Abstract
Patients with chronic anemia develop compensatory ventricular dilation, even when maintained on chronic transfusion regimens. It is important to characterize these effects to interpret pathological changes in cardiac dimensions and function introduced by iron overload and sickle cell vasculopathy. Our primary goal was to compare biventricular dimensions and function assessed by cardiovascular magnetic resonance (CMR) in pediatric, chronically-transfused sickle-cell disease (SCD) and thalassemia major (TM) patients who had normal cardiac iron levels. Moreover, we explored systematic sex differences in ventricular dimensions in both populations. We identified 261 studies suitable for analysis from 64 patients with SCD (34 females) and 49 patients with TM (20 females). All demographic and CMR parameters were inversely weighted by the number of exams. In both populations, males had larger left and right ventricular dimensions than females, with a more marked effect observed in patients with SCD. Compared to patients with TM, patients with SCD showed significantly greater biventricular dilation and left ventricular hypertrophy. This difference could not be explained by different hemoglobin levels, cardiac iron overload, and systolic blood pressure. The left ventricular (LV) ejection fraction (EF) for the males and the right ventricular (RV) EF for both the sexes were comparable between SCD and TM groups, while females with SCD had significantly lower LV EF than females with TM. Our results represent important baseline findings that place changes introduced by iron overload as well as systemic and pulmonary vasculopathy in proper context.
Introduction
Sickle cell disease (SCD) and thalassemias represent the most common forms of inherited disorders of hemoglobin. Although SCD results from qualitative abnormalities in the structure of the β-globin chains [1], thalassemias result from quantitative abnormalities of the expression of one or more of the globin chains of hemoglobin [2], with β-thalassemia major (TM) being the most severe form. Despite different pathophysiologies, both SCD and TM undergo intramedullary or intravascular hemolysis with severe anemia as a common feature. The hemodynamic consequence of a reduced oxygen-carrying capacity due to anemia is the need for an increased cardiac output [3]. In fact, even at hemoglobin levels approaching 10 g/100 ml, the increase in cardiac output necessary to maintain a given level of exercise is greater in patients with anemia than in normal subjects [4]. Increased cardiac output can be achieved either by increased ventricular stroke volume index or by increased heart rate. However, cardiac chamber dilation may be difficult to interpret in patients on chronic transfusion therapy. Chronic transfusions have been used for decades in patients with TM and many patients with SCD are now managed with chronic transfusions to prevent de novo or recurrent neurovascular complications and other serious morbidities [5]. Because humans lack any effective means to excrete excess iron, a possible drawback of this treatment is a secondary state of iron overload in the vital organs of the body [6]. Cardiac and vascular iron overload may reduce ventricular dimensions initially through vascular and ventricular stiffening [7–9] but may increase ventricular dimensions and decrease systolic function in end stage disease [10,11].
In this context, cardiovascular magnetic resonance (CMR) emerges as a key technique in the serial asessment of patients with chronic transfused SCD and TM. CMR is the gold-standard to assess biventricular size and function, thank to its excellent accuracy and reproducibility [12]. Although normative CMR values have been reported for adults with thalassemia [13,14] there is no published data about patients with pediatric SCD or TM. Moreover, CMR allows to non-invasively assess cardiac iron loading by the measurement of R2* values [15,16]. Importantly, cardiac iron burden has found to be common in TM but quite rare in patients with SCD [17,18].
Hence, the primary goal of this study was to compare right and left ventricular dimensions and function in pediatric SCD and TM subjects lacking cardiac iron to characterize differences in cardiovascular compensation for chronic anemia. Sex differences exist in these parameters in normal subjects, so our secondary goal was to explore systematic sex differences in ventricular dimensions in both populations.
Methods
Study population
We reviewed all CMRs performed in patients with SCD and TM regularly transfused referred to the Children’s Hospital Los Angeles (CHLA) between April 2004 and June 2011 and we selected those satisfying the following conditions:
patient’s age ≤ 18 years;
access to the pretransfusion hemoglobin value of the patient assessed within the space of 3 months;
cardiac R2* < 50 Hz, corresponding to a T2* ≥ 20 msec and indicating no significant myocardial iron burden [11]. In case of multiple CMRs for the same patient, if one exam showed cardiac iron, the patient was not considered in the analysis.
