Background
Thalassemia major (TM) is one of the most common inherited hemoglobin disorders. Ineffective erythropoiesis results in hemolytic anemia, and the patient is in need of lifelong transfusion therapy that ultimately leads to iron overload. Despite advances in the field of iron chelator therapies for TM, iron continues to accumulate in heart tissue, and subsequent cardiomyopathy remains the leading cause of death for these patients [
1‐
3].
The pathophysiology of the cardiomyopathy that develops in the modern era of iron chelator therapies is more complicated. Even though iron overload is still considered the leading cause of the occurrence of heart failure in patients with TM, the production of free radical oxygen, immunogenetic mechanisms and viral infections are being increasingly recognized [
2,
4,
5].
The complex etiopathogenetic milieu of thalassemia cardiomyopathy requires identifying the dysfunctional myocardium at an early stage to make the most of the early implementation of medical therapies. While cardiac magnetic resonance imaging (MRI) provides the ability to directly and noninvasively measure cardiac iron [
6,
7], echocardiography is still the first-line imaging tool for the assessment of myocardial function. Notably, however, conventional echocardiography parameters may still be within the normal range before the development of overt heart dysfunction. Newer echocardiographic techniques focusing on the analysis of myocardial deformation have been proven to be potentially useful tools for the early identification of myocardial dysfunction [
8].
The purpose of our study was to evaluate the changes in left ventricular global longitudinal strain and circumferential strain in patients with thalassemia major with and without myocardial iron overload.
Methods
We retrospectively studied 70 consecutive patients diagnosed with TM who were receiving blood transfusions and chelation therapy. Patients were eligible for enrollment when they fulfilled all the following criteria: (a) cardiac MRI for T2* measurement at the beginning of the study and during the follow-up period and (b) echocardiographic study at the initiation of the study and during the follow-up period. Patients were included in the study if their initial LVEF was more than 50%, (c) their transthoracic echocardiographic images for the measurement of left ventricular longitudinal and circumferential strain were of adequate quality, and (d) they had undergone a follow-up period between 12 and 36 months.
Patients with heart failure, valvulopathy, or history of myocardial infarction were not included in the study.
Echocardiography
Patients were studied in the left lateral decubitus position with a commercially available system (GE E9, Horten, Norway) coupled with a 3.5 MHz (M3S) transducer. Analysis was performed offline.
Echocardiographic study protocol
Cardiac MRI was performed no later than 1 month after the echocardiography study. Echocardiographic studies included complete 2D and Doppler examinations. Assessment of LVEF was performed using apical 4- and 2-chamber views. Doppler evaluation included the assessment of mitral inflow velocities. The mitral inflow parameters evaluated were early mitral inflow velocity (E-wave) and late or atrial mitral inflow velocity (a-wave). Peak right ventricular systolic myocardial velocity (RVSm) was obtained by placing the sample volume of tissue Doppler imaging at the lateral tricuspid valve annulus. All measurements were performed by following the current European Society of Echocardiography guidelines [
9,
10].
For both global longitudinal strain (GLS) and circumferential strain (CS) measurements, analysis was performed with dedicated software (EchoPAC v11, General Electric Medical Systems).
Longitudinal strain
Grayscale two-dimensional apical images of the LV (4-, 2-, and 3-chamber views) were obtained, and global longitudinal strain (GLS) analysis of the LV was performed by speckle-tracking imaging [
9,
10]. Three consecutive beats in each view were stored digitally for offline analysis. Mean values were calculated for all measured parameters. The frame rate was set between 50 and 100 frames/s, sector width was set as narrow as possible, and gain settings were optimized. For each view, three consecutive beats were analyzed, and mean values were calculated for all parameters derived. After the cardiac cycle was selected, the software prompted the operator to apply a region of interest in a ‘‘click-to-point approach'' to delineate the endocardium. Subsequently, the software automatically defined an epicardial and midmyocardial line and processed all frames of the selected cardiac cycle. Global longitudinal strain (GLS) was calculated as the average LS from all segments generated by software analysis of the 3 apical views.
Additionally, we separately studied non-iron–overloaded TM patients who demonstrated a relative percentage of worsening in the GLS > 15% during the follow-up period. We considered the relative percentage reduction in GLS > 15% to be a clinically meaningful reduction, similar to the cases of oncology patients where the changes in GLS are used for the early identification of myocardial damage following the administration of agents with potential cardiotoxic properties [
11].
Circumferential strain
Circumferential strain (CS) of the mid-LV was calculated using the short-axis view at the level of the papillary muscles. Peak CS was defined as the average CS of all 6 segments (generated as previously described) in the short-axis view.
Magnetic resonance imaging for cardiac and liver iron measurement
Magnetic resonance imaging examinations were performed with a 1.5-T scanner (Symphony, Siemens, Erlangen, Germany). The scans included measurements of the liver T2* and myocardial T2* values. The T2* of the heart was assessed by a cardiac-gated single breath-hold multiecho technique (FOV, 400 mm; TR, 135 ms; TE, 2.6–22.3 ms (8 echo times); flip angle, 20; slice thickness, 10 mm; matrix, 192 × 75; number of averages, 1; bandwidth in Hz/pixel, 810).
Statistical analysis
Values are reported as the mean ± SD. The paired t-test was used for intragroup comparisons of the echocardiography measurements at baseline and during follow-up. Independent-sample t-tests were used to compare baseline parameters of the TMio and Group 2 patients. For values with a nonnormal distribution, the Mann–Whitney U test was used. The χ2 test was used to compare the noncontinuous characteristics between the 2 groups. Spearman’s correlation coefficient was used to assess correlations of the measured parameters. The statistical software package SPSS for Windows was used for the analysis (SPSS 18 Inc., Chicago, IL, USA). A p-value < 0.05 was considered statistically significant.
