Background
Systolic and diastolic dysfunction of the left ventricle are present in patients with cardiac involvement in Fabry disease. The aim of this study was to investigate the diagnostic value of the Tei index, a marker for combined diastolic and systolic function, in patients with Fabry disease.
Methods
A total of 66 consecutive patients with genetically confirmed Fabry disease were included in this study. Standard echocardiography, including the Tei index, and magnetic resonance imaging were performed. Patients were followed for 2.9 ± 1.9 years; 56 patients received enzyme replacement therapy, and 10 patients had natural history follow-up. Patients were subdivided into three groups: (1) those without cardiac involvement, (2) those with left ventricular (LV) hypertrophy and without late enhancement on magnetic resonance imaging, and (3) those with late enhancement on magnetic resonance imaging.
Results
The Tei index was significantly higher in the groups 2 (0.56 ± 0.10) and 3 (0.60 ± 0.16) compared with patients without cardiac involvement (0.44 ± 0.10) ( P < .001). All patients with Tei indexes > 0.64 showed signs of cardiomyopathy. In contrast, ejection fractions were normal in all three patient groups and therefore not useful for the detection of cardiac involvement. A significant positive correlation was observed between LV wall thickness and the Tei index in the complete patient cohort. Moreover, receiver operating characteristic analysis revealed a large area under the curve for Tei index and hypertrophy, while the area under the curve for fibrosis was small. The Tei index remained unchanged in the natural history and enzyme replacement therapy groups during follow-up.
Conclusions
In this cohort, the Tei index was of limited value to detect myocardial fibrosis and monitor enzyme replacement therapy. However, the progression of cardiomyopathy toward LV hypertrophy seems to be paralleled by global functional impairment, which can be assessed by the Tei index but not by ejection fraction. Thus, the Tei index seems to be a global parameter that can detect LV functional reduction in patients with Fabry disease.
Fabry disease (FD), an X-linked lysosomal storage disorder caused by the deficiency of α-galactosidase A, is a multisystemic disease, and the main systems involved are the kidneys, the nervous system, and the heart. The progressive accumulation of globotriaosylceramides in the heart leads to left ventricular (LV) hypertrophy and ultimately myocardial replacement fibrosis. Finally, this progression results in heart failure and arrhythmias, and cardiac complications are the major causes of death in patients with FD. Despite progressive myocardial hypertrophy and the development of myocardial fibrosis, LV ejection fraction (EF) and fractional shortening remain normal until the late stages of the disease. We and others have shown that regional systolic function (assessed by strain rate imaging) is decreased in early stages of the disease, and fibrosis, assessed by late enhancement (LE) imaging, occurs even before EF is reduced. Besides systolic strain rate, an alternative parameter (easy to perform to assess systolic function) that might be superior to EF in patients with FD remains to be validated.
Apart from systolic impairment, diastolic dysfunction can be observed in patients with FD. Formerly classified as a restrictive cardiomyopathy, it has been demonstrated in recent years that diastolic function is often only mildly to moderately altered in patients with FD. Nevertheless, it has been clearly elucidated that FD affects both systolic and diastolic performance of the left ventricle. The Tei index, also known as the myocardial performance index, assessing both systolic and diastolic function, was proposed as a marker for end-stage cardiomyopathy in a cohort of 11 patients with the cardiac variant of FD.
The aim of this study was to evaluate the association between the Tei index and cardiac involvement in a large cohort of patients with the classic phenotype. In addition, the ability of the Tei index to monitor changes of cardiac involvement during natural history and under enzyme replacement therapy (ERT) was investigated.
Methods
Study Population and Study Protocol
In total, 66 consecutive patients with genetically confirmed FD (38 men) admitted to our center between 2001 and 2008 were included in the study. The patients were followed for 2.9 ± 1.9 years; 56 patients received ERT (Fabrazyme [agalsidase-β]; Genzyme Corporation, Cambridge, MA) (1 mg/kg body weight intravenously every 2 weeks), and the remaining 10 patients served as the natural history group. LE imaging using magnetic resonance imaging (MRI) and echocardiography was performed in all patients at yearly intervals. The clinical investigation program was carried out as described elsewhere, including measurement of N-terminal pro–brain-type natriuretic peptide. The study conformed to the principles outlined in the Declaration of Helsinki, and the locally appointed ethics committee approved the research protocol. Written informed consent was obtained from all patients or their guardians.
