Author + information
- Received May 6, 2002
- Revision received August 2, 2002
- Accepted August 19, 2002
- Published online February 5, 2003.
- Damien Garcia, Eng*,
- Jean G Dumesnil, MD, FACC†,
- Louis-Gilles Durand, Eng, PhD*,
- Lyes Kadem, Eng† and
- Philippe Pibarot, DVM, PhD, FACC*,†,* ()
- ↵*Reprint requests and correspondence:
Dr. Philippe Pibarot, Quebec Heart Institute/Laval Hospital, Laval University, 2725 Chemin Sainte-Foy, Sainte-Foy, Quebec, Canada, G1V-4G5.
Objectives We sought to obtain more coherent evaluations of aortic stenosis severity.
Background The valve effective orifice area (EOA) is routinely used to assess aortic stenosis severity. However, there are often discrepancies between measurements of EOA by Doppler echocardiography (EOADop) and those by a catheter (EOAcath). We hypothesized that these discrepancies might be due to the influence of pressure recovery.
Methods The relationship between EOAcathand EOADopwas studied as follows: 1) in an in vitro model measuring the effects of different flow rates and aortic diameters on two fixed stenoses and seven bioprostheses; 2) in an animal model of supravalvular aortic stenosis (14 pigs); and 3) based on catheterization data from 37 patients studied by Schöbel et al.
Results Pooling of in vitro, animal, and patient data showed a good correlation (r = 0.97) between EOAcath(range 0.3 to 2.3 cm2) and EOADop(range 0.2 to 1.7 cm2), but EOAcathsystematically overestimated EOADop(24 ± 17% [mean ± SD]). However, when the energy loss coefficient (ELCo) was calculated from EOADopand aortic cross-sectional area (AA) to account for pressure recovery, a similar correlation (r = 0.97) with EOAcathwas observed, but the previously noted overestimation was no longer present.
Conclusions Discrepancies between EOAcathand EOADopare largely due to the pressure recovery phenomenon and can be reconciled by calculating ELCo from the echocardiogram. Thus, ELCo and EOAcathare equivalent indexes representing the net energy loss due to stenosis and probably are the most appropriate for quantifying aortic stenosis severity.
According to the American College of Cardiology/American Heart Association (ACC/AHA) recommendations (1), the aortic valve effective orifice area (EOA) can be used to grade aortic stenosis severity as follows: mild at >1.5 cm2; moderate at >1.0 to ≤1.5 cm2; and severe at ≤1.0 cm2. In the clinical situation, the valve EOA is routinely determined by using either the Gorlin formula during cardiac catheterization or the continuity equation during Doppler echocardiography (2–6). However, there are often discrepancies between catheter- and Doppler-derived valve EOAs, resulting in potentially divergent estimations of aortic stenosis severity. Because catheter pressure measurements are generally performed a few centimeters downstream from the aortic valve, we hypothesized that these discrepancies might be due to the pressure recovery phenomenon, which mainly depends on the size of the ascending aorta (7–16). Thus, the objective of this study was to more closely examine the relationship between catheter (EOAcath) and Doppler (EOADop) measurements of EOA to reconcile such measurements and present more coherent evaluations of aortic stenosis severity.
The transvalvular pressure gradient through a stenotic valve is maximal (TPGmax) at the level of the vena contracta. However, it is generally difficult to obtain an adequate measurement of TPGmaxby a catheter because of the difficulty in adjusting and maintaining the position of the pressure sensor or pressure lumen orifice at the level of the vena contracta, as well as the position instability caused by flow-jet turbulences. Nonetheless, when TPGmax(mm Hg) is successfully measured by a catheter, the EOA at the vena contracta (cm2; EOAcath/max) can be calculated as follows using the Gorlin formula: (1)where Q is the flow rate in ml/s. Previous studies have demonstrated that the original Gorlin formula contains several errors that can be corrected by using a constant of 50 instead of 44.3 (17). To ensure the coherence of the results from both a theoretical and physiologic standpoint, we thus elected to use a constant of 50 in Equation 1. Because the EOA measured by Doppler using the continuity equation (EOADop) is also meant to represent the EOA at the vena contracta, there should theoretically be a close agreement between EOAcath/maxand EOADop.