Two-hundred and sixty-one CMR studies (131 from SCD and 130 from TM) suitable for analysis were identified from 64 patients with SCD (34 females) and 49 patients with TM (20 females).
Waiver of informed consent to analyze retrospective data was obtained in accordance with the Committee on Clinical Investigation at CHLA.
CMR
CMR was performed on one of the two available 1.5T scanners: the Philips Achieva (Philips Medical Systems, Best, The Netherlands) running system 2.5.1 and the GE Signa CVi (GE Healthcare, Waukesha, WI) running system 9.1. Similar phased array torso coils were used on both scanners and ECG-gating was used for cardiac acquisitions.
For cardiac iron overload assessment, a single mid-ventricular short axis slice was acquired using a multiecho gradient echo sequence, with a number of echoes dependent on the scanner (Philips: 16 echo times starting from 1.28 msec with an echo spacing of 1.09 msec; GE: 8 echo times starting from 1.40 msec with an echo spacing of 2.38 msec). Signal-decay curves were measured from a region of interest encompassing the interventricular septum and fit to mono exponential + constant model [19]. The constant was used to compensate for the signal “offset” due to rectified noise and unsuppressed blood signal. Since we used a pixelwise approach (fitting performed for all the pixels in the ROI), the truncation model, consisting in manually “discarding” the later echoes where the plateau becomes evident, was not applicable. R2*, equal to 1,000/T2*, was taken into account because it is proportional to tissue iron [20,21].
For the evaluation of biventricular volumes and ejection fractions (EF), 15 contiguous short-axis steady-state free precession cine images were acquired during 8-sec breath holds with slice thickness adjusted to span the heart, resulting in thinner slices for smaller hearts. Twenty to 30 cardiac phases were acquired per heart-beat. Image analysis was performed using semiautomatic recognition of the endocardial and epicardial borders of the left ventricular (LV) wall and of the right ventricular (RV) borders in end-diastole and end-systole, with manual correction by a single cardiologist (JCW). For the calculation of end-diastolic and end-systolic volumes (EDV and ESV, respectively), no geometric assumption of the shape of ventricles was needed; volume was calculated by sum of disks over all the relevant slices. The LV mass was given by the volume of the myocardium multiplied by its specific weight of 1.05 g/cm3. Papillary muscles were excluded when measuring volumes (blood pool techniques). Cardiac output was measured by phase contrast flow assessment from an imaging plane placed at the sino-tubular junction in the ascending aorta, using a velocity encoding of 200 cm/sec; all acquisitions were acquired free breathing with respiratory averaging (three excitations). Biventricular volumes and LV mass were indexed to the body surface area (BSA) derived using the variation of the Dubois and Dubois formula [22]. Similarly, the cardiac index (CI) was calculated as the ratio between the cardiac output and the BSA.
Statistical analysis
All analyses were performed using SAS JMP v. 5.1 statistical software (SAS Institute, Cary, NC).
All demographic and CMR parameters were inversely weighted by the number of exams such that each patient contributed equally to correlation and group-wise analyses.
All continuous variables were expressed as the mean ± standard deviation (SD). The Shapiro–Wilks W-test was applied to check the normality of the residual distributions.
For continuous values with normal distribution, comparisons between two groups were made by independent-samples t-test. First, Levene’s test was applied to verify the homogeneity of variances (homoscedasticity). When the significance level of Levene’s test was <0.05 and homoscedasticity could not be assumed, the Welch statistic was used. The Wilcoxon signed-rank test was applied for continuous values with non normal distribution.
Analysis of covariance (ANCOVA) models was used to evaluate the impact of potential covariates on group differences in CMR parameters. Covariates were included if a variable was significantly different between groups and associated with the outcome being assessed. When necessary, outcomes were log transformed to normalize the residual distributions and to equalize the residual variance. In the ANCOVA, the interaction term involving the covariate was first tested for heterogeneity of slopes [23]. When homogeneity of slopes was confirmed, the subsequent model omitted the interaction and we proceeded with the standard ANCOVA model. In case of multiple covariates, the effect of each covariate was first examined separately.
