Background
Fabry cardiomyopathy is characterized by progressive left ventricular hypertrophy (LVH) associated with diastolic dysfunction and is the most common cause of death in Fabry disease (FD). However, LVH is not present in all subjects, particularly early in disease progression and in female patients. Direct assessment of myocardial deformation by strain and strain rate (SR) analysis may be sensitive to detect subclinical Fabry cardiomyopathy independent of the presence of LVH.
Methods
Systolic (longitudinal, circumferential, and radial systolic strain and SR) and diastolic (SR during isovolumic relaxation [SR IVR ] and early diastole and strain at peak transmitral E wave) function was assessed in 16 patients with FD using two-dimensional speckle-tracking echocardiography. In addition, mean S′ and E′ mitral annular velocities by Doppler tissue imaging were measured. Diastolic filling indices, including E/SR IVR and E/E′ ratios, were calculated. The patients were compared with 24 healthy age-matched and gender-matched controls.
Results
All 16 patients with FD had normal left ventricular ejection fractions, and nine patients had LVH. Compared with controls, patients with FD had reduced longitudinal systolic strain ( P < .001) and systolic SR ( P = .007), while there were no differences in circumferential systolic strain and S′. Diastolic function assessment showed reduced longitudinal early diastolic SR ( P = .001), SR IVR ( P < .001), and E/SR IVR ( P < .001), while radial and circumferential diastolic function was not affected. Of the conventional diastolic function indices, reductions were seen in E ( P = .006), E′ ( P = .021), and E/E′ ratio ( P < .001). After correcting for LVH, only SR IVR ( P < .001) and E/SR IVR ( P = .025) remained significantly different between patients with FD and controls, with sensitivity of 94% and specificity of 92% for SR IVR of 0.235 sec −1 (area under the receiver operating characteristic curve, 0.953).
Conclusions
Strain and SR analysis is useful in identifying patients with FD with reduced myocardial function, with longitudinal systolic strain and diastolic isovolumic SR being superior to the other echocardiographic measurements of myocardial contraction and relaxation and independent of LVH.
Fabry disease (FD) is an X-linked glycolipid storage disease caused by a deficiency of α-galactosidase A enzyme resulting in progressive intracellular accumulation of glycosphingolipids in different tissues. Cardiac involvement is characterized by progressive left ventricular (LV) hypertrophy (LVH), heart failure, and arrhythmias. Importantly, subclinical cardiac involvement may represent the first sign of organ damage, particularly in female carriers. Large screening studies have shown a high incidence of late-onset FD. Fabry cardiomyopathy is one of the most common causes of death in these patients. Enzyme replacement therapy has proved effective in reducing glycosphingolipid accumulation in the myocardium and improving cardiac function. Echocardiography is a standard noninvasive screening test for Fabry cardiomyopathy. Because most patients with FD have normal LV systolic function, and up to 40% do not have LVH at the time of the diagnosis, using conventional echocardiography makes it challenging to recognize cardiac involvement. Strain and strain rate (SR) imaging has emerged as a potentially useful technique for detecting early subclinical systolic dysfunction in patients with FD and for monitoring the efficacy of the enzyme replacement therapy. Similarly, diastolic function assessment by Doppler tissue imaging (DTI) has been shown to be useful in diagnosing early stages of Fabry cardiomyopathy before the development of LVH. Recently, novel global diastolic function indices derived from strain and SR analysis have been more directly related to invasively measured myocardial relaxation than the conventional Doppler methods. The purpose of this study was to assess systolic and diastolic myocardial function in patients with FD by strain and SR analysis using two-dimensional (2D) speckle-tracking imaging. We hypothesized that strain and SR parameters, particularly those measured during early diastole, would allow differentiation between patients with FD and healthy controls, independent of LVH and superior to tissue Doppler measurements.
