Background
Reflux of the aortic regurgitation (AR) causes an increased diastolic reverse flow in the aorta and its branching vessels. We aimed to evaluate the feasibility and accuracy of Doppler measurements in the left subclavian artery (LSA) for quantification of AR in a cardiovascular magnetic resonance imaging (CMR) validation study.
Methods
Systolic and diastolic flow profiles of the LSA (subclavicular approach) were evaluated prospectively by use of pulsed wave Doppler in 59 patients (55.5 ± 15 years; 44 men), 47 with a wide spectrum of AR and 12 as control group. Using CMR phase-contrast sequences (performed 1 cm above the aortic valve), the AR was divided into three groups: mild, regurgitant fraction (RF) < 20% ( n = 17); moderate, RF 20%-40% ( n = 10); and severe, RF > 40% ( n = 20). The LSA Doppler-derived RF was calculated as the ratio between diastolic and systolic velocity-time integrals (VTI).
Results
Quality LSA Doppler signal could be obtained in all cases. Patients with CMR severe AR had higher values of LSA Doppler-derived RF (51% ± 9% vs 36% ± 11% vs 16% ± 8%; P < .0001). LSA Doppler showed a good correlation with CMR, with a sensitivity of 95%, specificity of 89%, and diagnostic accuracy for severe AR of 91.5%. Finally, Bland-Altman plots showed agreement in the group with moderate to severe AR (mean bias = −2.2% ± 8%, 95% CI, −17.7 to 13.3; P = .145) but differed in mild AR.
Conclusions
Measurements of the RF for quantification of AR using LSA Doppler are comparable to those of CMR, highlighting the potential role of LSA Doppler as an adjunctive technique to assess the severity of AR.
Highlights
- •
Echocardiography is the method of choice to evaluate aortic valve regurgitation.
- •
It is a challenge in patients without optimal acoustic windows or with multiple jets.
- •
Pulsed Doppler examination of the left subclavian artery velocity contour.
- •
Alternative method for quantitative assessment of regurgitant fraction.
- •
Feasible, reproducible, and comparable to cardiovascular magnetic resonance imaging.
Echocardiography is the method of choice to evaluate the aortic valve regurgitation (AR), but it is still challenging especially in patients without optimal acoustic windows or in those with multiple or eccentric jets. Qualitative assessment of AR severity by echocardiography has been shown to be unreliable when compared with cardiovascular magnetic resonance (CMR) measurements of regurgitant fraction (RF). Current echocardiographic guidelines recommend a multiparametric approach. Efforts should be made to use semiquantitative and quantitative parameters (i.e., vena contracta width and proximal isovelocity surface area [PISA] method) whenever possible. Adjunctive parameters help to consolidate the evaluation of the severity of AR, particularly when there is discordance between the quantified degree of AR and the clinical context. AR leads to diastolic flow reversal in the descending aorta (DA) and its branches. As the degree of the regurgitation increases, the duration and the velocity of the reversal flow during diastole increases. Thus, the measurement of the diastolic flow reversal in the aorta is recommended, when assessable, and should be considered as the strongest additional parameter for evaluating the severity of AR. However, Doppler examinations are often restricted to the DA, where a high-quality Doppler signal acquisition is usually challenging, allowing only a qualitative or a semiquantitative assessment. On the other hand, the examination of the left subclavian artery (LSA) by pulsed Doppler can be easily performed, and a quantitative assessment of the RF by examination of the velocity contour can be attempted, as was suggested for the first time by Boughner. Nevertheless, this method has never been validated and compared with the highly accurate and reproducible CMR measurements of aortic RF. Therefore, we sought to evaluate the feasibility and accuracy of LSA Doppler measurements for AR quantification in a prospective study using CMR as a reference standard.
Methods
Study Population
We prospectively enrolled 59 patients (55.5 ± 15 years; 44 men), 47 with a wide spectrum of AR of the native valve referred to our center for evaluation of the pathology (AR group) and 12 patients without any apparent AR on transthoracic echocardiography (TTE) or CMR with clinical indication for a CMR study other than heart valve disease (control group). Patients were eligible if they had sinus rhythm during the study. Patients with an associated cardiac valve lesion of more than moderate, with a significant aortopathy (i.e., ascending aorta diameter ≥55 mm or aortic coarctation), or with typical contraindications for CMR imaging were excluded.