Correlation analysis was performed using Pearson’s test or Spearman’s where appropriate.
Approximate age, sex, and disease specific Z-scores for the CMR functional parameters can be derived from the regression equations in the supplemental data table by dividing the difference of the observed and the predicted value by the mean-squared error of the regression. When the regression equations are not statistically significant, a Z-score can be obtained by substituting the mean and standard deviation accordingly.
In all tests, a two-tailed probability value of 0.05 was considered statistically significant.
Results
Sex differences in patients with SCDs
Demographics of patients with SCD can be extracted from Table I. The sexes were comparable for age (P = 0.063), hemoglobin values (P = 0.213), and serum ferritin levels (P = 0.291). No difference was present in systolic (P = 0.666), diastolic (P = 0.173), and mean (P = 0.725) blood pressures (BP). The heart rate (HR) was significantly lower in males than in females (P = 0.0009). Cardiac R2* values were comparable between the sexes (P = 0.408), and the distribution was also similar.
TABLE I.
Comparison of Demographics between Patients with SCD and TM Without Cardiac Iron Overload with the Differentiation for Sex
| SCD | TM | P-value | |
|---|---|---|---|
| Males | |||
| Age (years) | 12.3 ± 3.4 | 11.6 6 3.2 | 0.167 |
| Hemoglobin (g/dl) | 9.7 ± 0.6 | 10.0 6 0.9 | 0.554 |
| Serum ferritin (ng/ml) | 3481.1 ± 1755.2 (26/34 pts; 46 MRIs) | 1607.7 6 742.7 (26/29 pts; 74 MRIs) | <0.0001 |
| Systolic blood pressure (mmHg) | 106.2 ± 9.7 (31/34 pts; 54 MRIs) | 101.5 6 8.1 (26/29 pts; 75 MRIs) | 0.048 |
| Diastolic blood pressure (mmHg) | 56.5 ± 8.4 (31/34 pts; 54 MRIs) | 53.7 6 5.8 (26/29 pts; 75 MRIs) | 0.132 |
| Mean blood pressure (mmHg) | 76.5 ± 9.4 (31/34 pts; 54 MRIs) | 74.6 6 6.4 (26/29 pts; 75 MRIs) | 0.368 |
| Heart rate (bpm) | 76.4 ± 7.8 | 84.1 6 10.0 (27/29 pts; 75 MRIs) | 0.015 |
| R2* heart (Hz) | 28.6 ± 3.3 | 31.1 6 3.8 | 0.035 |
| Females | |||
| Age (years) | 11.4 ± 2.6 | 11.9 6 3.1 | 0.375 |
| Hemoglobin (g/dl) | 9.7 ± 0.6 | 9.9 6 0.9 | 0.615 |
| Serum ferritin (ng/ml) | 2863.8 ± 983.4 (28/30 pts; 68 MRIs) | 1954.1 6 233.6 (18/20 pts; 49 MRIs) | <0.0001 |
| Systolic blood pressure (mmHg) | 105.6 ± 6.3 (72 MRIs) | 99.9 6 7.4 | 0.004 |
| Diastolic blood pressure (mmHg) | 59.2 ± 7.1 (72 MRIs) | 59.6 6 6.9 | 0.873 |
| Mean plood pressure (mmHg) | 77.2 ± 6.5 (72 MRIs) | 74.9 6 7.2 | 0.255 |
| Heart rate (bpm) | 84.1 ± 8.4 (29/30 pts; 71 MRIs) | 85.5 6 11.3 (52 MRIs) | 0.784 |
| R2* heart (Hz) | 30.2 ± 4.0 | 32.9 6 3.9 | 0.002 |
The LV parameters for males and females with SCD are shown in Fig. 1A. Compared with the females, the males exhibited significantly higher LV EDVI (11%) and LV SVI (16%), but comparable LV end-systolic volume index (ESVI). Males showed higher LV EF than females and 14% higher LV mass index (MI). The cardiac index was comparable between the sexes (4.8 ± 0.6 l/min/m2 vs. 4.6 ± 0.6 l/min/m2; P = 0.094).