Discussion
The main findings of the present study suggest that in patients with TM, left ventricular longitudinal strain (GLS) can detect LV subclinical dysfunction due to high cardiac iron load better than LVEF. There was a good correlation of cardiac T2* with both LVEF and GLS, both at baseline and follow-up. At the mid-period of follow-up, we detected no changes in conventional or deformational echocardiographic parameters in either iron– or non-iron–overloaded TM patients. Interestingly, during the follow-up period, we detected a relative worsening in the left ventricular longitudinal strain by more than 15%, with a concomitant marginal reduction in their corresponding LVEF, in eight patients in the non-iron—overloaded patient group.
Two-dimensional and three-dimensional LVEF measurement using echocardiography is the standard method for evaluating systolic function in TM patients. Serial studies have identified that a reduction in LVEF ≥ 7% is a strong predictive factor for cardiac death in TM populations [
14,
15]. Once heart failure symptoms occur, survival decreases substantially [
16]. The central role of iron-mediated cardiac toxicity in TM patients established the role of CMR for cardiac iron load quantification, guiding the usage of iron chelator therapy and thus improving patient outcome [
17,
18].
Early impairment of other indices of left ventricular systolic function using tissue Doppler imaging, before any remarkable reduction in left ventricular ejection fraction, has already been reported [
16,
19,
20]. In this direction, echocardiographic deformational analysis [
21] in TM patients aims to detect early changes in cardiac mechanics before any observed reduction in LVEF. This strategy could potentially prompt a change in the patient’s therapeutic strategy. Previous echocardiographic studies have reported mixed results in correlating T2* values with cardiac deformational parameters. Monte et al. [
22] and Li et al. [
23] found no correlation between left ventricular deformational parameters and T2* signals. Our study is in agreement with others [
24,
25], where they found a correlation of longitudinal strain with the T2* signal. It is noteworthy that in our study, iron–overloaded patients had similar LVEFs compared to non-iron–overloaded patients. Furthermore, a previous study revealed that cardiac and liver iron–overloaded patients have impaired left ventricular global longitudinal strain compared to non-iron–overloaded patients [
25]. Future studies could evaluate whether T2* signal measurements combined with deformational parameters obtained by cardiac MRI can potentially provide better risk stratification for TM patients. Particularly in young patients, the identification of myocardial fibrosis by MRI could be evidence of previous iron overload episodes.
During the follow-up period, iron load status according to cardiac T2* did not change in either iron– or non-iron–overloaded patients; accordingly, we did not find significant changes in conventional or deformational echocardiographic parameters, while a statistically significant correlation between T2* signal and GLS remained. This suggests that left ventricular deformational analysis can detect early myocardial systolic dysfunction and could be a sensitive tool for serial follow-up measurements of cardiac function. The absence of improvement in deformational parameters and T2* values during follow-up in iron–overloaded patients could be attributed to the patient’s compliance issues with chelation therapies, and irreversible myocardial damage secondary to repeated exposure of the myocardium to toxic factors related to thalassemia major may also play a role.
In our study, circumferential strain (CS) was also measured as an additional index of myocardial deformation, but we identified no correlation between CS and CMR T2* signals. Previous studies have shown that GLS is a more sensitive parameter than CS in identifying early impairment of LV function [
26,
27].
Impairment of left ventricular longitudinal strain has been identified as an early marker of left ventricular dysfunction in patients undergoing chemotherapy [
11]. In our study, in the group of non-iron–overloaded patients, we identified 8 of 32 patients who revealed a more than 15% worsening in their left ventricular LS deformational parameter with a marginal reduction in the EF. Consistent with our findings, the results from Marsella et al. [
28] reported that up to 10% of TM patients without cardiac iron overload could still develop heart failure.
The pathogenesis of systolic dysfunction in TM is complex. Apart from the toxic effect of iron load, there is a significant contribution of the immunoinflammatory and inherited components [
29‐
31]. Chronic tissue hypoxia in addition to chronic anemia, nutritional deficiencies and viral infections may contribute to the different susceptibilities to iron overload and cardiac damage. Moreover, myocardial fibrosis has been revealed in CMR studies in both iron– and non-iron–overload patients [
32]: long-lasting consequences of previous damage could explain the contradictory results on the correlation between the actual CMR T2* signal and left ventricular deformational parameters [
22‐
25,
32]. Our results demonstrated a significant correlation between liver and heart iron load assessed by MRI-T2*, while this correlation was lost at follow-up. These contradictory findings are consistent with previous reports [
33,
34]. These results could be attributed to differences in iron transport and deposition in these organs [
33]. Furthermore, we identified good correlations between ferritin and myocardial and liver T2* measurements. However, during follow-up, the ferritin-liver T2* correlation was persistent, but the ferritin-heart T2* correlation was not. Other studies have also reported contradictory findings in the correlation between ferritin and iron deposition on the heart and liver measured by MRI [
35‐
37], indicating that serum ferritin cannot reliably predict the liver and heart iron content (Table
4).
The small number of patients in our study indicates that our results need validation with a larger group. Additionally, longer and less variable follow-up periods could potentially detect subtle differences in the deformational mechanics in patients with thalassemia major myocardial iron overload. The cutoff point value of a 15% reduction in GLS (validated in other patient groups) is arbitrary in patients with TM, and more studies are needed to define its clinical significance. Finally, late enhancement studies with gadolinium are not performed routinely during CMR to search for myocardial fibrosis. These studies could further characterize the myocardium in TM patients.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.