Disagreement between Cardiovascular Magnetic Resonance and Echo-Doppler Transvalvular Pressure Gradients

Introduction: Based on current guidelines for the evaluation of aortic stenosis (AS) severity, patients are candidates for valve replacement surgery if they have a valve effective orifice are (EOA) < 1.0 cm, transvalvular pressure gradient (TPG) ≥ 40 mmHg and symptoms [1]. Transthoracic echocardiography (TTE) is widely used to evaluate AS severity. However, there are often discrepancies among the TTE measurements. Cardiovascular magnetic resonance (CMR) may be used to corroborate stenosis severity. The aim of this study is to examine the agreement of TTE and CMR for the estimation of TPG and EOA in patients with AS. Methods: Seven (7) healthy subjects and 31 patients with mild to severe AS (0.72 cm ≤ EOA ≤ 1.73 cm) were included in this study. TTE measurements were performed according to the ASE guidelines [2]. CMR study was performed within 4 weeks after TTE study with the use of a 1.5 Tesla scanner. A standard ventricular function examination was performed for acquisition planning. Phase-Contrast (sQFlow Phase SENSE) retrospective examination was performed in standard short-axis planes in the left ventricular outflow tract (LVOT) at -12 mm upstream from the aortic valve annulus and in the ascending aorta at +6 mm and +10 mm downstream of the annulus. CMR imaging parameters consisted of: ET (2.76-3.05ms), flip angle (15°), phase (24), pixel spacing (1.32-2.07 mm), RT (4.6-4.92ms), thickness (10mm), matrix (256x208). Aortic TPG and valve EOA were computed using Bernoulli’s equation and continuity equation [1]. We calculated the corrected mean pressure gradient (MPG) by including CMR measurements in the following formula (combination of Bernoulli formula and continuity equation): EOA=(CO/SEP*HR)/ (44.3*MPG^0.5) [3,4]. Results: Thirty-one patients with mild to severe AS (77% men, age 67±12 years) and seven healthy subjects (71% men, age 34±8 years) were studied using TTE and CMR, Table 1. TTE overestimated VTILVOT (21±4 vs. 15±4 cm, p<0.001) and there was a good concordance between TTE and CMR for estimation of VTIAo (61±22 vs. 57±20cm and 61±22 vs. 53±19cm, p=0.02). Overall there was a good correlation and concordance between TTE-derived and CMR-derived EOA (1.52±0.68 vs. 1.60±0.74, r=0.92, bias=0.07, limits of agreement:-0.483 to 0.623 and 1.52±0.68 vs. 1.70±0.74, r= 0.88, bias=0.17, limits of agreement:-0.537 to 0.877 for 6mm and 10mm planes respectively). Figure 1 shows the fitted curves for TTE, CMR with measures at 6 mm and CMR with measures at 10 mm. Aortic TPG and mean/peak TPGs ratio were underestimated by CMR compared to TTE (Fig 2). The MPG predicted with the proposed formula correlated well with the MPG measured by TTE, r=0.76 for MPGAo06 and r=0.722 for MPGAo10 (Figure 3) and had a good concordance bias=-2.0mmHg, limits of agreement:8.3 to 4.3 and bias=-3.4mmHg, limits of agreement:-10.0 to 3.3, for MPGAo06 and MPGAo10 respectively. Discussion and Conclusion: EOA and TPG are the two main parameters used to assess AS severity. There is a good concordance between EOA measured by CMR and that measured by TTE. However, CMR underestimates the TPG compared to TTE. EOA measured by CMR can be used to confirm AS severity grading by TTE in case of inconsistencies. However, CMR underestimates TPG. The proposed model could be an issue to manage this difference. Acknowledgment: We thank CONACYT (grant 208171) for JG’s PhD scholarship, NSERC grant (343165-07) and Canadian Institutes of Health Research. References: 1. Bonow RO et al. Circulation 2006; 114:e84-e231. 2. Quinones et al. JASE 2002; 167-184. 3. Dumesnil JG et al. AJC 1991; 67: 1268-1272. 4. Minners J et al. Heart 2010; 96: 1463-1468. Table 1. Patients’ Characteristics Mean ± SD