Figure 1.
Comparison of (A) LV and (B) RV parameters between males and females with SCD and comparison of (C) LV and (D) RV parameters between males and females with TM.
The RV parameters for males and females with SCD are shown in Fig. 1B. All RV indexed volumes were significantly larger (~22%) in males than in females but the RV EF was not different between the sexes.
In both sexes, there was no correlation between CMR findings and pretransfusion hemoglobin levels.
With the exception of the ESVI for females, biventricular indexed volumes increased significantly with age in both males and females. The association was weaker in females. The scatter plot suggested a clear separation between the sexes after 9–10 years of age (Fig. 2A), suggesting that the phenomenon may be associated with sex steroids. The LV MI increased significantly with age for males but not for females. The biventricular EFs and the cardiac index did not show significant changes with age in either males or females.
Figure 2.
LV EDVI versus age for (A) patients with SCD and (B) patients with TM. Data were fitted with a spline (lambda of 100) separately for males (gray) and females (black). LV EDVI values are comparable and rise at a similar rate in males and females until 10 years of age. LV EDVI values continue to rise until age 18 in males, while LV EDVI values remain unchanged between 10 and 18 years of age in females.
Sex differences in patients with TM
Demographics of patients with TM are also summarized in Table I. Boys and girls with TM were comparable for age (P = 0.287), hemoglobin values (P = 0.647), and serum ferritin levels (P = 0.757). No difference was present in systolic (P = 0.486) and mean (P = 0.725) blood pressure, while the diastolic blood pressure was significantly lower in males than in females (P = 0.004). Heart rate and cardiac R2* were comparable in males and females (P = 0.659 and P = 0.110, respectively).
The LV parameters for males and females with TM are shown in Fig. 1C; sex differences were nearly identical to those observed in patients with SCD. Males demonstrated significantly higher LV EDVI (9%) and LV SVI (10%), but comparable LV ESVI. The LV EF was not significantly different between the sexes. The LV MI was 7% higher in males. Males showed 7% higher CI (4.5 ± 0.5 l/min/m2 vs. 4.2 ± 0.5 l/min/m2; P = 0.013).
The RV parameters for males and females with SCD are shown in Fig. 1D. All RV volume indexes were significantly larger (>14%) in males than in females but the RV EF was not different.
In both sexes there was no correlation between CMR findings and pretransfusion hemoglobin levels.
With the exception of the LV ESVI and the RV SVI for females, biventricular volumes increased significantly with age for both the sexes. The correlation was always lower for the females. The scatter plot suggested a more pronounced difference between the sexes in the rates of change with age after about 9 years (Fig. 2B). The LVMI increased significantly with age for males but not for females. The biventricular EFs and the cardiac index did not show significant changes with age in either males or females.
Demographic differences between patients with SCD and TM
Demographics of the study populations are summarized in Table I, separated by sex. Patients with SCD and TM were comparable for age and pretransfusion hemoglobin levels, but the serum ferritin levels were significantly higher in patients with SCD (more than 1800 ng/ml for males and more than 900 ng/ml for females). Patients with SCD had significantly lower R2* values even though patients having overtly pathologic R2* were excluded from the study. Among the males, the median of R2* values and the upper quartile were, respectively, 28.7 Hz and 31.0 Hz for the SCD group, whereas 30.3 Hz and 35.0 Hz for the TM group. Among the females, the median of R2* values and the upper quartile were, respectively, 28.4 Hz and 33.3 Hz for the SCD group, whereas 32.4 Hz and 37.6 Hz for the TM group. Systolic blood pressure was significantly higher in SCD than in patients with TM, while the diastolic and the mean blood pressures were comparable. HR was significantly lower in SCD than in TM, but only in male subjects.
For both the sexes, three variables emerged as possible covariates: ferritin levels, systolic blood pressure, and cardiac R2* values.