Usually, patients underwent both CMR imaging and TTE within 12 hours (in 38 patients of the AR group, the period was <12 hours, with an overall median delay between TTE and CMR examination <1 day [95% CI, 0.14 to 2.6]). All baseline characteristics were prospectively collected. Finally, on the basis of CMR quantification, the AR group was divided into three groups (i.e., mild, RF < 20%; moderate, RF 20%-40%; and severe, RF > 40%), and comparison with LSA Doppler-derived quantification was carried out using the identical severity grading.
Pulsed Doppler of the LSA Velocity Contour
The systolic and diastolic flow profiles of the LSA were evaluated by use of pulsed wave Doppler ultrasound (3.4-9 MHz linear probe). Patients were examined in a supine position with a subclavicular approach. Higher frequencies (>7 MHz) were used for assessment of the morphology, and lower frequency (<7 MHz) was preferred for Doppler examination. LSA was documented with gray-scale imaging and color Doppler to rule out relevant stenosis. The depiction of the LSA was modified to align the Doppler angle parallel to the vector of blood flow and to avoid the Doppler signal of the adjacent vein. The sample volume was placed just near the origin of the LSA. Patients with vascular shunts of the left upper arm were excluded.
We routinely used angle correction applying the following formula: Δ f = 2 f 0 V cos θ/ C (Δ f = Doppler shift frequency, f 0 = transmitted ultrasound frequency, V = the velocity of red blood cells, θ = angle between the transmitted beam and the direction of blood flow within the blood vessel, and C = speed of sound in the tissue). To avoid spectral broadening, the Doppler angle should not exceed 60° (the preferred angle was 50° ± 5°), the position of the sample volume box should be in the mid lumen parallel to the vessel wall, and the size of the sample volume box should be 2-3 mm. Color velocity scale was set between 30 and 40 cm/sec, sweep speed between 50 and 100 mm/sec, and forward and backward velocity-time integrals (VTI) profiles were traced using an adjusted flow velocity scale for more precise measurements. The outer edge of the dense (bright) envelope of the spectral recording (i.e., modal velocity) was used to measure the VTI. The RF was calculated as follows: RF(%) = LSA-derived diastolic reversed flow VTI × 100/LSA systolic forward flow VTI ( Figure 1 ). At least two measurements were performed. The LSA Doppler examination was carried out blinded to the results of echocardiography and the CMR evaluation of AR.
Standard Echocardiography
All TTE were performed by experienced echocardiographers using commercially available ultrasound machines (Vivid E9, General Electric Healthcare, Wauwatosa, WI; or Acuson SC2000, Siemens Healthcare GmbH, Erlangen, Germany) equipped with M5S or 4V1c two-dimensional TTE probes. All recordings were stored digitally for offline analysis. Left ventricular (LV) volumes and ejection fraction (EF) were calculated using the biplane Simpson disk method. Doppler measurements were evaluated as the average of at least three cycles. An effort was made to perform the flow convergence (or PISA) method, from apical views or, in the case of eccentric jets, from parasternal long-axis views. The AR severity was graded according to current recommendations, with a special focus on semiquantitative parameters like diastolic flow reversal in the DA measured by pulsed wave Doppler. Additionally, the same LSA Doppler scale and method described above to calculate the RF were used at the level of the DA. Finally, an integrative approach was applied to grade the AR, which was classified into one of three grades: mild, moderate, or severe. Evaluation of AR was performed by an experienced echocardiographer (R.A.S.) with >10 years of experience in echocardiography with European Society of Cardiology certification, who was blinded to the results of the CMR exam.