Methods
Patient Population
Sixteen consecutive patients with genetically confirmed FD were included. The patients were compared with 24 healthy age-matched and gender-matched controls recruited from hospital staff members and their relatives. All study participants had histories and physical examinations performed, and baseline clinical characteristics of the patients were obtained from the computerized hospital database. All study participants underwent electrocardiography and standard transthoracic echocardiography. LVH on electrocardiography was assessed using Sokolow-Lyon and Cornell criteria. The study was approved by the health ethics research board, and consent was obtained from all study subjects.
Standard Echocardiography
Echocardiography was performed using commercially available ultrasound equipment (M3 S probe, Vivid 7; GE Vingmed Ultrasound AS, Horten, Norway). All images were digitally stored for offline analysis using EchoPAC version BT07.0.0 (GE Vingmed Ultrasound). Complete 2D and Doppler images were acquired according to standard techniques. Maximal left atrial volumes were calculated using the biplane area-length formula and indexed to body surface area. Severity of mitral regurgitation was determined using an integrative approach including semiquantitative and quantitative color Doppler-based parameters as recommended by current guidelines. LV mass was derived from the LV linear dimensions and calculated using Devereux’s formula, with subsequent identification of patients with mild, moderate, and severe LVH using the cutoff values as per American Society of Echocardiography guidelines. LV end-systolic and end-diastolic volumes were calculated using Simpson’s biplane method of disks and indexed to body surface area. LV ejection fraction was subsequently derived. Evaluation of diastolic function included conventional Doppler-based measurements. First, pulsed-wave Doppler velocities were measured from the apical four-chamber view placing a 2-mm sample volume at the tips of the mitral leaflets. Transmitral E and A diastolic velocities, and the E-wave deceleration time were measured. Tissue Doppler was applied in the pulsed Doppler mode to record S′ and E′ mitral annular velocities at the septal and lateral corners, with the measurements being averaged (mean S′ and mean E′, respectively). The E/E′ ratio was derived as a measure of LV filling pressures. An E/E′ ratio < 8 represents normal LV filling pressures, whereas an E/E′ ratio > 13 indicates increased filling pressures, and E/E′ ratios of 8 to 13 are indeterminate.
Speckle-Tracking Imaging
To evaluate longitudinal LV function, 2D images of apical four-chamber, two-chamber, and three-chamber views were obtained using 2D-speckle-tracking imaging with the highest possible frame rates. Parasternal short-axis views at the papillary muscle level were used to assess radial and circumferential function. During 2D speckle-tracking analysis, the endocardial border was manually traced, and the region of interest width was adjusted to include the entire myocardium. The software package (EchoPAC version BT07.0.0) automatically tracks and accepts segments of good tracking quality, while allowing the observer to manually override its decisions. Individual global strain and SR curves were obtained automatically from each view. Peak global longitudinal, radial, and circumferential strain and SR were measured during the ejection phase, isovolumic relaxation period, and early diastole. Data for longitudinal function are presented as the mean of the three apical views. Time from the R wave of the QRS complex to the peak transmitral E wave was used to identify early diastolic strain.
In summary, the following global indices of LV function were obtained from 2D speckle-tracking ( Figure 1 ): (1) systolic strain (ε), (2) systolic SR (SR S ), (3) early diastolic strain, (4) early diastolic SR (SR E ), and (5) SR during the isovolumic relaxation period (SR IVR ). In addition, an index of LV filling pressure was calculated as the ratio of the peak transmitral E wave to SR IVR (E/SR IVR ). To determine intraobserver and interobserver variability, 2D speckle-tracking measurements were repeated in 10 randomly selected subjects ≥4 weeks apart by the same observer on the same echocardiographic images and by a second independent observer. To determine reduced myocardial relaxation, abnormal E′ and SR IVR were classified as E′ and SR IVR below that of the mean minus 2 standard deviations for each parameter of the healthy control group. To determine increased LV filling pressures, abnormal E/E′ ratio was classified as E/E′ ratio > 13, and abnormal SR IVR was classified as SR IVR above that of the mean plus 2 standard deviations of the healthy control group measurements.
