You are on page 1of 10

DIAGNOSTIC REVIEW

Aortic Stenosis: Physics and


Physiology—What Do the
Numbers Really Mean?
Arthur E. Weyman, MD, Marielle Scherrer-Crosbie, MD, PhD
Cardiology Division, Massachusetts General Hospital, Harvard Medical School, Boston, MA

Cardiac catheterization and Doppler echocardiography are two methods used to meas-
ure transvalvular gradients and valve area in the assessment of aortic stenosis severity.
Although both approaches are based on the same hemodynamic concepts and report
data using the same units of measure, each method measures pressure drop or gradient
at a different place; hence they produce fundamentally different quantities. Likewise,
cardiac catheterization formulas for valve area attempt to obtain the anatomic area
whereas the Doppler continuity equation reports the area to which flow is constricted.
To use these two methods appropriately, it is necessary to understand the underlying
hemodynamic principles and the effects of the methods of measurement on the values
obtained. This article examines these variables and shows how they affect the reported
gradients and valve areas and how differences can affect clinical application.
[Rev Cardiovasc Med. 2005;6(1)23-32]
DOWNLOAD
© 2005 MedReviews, LLC POWERPOINT FIGURES @
www.medreviews.com

Key words: Aortic stenosis • Catheterization • Doppler echocardiography


• Pressure gradients • Valve area

A
ssessment of the severity of aortic stenosis depends on measurement of
the transvalvular gradients and valve area. Historically these parameters
were determined by cardiac catheterization based on direct measurement
of pressures in the left ventricle and aorta and estimation of transvalvular flow
using the Fick, dye dilution, or thermodilution method. At present, however, gra-
dients and valve areas are generally determined using noninvasive Doppler
echocardiographic methods. Although both of these approaches derive from

VOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINEºººº23


Aortic Stenosis continued

similar hemodynamic concepts and


report their results in the same units
of measure, the catheter and Doppler
measurements of pressure drop or
gradient are not made at the same Cc = Effective orifice area
place and therefore are fundamen- Anatomic area
tally different quantities. Likewise,
catheterization formulas for valve
area attempt to derive the anatomic Flow
area whereas the Doppler continuity
equation reports the area to which Ao LV
flow is constricted or the effective
valve area. The effect of these differ-
ences is exaggerated because impor- Effective Area
tant changes in the classification of
Anatomic Area
severity are based on differences in
valve areas of only tenths of a square
centimeter. This is further complicat-
ed by the fact that although clinical Figure 1. The convergence of flow distal to a stenotic valve orifice produces an effective orifice or area to which
flow is constricted. The ratio of the effective area to the anatomic area is the coefficient of contraction (Cc). The
decisions are now commonly made
Cc is determined by the valve shape and the orifice size and shape. Ao, aorta; LV, left ventricle. Reproduced with
based on Doppler echocardiographic permission from Gilon et al.16 www.medreviews.com
data, criteria for severity are based on
more traditional catheterization data
in which outcomes have been more of gradients and valve areas, it is first called the coefficient of contraction
extensively studied.1 The appropriate necessary to appreciate the hydrody- (Figure 1). This ratio is affected by
use of these frequently differing data namic pattern of flow through a the size and shape of the orifice and
therefore depends on an understand- stenosis. In any closed system, the by the inlet geometry. To determine
ing of the underlying hemodynamic law of conservation of mass requires the anatomic orifice size, some cor-
principles and effects of the methods the volume flow to be constant at rection must be introduced to
of measurement on the resulting val- all points. Because flow equals the account for this coefficient of con-
ues. The purpose of this review is to product of mean velocity and vessel traction, which is unknown in most
examine these variables and to show area, as the area of the flow stream cases. Such calculations are facilita-
how they affect the reported gradi- decreases as blood approaches a ted by a term used in engineering,
ents and valve areas and how differ- stenosis, the velocity must increase the coefficient of discharge, which
ences can affect clinical application. proportionately in order to maintain compares the actual flow through
Although differences between constant flow. As flow exits the the valve with the predicted flow if
Doppler echocardiographic and stenotic orifice, the streamlines con- no contraction or viscous losses
catheterization data are often a result tinue to converge for a short dis- occur.2 The coefficient of discharge is
of technical errors in data acquisition, tance (the vena contracta effect) specific for an individual valve shape
inherent errors in the component because inertia prevents the periph- and size and converts the effective
methods, or simply differences in the eral streamlines entering the steno- area to the anatomic area.
parameter measured (eg, the differ- tic orifice from the side from Distal to the stenosis, there is an
ence between peak instantaneous and changing directions instantly. By the abrupt separation of the jet from the
peak-to-peak gradients), for the pur- time the flow lines have again vessel wall, resulting in the jet shear-
poses of this review, we assume that become parallel, the cross-sectional ing against the stagnant blood in the
all the data are appropriately recorded area of the jet will be smaller than sinuses of Valsalva, causing vortex
and instantaneously compared. the orifice area and the velocity of formation and a zone of recirculation
the jet will be proportionately higher. (Figure 2). The abrupt change in ves-
Characteristics of Flow The ratio of the jet area at the vena sel diameter (orifice diameter to aor-
Through Stenotic Valves contracta (the area to which flow is tic diameter) suddenly increases the
To understand the differences in constricted or effective orifice area ratio of kinetic to viscous forces, rep-
catheter and Doppler measurements [Aeff]) to the anatomic area (Aanat) is resented by the Reynolds number, up

24ººººVOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINE


Aortic Stenosis

result, its potential energy (lateral


pressure) must decrease to maintain
total energy constant. In pulsatile
systems, additional energy may be
required to overcome inertia and
accelerate blood to its peak velocity.
Energy may also be lost as heat as a
result of viscous friction. These rela-
tionships can be expressed mathe-
matically using the Bernoulli
equation,
P = 1 (22 – 21) + ∫ 2
1
dv ds + R(, )
2 dt
Eq.1

where P is the difference in the


pressures proximal and distal to the
stenosis, 1 and 2 are the velocities
proximal to the stenosis and at the
vena contracta, s is the distance over
which flow accelerates, R is viscous
resistance,  is the mass density of
blood, and  is the viscosity. The
first term in the equation accounts
for convective acceleration of flow
through the stenosis, the second
Orifice Flow term for flow acceleration, and the
third for viscous friction. In the clin-
ical situation, viscous friction has
been shown to be negligible for dis-
crete orifices ≥ 0.25 cm2, because
blood velocity is approximately con-
stant across the orifice, and as a
result, there is no frictional loss
Laminar core
between adjacent fluid layers. In
Turbulent eddies
addition, although the need for flow
Figure 2. Flow visualization study illustrating the pattern of flow distal to a stenosis. Immediately beyond the to accelerate from zero delays the
orifice, the jet shears against the stagnant blood in the parajet region, causing turbulent eddies that erode the velocity waveform slightly relative
laminar core, resulting in the loss of energy as heat. Reproduced with permission from Levine et al.7
to the pressure waveform, it does not
www.medreviews.com
significantly alter the calculation of
the peak gradient (because at peak
to fivefold. The Reynolds number valvular pressure drop or gradient is velocity, dv/dt [acceleration] = 0).
(Re = blood density velocity x vessel based on the law of conservation of The mean gradient is also unaffected
diameter/blood viscosity) is a meas- energy, which states that for flow in because the lag in velocity is rough-
ure of the tendency of the blood to a closed system, the total energy at ly symmetric during acceleration
become turbulent, and this abrupt all points must remain constant. and deceleration. At other points in
increase is associated with flow insta- Because the linear velocity in cen- the cardiac cycle, the acceleration
bility and turbulence. In this turbu- timeters per second through a term produces small discrepancies
lent zone, kinetic energy is lost as stenotic valve must increase as the between pressure gradient and
heat that is nonrecoverable. valve area decreases (law of conser- velocity, but these are not clinically
vation of mass), its kinetic energy, important. Because the viscous fric-
Pressure Gradients which is proportional to the square tion and flow acceleration terms are
Doppler measurement of the trans- of the linear velocity, increases. As a negligible,3,4 they can be ignored,

VOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINEºººº25


Aortic Stenosis continued

and equation 1 simplifies to kinetic energy is reconverted to pres- Doppler and catheter measurements
P = 1 (22 – 21) sure. It is this downstream loss of of pressure occur at different places,
2
Eq. 2 energy that is the important effect of they measure fundamentally differ-
the stenosis, because if velocity were ent quantities and may report signifi-
In addition, in most stenotic simply reconverted to pressure (pres- cantly different results when pressure
lesions, 22 >> 12, so that v1 can be sure recovery), there would be no recovery is present.
ignored, and the pressure gradient energy loss to the system and the
after correction for different units  stenosis would have no significant Pressure Recovery
= 1.06/981 g s2/mL x 1/1.36 (to con- hemodynamic effect. This lack of Although clinical studies comparing
vert dyne centimeters to millimeters effect would occur despite the fact Doppler and catheter gradients gen-
of mercury [mm Hg]) x 1/2 , which that Doppler would record an erally show excellent correlations,
after appropriate conversion of increase in velocity at the stenosis the slopes of the regression lines
measurement units (x 104) = 3.972, and pressure taps proximal to the often differ significantly, with the
or roughly 4, so that stenosis and at the vena contracta Doppler gradients frequently over-
AP = 42
Eq. 3 It is this downstream loss of energy that is the important effect of the
The simplified Bernoulli equation stenosis, because if velocity were simply reconverted to pressure (pressure
permits calculation of the trans- recovery), there would be no energy loss to the system and the stenosis
valvular gradient at each instant in would have no significant hemodynamic effect.
systole; however, two measures of
transvalvular pressure gradient are would register a corresponding fall in estimating those reported at
usually reported from Doppler aor- pressure. In practice, the conversion catheterization. Some of this overes-
tic velocity profiles: the peak gradi- of kinetic energy to heat is not always timation has been attributed to fail-
ent and the mean gradient. The complete and some pressure can be ure to account for the proximal
peak gradient is determined from recovered (see below). velocity in the Bernoulli equation in
the peak velocity, whereas the mean high-flow states. However, a num-
gradient is the mean of the squared Catheter Measures of ber of experimental studies have
instantaneous velocities recorded Transvalvular Gradients demonstrated that as the stream-
during the systolic ejection period. Experimental studies comparing lines of flow reattach to a vessel wall
It is important to remember that Doppler gradients measured using downstream from a stenosis, some
continuous-wave Doppler measures the peak velocity at the vena contrac- of the momentum in the jet is
the change in velocity from a point ta with catheter values obtained by reconverted to lateral or pressure
energy, a phenomenon known as
A number of experimental studies have demonstrated that as the stream- pressure recovery.5-7 Because the
lines of flow reattach to a vessel wall downstream from a stenosis, some energy loss in aortic stenosis is the
result of flow separation and vortex
of the momentum in the jet is reconverted to lateral or pressure energy,
formation (turbulence), the extent
a phenomenon known as pressure recovery. of this phenomenon depends on
the size relationship between the
proximal to the onset of convective positioning a catheter at the level of orifice and aorta.5,6,8 The smaller the
acceleration toward the stenosis to the vena contracta have uniformly valve orifice relative to the size of
the peak velocity at the vena contrac- reported slopes close to unity. In clin- the aorta, the more turbulence will
ta. A critical but unstated assumption ical practice, however, it is very diffi- occur, and because turbulence
in using this increase in velocity as a cult to position a catheter in the vena results in the conversion of kinetic
measure of the pressure gradient contracta because it cannot be seen energy to heat that is nonrecover-
across the valve is that all of the pres- and the jet itself tends to displace the able, less energy will be available to
sure that is converted to kinetic ener- catheter. As a result, catheter pres- be recovered as pressure. Conversely,
gy (velocity) is then lost as heat in the sures are usually recorded in the the larger the valve orifice or the
turbulent eddies downstream from ascending aorta after pressure recov- smaller the aorta, the less turbulence
the stenosis and that none of the ery has occurred. Because clinical will occur and the greater the

26ººººVOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINE


Aortic Stenosis

pressure recovered. Although the


distance required for maximal pres- 200 125

Mean Doppler Gradient (mm Hg)


Peak Doppler Gradient (mm Hg)
sure recovery varies with orifice size
and aortic diameter, the majority of 100
150
the pressure is usually recovered
75
within a few centimeters (~5 cm)
100
beyond the vena contracta.5,6
Aorta 1.8 50 Aorta 1.8
When pressure recovery occurs, Aorta 2.4 Aorta 2.4
the actual pressure loss to the sys- 50 Aorta 3.0 Aorta 3.0
25
tem, termed the head loss, will be Aorta 4.0 Aorta 4.0
Aorta 5.0 Aorta 5.0
less than that reflected by the
0 0
increase in kinetic energy at the 0 50 100 150 200 0 25 50 75 100 125
vena contracta, and the pressure Peak Catheter Gradient (mm Hg) Mean Catheter Gradient (mm Hg)
gradient between the left ventricle
and aorta (the net pressure gradient) Figure 3. Effect of aortic size on the correlation between peak Doppler and catheter gradients (left) and mean
Doppler and mean catheter gradients (right) for a valve area of 1.25 cm2 with aortas 1.8, 2.4, 3.0, 4.0, and 5.0 cm
will be less than the gradient in diameter. Slopes vary from 1.1 to 1.86. Reproduced with permission from Niederberger et al.8
between the left ventricle and the www.medreviews.com
vena contracta. Figure 3 compares
the effects of aortic size on the cor- 200 150

Mean Doppler Gradient (mm Hg)


relations between peak and mean
Peak Doppler Gradient (mm Hg)