After the vena contracta, part of the jet kinetic energy is recovered in pressure, resulting in a net pressure gradient (TPGnet) lower than TPGmax, and the magnitude of TPGmax− TPGnet(i.e., pressure recovery) is dependent on the valve EOA and the size of the ascending aorta (7–16).
It should be noted that the measurement generally recorded during cardiac catheterization is TPGnet, and consequently, the EOA reported corresponds to: (2)Recently, we proposed a new index based on EOADopand aortic cross-sectional area (AA) that takes into account the pressure recovery phenomenon. Hence, the energy loss coefficient (ELCo) provides an accurate estimation of the energy loss (EL) due to aortic stenosis (16), as demonstrated by this equation: (3)where AAis in cm2and EL is in mm Hg. It should be noted that ELCo can be calculated from the echocardiogram using measurements of EOADopand AA. Because the transvalvular flow rate at rest is mainly dependent on body size, ELCo can also be indexed for body surface area to take into account the cardiac output requirements of the patient. In a previous study (16), we found that the indexed ELCo (i.e., EL index) was superior to either EOADopor indexed EOADopin predicting adverse outcomes in patients with aortic stenosis.
It is interesting to note that Equation 3is very similar to the traditional Gorlin equation. However, instead of the valve EOA, the left-hand side of the equation represents ELCo, and the right-hand side represents EL in terms of pressure instead of TPGnet.
The EL is the sum of TPGnetand the dynamic pressure gradient: (4)where VVand VAare the blood velocities (expressed in m/s) in the left ventricular outflow tract and ascending aorta, respectively. In patients with aortic stenosis, the dynamic pressure gradient 4(VV2− VA2) is negligible compared with TPGnet, so that EL ∼ TPGnetand thus ELCo ∼ EOAcath, according to Equations 2 and 3. Hence, it should theoretically be possible to estimate EOAcathfrom Doppler echocardiographic data by calculating ELCo using the left-hand side of Equation 3.
In vitro study
The pulse duplicator used for the in vitro study has been previously described in detail (16,18). Two fixed stenoses (2 plates with circular orifices of 1.0 and 1.5 cm2) and seven aortic bioprosthetic heart valves (Medtronic Intact 19, 21, 23, and 25 mm, and Medtronic Mosaic 21, 23, and 25 mm) were tested in this model under 10 levels of flow rate ranging from 90 to 430 ml/s and using two aortic sizes: 2.54 cm (cross-sectional area: 5.07 cm2) and 3.8 cm (11.34 cm2).
Flow rate was measured with an electromagnetic flowmeter, and pressure measurements were performed using fluid-filled, side-hole catheters. Ventricular pressure was measured 20 mm upstream from the valve, and aortic pressures at 5 and 100 mm downstream of the valve to calculate TPGmaxand TPGnet, respectively. EOAcathwas calculated from TPGnetand mean flow rate using Equation 2, and EOAcath/maxwas calculated from TPGmaxand mean flow rate using Equation 1.
An Ultramark 9 HDI (Philips Medical Systems/ATL, Bothell, Washington) was used for Doppler velocity measurements. EOADopwas determined by the standard continuity equation using stroke volume measured by the electromagnetic flowmeter and the velocity–time integral of the continuous-wave Doppler aortic jet signal. The ELCo was calculated using the left-hand term of Equation 3.
Animal care and experiments were conducted in accordance with the Guidelines of the Canadian Council for Animal Care. The protocol was approved by the institutional Animal Care Committee of Laval University, Sainte-Foy, Quebec, Canada. Fourteen pigs weighing between 27 and 35 kg were anesthetized, and a lateral thoracotomy was performed in the fourth left intercostal space. A supravalvular aortic stenosis was created using umbilical tape tightened around the aorta ∼2 cm downstream from the aortic valve annulus (19).