Statistical Analysis
Continuous variables are presented as mean ± SD. Categorical data are summarized as frequencies and percentages. Unpaired Student’s t tests were used to compare continuous data between patients with FD and healthy controls and from the subgroups of patients with FD with and without LVH. Pearson’s correlation was used to examine the linear associations between continuous variables. Multiple linear regression analysis was used to determine associations of echocardiographic diastolic parameters with FD, controlling for the presence of LVH. Statistical significance was taken as P < .05. Intraobserver and interobserver agreement for 2D speckle-tracking measurements was evaluated by calculating intraclass correlation coefficients, with good correlation being defined as an intraclass correlation coefficient > 0.8. Receiver operating characteristic (ROC) curves were plotted to determine the area under the curve, and the net reclassification index was determined using the computing environment R (R Foundation for Statistical Computing, Vienna, Austria) and R packages rms and Hmisc. All other statistical analyses was performed using SPSS version 20 for Windows (SPSS, Inc, Chicago, IL).
Results
The study cohort consisted of 16 patients with FD and 24 controls, with 10 male (62%) and six female (38%) patients in the FD group and 15 male (62%) and nine female (38%) patients in the control group ( Table 1 ). Table 2 summarizes the baseline clinical characteristics of patients with FD. The mean systolic and diastolic blood pressures in our FD cohort were 117 ± 3.5 and 69 ± 3 mm Hg, respectively, while in our four patients with FD with treated hypertension, mean blood pressures were 119 ± 5.5 and 71 ± 4.5 mm Hg, respectively. Thirteen patients (82%) received enzyme replacement therapy (mean treatment duration, 7.0 ± 2.6 years). Twelve (75%) and four (25%) patients with FD were in New York Heart Association functional classes I and II, respectively. Table 3 summarizes echocardiographic characteristics of the study subjects. All study participants had no or trace mitral regurgitation. The mean LV mass was 93.2 ± 26.3 and 129.7 ± 33.4 g/m 2 in female and male patients, respectively. Overall, patients with FD had significantly increased LV mass compared with controls (116.0 ± 35.1 vs 70.8 ± 16.0 g/m 2 , P < .001), with nine patients (56%) having LVH on echocardiography. There were no differences in LV and left atrial volumes, LV ejection fraction, and S′ between the two groups. Diastolic function assessment revealed significantly higher E velocities (91.3 ± 21.9 vs 74.4 ± 15.1 cm/sec, P = .006) in patients with FD but no difference in E/A velocity ratio between the two groups. In addition, patients with FD had significantly lower mitral annular E′ velocities (9.1 ± 3.0 vs 11.1 ± 2.3 cm/sec, P = .021) and significantly increased E/E′ ratio (10.9 ± 4.4 vs 6.9 ± 1.9, P < .001) compared with controls. On further analysis, four patients (27%) with FD had E/E′ ratios ≥ 13, suggesting elevated LV filling pressures, while 11 patients (73%) with FD had either normal (four patients) or indeterminate (eight patients) filling.
