VALVULAR
HEART DISEASE
Balloon Dilatation of the Stenosed Aortic Valve: How Does It Work? Why Does It Fail? Francis Robicsek, MD, Norris B. Harbold, Jr., MD, Lawrence N. Scotten, DiplT, and David K. Walker, PhD
The hemodynamic changes that may occur in patients undergoing aortic balloon valvuloplasty were examined in the circulatory model. Four conclusions were reached. (1) Significant transvalvular pressure gradients appear only if the orifice is severely narrowed. (2) The magnitude of this gradient is highly flow dependent. (3) At critical narrowings, minute alterations in orifice size may induce most significant changes in the transvatvular gradient. (4) In low flow states significant gradients appear only if the stenosis is extreme. In patients with aortic stenosis, especially those with failing hearts and low cardiac output, the pressure gradient may be effectively decreased by minimal dilatation of the aortic orifice. These patients, however, remain in jeopardy because recurrent narrowing may cause a gradient incompatible with life. (Am1 Cardiol 1999;65:761-766)
orlinl wrote, “What precisely the obstructive lesion contributes to the clinical picture may be hard to assess except in anatomical terms.” Since percutaneous balloon dilatation was introduced for the treatment of patients with aortic valvular stenosis in 1986,* this technique has gained considerable popularity both in the United States and abroad. The mechanism how this procedure may work, however, is still poorly understood and the prognosis of patients in whom this method has been applied remains in doubt. The debate surrounding the clinical merits of aortic balloon valvuloplasty is centered around 2 seemingly contradictory findings. (1) As we have previously demonstrated, the procedure fails to alter significantly the anatomy of the stenosed and calcified aortic valve (Figure 1).3.4 (2) Despite these very modest anatomic changes, several investigators report significant decrease in the transvalvular aortic pressure gradient and appreciable clinical improvement in patients in whom this method has been applied. 5.6 This controversy has been examined in several in vitro hydrodynamic experiments.
G
METHODS
From the Department of Thoracic and Cardiovascular Surgery and the Heineman Medical Research Laboratory at the Carolinas Heart Institute, Charlotte, North Carolina, and the Cardiac Development Laboratory at the Royal Jubilee Hospital, Victoria, British Columbia, Canada. This study was supported in part by a grant from the Heineman Medical Research Center, Charlotte, North Carolina. Manuscript received August 8, 1989; revised manuscript received November 10, 1989, and accepted November 15. Address for reprints: Francis Robicsek, MD, Heineman Medical Research Center, Box 35457, Charlotte, North Carolina 28235.
Our observations were made using the Vivitro Model Left Heart System (Vivitro Systems, Inc.) modified for the present experiments (Figure 2). In this model, a 3 1-mm pericardial trileaflet xenograft (Symbion, Inc.) represented the aortic and a 29-mm pericardial trileaflet xenograft the mitral valve. Output from a waveform generator to a pumping element provided left ventricular action. The generator was programmed in a way to produce variable fluid flows. Twenty central-orifice plates varying from 0.1 to 2.5 cm2 were used individually to create aortic stenoses of various severity. The net ventricular flow rate was set to provide a systole/diastole ratio of 0.54. Intervals during the cycle were defined as shown in Figure 3. Flow rates were calculated either as mean during forward flow or as root mean square value (RMS) during forward flow. All tests were made at a heart rate of 70 beats/min. The test fluid was physiologic saline in most, and a mixture of glycerol and saline with a viscosity of 3.55 mPas in some experiments. The end-diastolic aortic pressure was set to 80 mm Hg in all experiments. This resulted in a mean aortic pressure of approximately 100 mm Hg. Mean transvalvular pressures were computed over the interval bounded by zero transvalvular pressures. A calculated orifice area similar to the clinical Gorlin formula showed close correlation to the geometricalTHE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
15. 1990
761
ly measured size of the aortic orifices used. The cardiac output was computed in liters/min as forward minus regurgitant aortic flow. Data were averaged over 10 cycle periods. Signals were recorded through an analog to digital converter into a disc using a Digital Equipment Corp. computer system. Ten cycles of each waveform were stored using 256 points/cycle. Custom software
FIGURE
1. The stemmed
and cakitted
aortic
valve
before
and after
was used to allow efficient and accurate control of the acquisition, analysis, display and management data. In the first series of experiments using normal saline solution as test fluid, aortic pressure gradients were measured for the 16 largest orifices with a cardiac output of 5.1 liters/min and for the 4 smallest orifices with a cardiac output of 1.3 liters/min. With these results in
balloon
dilatatkm3
adjustable periptwal resistance
characteristic
FIGURE 2. The circulatomymodslappliedin cur experiments. interchang,eable orifice plate to simulate stenosis
blood
analog
hydraulic
fluid
air water
762
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
65
f,h?r
fluid
hand, a second series of experiments was done using saline solution with 5 different cardiac outputs ranging from 1.3 to 4.1 liters/min, with 5 orifices chosen for each output so that transaortic valve pressure gradient measurement ranged between 20 and 335 mm Hg. A limited number of experiments were performed using a mixture of glycerol and saline instead of normal saline solution as test fluid.