Doppler and catheter gradients for a


150
1.25 cm2 orifice with aortas of vary-
100
ing diameters. As illustrated, the
smallest aortic diameters were asso- 100
ciated with the greatest discrepan-
cies between the Doppler and AVA 1.25 50 AVA 1.25
50 AVA 1.0 AVA 1.0
catheter gradients. Figure 4 illus-
AVA 0.75 AVA 0.75
trates the effect of valve area on AVA 0.5
AVA 0.5
pressure recovery for a fixed aortic 0 0
0 50 100 150 0 25 50 75 100 125
size. As predicted, the larger the ori-
Peak Catheter Gradient (mm Hg) Mean Catheter Gradient (mm Hg)
fice, the greater the amount of pres-
sure recovery. In small clinical
Figure 4. Effect of orifice size on the correlations between peak Doppler and peak catheter gradients (left) and
studies, an increase in peak down- between mean Doppler and mean catheter gradients (right) for an aorta with a 1.8 cm diameter and 0.5, 0.75,
stream pressure of between 4 and 82 1.0, and 1.25 cm2 orifices. Slopes range from 1.28 to 1.86. AVA, aortic valve area. Reproduced with permission
from Niederberger et al.8 www.medreviews.com
mm Hg has been reported. In
patients with aortas larger than 3
cm, only small differences in correction factor is given as peak and mean catheter gradients in
catheter and Doppler gradients were a group of patients with valvular aor-
P = 4V (1 – C)
2

observed (peak, 7.3 ± 8.7 mm Hg; with tic stenosis. Using this correction
mean, 2.6 ± 6.1 mm Hg), whereas in decreased the slopes of the peak and
C = 2[(Aeff /AA) – Aeff2 /AA2 ],
the subgroup with aortas ≥ 3 cm, Eq. 4 mean Doppler gradients from 1.36
greater degrees of pressure recovery and 1.25, respectively, for the peak
were observed (peak, 24.8 ± 19.7 where Aeff is the vena contracta area and mean uncorrected values to 1.03
mm Hg; mean, 16.2 ± 13.2 mm Hg).9 and AA is the area of the aorta. The and 0.96 for the corrected Doppler
On the basis of fluid mechanics calculated Aeff assumes a circular ori- values when compared with catheter-
theory, the Doppler-predicted pres- fice, which is appropriate given that derived gradients.
sure drop can be corrected for the pressure recovery is not affected by For any given valve area and aortic
size of the aorta in order to derive the shape of the orifice.10 Figure 5 diameter, the orifice velocity and
the actual pressure drop (head loss) compares peak and mean Doppler therefore the gradient will depend on
or net pressure gradient after pres- gradients with the Doppler gradients the flow rate. The absolute amount
sure recovery. This area-based predicted using equation 4 with the of pressure recovery increases

VOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINEºººº27


Aortic Stenosis continued

valve orifice, the streamlines almost


200 150
immediately reattach to the vessel

Mean Catheter Gradient (mm Hg)


Peak Catheter Gradient (mm Hg)

y = 1.02x – 1.0 y = 1.04x + 0.2


SEE = 6 mmHg SEE = 3 mmHg wall and the head loss is only 15%
150 r = 0.99 r = 0.99 (pressure recovery 85%).