Cardiovascular Magnetic Resonance
All CMR examinations were performed in our cardiology department on a 1.5-T magnetic resonance imaging system (Ingenia, Philips Healthcare, Best, the Netherlands) equipped with a 28-element array coil with full in-coil signal digitalization combined with optical transmission. All scans were accomplished without sedation. Image data acquisition and subsequent analysis were carried out according to current guidelines. For cine imaging, a balanced steady-state free precession (b-SSFP) sequence with retrospective gating was used during short periods of breath holding. All standard cardiac geometries were acquired (multiple, gapless short-axis slices covering the entire left ventricle and two-, three- and four-chamber views). Imaging parameters were chosen as follows: echo time (TE) and repetition time (TR) were set to shortest, resulting in an average TR of around 4 msec and a TE of 2 msec, with a reconstructed in-plane resolution of 1.0 × 1.0 mm 2 ; the slice thickness was 8.0 mm. The typical temporal resolution of the cine b-SSFP sequences was 30-25 msec depending on the heart rate. The imaging plane for the through-plane phase-contrast flow measurement was placed in the ascending aorta approximately 10 mm above the aortic valve and positioned perpendicular to the flow direction. On a coronal image of the aorta together with the three-chamber view, the CMR operator checked that the image plane was truly perpendicular to the aortic flow direction. To avoid aliasing, velocity encoding was individually adapted, starting at 200 cm/sec, and if aliasing occurred, the maximum velocity was increased by 50 cm/sec steps until aliasing disappeared. Image data acquisition was gated to the electrocardiogram signal with a temporal resolution of 35 phases per cardiac cycle and acquired during a 12-15 second breath hold. Outlining the region of interest within the aortic lumen for each cardiac phase, the instantaneous flow volume (cm 3 /sec) was calculated and graphically displayed over the entire cardiac cycle.
Forward and reversed flow volumes were measured, and the RF was calculated as follows: aortic RF (%) = diastolic reversed flow volume × 100/systolic forward flow volume. As previous comparisons suggested, the AR severity was classified with RF <20% as mild, between 20% and 40% as moderate, and >40% as severe ( Figure 2 ).
Statistical Analysis
Data are presented as mean (SD), median (25th to 75th percentile) or frequency (%) as appropriate. Statistical differences between groups were assessed using Student’s t -test for continuous variables or Fisher’s exact test for categorical variables. Multigroup comparisons of continuous variables were performed using an analysis of variance. Spearman correlation coefficient, Bland-Altman plots, and the intraclass correlation were used to assess correlations and agreements between CMR and TTE parameters. Rate of agreement for AR grading was evaluated by calculating a κ statistic. Inter- and intraobserver variabilities were assessed by analysis of 24 patients (10 patients with severe and 14 patients with less than severe AR) for LSA Doppler-derived RF; of 22 patients (10 with severe AR and 12 with less than severe AR) for DA Doppler-derived RF; and of 24 patients (12 with severe AR and 12 with less than severe AR) for CMR imaging. Two observers independently measured the RF to assess interobserver variability for LSA and DA Doppler method (F.B. and R.A.S.) and for CMR imaging (C.J. and R.A.S.). These same studies were reexamined by one observer (R.A.S.) >4 weeks apart to determine intraobserver variability. Inter- and intraobserver coefficients of variation on RF measurements were determined by analysis of the absolute difference between (re)-measurements divided by the mean of both measurements. Additionally, the mean difference with the 95% CI, Bland-Altman plots, and the intraclass correlation coefficient (ICC) were also determined. Two-tailed P values < .05 were considered statistically significant. Analyses were performed using SPSS software (IBM-SPSS Statistics, ver. 20, IBM Corp., Armonk, New York, USA). The study was conducted in accordance with the Declaration of Helsinki and was approved by the local research ethics committee (067/17-ek). All patients gave informed consent.
Results
Demographic and baseline patient characteristics are presented in Table 1 . There were no significant differences between controls and AR group, with the exception of a higher rate of hypertension in the AR group. A high-quality Doppler signal of the LSA could be obtained in all patients. PISA method and a Doppler signal in the DA could be obtained in 91.5% and 81% of the patients with AR, respectively. Echocardiographic characteristics of all groups are shown in Table 2 . The AR group had higher TTE- and CMR-derived LV end-diastolic and stroke volume compared with control group, without differences in LV EF. Patients with CMR-determined severe AR had significantly higher diastolic VTI in the LSA than patients with moderate or mild AR and the control group (11 ± 3 vs 6 ± 2 vs 3.3 ± 2 vs 1.3 ± 0.6 cm, respectively; P < .001 for all) and also showed higher values of LSA-derived RF (51% ± 9% vs 36% ± 11% vs 16% ± 8% vs 7.6% ± 3%; P < .0001).