| Variable | Patients | Controls | P |
|---|---|---|---|
| ( n = 16) | ( n = 24) | ||
| Age (y) | 45.0 ± 11.4 | 47.00 ± 10.2 | .566 |
| Male/female | 10/6 | 15/9 | |
| Height (cm) | 171.9 ± 10.7 | 171.3 ± 10.2 | .853 |
| Weight (kg) | 78.0 ± 17.1 | 74.8 ± 15.3 | .536 |
| Body surface area (cm 2 ) | 1.91 ± 0.21 | 1.87 ± 0.20 | .556 |
| Variable | Value |
|---|---|
| Cardiovascular history | |
| Hypertension | 4 (25%) |
| Dyslipidemia | 4 (25%) |
| Diabetes mellitus | 1 (6%) |
| Current smoking | 3 (15%) |
| Stroke/transient ischemic attack | 5 (31%) |
| Coronary artery disease | 1 (6%) |
| Paroxysmal atrial fibrillation | 4 (25%) |
| Renal involvement | |
| Creatinine (μmol/L) | 104 ± 31 |
| eGFR (mL/min/1.73 m 2 ) | 73.8 ± 18.4 |
| Proteinuria | 11 (69%) |
| Dialysis | 1 (6%) |
| Kidney transplantation | 1 (6%) |
| Electrocardiography | |
| LVH | 7 (44%) |
| Right ventricular hypertrophy | 2 (13%) |
| Left bundle branch block | 1 (6%) |
| Right bundle branch block | 1 (6%) |
| PR interval (msec) | 163 ± 22 |
| Medications | |
| Antiplatelet agents | 12 (75%) |
| β-blockers | 3 (15%) |
| ACE inhibitors | 11 (69%) |
| ARBs | 5 (31%) |
| Calcium channel blockers | 2 (13%) |
| Statins | 11 (69%) |
| Variable | Patients | Controls | P |
|---|---|---|---|
| ( n = 16) | ( n = 24) | ||
| LV mass (g/m 2 ) | 116.0 ± 35.1 | 70.8 ± 16.0 | <.001 |
| Normal | 7 (44%) | 24 (100%) | |
| Mild LVH | 3 (19%) | 0 (0%) | |
| Moderate LVH | 3 (19%) | 0 (0%) | |
| Severe LVH | 3 (19%) | 0 (0%) | |
| LV ejection fraction (%) | 63.0 ± 5.9 | 60.9 ± 6.5 | .313 |
| End-systolic volume index (mL/m 2 ) | 21.9 ± 9.1 | 20.5 ± 6.1 | .545 |
| End-diastolic volume index (mL/m 2 ) | 57.9 ± 20.7 | 51.7 ± 11.1 | .223 |
| Left atrial volume index (mL/m 2 ) | 29.4 ± 10.7 | 25.0 ± 4.9 | .084 |
| Mean mitral annular S′ velocity (cm/sec) | 8.0 ± 1.6 | 8.8 ± 1.5 | .118 |
| Mitral E velocity (cm/sec) | 91.3 ± 21.9 | 74.4 ± 15.1 | .006 |
| Mitral A velocity (cm/sec) | 64.4 ± 28.6 | 66.2 ± 14.7 | .797 |
| Transmitral E/A velocity ratio | 1.98 ± 2.0 | 1.17 ± 0.3 | .058 |
| Deceleration time (msec) | 229 ± 47.2 | 207 ± 44.1 | .141 |
| Mean mitral annular E′ velocity (cm/sec) | 9.1 ± 3.0 | 11.1 ± 2.3 | .021 |
| E/E′ velocity ratio | 10.9 ± 4.4 | 6.9 ± 1.9 | <.001 |
Representative traces of global 2D speckle-tracking in a healthy control and a patient with FD are shown in Figure 1 , and the summarized global strain and SR parameters of LV systolic and diastolic function are outlined in Table 4 . The mean 2D speckle-tracking frame rates were 87.3 ± 12.2 frames/sec in patients with FD and 92.6 ± 14.7 frames/sec in controls. The overall feasibility for strain and SR parameters was 100%. Of the systolic function indices, patients with FD had significantly reduced longitudinal ε (−16.0 ± 3.8% vs −19.5 ± 1.8%, P < .001) and SR S (−0.94 ± 0.16 vs −1.06 ± 0.11 sec −1 , P = .007). Of the diastolic function indices, patients with FD had significantly reduced longitudinal SR E (1.01 ± 0.31 vs 1.33 ± 0.27 sec −1 , P = .001) and SR IVR (0.15 ± 0.08 vs 0.35 ± 0.11 sec −1 , P < .001). In addition, E/SR IVR ratio, reflecting diastolic filling pressures, was significantly increased in patients with FD compared with healthy controls (801.8 ± 483 vs 232.5 ± 88.3 cm, P < .001). Importantly, patients with FD without LVH ( n = 7) had significantly decreased SR IVR (−0.16 ± 0.06 vs −0.35 ± 0.1 sec −1 , P < .001) compared to the healthy controls, while there was no difference in the mitral annular E′ velocity between the two groups.