cardiac workload (Figure 6). (3) In low flow states significant gradients appear only if the stenosis is very severe and minimal increase or decrease in orifice size may lead to dramatic changes both in transvalvular pressure gradient and in cardiac work (Figure 7). (4) The outcome of the experiments varied only minimally whether they were performed with blood analog test fluid or with physiologic saline solution.
RESULTS
DISCUSSION
The results of our observations are shown in Figures 4 to 7. From these, 4 conclusions are valid. (1) Significant transvalvular pressure gradients appear only at severely narrowed orifices and the magnitude of the pressure gradient is highly flow-dependent (Figures 4 and 5). (2) At critical narrowings the orifice size needs to be increased only minimally to achieve a most significant decrease in transvalvular gradient and to decrease the
Although we describe results obtained for the measurement of pressure across simulated aortic stenosis in vitro, we believe that the clinical implications that may be inferred are significant and multifold. First, our study objectively documents the fact that significant aortic transvalvular gradients appear only at critical stenoses and that their gradient for a given cardiac output may be well alleviated even by a minimal
FIGURE intervals cycle.
3. Flow rate during a
Forwud
Ckrod
ClO8irrg
300
Ordice
Ares
km’)
0.3
28
4
FIGURE 4. thrdath behveen valvular premure gradient, flow ofllice size I.
transand
0
‘, 1.2
I 2.2
I 3.2 FLOW
THE AMERICAN
JOURNAL
RATE
I 4.2
1 5.2
(L/mm)
OF CARDIOLOGY
MARCH
15. 1990
763
increase in the aortic orifice. Patients with low cardiac outputs are especially sensitive to this process. This reconciles the controversy as to why aortic balloon valvuloplasty in some cases may induce dramatic and immediate decreases in the transvalvular pressure gradient despite very modest increases in orifice size.
Also, in some patients even a moderate decrease in cardiac output can cause a significant decrease of transvalvular pressure gradients. Therefore, it appears to be mandatory that during balloon valvuloplasty as well as in the late evaluation both pressures and blood flow should be simultaneously monitored. These observations
FIGURE 5. Car-relation vaivular pressure orifice size II.
1 1.2
11.5 1 1.1
11.3 18 1.0
10.7 1I 0.0
10.1 I. 0.5
ORIFICE 9.4 I1 0.7
DIAMETER (mm) 5.7 5.0 7.1 I 1I ,1 0.6 0.5 0.4
OWFCE
AREA
5.2 I 0.3
5.0 I 0.2
3.6 I 0.1
gradii,
between transttow and
0 1 0
(cd)
flGURE 6. Decrease in orifice size necessaary to increase the transvaivular pressure gradient from 36 to 150 mm Hg.
764
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
65
ORIFICE DIAMETER
1 AREA
(mm)
(cm’)
0.0 --
0
3.6 --0.1 5.0 --0.2
6.2
--0.3
7.1 --0.4
0.0
flGURE 7. Condation ohe, lbw end cardiac
between work.