| Controls ( n = 12) | AR group ( n = 47) | |
|---|---|---|
| Age, years | 51.3 ± 17 | 56.5 ± 14 |
| Male, n (%) | 7 (58) | 37 (79) |
| Coronary artery disease, n (%) | 0 | 11 (23.5) |
| Hypertension, n (%) | 4 (33) | 35 (74.5) ∗ |
| Diabetes, n (%) | 0 | 6 (13) |
| Dyslipidemia, n (%) | 2 (17) | 17 (36) |
| Body surface area, m 2 | 1.91 ± 0.2 | 1.98 ± 0.2 |
| Aorta diameter, mm | 31 ± 4 | 36 ± 6 |
| Aortic valve morphology | ||
| Tricuspid, n (%) | 12 (100) | 30 (63.8) |
| Bicuspid, n (%) | 0 | 13 (27.7) |
| Quadricuspid, n (%) | 0 | 1 (2.1) |
| Undefined, n (%) | 0 | 3 (6.4) |
| Control ( n = 12) | AR group ( n = 47) | Mild ( n = 17) | Moderate ( n = 10) | Severe ( n = 20) | P value ∗ | |
|---|---|---|---|---|---|---|
| Echocardiographic parameters | ||||||
| EDV, mL | 116 ± 39 | 201 ± 81 † | 140 ± 33 | 188 ± 51 † | 260 ± 81 † | <.0001 |
| ESV, mL | 53 ± 31 | 90 ± 47 | 62 ± 19 | 85 ± 39 | 115 ± 55 † | .002 |
| LVSD, mm | 36 ± 7 | 43 ± 8 | 37 ± 6 | 41 ± 7 | 48 ± 6 † | <.0001 |
| EF, % | 56 ± 14 | 56 ± 9 | 55 ± 9 | 56 ± 10 | 57 ± 9 | .850 |
| Stroke volume, mL | 70 ± 20 | 123 ± 43 † | 88 ± 20 | 118 ± 25 † | 155 ± 40 † | <.0001 |
| EROA, cm 2 ‡ | NA | 0.27 ± 0.18 | 0.09 ± 0.07 | 0.25 ± 0.1 | 0.42 ± 0.14 | <.0001 |
| RV, mL ‡ | NA | 53 ± 34 | 18 ± 10 | 53 ± 19 | 82 ± 23 | <.0001 |
| RF, % ‡ | NA | 39 ± 17 | 20 ± 9 | 44 ± 12 | 53 ± 7 | <.0001 |
| Vena contracta, mm | NA | 4.4 ± 1.8 | 2.5 ± 0.5 | 4.5 ± 1.0 | 6.2 ± 0.8 | <.0001 |
| PHT, msec | NA | 425 ± 185 | 605 ± 155 | 405 ± 102 | 283 ± 90 | <.0001 |
| QP/QS | 1.02 ± 0.06 | 0.64 ± 0.2 † | 0.86 ± 0.08 † | 0.62 ± 0.1 † | 0.46 ± 0.08 † | <.0001 |
| DA ( n = 48/59) | ||||||
| End-diastolic flow velocity, cm/sec | 0 | 19.5 ± 14 † | 3 ± 5 | 13 ± 7 † | 30.5 ± 9 † | <.0001 |
| Backward VTI, cm | 2.4 ± 0.7 | 11.4 ± 8 † | 3.5 ± 2 | 8.6 ± 3 † | 17.5 ± 5 † | <.0001 |
| Backward/forward VTI ratio | 14 ± 3.5 | 47 ± 22 † | 23 ± 13 | 51 ± 24 † | 61 ± 11 † | <.0001 |
| CMR parameters | ||||||
| EDV, mL | 161 ± 39 | 247 ± 97 † | 167 ± 35 | 238 ± 59 † | 319 ± 94 † | <.0001 |
| ESV, mL | 78 ± 42 | 118 ± 61 | 78 ± 23 | 114 ± 56 | 154 ± 65 † | <.0001 |
| EF, % | 54 ± 11 | 54 ± 8 | 54 ± 8 | 54 ± 10 | 53 ± 8 | .953 |
| RF, % | 3 ± 2 | 32 ± 20 † | 10 ± 5 † | 31 ± 8 † | 51 ± 8 † | <.0001 |
| Aorta forward flow, mL | 83 ± 22 | 125 ± 49 † | 90 ± 21 | 112 ± 26 | 161 ± 50 † | <.0001 |
| Aorta backward flow, mL | 2 ± 1.8 | 45 ± 38 † | 8 ± 5 † | 35 ± 12 † | 81 ± 27 † | <.0001 |
| PA forward flow, ml | 91 ± 24 | 84 ± 30 | 79 ± 37 | 83 ± 25 | 86 ± 31 | .928 |
| PA backward flow, mL | 2.2 ± 2 | 2.5 ± 3 | 4 ± 6 | 2.6 ± 1.6 | 2 ± 1.4 | .461 |
| QP/QS | 1.0 ± 0.1 | 0.64 ± 0.16 † | 0.87 ± 0.15 | 0.72 ± 0.05 † | 0.53 ± 0.1 † | <.0001 |
| LSA Doppler | ||||||
| Forward VTI, cm | 16.4 ± 3 | 21 ± 7 | 21 ± 8 | 18 ± 6 | 22 ± 6 | .265 |
| Backward VTI, cm | 1.3 ± 0.6 | 7 ± 4 † | 3.3 ± 2 † | 6 ± 2 † | 11 ± 3 † | <.0001 |
| Peak backward velocity, cm/sec | 24 ± 6 | 37 ± 12 † | 31 ± 12 | 36 ± 7 | 43 ± 11 † | .005 |
| RF, % | 7.6 ± 3 | 35 ± 18 † | 16 ± 8 § | 36 ± 11 † | 51 ± 9 † | <.0001 |
† P < .05 compared with control group.