Standard Echocardiographic Measurements
LV septal and posterior wall thickness at end-diastole were measured using M-mode echocardiography in parasternal long-axis images (Vivid 7 [3.5 MHz]; GE Vingmed Ultrasound AS, Horten, Norway), and the mean was calculated and defined as LV wall thickness. The same parasternal LV long-axis images were used to extract LV end-diastolic and end-systolic dimensions using M-mode echocardiographic methods. Fractional shortening was calculated. The LV EF was measured using the modified biplane Simpson’s method. Blood-pool pulsed Doppler of the mitral valve inflow was used to extract the ratio of early-to-late diastolic flow velocity (E/A) and the deceleration time. To measure transmitral flow, the pulsed Doppler sample window was positioned between the tips of the mitral valve leaflets, and peak flow velocities in early diastole (E wave) and during atrial contraction (A wave) were measured. From these, the E/A ratio was calculated, as well as the deceleration time of the E wave. In addition, the duration of blood flow into the left ventricle after atrial contraction was measured. Next, the pulsed Doppler sample volume was positioned 1 cm into the pulmonary veins guided by color flow data for visualization of the blood returning into the left atrium. Using these traces, the duration of blood flow after atrial contraction backward into the pulmonary veins was measured. In all unclear patients (those in whom deceleration times and E/A ratios were insufficient for the assessment of diastolic function), the difference between the duration of blood flow after atrial contraction and the duration of blood flow into the left ventricle after atrial contraction was calculated, and a value > 30 msec indicated elevated filling pressure as typically seen in pseudonormal or restrictive diastolic dysfunction. In addition, tissue color Doppler was performed with standard presets optimized to eliminate background noise and enhance tissue signals. Using the four-chamber view, a color Doppler trace was extracted from the septal and lateral mitral annulus. The peak early tissue Doppler velocity (E′) was measured, and the E/E′ ratio was calculated. Measurements were averaged over three cycles. Using all these echocardiographic parameters, diastolic function was graded according to actual guidelines for the assessment of diastolic function either as normal (grade 1), relaxation abnormalities (grade 2), pseudonormal (grade 3), or restrictive (grade 4). The Tei index was obtained by mitral filling and aortic ejection flow. As described above, the mitral inflow pattern was recorded from the apical four-chamber view with the pulsed-wave Doppler sample volume positioned between the tips of the mitral leaflets. The LV outflow velocity pattern was recorded from the apical long-axis view with the pulsed-wave Doppler sample volume positioned just below the aortic valve. LV ejection time was determined as the interval between the onset and cessation of LV outflow. The time interval from the end to the start of transmitral flow was defined as interval a and included the isovolumetric contraction time, the ejection time, and the isovolumetric relaxation time. The ejection time was defined as interval b . The Tei index was then calculated as ( a − b )/ b .
Assessment of Fibrosis
Magnetic resonance–guided LE imaging was carried out with the injection of gadopentetate dimeglumine 0.2 mmol/kg (Magnevist; Schering AG, Berlin, Germany) on a 1.5-T scanner (Magnetom Vision; Siemens Medical Systems, Erlangen, Germany). The LE technique (8-mm slice thickness, breath hold, short heart axis) was applied to detect changes of tissue integrity in the LV myocardium. Images were acquired using an inversion recovery sequence (field of view, 240 × 320 mm 2 ; matrix size, 165 × 256; repetition time, 7.5 ms; echo time, 3.4 ms; flip angle, 25°; inversion time determined individually). Short-axis views at the basal, middle, and apical segments were used for the assessment of the appearance of LE and every LV (17 segment model) segment was evaluated for the occurrence of myocardial replacement fibrosis (LE present or absent). In addition, the pattern of LE in each segment (epimyocardial, midmyocardial, or endomyocardial LE) was analyzed.
