Objective To evaluate the usefulness of velocity ratio (VR) in patients with low gradient severe aortic stenosis (LGSAS) and preserved EF.
Background LGSAS despite preserved EF represents a clinically challenging entity. Reliance on mean pressure gradient (MPG) may underestimate stenosis severity as has been reported in the context of paradoxical low flow, LGSAS. On the other hand, grading of stenosis severity by aortic valve area (AVA) may overrate stenosis severity due to erroneous underestimation of LV outflow tract (LVOT) diameter, small body size or inconsistencies in cut-off values for severe stenosis. We hypothesised that VR may have conceptual advantages over MPG and AVA, predict clinical outcomes and thereby be useful in the management of patients with LGSAS.
Methods Patients from the prospective Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) study with an AVA<1.0 cm2, MPG≤40 mm Hg and EF≥55% and asymptomatic at baseline were stratified according to VR with a cut-off value of 0.25. Outcomes were evaluated according to aortic valve-related events and cardiovascular death.
Results Of 435 patients with LGSAS, 197 (45%) had VR<0.25 suggesting severe and 238 (55%) had VR≥0.25 suggesting non-severe stenosis. Aortic valve-related events (mean follow-up 42±14 months) were more frequent in patients with VR<0.25 (57% vs 41%; p<0.001) as was cardiovascular death within the first 24 months (p<0.05). In multivariable Cox regression analysis, MPG was the strongest independent predictor of aortic valve events (p<0.001) followed by VR (p<0.02). Adjusting AVA by VR increased predictive accuracy for aortic valve events (area under the receiver operating curve 0.62 (95% CI 0.57 to 0.67) vs 0.56 (95% CI 0.51 to 0.61) for AVA, p=0.02) with net reclassification improvement calculated at 0.36 (95% CI 0.17 to 0.54, p<0.001). VR did not improve the prediction of clinical events by MPG.
Conclusions In the difficult setting of LGSAS, VR shows a strong association with valve-related events and—although not outperforming MPG—may be particularly useful in guiding clinical management.
Trial registration number NCT00092677.
Statistics from Altmetric.com
Low gradient severe aortic stenosis (LGSAS) is defined by a severely stenotic aortic valve area (AVA<1.0 cm2) but a non-severe mean pressure gradient (MPG≤40 mm Hg) despite preserved EF and represents a clinically challenging entity with conflicting reports regarding prevalence, management and outcomes in these patients.1–8 LGSAS may be due to ‘paradoxical’ low flow,1 a concept suggesting that despite normal EF there is a low flow state that—in analogy to patients with reduced LV function—requires estimation of stenosis severity by AVA rather than flow-dependent pressure gradients. On the other hand, the same constellation of LGSAS may be due to other reasons including measurement errors (in general underestimation) in the assessment of LV outflow tract (LVOT) diameter critical in the calculation of AVA and stroke volume,2 ,7 ,9 inconsistencies in the current cut-off values of AVA and MPG10 ,11 and small body size,12 ,13 resulting in an overestimation of stenosis severity by AVA. Since calculation of flow is complex irrespective of the method employed, we hypothesised that a flow- (and LVOT-diameter-) independent parameter of stenosis severity may differentiate between truly severe and non-severe aortic stenosis and thereby be helpful in the management of patients with LGSAS. We therefore evaluated the predictive accuracy of velocity ratio (VR=velocity time integral (VTI) in the LVOT/VTI within the valve) with a partition value of 0.2514 for aortic valve events (AVE) and cardiovascular death in 435 patients with LGSAS from the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) study.
Patients and methods
Patients were recruited from the SEAS study (NCT00092677), which enrolled 1873 patients with asymptomatic aortic valve stenosis, defined by echocardiography at local study centres as aortic valve thickening and transaortic Doppler jet velocity ≥2.5 and ≤4.0 m/s. Patients were randomised from January 2001 to February 2004 to at least 4-year placebo-controlled combined treatment with ezetimibe 10 mg/day and simvastatin 40 mg/day. Patients with coronary heart disease, heart failure, diabetes, history of stroke or peripheral vascular disease, clinically significant mitral valve disease, severe or predominant aortic regurgitation, rheumatic valvular disease, aortic valve prosthesis or renal insufficiency and patients already on lipid-lowering therapy or having an indication for lipid lowering according to guidelines were excluded. The primary result of the SEAS study showed no difference in major cardiovascular events between treatment and placebo group but a significant reduction in ischaemic events in patients treated with simvastatin/ezetimibe.15 The present study population comprises 1525 (81.4%) of the 1873 patients recruited in the SEAS study with an EF ≥55% in whom a detailed echocardiographic assessment by the core laboratory was available.