| Variable | Patients | Controls | P |
|---|---|---|---|
| ( n = 16) | ( n = 24) | ||
| Mean ε (%) | |||
| Longitudinal | −16.0 ± 3.8 | −19.5 ± 1.8 | <.001 |
| Radial | 51.6 ± 21.9 | 46.6 ± 15.5 | .407 |
| Circumferential | −20.2 ± 6.1 | −17.8 ± 2.9 | .107 |
| Mean SR S (sec −1 ) | |||
| Longitudinal | −0.94 ± 0.16 | −1.06 ± 0.11 | .007 |
| Radial | 2.53 ± 0.94 | 2.42 ± 0.66 | .646 |
| Circumferential | −1.29 ± 0.27 | −1.14 ± 0.28 | .115 |
| Mean early diastolic strain (%) | |||
| Longitudinal | −12.8 ± 2.9 | −13.5 ± 2.2 | .337 |
| Radial | 34.7 ± 20.7 | 28.4 ± 9.8 | .200 |
| Circumferential | −12.2 ± 4.0 | −11.1 ± 3.2 | .370 |
| Mean SR E (sec −1 ) | |||
| Longitudinal | 1.01 ± 0.31 | 1.33 ± 0.27 | .001 |
| Radial | −2.59 ± 1.26 | −2.68 ± 0.72 | .766 |
| Circumferential | 1.38 ± 0.42 | 1.38 ± 0.40 | .971 |
| SR IVR (sec −1 ) | |||
| Longitudinal | 0.15 ± 0.08 | 0.35 ± 0.11 | <.001 |
| Radial | −0.91 ± 0.65 | −0.90 ± 0.51 | .961 |
| Circumferential | 0.55 ± 0.43 | 0.45 ± 0.20 | .340 |
| E/SR IVR ratio (cm) | 801.8 ± 483.6 | 232.5 ± 88.3 | <.001 |
Table 5 describes the clinical and echocardiographic characteristics of the individual patients. The individual conventional and novel measurements of diastolic function displayed as data points are shown in Figure 2 . DTI and speckle-tracking imaging were compared for their ability to detect impaired myocardial relaxation and increased LV filling pressures in patients with FD. Using the cutoffs of <6.5 cm/sec for mitral annular E′ velocities and <0.13 sec −1 for SR IVR , derived from the mean minus 2 standard deviations in the healthy controls, three patients (19%) had impaired myocardial relaxation by conventional Doppler echocardiography, while it was identified in seven patients (44%) using speckle-tracking imaging ( Table 5 ). Using the cutoffs of >13 for E/E′ ratio as per recommendations and >410 for E/SR IVR ratio, only four patients (25%) had abnormal filling pressures by E/E′ ratio, compared with 11 patients (69%) on the basis of E/SR IVR ratio. Interestingly, a significant association with FD after correcting for LVH was present only for SR IVR ( P < .001) and E/SR IVR ratio ( P = .025) ( Table 6 ). On the basis of the healthy controls, the ROC curves for the LVH, diastolic DTI, and SR measurements were also evaluated and plotted ( Figure 3 ). The largest areas under the curve were for SR IVR (0.953) and E/SR IVR ratio (0.947), compared with 0.851 for E/E′ ratio, 0.753 for E′, and 0.781 for LVH ( Figure 3 ). The diastolic parameter considered the best predictor for FD was SR IVR. We tested the ability of the SR IVR ROC to improve the classification of FD compared with E′ and E/E′ ratio, which resulted in net reclassification index values of 1.29 (95% confidence interval, 0.66–1.92) and 0.92 (95% confidence interval, 0.29–1.55). ROC analysis identified an optimum cutoff of 0.235 sec −1 , with sensitivity of 94% and specificity of 92%, to discriminate patients with FD with cardiac involvement. Although renal failure can contribute to cardiac disease, estimated glomerular filtration rate or the degree of proteinuria in the FD cohort did not correlate with SR IVR ( P = .31 and P = .49, respectively) or E/SV IVR ratio ( P = .25 and P = .64, respectively).
