The purpose of this study was to analyze the relation of aortic valve weight to transvalvular gradient and area, with special regard to valve anatomy, size of calcific deposits, gender, and body size. Two hundred forty-two surgically excised stenotic aortic valves of patients (139 men, mean age 72 ± 9 years) who had undergone preoperative cardiac catheterization and echocardiography were weighed and examined with respect to number of cusps (tricuspid vs bicuspid), size of calcium deposits (microaggregates vs nodular macroaggregates), and presence of cholesterol clefts. The relation among valve weight, gradient, and area was studied. Transvalvular gradient was independent of gender or valve anatomy and was linearly correlated with valve weight absolutely (r = 0.33, p <0.01) or normalized by body surface area (r = 0.40, p <0.01). No correlation was evident between valve area and weight. Calcium macroaggregates were mainly present in men (51%) and in bicuspid valves (67%) and were seen to be strong determinants of valve weight (2.84 ± 1.03 g with macroaggregates vs 1.63 ± 0.56 g with microaggregates, p <0.001) but not of transvalvular gradient. Calcium microaggregates characterized tricuspid valves (62%), where transvalvular gradient was determined by valve weight (p = 0.0026). In conclusion, the heavier the valve, the less frequent were hypercholesterolemia, valve cholesterol clefts, hypertension, and diabetes mellitus.
Aortic stenosis in Western countries is mostly owing to dystrophic calcification of bicuspid or tricuspid aortic valves, which are characterized by increased leaflet thickness and calcium deposition, without commissural fusion. Recently, it has been demonstrated that the weight of the aortic valve, excised during surgical replacement, is linearly related to transvalvular gradient and not to valve area. The aim of our study was to ascertain whether this relation holds true irrespective of the anatomy of the valve (bileaflets vs trileaflets), gender, and body size. Moreover, we searched for a correlation between valve weight and hypercholesterolemia, valvular cholesterol, hypertension, and diabetes mellitus.
Methods
In the University of Padua (Padua, Italy) all aortic valves excised during valve replacement are sent to the pathology department. Two of the authors (S.R., C.B.) weighed and evaluated pathologically 269 stenotic aortic valves of patients operated because of aortic valve stenosis. Surgical pathologic evaluation aimed to determine number of cusps (trileaflets vs bileaflets), size of calcium deposits (diffuse microaggregates, ≤4 mm, vs nodular macroaggregates, >4 mm), and cause of disease (dystrophic calcification vs rheumatic). Moreover, histologic study addressed the presence of cholesterol cleft deposits within leaflets.
Our study focused on 242 valves with dystrophic calcification, among which 180 were trileaflets and 62 were bileaflets. All patients had undergone cardiac catheterization and echocardiography within 1 month before surgery. Catheter gradients were measured as peak-to-peak gradient during pull-back from the left ventricle. Aortic valve area was calculated with the Gorlin equation. Echocardiographic gradients were measured with continuous Doppler recordings and valve areas were computed with a continuity equation.
We compared valve weight between groups and evaluated correlations between valve gradients and areas in all cases and within each group. Comparison between groups was carried out with analysis of variance, and correlations were computed with the least squares method. The statistical package STATISTICA 7.1 was used.
Results
General characteristics of patients are presented in Table 1 . Of 242 patients, 139 were men and 103 were women. Women were significantly older (75 ± 7 vs 70 ± 10 years old, p <0.001). There were 180 trileaflet and 62 bileaflet valves. Bileaflet valves weighed much more than trileaflet valves (2.77 ± 1.3 vs 1.98 ± 0.8 g, p <0.001), and patients with bileaflet valves were younger than patients with trileaflet valves (64 ± 12 vs 75 ± 6 years, p <0.001). Plasma cholesterol levels were 184 ± 38 mg/dl in men and 201 ± 34 mg/dl in women.