Echocardiography was performed following the guidelines for the clinical application of echocardiography16 and has been described in detail previously.10 ,15 Briefly, maximal jet velocity within the valve (v2) was recorded by aligning the continuous wave beam parallel to the stenotic jet. The velocity curve was traced and MPG was calculated automatically from the mean of a series of instantaneous velocities (vi) of a single beat measured during the systolic ejection period using the simplified Bernoulli equation (). In patients with atrial fibrillation, MPG was calculated from a representative (average) beat. AVA was calculated from the continuity equation using VTIs (, ie, ). VR was calculated as and in a second analysis as with v1 representing maximal jet velocity in the LVOT measured by pulsed wave Doppler just below the aortic valve. LVOT area A1 was calculated as A1=π×r2. LVOT diameter (D=2×r) was measured at end-diastole in the two-dimensional parasternal long-axis view at the aortic valve annulus by an inner-edge-to-inner-edge method.15 Stroke volume was calculated either as () or based on LV end-diastolic (EDV) and end-systolic (ESV) volumes derived from the Teichholz formula as ().
The primary end point of the outcomes analysis of the present study was aortic valve-related events (defined as aortic valve replacement, congestive heart failure due to aortic stenosis or death from cardiovascular causes), secondary end point was cardiovascular death. All end points had been adjudicated with a predefined end point protocol by an end point committee blinded for study conduct and results in the original SEAS study.15
Continuous variables are presented as mean±SD and categorical variables as percentages. Percentages were compared with the use of the Pearson χ2 test. Kaplan–Meier method was used to assess event-free survival with differences checked by means of the log-rank test. A univariable Cox proportional hazards model (after standardisation of predictors by calculation of z-scores) was used to analyse the relationship between baseline covariables and AVEs. Results are reported as HRs with 95% CIs. Forward stepwise multivariable Cox regression analysis was performed to identify independent predictors of outcome. All variables with a p value of <0.05 on univariable analysis entered the multivariable model. Receiver operating characteristic (ROC) curves (and their 95% CIs) were constructed to assess sensitivity and specificity of echocardiographic parameters of stenosis severity for the prediction of aortic valve-related events. Comparison of areas under the ROC curves was performed as recommended by DeLong et al.17 Net reclassification improvement (NRI) and the integrated discrimination improvement (IDI) were calculated with AVEs as the categorised outcome variable (event vs no event) and the echocardiographic variables MPG, AVA, VR and VR-adjusted AVA in continuous format using algorithms developed by Frank Harrell based on the method proposed by Pencina.18 All hypothesis testing was two-tailed, and a p value of 0.05 was considered statistically significant. Statistical analyses were performed using SPSS for Windows V.21 (SPSS Inc, Chicago, Illinois, USA), MedCalc V.184.108.40.206 (MedCalc Software, Mariakerke, Belgium) and the R statistical package (http://www.R-project.org).
Of the 1525 patients from the SEAS study with available echocardiographic data, 435 (29%) had the echocardiographic constellation of a LGSAS (AVA<1 cm2, MPG≤40 mm Hg and a normal EF≥55%) and form the basis of the current analysis. Of the patients with LGSAS, 197 (45%) had a VR<0.25 indicative of severe and 238 (55%) had a VR≥0.25 suggesting non-severe stenosis based on this parameter. There were marked differences between the two groups with patients with VR<0.25 having a smaller AVA, higher MPG and increased LV mass consistent with more severe disease. In addition, patients with VR<0.25 were more likely male, slightly larger, with a lower EF and heart rate. Age was comparable between groups. Baseline characteristics for the two groups are summarised in table 1.
During a median follow-up time of 42±14 months, there were 211 first AVEs, with a total of 183 aortic valve replacements, 17 heart failures due to aortic stenosis and 34 cardiovascular deaths with a median time to event of 33±13 months. Aortic valve-related events were more frequent in patients with VR<0.25 than in patients with VR≥0.25 (p<0.001, figure 1A, table 2).
Within the first 24 months of follow-up, there were 14 cardiovascular deaths, 10 in patients with a VR<0.25 and 4 in patients with VR≥0.25 (p<0.05 by the log-rank test). After 42±14 months, the difference between the two groups was no longer statistically significant (figure 1B). Ischaemic events and all-cause mortality were not statistically different between groups (figure 1C, D).
Calculating VR from peak jet velocities rendered similar results. The correlation between VR from VTIs and VR calculated from peak jet velocities was 0.825 (p<0.001). Based on jet velocities, 193 patients (44%) had a VR<0.25 and 242 (56%) a VR≥0.25. AVEs occurred in 122 (63%) patients with VR<0.25 and 89 (37%) with a VR≥0.25 (p<0.001); cardiovascular death at 2 years in 10 (5.2%) vs 4 (1.7%, p<0.04).