Doppler Predicted
Doppler Predicted

100 Importantly, when pressure recov-


100
ery occurs, the head loss (which is
the loss of energy to the system and
50 is measured at catheterization where
50 the gradient after pressure recovery is
Line of Identity Line of Identity
recorded) is the appropriate measure
0 0 of energy loss to the system and
0 50 100 150 200 0 50 100 150
determines the left ventricular pres-
Peak Catheter Gradient (mm Hg) Mean Catheter Gradient (mm Hg)
sure required to maintain a given
aortic pressure. However, the
Figure 5. Correlation between Doppler-predicted catheter gradients (Doppler gradient minus predicted pressure Doppler gradient, which measures
recovery) and observed catheter gradients for central jets (left, peak gradients; right, mean gradients). The dashed line
represents the line of identity. SEE, standard error of estimate. Reproduced with permission from Niederberger et al.8 the conversion of pressure to kinetic
www.medreviews.com energy induced by the stenosis, is the
appropriate gradient to use to calcu-
with flow rate (orifice velocity). The pressure gradient (catheter after pres-
sure recovery) may differ significant- late the effective orifice area.
percent overestimation, however,
remains approximately the same and ly and therefore affect management,
particularly when the gradients are Valve Area Determination
therefore is independent of flow.
used to calculate valve area. Because the pressure drop or gradi-
Jet eccentricity also affects pressure
In addition to the ratio of the ent across a stenotic valve varies
recovery, with more eccentric jets
stenotic orifice to the aorta, both with flow, it has become common
showing decreasing Doppler-catheter
the inlet and outlet geometry of the practice to calculate the valve area,
differences (less pressure recovery
stenosis importantly affect the which is a flow-independent meas-
because momentum is lost when the
amount of pressure recovery and thus ure of severity.
jet strikes the vessel wall). In this
case, the correction factor described the relationship between the catheter-
Doppler Echocardiographic Estimation
above is no longer applicable. and Doppler-measured gradients.
of Valve Area: the Effective Valve Area
Clinically, pressure recovery is Valvular stenoses are usually discrete,
The calculation of the aortic valve
most relevant in patients with mod- with abrupt narrowing and expan-
area from Doppler recordings is
erate aortic stenosis, small aortas, sion, and although pressure recovery
based on the law of conservation
and high flow rates. In these cases, is present, it is usually not great. For
of mass, which states that for an
the maximal pressure drop at the more tapering stenoses—such as
incompressible fluid in a closed sys-
vena contracta (Doppler) and net those that often characterize sub-
tem, flow (Q) at all points must
valvular7 and supravalvular obstruc-
Figure 6. Pressure recovery-flow through a Venturi tube. remain constant. As illustrated in
The solid line gives the pressure distribution along the tion and coarctation—the shape of
center line; the dashed curve gives the pressure distribu- the outlet becomes more important
tion along the wall. A, tube area upstream from the Figure 7. Calculation of aortic valve area using the
stenosis; a, stenosis area at the vena contracta; po , pres- than the simple orifice diameter to continuity principle. Aortic outflow is calculated as
the product of the subvalvular area A1 and velocity
sure proximal to the stenosis; p, pressure at the vena downstream vessel diameter ratio,
contracta. Reproduced with permission from Prandt V1. The stenotic valve area is then equal to A1 x V1
and Teitjens.17
and pressure recovery may cause a divided by the peak transvalvular velocity V2. LA, left
greater disparity between Doppler- atrium; Ao, aorta; LV, left ventricle.
and catheter-measured gradients than
would be predicted simply for the ori-
A a
fice aortic ratio. Figure 6 illustrates A2•V2
Ao
p this effect for a Venturi tube with a
Po A1•V1
gradually tapering outlet, which per- LV
po – p 0.15 (po – p) mits almost immediate reattachment LA
of the flow streamlines to the vessel
wall. In this example, although there
www.medreviews.com is a significant pressure loss at the www.medreviews.com

28ººººVOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINE


Aortic Stenosis

Figure 7, the flow through the out- and V2 are used, it can be assumed Thus, substituting (1 mm Hg = 1333
flow tract must be the same as the that the maximal A2 is calculated dyne/cm2) and  = 1.05 g/mL gives
flow through the valve (Q1 = Q2) at because the gradient is greatest at
V = (2*1.333P/1.05)1/2 =
any point in time. In addition, this point, forcing the valve to open
(2,539P)1/2 = 50.4 P
because flow equals mean velocity maximally. Conversely, using the
Eq. 7
times area at any point, mean velocity will give the average
area occurring throughout systole. and
Q1 = Q2 = A1 *V1 = A2 V2
Eq. 5 One may also use the instantaneous Aeff = Q
velocities throughout systole to cal- 50.4 P
If flow through the valve is Eq. 8
culate the instantaneous valve areas
known or can be determined from and thereby detect any flow-related This equation calculates the effective
the product of area and velocity at a changes. orifice area and yields the same result
reference level, and the velocity at Because the peak Doppler velocity as the Doppler continuity equation.
the stenosis can be recorded, then is the velocity at the vena contracta, Because the effective orifice area, or
the area at the point of stenosis can the calculated valve area will be the the area to which flow is constricted,
be calculated as follows: smallest area to which the flow is related to the anatomic area of the
A2 = A1 *V1 stream is reduced, which will be valve by the coefficient of discharge:
V2 equal to the anatomic area reduced
Eq. 6 Aeff = Aanat •CD
by the coefficient of discharge. The
or
where V2 is the velocity and A2 the effective area is the appropriate
Aanat = Aeff
area at the vena contracta and V1 hydrodynamic area, but should be CD
Eq. 9