Assessment of Cardiomyopathy
The patients were divided into three groups: (1) those without signs of cardiac involvement on echocardiography and MRI, (2) those with hypertrophy but without fibrosis, and (3) those with fibrosis on MRI. Hypertrophy was defined as a septal wall thickness ≥ 12 mm. The septal wall was chosen because of the known thinning of the posterior wall in end-stage patients.
Statistical Analysis
Data are expressed as mean ± SD or as absolute patient numbers. Differences between groups were tested using unpaired t tests or one-way or two-way analysis of variance as appropriate. Receiver operating characteristic analysis was conducted to show the performance of the Tei index in the characterization of hypertrophy detection and the detection of patients with LE on MRI by the Tei index. Differences between baseline and latest follow-up data in the groups were tested using paired t tests. P values < .05 were considered as indicating statistical significance. Statistica version 8.0 (StatSoft, Inc., Tulsa, OK) was used.
Results
The clinical data of female and male patients are presented in Table 1 . Glomerular filtration rate (measured by diethylenetriamine pentaacetic acid clearance) was significantly higher in women compared with men, while no significant difference was seen in the number of patients with proteinuria. Systolic and diastolic blood pressures were similar between the two groups.
| Variable | Men ( n = 38) | Women ( n = 28) | P |
|---|---|---|---|
| Age (y) | 43 ± 12 | 43 ± 16 | .92 |
| BMI (kg/m 2 ) | 22.6 ± 3.2 | 24.8 ± 4.8 | .04 |
| Systolic BP (mm Hg) | 126 ± 19 | 121 ± 15 | .31 |
| Diastolic BP (mm Hg) | 83 ± 9 | 79 ± 11 | .18 |
| AT | 13 | 7 | .59 |
| β-blockers | 6 | 5 | .99 |
| ACE inhibitors | 14 | 6 | .28 |
| Calcium antagonists | 1 | 0 | .99 |
| GFR (mL/min) | 86 ± 47 | 115 ± 57 | .04 |
| Proteinuria ∗ | 26 | 15 | .31 |
| Creatinine (mg/dL) | 1.6 ± 1.6 | 0.7 ± 0.2 | .01 |
| Dialysis | 6 | 0 | .03 |
| Kidney transplantation | 3 | 0 | .26 |
| Hemoglobin (mg/dL) | 13.6 ± 1.5 | 13 ± 0.9 | .07 |
| NYHA class | |||
| I | 27 | 20 | .99 |
| II | 7 | 7 | .56 |
| III | 4 | 1 | .39 |
| IV | 0 | 0 | .99 |
| NT-proBNP (mg/dL) | 872 ± 2014 | 377 ± 546 | .24 |
The echocardiographic standard parameters of the patients subdivided by severity of cardiac involvement are shown in Table 2 . EFs were normal in all groups, as was fractional shortening. LV diameters were in the normal range, while LV wall thickness was normal in patients without cardiomyopathy and significantly increased in the two cardiomyopathy subgroups. Diastolic function was normal in all patients without cardiac involvement and in 22 patients (44%) with cardiomyopathy. Relaxation abnormalities were observed in 19 (38%) and pseudonormalization in nine (18%) of the cardiomyopathy patients. Restriction was not present in any of the patients. Interestingly N-terminal pro–brain-type natriuretic peptide measurements showed no difference between the group without cardiac involvement and the group without LE but with hypertrophy. In contrast, a significant difference was observed between the patients with LE and the two other groups ( Table 2 ).