Value of VR for the prediction of AVEs
In univariable analysis, patients with VR<0.25 were at increased risk of an AVE compared with patients with VR≥0.25 (HR 1.65, 95% CI 1.26 to 2.17, p<0.001). Other variables associated with an increased risk for an AVE were male gender, height, EF, LV mass index and the measures of stenosis severity jet velocity, MPG, AVA, AVAindex and VR (table 3). All parameters significantly associated with AVE in univariable analysis were included in the multivariable model. Proportional hazards assumption testing demonstrated no evidence of non-proportional hazards (global model: χ2=4.21, p=0.76). Jet velocity, MPG and VR emerged as the only independent predictors of an AVE (p<0.001, p<0.001 and p<0.02, respectively). Since there was a significant correlation between VR and LVOT diameter (r=−0.4)7 multivariate analysis was repeated with LVOT-adjusted echocardiographic parameters, which did not change the results. Due to colinearity between jet velocity and MPG (r=0.934), multivariable models were run separately with either variable. VR was significantly associated with AVEs after substituting jet velocity for MPG (HR 0.817, 95% CI 0.701 to 0.952, p=0.01). None of the echocardiographic parameters of stenosis severity was a predictor of cardiovascular death or all-cause mortality in multivariable Cox regression analysis (table 3).
In ROC curve analysis of the echocardiographic parameters of stenosis severity, MPG showed the largest area under the curve (AUC) for the prediction of AVEs during follow-up (0.68; 95% CI 0.63 to 0.73, figure 2 and table 4). Adjusting MPG by VR had no significant effect (AUC of VR-adjusted MPG 0.69; 95% CI 0.64 to 0.74, p=0.45 vs MPG). In contrast, AVA demonstrated an AUC of 0.56 (95% CI 0.51 to 0.61), which was significantly improved by adjusting AVA by VR (AUC 0.62; 95% CI 0.57 to 0.67, p=0.02). Combining all three parameters (MPG, AVA and VR) was not superior to MPG (AUC 0.69; 95% CI 0.64 to 0.74 vs AUC 0.68; 95% CI 0.63 to 0.73, respectively, p=0.84). Stroke volume index calculated either from VTI or from the Teichholz formula demonstrated AUCs of 0.48 (CI 0.42 to 0.53) and 0.49 (CI 0.43 to 0.54), respectively (table 4).
NRI for VR-adjusted AVA compared with AVA was calculated at 0.357 (95% CI 0.173 to 0.542, p<0.001) with a highly significant IDI (p=0.003). In line with the findings from ROC analysis (table 4), there was no further improvement of the predictive accuracy of MPG when adjusting MPG by VR (NRI −0.009, IDI p=0.41).
To our knowledge, this study investigated for the first time the usefulness of VR in patients with low gradient severe aortic valve stenosis (LGSAS) and preserved EF. VR was not superior to MPG but predicted AVEs, improved predictive accuracy of AVA and, therefore, may be useful in the management of patients with the challenging echocardiographic constellation of severe stenosis based on AVA and non-severe stenosis based on MPG.
VR in LGSAS
VR has been included in the 2012 European guidelines on the management of valvular heart disease14 as a fifth parameter for the assessment of aortic valve stenosis severity next to jet velocity, MPG, AVA and AVAindex following a level 2 recommendation in the EAE/ASE publication on echocardiographic assessment of valve stenosis issued in 2009.9 VR is considered useful in the assessment of stenosis severity for at least two reasons. First, it is highly reproducible and by obviating the need to measure LVOT diameter simplifies data acquisition and reduces the error related to LVOT diameter measurements required in the calculation of AVA by the continuity equation. Second, it normalises for body size since it represents the ratio of actual to expected valve area in each patient. Finally, VR should be close to 1 in healthy individuals and consequently a cut-off value for severe stenosis has been set to 0.25 corresponding to a valve area of 25% of normal.19 The characteristics of VR make it an ideal candidate in the further evaluation of patients with LGSAS (AVA<1.0 cm2 and MPG≤40 mm Hg) despite preserved EF. There have been conflicting reports on prevalence, management and outcomes in these patients,1–6 and a unifying hypothesis to explain discrepancies in the assessment of stenosis severity in the individual patient is still lacking. However, the data presented here indicate that consideration of VR with a cut-off of 0.25 in the assessment of patients with LGSAS may be clinically useful. Particularly in case of uncertainty about the trustworthiness of flow-dependent parameters MPG and jet velocity, as may be the case in suspected LV-compromise despite normal EF or concomitant mitral regurgitation, VR may help to distinguish truly severe (AVA<1.0 cm2 in the presence of a VR<0.25) from non-severe stenosis (AVA<1.0 cm2, VR≥0.25).
As outlined above, the mechanisms by which calculation of VR improves predictive accuracy for clinical events include increased reproducibility, abolition of the need to measure LVOT diameter and the adjustment of interindividual differences in body size and (more importantly) stroke volume.