| Patient | Age/sex | Comorbidities | NYHA class | ERT (y) | LVH | Mean E′ (cm/sec) | Mean E/E′ ratio | E/A ratio | DT (msec) | Longitudinal | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SR E (s −1 ) | SR IVR (sec −1 ) | E/SR IVR ratio (cm) | ||||||||||
| 1 | 46/F | HTN | I | None | Yes | 10.2 | 9.8 | 1.02 | 160 | 1.18 | 0.15 | 668.7 |
| 2 | 51/F | Proteinuria, PAF | I | 5 | Yes | 7.95 | 12.2 | 1.22 | 230 | 0.90 | 0.10 | 969.0 |
| 3 | 52/M | Stroke, PAF, kidney transplantation | I | 9.5 | Yes | 5.8 | 22.2 | 1.15 | 270 | 0.33 | 0.15 | 858.7 |
| 4 | 35/M | Proteinuria | I | 9 | No | 12.8 | 6.3 | 1.93 | 180 | 1.27 | 0.22 | 363.6 |
| 5 | 68/F | HTN, DM, TIA | II | 6.5 | No | 6.65 | 7.2 | 0.92 | 210 | 0.61 | 0.16 | 299.4 |
| 6 | 47/M | Dyslipidemia, proteinuria | I | 6.5 | Yes | 8.15 | 14.3 | 1.82 | 230 | 0.99 | 0.17 | 686.5 |
| 7 | 50/M | TIA, PAF | II | 9.5 | Yes | 9.05 | 7.4 | 1.32 | 350 | 1.08 | 0.11 | 604.6 |
| 8 | 33/M | Proteinuria | I | 8.5 | No | 8.90 | 9.3 | 1.28 | 210 | 1.21 | 0.23 | 359.6 |
| 9 | 30/M | Dyslipidemia, Proteinuria | I | 5.5 | No | 14.5 | 8.0 | 1.35 | 240 | 0.96 | 0.20 | 582.5 |
| 10 | 34/M | HTN, dyslipidemia, proteinuria | I | 9 | Yes | 13.0 | 9.6 | 1.05 | 190 | 1.21 | 0.36 | 311.9 |
| 11 | 48/M | Proteinuria | I | 8.5 | Yes | 5.0 | 17.5 | 1.33 | 190 | 0.64 | 0.04 | 2192.5 |
| 12 | 26/F | None | I | None | No | 13.61 | 7.3 | 1.57 | 280 | 1.54 | 0.08 | 1237.5 |
| 13 | 50/M | HTN, CAD, dyslipidemia, proteinuria | II | 8.5 | No | 6.05 | 14.9 | 1.06 | 220 | 0.91 | 0.10 | 871.0 |
| 14 | 50/M | PAF, proteinuria | II | 0.2 | Yes | 7.65 | 8.8 | 1.08 | 220 | 0.80 | 0.04 | 1502.3 |
| 15 | 67/F | TIA, proteinuria | I | 5 | No | 7.15 | 10.1 | 0.94 | 220 | 1.27 | 0.20 | 350.5 |
| 16 | 39/F | Proteinuria | I | None | Yes | 9.15 | 10.9 | 1.58 | 260 | 1.30 | 0.09 | 1153.9 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