The pressure measurements were performed using a Millar catheter (customized model, Millar Instruments, Houston, Texas) with a distal (P1), intermediary (P2), and proximal (P3) sensor. The P2 was positioned at the level of the vena contracta (minimal pressure downstream from the stenosis). The P1, which is at 1.5 cm of the intermediary sensor, was therefore located ∼1 cm upstream from the stenosis. The P3, located at 4 cm of the intermediary sensor, was used to measure the aortic pressure after recovery. Cardiac output was measured using an ultrasonic flowmeter (T206, Transonic Systems, Ithaca, New York), with the probe positioned around the main pulmonary artery. The electrocardiogram, the three pressure signals, and the flow signal were simultaneously recorded and digitized (Digidata 1322, Axon Instruments, Foster City, California). The systolic trans-stenotic pressure gradients were calculated as follows: TPGmax= P1 − P2; and TPGnet= P1 − P3. EOAcathand EOAcath/maxwere calculated as described in the in vitro study.
The Doppler echocardiographic measurements were performed with a Sonos 5500 (Philips Medical Systems/Agilent Technologies, Andover, Massachusetts). An upper laparotomy was performed, and the ultrasound probe was introduced in the abdominal cavity and positioned on the diaphragm at the level of the cardiac apex. This window allowed the visualization of high-quality apical five-chamber images and optimal recording of the left ventricular outflow tract pulsed-wave velocity and aortic jet continuous-wave velocity. EOADopwas calculated using the standard continuity equation. The diameter of the ascending aorta was measured at 2 to 3 cm downstream of the stenosis by epicardial bi-dimensional echocardiography, using a 12-MHz probe. The AAwas calculated assuming a circular shape. The ELCo was calculated using Equation 3.
These measurements were obtained under the following experimental conditions: 1) moderate stenosis; 2) severe stenosis; 3) severe stenosis plus a mild increase in systemic resistance; 4) severe stenosis plus a moderate increase in systemic resistance; and 5) severe stenosis plus a marked increase in systemic resistance. The increase in resistance was obtained by constriction of the descending thoracic aorta. The objective of this intervention was to increase the aortic pressure downstream of the stenosis to produce dilation of the ascending aorta and thus an increase in AA.
To further validate the results obtained in the pulsed duplicator and in the animals, we used the raw data published by Schöbel et al. (14). Their study was performed in 37 patients with aortic stenosis and no significant regurgitation. They simultaneously recorded the pressures within the left ventricle, at the vena contracta, and in the aorta at the site after pressure recovery. Cardiac output was determined by thermodilution, and the mean transvalvular flow rate was calculated. EOAcath(noted as AV-Ain their report) and EOAcath/max(noted as AV-X) were determined from the Gorlin formula using TPGnetand TPGmax, respectively. However, Schöbel and colleagues used the original Gorlin formula with a constant of 44.3. Schöbel’s raw EOA data were therefore corrected by multiplying by 0.89 (44.3/50). In their study, the AAwas derived from angiographic images in the middle part of the ascending aorta.
Data are expressed as the mean value ± SD. The EOA values obtained from different methods (EOADop, EOAcath/max, and EOAcath) were compared within each data subset (in vitro, animals, and patients) using one-way analysis of variance for repeated measures. Statistical analysis of the association between variables was performed with the Pearson correlation coefficient, and graphs were constructed with the corresponding regression equation. Values of p < 0.05 were considered significant.
Table 1presents the Doppler echocardiographic and catheter data obtained in vitro and in the animals, as well as the catheter data from the patients studied by Schöbel et al. (14). It should be noted that EOADopwas not available in the latter study. In the in vitro model and the animals, EOADopwas, on average, 24 ± 17% lower than the EOAcathvalues (p < 0.001), whereas EOADopand EOAcath/maxwere in close agreement and correlated strongly (in vitro data: y = 1.12x − 0.09, r = 0.97; animal data: y = 1.06x + 0.03, r = 0.92; all data: y = 1.02x + 0.03, r = 0.98) (Fig. 1). These experimental results confirm that EOAcath/max(EOA determined by a catheter using TPGmax) and EOADopare equivalent parameters reflecting EOA at the vena contracta.