| Variable | No Cardiomyopathy ( n = 16) | Hypertrophy, No Fibrosis ( n = 15) | Fibrosis ( n = 35) |
|---|---|---|---|
| LVEDD (mm) | 46 ± 5 | 50 ± 6 | 49 ± 6 |
| LVESD (mm) | 29 ± 5 | 32 ± 3 | 30 ± 5 |
| FS (%) | 36 ± 8 | 34 ± 9 | 38 ± 7 |
| IVSD (mm) | 8.7 ± 1.8 | 13.0 ± 1.1 ∗ | 14.1 ± 2.7 ∗ |
| LVPWD (mm) | 8.6 ± 1.9 | 12.6 ± 1.1 ∗ | 14.5 ± 2.3 ∗ |
| EF (%) | 65 ± 5 | 63 ± 4 | 64 ± 5 |
| E/A ratio | 1.5 ± 0.4 | 1.3 ± 0.4 | 1.3 ± 0.4 |
| DT (msec) | 192 ± 38 | 211 ± 65 | 229 ± 61 |
| Diastolic function | |||
| Normal | 16 | 9 ∗ | 13 ∗ |
| Abnormal relaxation | 0 | 4 ∗ | 15 ∗ |
| Pseudonormal | 0 | 2 | 7 |
| Restriction | 0 | 0 | 0 |
| NT-proBNP | 100 ± 98 | 94 ± 88 | 1139 ± 2035 ∗,† |
| Tei index | |||
| Women | 0.46 ± 0.10 ( n = 13) | 0.54 ± 0.14 ( n = 3) | 0.56 ± 0.10 ( n = 12) |
| Men | 0.38 ± 0.07 ( n = 3) | 0.57 ± 0.10 ( n = 12) ∗ | 0.62 ± 0.18 ( n = 23) ∗ |
∗ P < .05 versus no cardiomyopathy.
Tei Index Versus Cardiomyopathy
The mean Tei index was 0.51 ± 0.11 in female patients and 0.59 ± 0.17 in male patients ( P = .04). When analyzing the Tei index in subgroups according to no cardiac involvement (group 1), hypertrophy without fibrosis (group 2), and fibrosis (group 3), the Tei index was significantly higher in groups 2 (0.56 ± 0.10) and 3 (0.60 ± 0.16) compared with group 1 (0.44 ± 0.10) ( P < .001). When stratified by gender, no significant differences were seen between female and male patients in the same subgroups ( Table 2 ), while the analysis within genders showed a continuous Tei index increase of the mean in these three groups, although not significant in female patients. More men had severe cardiomyopathy, while nearly half of women showed no cardiomyopathy. The Tei index ranged from 0.40 to 1.22 in patients with cardiac involvement and from 0.30 to 0.64 in patients without cardiomyopathy. LV hypertrophy or LE was evidenced in all patients with Tei indexes > 0.64 ( n = 15 [23%]).
For the association between myocardial performance and LV morphology, a significant positive correlation was observed between LV wall thickness (measured at end-diastole in the septal wall because of the known thinning of the posterior wall in patients with end-stage FD; r = 0.66, P < .001) and the Tei index in the complete patient cohort. Thus, patients with the highest amount of LV hypertrophy presented the highest Tei indexes ( Figure 1 ).
LE was detected in 43% of the women ( n = 12) and 61% of the men ( n = 23) ( P = .21). On MRI, LE in women consistently occurred in the basal posterolateral wall ( n = 12). In four women, additional segments showed LE; in the others, LE was limited to the typical location (posterolateral). In men, all but one patient showed LE in the basal posterolateral wall, one in the anterior basal to apical segments but not posterolateral, and eight patients showed LE in additional segments apart from the typical location. When focusing on the extension of LE within the wall, 28 patients showed localized midmyocardial LE in the typical location. One patient had additional endocardial fibrosis lateral-basal, and one patient showed epicardial and endocardial fibrosis. In five patients, transmural LE was observed.
The Tei index distribution of the patients with and without LE in MRI is shown in Figure 2 . The figure shows that fibrosis is very unlikely with normal or near normal Tei indexes. In contrast, high Tei indexes do not necessarily predict cardiac fibrosis in FD. Figure 3 and 4 show the performance of the Tei index in the characterization of patients with LV hypertrophy and patients with LE on MRI. It is of note that the area under the curve is larger for hypertrophy than for LE.