--OS
orifice 0.7 --O.d
9.4 --0.7 Flow 10.1
--OS
10.7
--0.9 6
11.3
--1.0
11.5
--1.1
Al.2
d
0
p
I
I
I
I
0.0
0.5
1.0
1.5
2.0 CARDIAC
also make obsolete the stance7 that the severity of aortic stenosis may be graded by the transvaivular gradient alone. Patients with critical aortic stenoses, whose transvalvular aortic pressure gradient has been alleviated by moderately increasing the aortic orifice with valvuloplasty, may experience sudden return of the gradient if the cardiac output increases on exercise8 or due to improved ventricular function.9 Clinical data show that an overwhelming majority of patients who undergo balloon dilatation for aortic stenosis will end up with an orifice area of about 0.8 cm29 loI2 which clinically is still specified as “severely stenotic.” About 25% of these patients experience periprocedural major complications including death, and an equal number have to undergo valve replacement within a year. At 1 year the actuarial mortality rates have been reported as between 24 and 60%, similar to patients with aortic stenosis of similar degree who were managed medically. 13m16The dangers that a patient faces living with such a marginal valve have already been noted by Rahimtoola and othersi7J8 in connection with undersized (
HEART DISEASE
Balloon Dilatation of the Stenosed Aortic Valve: How Does It Work? Why Does It Fail? Francis Robicsek, MD, Norris B. Harbold, Jr., MD, Lawrence N. Scotten, DiplT, and David K. Walker, PhD
The hemodynamic changes that may occur in patients undergoing aortic balloon valvuloplasty were examined in the circulatory model. Four conclusions were reached. (1) Significant transvalvular pressure gradients appear only if the orifice is severely narrowed. (2) The magnitude of this gradient is highly flow dependent. (3) At critical narrowings, minute alterations in orifice size may induce most significant changes in the transvatvular gradient. (4) In low flow states significant gradients appear only if the stenosis is extreme. In patients with aortic stenosis, especially those with failing hearts and low cardiac output, the pressure gradient may be effectively decreased by minimal dilatation of the aortic orifice. These patients, however, remain in jeopardy because recurrent narrowing may cause a gradient incompatible with life. (Am1 Cardiol 1999;65:761-766)
orlinl wrote, “What precisely the obstructive lesion contributes to the clinical picture may be hard to assess except in anatomical terms.” Since percutaneous balloon dilatation was introduced for the treatment of patients with aortic valvular stenosis in 1986,* this technique has gained considerable popularity both in the United States and abroad. The mechanism how this procedure may work, however, is still poorly understood and the prognosis of patients in whom this method has been applied remains in doubt. The debate surrounding the clinical merits of aortic balloon valvuloplasty is centered around 2 seemingly contradictory findings. (1) As we have previously demonstrated, the procedure fails to alter significantly the anatomy of the stenosed and calcified aortic valve (Figure 1).3.4 (2) Despite these very modest anatomic changes, several investigators report significant decrease in the transvalvular aortic pressure gradient and appreciable clinical improvement in patients in whom this method has been applied. 5.6 This controversy has been examined in several in vitro hydrodynamic experiments.
G
METHODS
From the Department of Thoracic and Cardiovascular Surgery and the Heineman Medical Research Laboratory at the Carolinas Heart Institute, Charlotte, North Carolina, and the Cardiac Development Laboratory at the Royal Jubilee Hospital, Victoria, British Columbia, Canada. This study was supported in part by a grant from the Heineman Medical Research Center, Charlotte, North Carolina. Manuscript received August 8, 1989; revised manuscript received November 10, 1989, and accepted November 15. Address for reprints: Francis Robicsek, MD, Heineman Medical Research Center, Box 35457, Charlotte, North Carolina 28235.
Our observations were made using the Vivitro Model Left Heart System (Vivitro Systems, Inc.) modified for the present experiments (Figure 2). In this model, a 3 1-mm pericardial trileaflet xenograft (Symbion, Inc.) represented the aortic and a 29-mm pericardial trileaflet xenograft the mitral valve. Output from a waveform generator to a pumping element provided left ventricular action. The generator was programmed in a way to produce variable fluid flows. Twenty central-orifice plates varying from 0.1 to 2.5 cm2 were used individually to create aortic stenoses of various severity. The net ventricular flow rate was set to provide a systole/diastole ratio of 0.54. Intervals during the cycle were defined as shown in Figure 3. Flow rates were calculated either as mean during forward flow or as root mean square value (RMS) during forward flow. All tests were made at a heart rate of 70 beats/min. The test fluid was physiologic saline in most, and a mixture of glycerol and saline with a viscosity of 3.55 mPas in some experiments. The end-diastolic aortic pressure was set to 80 mm Hg in all experiments. This resulted in a mean aortic pressure of approximately 100 mm Hg. Mean transvalvular pressures were computed over the interval bounded by zero transvalvular pressures. A calculated orifice area similar to the clinical Gorlin formula showed close correlation to the geometricalTHE AMERICAN
JOURNAL
OF CARDIOLOGY
MARCH
15. 1990
761
ly measured size of the aortic orifices used. The cardiac output was computed in liters/min as forward minus regurgitant aortic flow. Data were averaged over 10 cycle periods. Signals were recorded through an analog to digital converter into a disc using a Digital Equipment Corp. computer system. Ten cycles of each waveform were stored using 256 points/cycle. Custom software
FIGURE
1. The stemmed
and cakitted
aortic
valve
before
and after
was used to allow efficient and accurate control of the acquisition, analysis, display and management data. In the first series of experiments using normal saline solution as test fluid, aortic pressure gradients were measured for the 16 largest orifices with a cardiac output of 5.1 liters/min and for the 4 smallest orifices with a cardiac output of 1.3 liters/min. With these results in
balloon
dilatatkm3
adjustable periptwal resistance
characteristic
FIGURE 2. The circulatomymodslappliedin cur experiments. interchang,eable orifice plate to simulate stenosis
blood
analog
hydraulic
fluid
air water
762
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
65
f,h?r
fluid
hand, a second series of experiments was done using saline solution with 5 different cardiac outputs ranging from 1.3 to 4.1 liters/min, with 5 orifices chosen for each output so that transaortic valve pressure gradient measurement ranged between 20 and 335 mm Hg. A limited number of experiments were performed using a mixture of glycerol and saline instead of normal saline solution as test fluid.