Figure 2shows the relationship (r = 0.97) found between EOAcathand EOADop. For the patient data, EOAcath/maxwas substituted for EOADop, as the latter was not available in the study of Schöbel (14). This substitution is nonetheless valid because, as shown in Figure 1, there is a strong agreement between EOAcath/maxand EOADop. Figure 2confirms that EOADopis systematically lower than EOAcath(in vitro data: y = 1.36x − 0.16, r = 0.96; animal data: y = 1.47x − 0.04, r = 0.81; patient data: y = 1.38x − 0.13, r = 0.95; all data: y = 1.26x − 0.02, r = 0.97) (Fig. 2). The discrepancy between EOADopand EOAcathtended to be more important in animals than in patients; this may be due to the fact that pigs have relatively smaller aortas compared with humans.
Finally, Figure 3shows that there is an excellent correlation and concordance (in vitro data: y = 0.93 + 0.10, r = 0.94; animal data: y = 0.87x + 0.12, r = 0.81; patient data: y = 1.15 − 0.04, r = 0.94; all data: y = 0.93x + 0.09, r = 0.97) between EOAcathand ELCo and that, in contrast to EOADop, ELCo does not systematically underestimate EOAcath. This result thus confirms that EOAcathcan be accurately predicted from Doppler echocardiographic data by calculating ELCo using Equation 3.
Because it is less flow-dependent than pressure gradients, the valve EOA is one of the main parameters on which clinicians have relied to assess aortic stenosis severity (1). However, although some investigators have found a relatively good agreement between Doppler and catheter EOA measurements (4,5,20,21), others have reported important discrepancies, and in the latter studies, the catheter EOAs were usually higher than the Doppler EOAs (9,22–27). Consistent with the latter studies, the present study also found catheter EOAs to be consistently higher than Doppler EOAs in the same individuals.
Moreover, the theoretical background shows that the systematic underestimation of EOAcathby EOADopis largely justified by the important following concepts: 1) EOADopis derived from the maximal velocity of the jet and reflects the cross-sectional area of the vena contracta. As confirmed by the present study, EOAcath/maxis a parameter equivalent to EOADopthat can also be used to estimate the area of the vena contracta. 2) The calculation of EOAcath/maxrequires the measurement of TPGmax, which is rarely performed during routine catheterization because of the difficulty in obtaining adequate pressure measurements within the vena contracta. The fact that EOADopunderestimates EOAcathis therefore not surprising, because EOADopreflects the area at the vena contracta, whereas EOAcathis derived from TPGnetrecorded after pressure recovery and thus downstream of the vena contracta. 3) Given 1) and 2), EOAcathwill thus necessarily be higher than EOADop, but in varying proportions depending on the size of the ascending aorta and the severity of the stenosis (12–16,28). 4) The discrepancies between EOADopand EOAcathcan be reconciled by calculating ELCo from the echocardiogram. This parameter takes into account pressure recovery, and, as shown by Equations 2 and 3, its formulation is very close to that of EOAcath. 5) Neither EOAcathnor ELCo represents the true EOA, but are rather dimensionless and relatively flow-independent indexes representing the relative loss of energy due to stenosis.
Our experimental results largely confirm these theoretical considerations. As shown in Figure 1, there is indeed a very good correlation and concordance between EOADopand EOAcath/max, as both parameters are a reflection of the cross-sectional area of the vena contracta. In contrast, EOAcathoverestimated both EOADopand EOAcath/max, but in varying proportions depending on the diameter of the aorta (Table 1). Finally, when ELCo was calculated from the echocardiogram to account for pressure recovery, there was an excellent agreement between EOAcathand ELCo, and the aforementioned discrepancy between echocardiographic and catheter measurements was no longer present (Fig. 3). It should be noted that according to fluid mechanics considerations and previous in vivo studies (13,16,29), the diameter of the aorta used to calculate ELCo should be measured at the sino-tubular junction (i.e., at the site where pressure recovery is ongoing). Schöbel et al. (14)also proposed an equation that incorporates EOAcath/maxand AAto predict EOAcath. Nonetheless, this equation is not readily applicable to reconcile the discrepancies between EOAcathand EOADop, as the latter was not measured in their study.