The impact of a stenosis on pressure and flow depends not only on the The Gorlin Equation
cross-sectional area of the orifice but also on the three-dimensional The first clinical application of these
geometry of the leaflets proximal to the orifice. concepts was by Gorlin and Gorlin
in 1951.11 In their original formula,
and A1 are the velocity in the out- smaller than the area calculated at they included a constant to account
flow tract proximal to the onset of catheterization using the Gorlin for- for the coefficient of contraction
convective acceleration. Because mula, which includes a constant to and thereby attempted to correct
both V1 and V2 increase proportion- account for the coefficient of dis- the flow area to the anatomic area,
ately with increasing flow, the calcu- charge and thus attempts to convert given that their standard of refer-
lated area should be independent of hydrodynamic area to anatomic area. ence was excised valves. Thus,
flow. When flow through the out- Aanat = Q
flow tract is used as a reference, it Calculation of Valve Area at C • 44.3 P
includes both forward and aortic Catheterization Eq. 10
regurgitant flow so that the valve Theoretical Background
areas calculated using the continuity Valve area can also be calculated where 44.3 = 2 • 981. The empiric
equation will be accurate whether using the pressure drop or gradient constant C in the original formula-
aortic regurgitation is present or not. across the valve. To do this requires tion included the coefficients of
The continuity equation is instan- use of both the continuity and contraction Cc and viscosity Cv as
taneously valid so that the velocity Bernoulli equations, where well as correction for the conversion
used in the equation may be either Q = Aeff *V of centimeters of water (cm H2O) to
the stroke velocity integral or the and mm Hg. Recognizing that blood vis-
peak velocity (assuming that the P = 1V2. cosity, turbulence, pulsatile flow,
2
valve area does not change). If the and the inconstant shape of
orifice is elastic and varies because of Solving for V gives deformed valves made it almost
changes in the gradient across the impossible to predict the discharge
V = Q/Aeff and V = (2P/)1/2.
obstruction during pulsatile flow, coefficient analytically, they deter-
the size of the valve area calculated In this equation, P is in metric mined an empiric coefficient from
by equation 6 will depend on when units (dyne/cm2), whereas clinically direct measurement of mitral valves
V1 and V2 are determined. If peak V1 it is usually expressed in mm Hg. at surgery or autopsy. For the mitral

VOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINEºººº29


Aortic Stenosis continued

cient. Because flow contraction by


Discharge coefficient itself should not cause these
changes, they have been attributed
to the viscosity,13 because for an
0.85 eccentric jet, the jet perimeter—
where viscous losses occur—is larger
relative to the cross-sectional area
0.80 than for a circular jet. Likewise, for a
small jet, the perimeter-to-area ratio
is greater than for a large jet.
0.75 The shape of the valve inlet also
affects the coefficient of discharge.13
Figure 9 illustrates the relationship of
0.70 valve shape to coefficient of contrac-
1:1 tion as a function of valve area
2:1
3:1 derived from three-dimensional mod-
0.65 5:1 ECC els of clinical echocardiographically
0.3 0.5 1 1.5 2 2.5
imaged aortic valves. As illustrated,
Area [cm2] the impact of a stenosis on pressure
and flow depends not only on the
Figure 8. Influence of valve area and eccentricity (ECC) on discharge coefficients. The discharge coefficient for each
area decreases with higher eccentricity; for each eccentricity, the discharge increases as area increases. Reproduced cross-sectional area of the orifice but
with permission from Flachskampf et al.13 www.medreviews.com also on the three-dimensional geome-
try of the leaflets proximal to the ori-
valve, they concluded that the mean pressure gradient in the Gorlin fice. The geometry determines the
anatomic valve area equals formula instead of the mathemati- pattern of flow convergence and thus
cally correct mean of the square the relationship of Aeff to Aeff. Patients
Aanat = Q/31 P1 – P2 with flat valves and steeper flow con-
roots of the instantaneous gradients,
Eq. 11 which systematically underestimates vergence have smaller Aeff than those
the valve area and results in a value with more gradually tapered domed
which corresponds to a discharge valves for the same anatomic area and
closer to the physiologic area.12,13
coefficient of roughly 0.62. flow rate. Corresponding pressure
Thus the Gorlin formula actually
For the aortic valve, they used the losses may be increased by up to 40%,
calculates the flow area (Aeff) offset by
constant C = 1, which results in increasing the ventricular workload
a small constant and will correspond
Aanat = Q to the anatomic area only by chance proportionately. Theoretically, hemo-
44.3 P (ie, when the empiric coefficient of dynamic severity may increase
Eq. 12
discharge happens to be appropriate because of changes in valve shape
Use of the constant 44.3 ignores the for the valve in question). without change in orifice area, as with
conversion of cm H2O to mm Hg Does the Gorlin constant bear a the progressive calcification of a bicus-
and the mass density of blood but consistent relationship with the pid valve.
differs from equation 6 only in the anatomic area? Recent studies have
constant (50.4 vs 44.3). Because shown that the true coefficient of
anatomic area can in theory be cal- discharge is not constant but rather Effects of Pressure Recovery
culated as varies with valve orifice size and on Valve Area Measurements
Aanat = Q/(CD50.4 P) shape, but, importantly, is inde- Pressure recovery also affects the
pendent of flow, except at very low measurement of valve area at
Eq. 13
flow rates.13 The discharge coeffi- catheterization because the net pres-
the Gorlin constant corresponds to cient increases with increasing valve sure gradient (measured between the
CD*50.4, which implies a discharge area and decreases with increasing left ventricle and aorta after pressure
coefficient of 0.879. eccentricity. Figure 8 illustrates the recovery) is smaller than the gradient
This correction is partially offset combined effects of orifice area and at the vena contracta, therefore
by the use of the square root of the eccentricity on the discharge coeffi- resulting in a larger calculated valve

30ººººVOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINE


Aortic Stenosis

area. The valve area calculated using


the net gradient does not correspond
to either the effective or anatomic Domed Intermediate Flattened
valve area but rather represents the
area that would have occurred had
there been no pressure recovery: a
theoretic value, but one that reflects Anatomic Area
the hydrodynamic impact of the
stenosis. Use of the net gradient fur-
ther increases the difference between 1.0 cm2 0.9 0.85 0.76
Doppler measures of the effective
valve area and the area calculated at
catheterization.
Because the amount of pressure 0.75 cm2 0.88 0.83 0.74
recovery is determined by the ratio
of the orifice area to the aortic area,
a correction factor has been pro-
posed to correct the Doppler conti- 0.5 cm2 0.85 0.81 0.71
nuity valve area for the effect of
pressure recovery.14 Using this cor- Figure 9. Coefficients of contraction according to valve shape and orifice size. Reproduced with permission from
Gilon et al.16 www.medreviews.com
rection, Doppler Aeff is multiplied by
Aa /Aa – Aeff = 1 + Aeff /Aa – Aeff
Cardiology guidelines.1 Further, larger than that measured using
where Aa is the area of the ascending although Aeff and CD • Aanat are meas- Doppler echocardiography when
aorta. As can be appreciated from ures of the severity of stenosis, Aanat these data are accurately recorded
this equation, in patients with tight is not because the hemodynamic and simultaneously compared, even
aortic stenosis in which the Aeff is severity of stenosis may increase in the absence of pressure recovery.
much smaller than the ascending with no change in the anatomic When pressure recovery is present,
aortic area, there will be little differ- area, as a result of progressive dila- this difference will increase and may
ence between the actual and cor- tion of the aorta with a correspon- result in significant misclassification
rected Aeff. Table 1 illustrates the ding decrease in pressure recovery. of severity. Understanding these dif-
relationship between the Gorlin- ferences allows appropriate correc-
derived effective areas and the Summary tions to be introduced and the total
Doppler areas for a series of aortic The area of a stenotic aortic valve effect of the stenosis on the system to
diameters.14 Relating these changes calculated using standard hemody- be appreciated. Alternatively, more
to current criteria for severity, it can namic methods will be consistently conservative standards may be more
be shown that a patient with an
ascending aortic diameter of 2.6 cm
and a Doppler Aeff of 0.9 cm2 would Table 1
have a valve area of 1.1 cm2 after Doppler-Derived Effective Orifice Area (EOA)
correcting for pressure recovery.1,15
This effect would be greater for a
Catheter- Catheter- Aortic Aortic Aortic
patient with the same aortic diame- Derived Diameter Diameter Diameter Derived
ter and a Doppler Aeff of 1.2 cm2, in EOA (cm2) EOA (cm2) = 2.0 cm (Area = 3.0 cm (Area = 4.0 cm (Area
which the corrected area would cor- (Constant 50) (Constant 44.3) = 3.14 cm2) = 7.07 cm2) = 12.6 cm2)
respond to a recovered catheter- 1.50 1.69 1.02 1.24 1.34
derived area of 1.6 cm2.15 In both of 1.00 1.13 0.76 0.88 0.93
these cases, the change would shift 0.75 0.85 0.61 0.68 0.71
the patient to a lower grade of sever- 0.50 0.56 0.43 0.47 0.48
ity based on current American Heart Adapted from Garcia et al.14
Association/American College of

VOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINEºººº31


Aortic Stenosis continued

appropriate for estimating severity 5. Clark C. The fluid mechanics of aortic steno- 12. Dumesnil JG, Yoganathan AP. Theoretical and
sis—I. Theory and steady flow experiments. J practical differences between the Gorlin for-
using Doppler echocardiographic Biomech. 1976;9:521-528. mula and the continuity equation for calcu-
data than those currently recom- 6. Clark C. The fluid mechanics of aortic steno- lating aortic and mitral valve areas. Am J
sis—II. Unsteady flow experiments. J Biomech. Cardiol. 1991;67:1268-1272.
mended for catheterization values. 1976;9:567-573. 13. Flachskampf FA, Weyman AE, Guerrero JL,
7. Levine RA, Jimoh A, Cape EG, et al. Pressure Thomas JD. Influence of orifice geometry
recovery distal to a stenosis: potential cause of and flow rate on effective valve area: an in
References gradient “overestimation” by Doppler echocar- vitro study. J Am Coll Cardiol. 1990;15:
1. Bonow RO, Carabello B, de Leon AC, et al. diography. J Am Coll Cardiol. 1989;13:706-715. 1173-1180.
ACC/AHA guidelines for the management of 8. Niederberger J, Schima H, Maurer G, 14. Garcia D, Dumesnil JG, Durand LG, et al.
patients with valvular heart disease. Executive Baumgartner H. Importance of pressure recov- Discrepancies between catheter and Doppler
summary. A report of the American College of ery for the assessment of aortic stenosis by estimates of valve effective orifice area can be
Cardiology/American Heart Association Task Doppler ultrasound. Role of aortic size, aortic predicted from the pressure recovery phenom-
Force on Practice Guidelines (Committee on valve area, and direction of the stenotic jet in enon: practical implications with regard
Management of Patients with Valvular Heart vitro. Circulation. 1996;94:1934-1940. to quantification of aortic stenosis severity. J
Disease). J Heart Valve Dis. 1998;7:672-707. 9. Baumgartner H, Stefenelli T, Niederberger J, et Am Coll Cardiol. 2003;41:435-442.
2. Fox RMT. Incompressible Viscous Flow. New al. “Overestimation” of catheter gradients by 15. Levine RA, Schwammenthal E. Stenosis is in
York: John Wiley and Sons; 1978. Doppler ultrasound in patients with aortic the eye of the observer: impact of pressure
3. Holen J, Waag RC, Gramiak R, et al. Doppler stenosis: a predictable manifestation of pressure recovery on assessing aortic valve area. J Am
ultrasound in orifice flow. In vitro studies of recovery. J Am Coll Cardiol. 1999;33:1655-1661. Coll Cardiol. 2003;41:443-445.
the relationship between pressure difference 10. Voelker W, Reul H, Stelzer T, et al. Pressure 16. Gilon D, Cape EG, Handschumacher MD, et
and fluid velocity. Ultrasound Med Biol. recovery in aortic stenosis: an in vitro study in al. Effect of three-dimensional valve shape on
1985;11:261-266. a pulsatile flow model. J Am Coll Cardiol. the hemodynamics of aortic stenosis: three-
4. Teirstein PS, Yock PG, Popp RL. The accuracy 1992;20:1585-1593. dimensional echocardiographic stereolithography
of Doppler ultrasound measurement of pres- 11. Gorlin R, Gorlin SG. Hydraulic formula for and patient studies. J Am Coll Cardiol. 2002;
sure gradients across irregular, dual, and tun- calculation of the area of the stenotic mitral 40:1479-1486.
nellike obstructions to blood flow. Circulation. valve, other cardiac valves, and central circu- 17. Prandt L, Teitjens OJ. Applied Hydro- and
1985;72:577-584. latory shunts. I. Am Heart J. 1951;41:1-29. Aeromechanics. New York: Dover; 1957.

Main Points
• To understand the differences in catheter and Doppler measurements of gradients and valve areas, it is first necessary
to appreciate the hydrodynamic pattern of flow through a stenosis.
• Doppler measurement of the transvalvular pressure drop or gradient is based on the law of conservation of energy,
which states that for flow in a closed system, the total energy at all points must remain constant. Catheter pressures
are usually recorded in the ascending aorta after pressure recovery has occurred. Because clinical Doppler and catheter
measurements of pressure occur at different places, they measure fundamentally different quantities and may report
significantly different results when pressure recovery is present.
• The calculation of the aortic valve area from Doppler recordings is based on the law of conservation of mass, which
states that for an incompressible fluid in a closed system, flow at all points must remain constant. Catheterization
calculates valve area by measuring the pressure drop or gradient across the valve using continuity and Bernoulli
equations.
• The area of a stenotic aortic valve calculated using standard hemodynamic methods will be consistently larger than
that measured using Doppler echocardiography when these data are accurately recorded and simultaneously com-
pared, even in the absence of pressure recovery. When pressure recovery is present, this difference will increase and
may result in significant misclassification of severity.

32ººººVOL. 6 NO. 1ºº2005ººººREVIEWS IN CARDIOVASCULAR MEDICINE