cardiac workload (Figure 6). (3) In low flow states significant gradients appear only if the stenosis is very severe and minimal increase or decrease in orifice size may lead to dramatic changes both in transvalvular pressure gradient and in cardiac work (Figure 7). (4) The outcome of the experiments varied only minimally whether they were performed with blood analog test fluid or with physiologic saline solution.
RESULTS
DISCUSSION
The results of our observations are shown in Figures 4 to 7. From these, 4 conclusions are valid. (1) Significant transvalvular pressure gradients appear only at severely narrowed orifices and the magnitude of the pressure gradient is highly flow-dependent (Figures 4 and 5). (2) At critical narrowings the orifice size needs to be increased only minimally to achieve a most significant decrease in transvalvular gradient and to decrease the
Although we describe results obtained for the measurement of pressure across simulated aortic stenosis in vitro, we believe that the clinical implications that may be inferred are significant and multifold. First, our study objectively documents the fact that significant aortic transvalvular gradients appear only at critical stenoses and that their gradient for a given cardiac output may be well alleviated even by a minimal
FIGURE intervals cycle.
3. Flow rate during a
Forwud
Ckrod
ClO8irrg
300
Ordice
Ares
km’)
0.3
28
4
FIGURE 4. thrdath behveen valvular premure gradient, flow ofllice size I.
transand
0
‘, 1.2
I 2.2
I 3.2 FLOW
THE AMERICAN
JOURNAL
RATE
I 4.2
1 5.2
(L/mm)
OF CARDIOLOGY
MARCH
15. 1990
763
increase in the aortic orifice. Patients with low cardiac outputs are especially sensitive to this process. This reconciles the controversy as to why aortic balloon valvuloplasty in some cases may induce dramatic and immediate decreases in the transvalvular pressure gradient despite very modest increases in orifice size.
Also, in some patients even a moderate decrease in cardiac output can cause a significant decrease of transvalvular pressure gradients. Therefore, it appears to be mandatory that during balloon valvuloplasty as well as in the late evaluation both pressures and blood flow should be simultaneously monitored. These observations
FIGURE 5. Car-relation vaivular pressure orifice size II.
1 1.2
11.5 1 1.1
11.3 18 1.0
10.7 1I 0.0
10.1 I. 0.5
ORIFICE 9.4 I1 0.7
DIAMETER (mm) 5.7 5.0 7.1 I 1I ,1 0.6 0.5 0.4
OWFCE
AREA
5.2 I 0.3
5.0 I 0.2
3.6 I 0.1
gradii,
between transttow and
0 1 0
(cd)
flGURE 6. Decrease in orifice size necessaary to increase the transvaivular pressure gradient from 36 to 150 mm Hg.
764
THE AMERICAN
JOURNAL
OF CARDIOLOGY
VOLUME
65
ORIFICE DIAMETER
1 AREA
(mm)
(cm’)
0.0 --
0
3.6 --0.1 5.0 --0.2
6.2
--0.3
7.1 --0.4
0.0
flGURE 7. Condation ohe, lbw end cardiac
between work.
--OS
orifice 0.7 --O.d
9.4 --0.7 Flow 10.1
--OS
10.7
--0.9 6
11.3
--1.0
11.5
--1.1
Al.2
d
0
p
I
I
I
I
0.0
0.5
1.0
1.5
2.0 CARDIAC
also make obsolete the stance7 that the severity of aortic stenosis may be graded by the transvaivular gradient alone. Patients with critical aortic stenoses, whose transvalvular aortic pressure gradient has been alleviated by moderately increasing the aortic orifice with valvuloplasty, may experience sudden return of the gradient if the cardiac output increases on exercise8 or due to improved ventricular function.9 Clinical data show that an overwhelming majority of patients who undergo balloon dilatation for aortic stenosis will end up with an orifice area of about 0.8 cm29 loI2 which clinically is still specified as “severely stenotic.” About 25% of these patients experience periprocedural major complications including death, and an equal number have to undergo valve replacement within a year. At 1 year the actuarial mortality rates have been reported as between 24 and 60%, similar to patients with aortic stenosis of similar degree who were managed medically. 13m16The dangers that a patient faces living with such a marginal valve have already been noted by Rahimtoola and othersi7J8 in connection with undersized (