The clinical implications of these findings are important because they may have a direct implication with regard to the criteria used to quantify aortic stenosis severity. As mentioned, measurements of TPGmaxand EOAcath/maxare rarely performed in the clinical setting because of the difficulty in obtaining adequate pressure measurements in the vena contracta, and the parameters routinely reported from catheter measurements are TPGnetand EOAcath. In this context, it should be emphasized that the ACC/AHA guidelines for defining aortic stenosis severity were established mainly based on data obtained from catheter measurements, as well as clinical outcomes in relation to these measurements (1,30–32). The same values for aortic stenosis severity (e.g., <1.0 cm2) were then extended to the echocardiographic data on the assumption that EOADopand EOAcathwere equivalent parameters, and indeed, the aforementioned guidelines do not distinguish between catheter and Doppler measurements.
A most important finding of this study is that the pressure recovery phenomenon may cause important discrepancies between EOAcathand EOADop, and that EOADopsystematically tends to overestimate aortic stenosis severity, compared with EOAcath. The practical implications of this finding are best evidenced by considering Table 2, where Equations 2 and 3are used to calculate the theoretical values of EOADopfor different values of EOAcathand aortic size. The range of aortic sizes used in this table is based on the study of Gjertsson et al. (29), performed in a large group of patients with aortic stenosis (range of aortic diameters at the sino-tubular junction 2.1 to 4.1 cm, mean 3.0 cm). As expected, the greatest discrepancies between EOAcathand EOADopare observed in patients with smaller aortas (diameter ≤3.0 cm), and when comparing Doppler and catheter EOAs in a given patient, it is therefore important to remember that these parameters are not equivalent and that discrepancies up to 50% may be observed depending on the size of the aorta and the severity of the stenosis. Overall, 10% (27/274) of the stenoses examined in the present study would have been classified as severe on the basis of EOADopand moderate on the basis of EOAcath. Of the 37 patients included in this study, three (8%) would have been misclassified. Furthermore, these discrepancies become even more important if one uses the Gorlin formula with a constant of 44.3, as routinely done in catheterization laboratories (Table 2). The present guidelines, based mostly on EOAs measured during catheterization, may therefore not be directly applicable to measurements made from EOADopand may result in overestimations of severity, thus affecting clinical management.
From a practical standpoint, there would appear to be two options. One would be to use a different threshold of severity (e.g., EOA ≤0.75 cm2) when EOA is measured by Doppler echocardiography. However, the underestimation of EOAcathby EOADopwill vary depending on the size of the aorta, and the large standard deviation (±17%) observed in this study suggests that it would not be appropriate to apply a single correction factor. The second alternative would be to use ELCo as the Doppler echocardiographic measurement of aortic stenosis severity. As shown in Figure 3, this would have the advantage of consistency with catheter measurements, and the present guidelines for severe aortic stenosis (EOA ≤1.0 cm2) could then be directly applicable to this parameter. From a conceptual standpoint, it should also be emphasized that ELCo and EOAcathboth reflect the net EL due to stenosis and, as such, are more representative of the increased burden imposed on the left ventricle, compared with EOADopand EOAcath/max. In contrast, uncorrected EOADophas major disadvantages for clinical use because it does not account for differences in actual left ventricular burden, as the latter varies markedly depending on the magnitude of pressure recovery (10–16).
For these reasons, it would appear logical to use ELCo rather than EOADopas the preferred echocardiographic parameter for quantifying aortic stenosis severity, in which case the severity criteria proposed in the ACC/AHA guidelines would become directly applicable (1). Also, as previously shown, ELCo can be indexed for the patient’s body surface area to better account for differences in cardiac output requirements due to differences in body size (16). Previous studies are consistent in suggesting that an indexed EOAcathor EL index ≤0.55 to 0.60 cm2/m2is indicative of severe aortic stenosis (16,32).
An obvious limitation of this study is the absence of a gold standard method for the direct measurement of EOA at the vena contracta. Nonetheless, the strong agreement between the experimental results and the theoretical equations derived from fluid dynamics confirms the conceptual validity of our results and conclusions.