Stroke volume versus VR
Measurement of stroke volume as an estimate of LV flow output has recently been proposed in the workup of patients with LGSAS.4–6 ,20 ,21 On the other hand, we have previously demonstrated that stroke volume index, in contrast to VR, was not predictive of aortic events or cardiovascular death in our prospectively followed cohort from the SEAS trial independently of the method of stroke volume calculation (VTI, LV volumes).2 Differences in study populations in particular with respect to comorbidities and risk factors for cardiovascular disease may account for the discrepant results. From a technical perspective, stroke volume is, in general, calculated from Doppler-derived VTI in the LVOT and LVOT diameter, which are the same measurements included in the continuity equation for the calculation of AVA. Therefore, underestimation of AVA will be associated with underestimation of stroke volume and may erroneously lead to the diagnosis of low flow, low gradient aortic stenosis. Despite the fact that stroke volume was predictive of outcome in some studies on patients with LGSAS as a group, reliable assessment of stroke volume by echocardiography may be problematic in the individual patient. In contrast, VR has high reproducibility, is less dependent of flow and independent of body size and may thereby overcome the simultaneous and erroneous underestimation of AVA and stroke volume.
This is a retrospective analysis of a prospectively followed cohort of low-risk patients with mild-to-moderate aortic stenosis. Patients were asymptomatic at baseline, and all comorbidity except hypertension was excluded as far as possible. Therefore, the results might not apply to symptomatic patients or patients with more advanced stage of the disease. However, from a conceptual point of view, VR should be even more effective in differentiating severe from non-severe aortic stenosis when alternative reasons or mechanisms of a low gradient (eg, low flow state in concomitant mitral valve regurgitation or reduced EF) interfere in a more complex manner. On the other hand, in strictly asymptomatic patients, there generally is no indication for valve replacement. Additional consideration of VR may therefore change clinical management in respect to timing of follow-up and the level of surveillance only. In patients with low EF, reduced opening forces may result in incomplete opening (‘pseudostenosis’) and increasing contractility, that is, by dobutamine, may disclose the real anatomic obstruction. However, it is unlikely that this issue might have played an important role in the investigated cohort of asymptomatic patients. The predictive value of VR for clinical events as assessed by ROC analysis, NRI and IDI was only moderate possibly due to the long time period between index echo and event. An analysis of the last available echo improved the overall predictive accuracy of all measures of stenosis severity without however demonstrating a further advantage of VR (data not shown). Another potential confounding factor relates to the negative correlation between LVOT diameter and VR (r=−0.4) in our population, suggesting a dependency of VR from individual LV geometry, which might limit its predictive value. Michelena et al7 report a similar finding in their patients and—based on a comparative analysis of cut-off values for severe stenosis—recommend LVOT-adjusted cut-off values for VR. The data presented in this paper demonstrate no improvement in the prediction of clinical events with the inclusion of the LVOT diameter in Cox regression models.
In the difficult setting of LGSAS (discordant grading of stenosis severity with a low MPG and a small AVA in the presence of a normal EF), VR shows a strong association with valve-related events and—although not outperforming MPG in the SEAS population of asymptomatic, low-risk patients—may be particularly useful in guiding clinical management.
What is known on this subject?
Velocity ratio (velocity time integral in the LV outflow tract/velocity time integral within the valve) has been included in the recent European Society of Cardiology (ESC) guidelines on valvular heart disease and has conceptual advantages over other parameters of stenosis severity since it is less dependent of flow or body size and does not require LV outflow tract diameter measurement. However, there is very little information in the literature on the prognostic accuracy of velocity ratio, particularly in the frequent and challenging situation of low gradient severe aortic stenosis, defined as the discrepant finding of a severely stenotic aortic valve area (AVA<1.0 cm2) and a non-severe pressure gradient (MPG≤40 mm Hg) in the presence of a normal EF (EF≥55%) encountered in up to 30% of patients with aortic valve stenosis.
What might this study add?
This study demonstrates that velocity ratio predicts aortic valve-related events and cardiovascular death in asymptomatic patients with low gradient severe aortic stenosis, thereby substantiating the role of velocity ratio as a useful additional measure of aortic stenosis severity.
How might this impact on clinical practice?
In asymptomatic patients with low gradient severe aortic stenosis, a velocity ratio of <0.25 warrants closer clinical and echocardiographic follow-up. When extended to symptomatic patients with comorbidities, velocity ratio may help to confirm stenosis severity and guide clinical management.
Contributors JM as corresponding author acts as guarantor. NJ, CG-B and JM were involved in the study conception and design; acquisition, analysis and interpretation of data; drafted the article and had final approval of the manuscript. BAK, EB, EG, KB, JBC, CAN, KW, TRP, WH and F-JN were involved in interpretation of data, revising the article and final approval of the manuscript
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.