‡ Calculated with the PISA method ( n = 43).
Agreement between Methods
Overall, the level of agreement between LSA Doppler and CMR grading of AR was strong (κ = 0.705, P < .001); 38 of 47 patients (81%) had the same AR grade on both methods. In patients with mild and moderate AR on CMR, LSA Doppler overestimated the severity of AR in approximately one-third of patients. Importantly, in patients with severe AR on CMR, there was only one misclassified patient by LSA Doppler, resulting in a sensitivity of 95%, specificity of 89%, and a diagnostic accuracy for severe AR of 91.5%.
In the whole cohort, the RF measurements assessed by LSA Doppler (30% ± 20%) and CMR imaging (26% ± 21%) were strongly correlated ( r = 0.931 and ICC = 0.964; P < .0001 for both; Figure 3 ). There was an overestimation of the RF determined by LSA Doppler only in the group with mild AR ( Figure 4 ). The small mean differences between LSA Doppler and CMR measurements (<2.5%) were not statistically significant, and the Bland-Altman plots showed a good limit of agreement in the group of patients with moderate to severe AR (mean bias = −2.2% ± 8%; 95% CI, −18 to 13; P = .145; Figure 5 ).
Table 3 shows the diagnostic value of Doppler-derived RF at the level of the LSA and DA and of the standard echocardiographic quantitative PISA method. There were comparable feasibility and overall agreement for assessment of AR severity using LSA Doppler and PISA methods. Nevertheless, κ values for the agreement between LSA Doppler and CMR were numerically higher than those of DA Doppler and CMR. In patients with mild AR on CMR, DA Doppler-derived RF overestimated the severity of AR in three of 12 patients, and one misclassified patient had more than one grade disagreement of AR, which was not observed with the LSA Doppler or PISA method.
| CMR | Sensitivity, specificity, and accuracy, % | |||||
|---|---|---|---|---|---|---|
| Mild | Moderate | Severe | n (%) | Severe AR | Moderate to severe AR | |
| PISA method ∗ | ||||||
| Mild | 13 | 13 (30.2) | 83 | 100 | ||
| Moderate | 2 | 7 | 3 | 12 (27.9) | 88 | 86.7 |
| Severe | 3 | 15 | 18 (41.9) | 86 | 95.3 | |
| Total (%) | 15 (34.9) | 10 (23.2) | 18 (41.9) | 35/43 (81.4) | ||
| LSA Doppler † | ||||||
| Mild | 12 | 12 (25.5) | 95 | 100 | ||
| Moderate | 5 | 7 | 1 | 13 (27.7) | 89 | 70.6 |
| Severe | 3 | 19 | 22 (46.8) | 91.5 | 89.4 | |
| Total (%) | 17 (36.2) | 10 (21.3) | 20 (42.5) | 38/47 (80.9) | ||
| DA Doppler ‡ | ||||||
| Mild | 9 | 1 | 10 (26.3) | 100 | 96.1 | |
| Moderate | 2 | 2 | 4 (10.5) | 73.7 | 75 | |
| Severe | 1 | 4 | 19 | 24 (63.2) | 86.8 | 89.5 |
| Total (%) | 12 (31.6) | 7 (18.4) | 19 (50) | 30/38 (78.9) | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