Ideally, it would also have been interesting to obtain simultaneous measurements of EOADopand EOAcathnot only in vitro and in animals, but also in patients. However, the measurement of EOAcathrequires complete left- and right-heart catheterization, a procedure that is not without risk for the patient. Indeed, the most recent ACC/AHA guidelines recommend that this procedure should be performed only if there is a discrepancy between the clinical and echocardiographic evaluations of aortic stenosis severity (1). Hence, systematic performance of such a procedure in patients would have been difficult to justify from an ethical standpoint, and for this reason, we elected to use the data previously published by Schöbel et al. (14). The fact that these data were collected independently and agree well with our own results further validates the conclusions of the present study.
Discrepancies between catheter and Doppler measurements of EOA are largely due to the pressure recovery phenomenon and can be reconciled by calculating ELCo from the Doppler echocardiogram. Although EOADopbetter represents the actual cross-sectional area of the vena contracta, ELCo and EOA measured from the catheter net gradient are equivalent indexes that primarily reflect the net EL due to stenosis rather than the EOA, per se. As such, the latter indexes better reflect the increased burden imposed by the stenosis upon the left ventricle and are probably the most appropriate for quantifying aortic stenosis severity.
We thank Guy Rossignol for his technical assistance in the realization of the animal study.
☆ This work was supported by the Canadian Institutes of Health Research (MA-10929), Ottawa, and the Canadian Foundation for Innovation, Ottawa, Ontario, Canada. Dr. Pibarot is the recipient of a research scholarship from the Heart and Stroke Foundation of Canada, Ottawa, Ontario, Canada.
- cross-sectional area of the aorta
- energy loss
- energy loss coefficient
- effective orifice area
- effective orifice area measured by catheter
- effective orifice area measured by catheter with use of maximal transvalvular pressure gradient
- effective orifice area measured by Doppler echocardiography
- net transvalvular pressure gradient
- maximal transvalvular pressure gradient
- Received May 6, 2002.
- Revision received August 2, 2002.
- Accepted August 19, 2002.
- American College of Cardiology Foundation
- Bonow R.O.,
- Carabello B.A.,
- de Leon A.C. Jr.,
- et al.
- Cannon S.R.,
- Richards K.L.,
- Crawford M.
- Skjaerpe T.,
- Hegrenaes L.,
- Hatle L.
- Otto C.M.,
- Pearlman A.S.,
- Comess K.A.,
- Reamer R.P.,
- Janko C.L.,
- Huntsman L.L.
- Otto C.M.
- Levine R.A.,
- Jimoh A.,
- Cape E.G.,
- McMillan S.,
- Yoganathan A.P.,
- Weyman A.E.
- Laskey W.K.,
- Kussmaul W.G.
- Voelker W.,
- Reul H.,
- Stelzer T.,
- Schmidt A.,
- Karsch K.R.
- Niederberger J.,
- Schima H.,
- Maurer G.,
- Baumgartner H.
- Baumgartner H.,
- Steffenelli T.,
- Niederberger J.,
- Schima H.,
- Maurer G.
- Schöbel W.A.,
- Voelker W.,
- Haase K.K.,
- Karsch K.R.
- Garcia D.,
- Pibarot P.,
- Dumesnil J.G.,
- Sakr F.,
- Durand L.G.
- Smith M.D.,
- Dawson P.L.,
- Elion J.L.,
- et al.
- Foster G.P.,
- Weissman N.J.,
- Picard M.H.,
- Fitzpatrick P.J.,
- Shubrooks S.M. Jr.,
- Zarich S.W.
- Come P.C.,
- Riley M.F.,
- McKay R.G.,
- Safian R.
- Chambers J.B.,
- Sprigings D.C.,
- Cochrane T.,
- et al.
- Burwash I.G.,
- Thomas D.D.,
- Sadahiro M.,
- et al.
- Vandervoort P.M.,
- Greenberg N.L.,
- Pu M.,
- Powell K.A.,
- Cosgrove D.M.,
- Thomas J.D.
- Turina J.,
- Hess O.,
- Sepulcri F.,
- Krayenbuehl H.P.
- Horstkotte D.,
- Loogen F.