| Variable | Values |
|---|---|
| Men | 139 (57%) |
| Age (years), mean ± SD | 72 ± 9 |
| Bileaflet valves | 62 (26%) |
| Trileaflet valves | 180 (74%) |
| Aortic mean gradient (mm Hg), mean ± SD | 57 ± 24 |
| Aortic valve area (cm 2 ), mean ± SD | 0.74 ± 0.21 |
| Valve weight (g), mean ± SD | 2.19 ± 1.01 |
| Normalized valve weight (g/m 2 ), mean ± SD | 1.2 ± 0.58 |
| Microscopic calcium deposits | 131 (54%) |
| Macroscopic calcium deposits | 111 (46%) |
| Valvular cholesterol deposits | 102 (42%) |
| Systolic/diastolic pressure (mm Hg), mean ± SD | 146 ± 27/88 ± 18 |
| Diabetes mellitus | 56 (23%) |
| Atherosclerosis | |
| Coronary | 133 (55%) |
| Carotid | 61 (25%) |
| Peripheral | 44 (18%) |
Transaortic gradient was very similar irrespective of gender or number of leaflets, whereas absolute valve area was slightly larger in men ( Table 2 ). The latter difference disappeared when valve area was normalized by body surface area (0.41 ± 0.09 cm 2 /m 2 in men vs 0.43 ± 0.13 cm 2 /m 2 in women, p = 0.41).
| Number | Age (years), Mean ± SD | Weight (g), Mean (range) | Gradient (mm Hg), Mean (range) | Peak Gradient (mm Hg), Mean (range) | Area (cm 2 ), Mean ± SD | |
|---|---|---|---|---|---|---|
| Valves | 242 | 72 ± 9 | 2.19 (0.56–6.5) | 57 (28–89) | 77 (23–194) | 0.74 ± 0.21 |
| Men | 139 | 70 ± 10 ⁎ | 2.5 (0.79–6.5) ⁎ | 56 (10–110) | 75 (23–155) † | 0.77 ± 0.20 † |
| Women | 103 | 75 ± 7 ⁎ | 1.7 (0.56–3.6) ⁎ | 58 (20–126) | 82 (32–194) † | 0.70 ± 0.22 † |
| Trileaflets | 180 | 75 ± 6 ⁎ | 1.98 (0.56–4.9) ⁎ | 56 (20–121) | 79 (34–194) | 0.75 ± 0.20 |
| Bileaflets | 62 | 64 ± 12 ⁎ | 2.77 (1.1–6.5) ⁎ | 60 (10–128) | 75 (23–136) | 0.70 ± 0.25 |
| Macroaggregates | 111 | 70 ± 10 | 2.8 (1–6.5) | 65 (10–128) | 85 (36–194) | 0.69 ± 0.2 |
| Microaggregates | 131 | 74 ± 7 | 1.6 (0.56–3.4) | 50 (20–115) | 71 (23–130) | 0.7 ± 0.2 |
⁎ p <0.001 for men versus women and trileaflets versus bileaflets.
A good linear correlation existed between aortic gradient measured during cardiac catheterization and valve weight by absolute weight or normalized by body surface area ( Figure 1 ) in men and women. In contrast, no correlation was evident between valve weight and valve area ( Figure 2 ).
At gross examination macroscopic calcium deposits (macroaggregates) were present in 111 (46%) aortic valves and microscopic (or microaggregates) in 131 (54%; Figures 3 and 4 ). Nodular calcium macroaggregates were mainly present in bileaflet valves (67%), whereas in trileaflets microscopic aggregates prevailed (62%, chi-square 17.08, p = 0.000036). When present, calcium macroaggregates were important determinants of valve weight (2.84 ± 1.03 g when macroaggregates are present vs 1.63 ± 0.56 g when microaggregates are present, p <0.001). In contrast, macroaggregates were not determinants of valve gradient or area. In fact, when calcium was present as diffuse microaggregates, valve gradient was determined by valve weight ( Figure 5 ), whereas no significant relation was present when calcium was macroaggregated.
Distribution in calcium macroaggregates was slightly unbalanced in favor of men (51%) compared to women (39%, chi-square 3.78, p = 0.052). This may be the reason why the slope of the relation between aortic gradient and valve weight was much steeper in women than in men ( Figure 1 ); therefore, the gradient increases faster in women than in men.
We compared normalized valve weight in different subgroups of patients according to plasma cholesterol concentration or cholesterol deposits within valve leaflets, hypertension, and diabetes. The heavier the valve, the less frequent were valvular cholesterol deposits and hypertension ( Table 3 ). Moreover, valve weight and area were independent of plasma cholesterol concentration ( Figure 6 ), valvular cholesterol, hypertension, and diabetes.
