spriggings D, Amarasena N, O Regan DJ, Parry AJ,

416 JIN ET AL Ann Thorac Surg LV MASS LATE AFTER STENTLESS AVR 1999;67:411 6 Westaby S. Stentless aortic bioprosthesis? The way forward: early experience with the Edwards valve. Ann Thorac Surg 1993;56:88 91. 12. Westaby S, Amarasena N, Long V, et al. Time-related hemodynamic changes after aortic replacement with the Freestyle stentless xenograft. Ann Thorac Surg 1995;60: 1633 9. 13. Sahn DJ, Demaria A, Kisslo J, Weyman A. The committee on M-mode standardization of American Society of Echocardiography: recommendations regarding quantitation in M- mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58:1072 83. 14. Dolan MS, Castello R, St. Vrian JA, Aguirre F, Labovitz AJ. Quantification of aortic regurgitation by Doppler echocardiography: a practical approach. Am Heart J 1995;129:1014 20. 15. Teichholz LE, Kreulen T, Herman MV, Gorlin R. Problems in echocardiographic volume determination: echocardiographic angiographic correlations in the presence or absence of asynergy. Am J Cardiol 1976;37:7 11. 16. Dujardin KS, Enriquez-Sarano M, Rossi A, Bailey KR, Seward JB. Echocardiographic assessment of left ventricular remodeling: are left ventricular diameters suitable tools? J Am Coll Cardiol 1997;30:1534 41. 17. Dodge HT, Baxley WA. Left ventricular volume and mass and their significance in heart disease. Am J Cardiol 1969;23: 528 37. 18. Minitab Inc. Minitab reference manual, release 11, Window version. Philadelphia: Minitab Inc, 1996:3.1 3.11. 19. Monrad ES, Hess OM, Murakami T, Nonogi H, Corin WJ, Krayenbuehl HP. Time course of regression of left ventricular hypertrophy after aortic valve replacement. Circulation 1988;77:1345 55. 20. Krayenbuehl H, Hess OM, Monrad S, Schneider J, Mall G, Turina M. Left ventricular myocardial structure in aortic valve disease before, intermediate, and later after aortic valve replacement. Circulation 1989;79:744 55. 21. Carroll JD, Gaasch WH, Zile MR, Levine HJ. Serial changes in left ventricular function after correction of chronic aortic regurgitation. Dependence on early changes in preload and subsequent regression of hypertrophy. Am J Cardiol 1983;51: 476 82. 22. Taniguchi KT, Nakano S, Kawashima Y, et al. Left ventricular ejection performance, wall stress, and contractile state in aortic regurgitation before and after aortic valve replacement. Circulation 1990;82:798 807. 23. Morris JJ, Schaff HV, Mullany CJ, Morris PB, Frye RL, Orszulak TA. Gender differences in left ventricular functional response to aortic valve replacement. Circulation 1994;90(5 Pt 2):II-183 9. 24. Kannel WB, D Agostino RB, Sillbershatz H. Blood pressure and cardiovascular morbidity and mortality rates in the elderly. Am Heart J 1997;134:758 63. 25. Smulyan H, Safar M. Systolic blood pressure revisited. J Am Coll Cardiol 1997;29:1407 13. 26. Ganau A, Devereux RB, Roman MJ, et al. Patterns of left ventricular hypertrophy and geometric remodeling in essential hypertension. J Am Coll Cardiol 1992;19:1550 8. 27. Sumimoto T, Mukai M, Murakami E, et al. Effect of age on left ventricular geometric patterns in hypertensive patients. J Hypertens 1995;13:1813 7. 28. Greenbaum RA, Gibson DG. Myocardial structure of the left ventricle in health and disease. In: Hamer J, Rowlands DJ, eds. Recent advances in cardiology, 8th ed., Edinburgh: Churchill Livingstone, 1981:207 31.