LV parameters: SCD versus TM
Table II shows the comparison of LV parameters between SCD and TM with the differentiation by sex. All LV indexed volumes and the LV mass index were significantly higher in SCD than TM patients. For the males the LV EF was comparable between SCD and TM groups, while females with SCD had significantly lower LV EF than females with TM. CI was significantly increased in the SCD versus the TM group in both sexes.
TABLE II.
Left (A) and Right (B) Ventricular Parameters for SCD and TM Pediatric Patient Without Cardiac Iron Overload, with Differentiation for Sex
| SCD | TM | P | P adjusted for syst. BP | P adjusted for cardiac R2* | P adjusted for both covariates | |
|---|---|---|---|---|---|---|
| (A) Left ventricular parameters | ||||||
| Males | ||||||
| LV EDVI (ml/m2) | 104.7 ± 15.2 [99.5–109.9] | 88.7 ± 11.0 [84.5–92.8] | <0.0001 | <0.0001 | <0.0001 | <0.0001 |
| LV ESVI (ml/m2) | 38.1 ± 7.4 [35.6–40.7] | 31.6 ± 5.5 [29.5–33.6] | <0.0001 | <0.0001 | 0.0003 | <0.0001 |
| LV SVI (ml/m2) | 66.6 ± 9.5 [63.3–6987] | 57.1 ± 6.5 [54.6–59.5] | <0.0001 | <0.0001 | <0.0001 | <0.0001 |
| LV EF (%) | 63.8 ± 3.5 [62.6–65.0] | 64.7 ± 2.8 [63.6–65.7] | 0.286 | |||
| LV MI (g/m2) | 83.5 ± 11.4 [79.1–87.9] | 66.9 ± 7.0 [64.0–69.8] | <0.0001 | <0.0001 | <0.0001 | <0.0001 |
| CI (l/min/m2) | 4.8 ± 0.6 [4.6–5.1] | 4.5 ± 0.5 [4.4–4.7] | 0.017 | |||
| Females | ||||||
| LV EDVI (ml/m2) | 93.5 ± 9.6 [90.0–97.0] | 81.4 ± 6.7 [78.4–84.4] | <0.0001 | <0.0001 | <0.0001 | 0.0001 |
| LV ESVI (ml/m2) | 36.0 ± 5.3 [34.0–37.9] | 29.4 ± 3.8 [27.6–31.1] | <0.0001 | <0.0001 | <0.0001 | <0.0001 |
| LV SVI (ml/m2) | 57.5 ± 6.3 [55.2–59.8] | 52.0 ± 4.2 [50.1–53.4] | 0.0006 | 0.009 | ||
| LV EF (%) | 61.6 ± 3.5 [60.4–62.9] | 64.1 ± 2.9 [62.8–65.4] | 0.008 | |||
| LV MI (g/m2) | 72.9 ± 8.9 [69.6–79.3] | 62.8 ± 7.5 [60.4–66.1] | 0.0001 | 0.0002 | ||
| CI (l/min/m2) | 4.6 ± 0.6 [4.4–4.8] | 4.2 ± 0.5 [3.9–4.4] | 0.006 | |||
| (B) Right ventricular parameters | ||||||
| Males | ||||||
| RV EDVI (ml/m2) | 103.3 ± 18.8 [96.8–109.8] | 87.3 ± 11.5 [82.9–91.6] | <0.0001 | <0.0001 | 0.0004 | 0.0002 |
| RV ESVI (ml/m2) | 37.9 ± 9.6 [34.7–41.2] | 32.4 ± 5.7 [30.2–34.5] | 0.0005 | 0.006 | 0.050 | 0.037 |
| RV SVI (ml/m2) | 65.4 ± 10.8 [61.6–69.1] | 54.9 ± 6.9 [52.3–57.5] | <0.0001 | 0.016 | <0.0001 | <0.0001 |
| RV EF (%) | 63.8 ± 4.7 [62.2–65.4] | 63.2 ± 3.4 [61.9–64.5] | 0.620 | |||
| Females | ||||||
| RV EDVI (ml/m2) | 84.7 ± 10.3 [80.9–88.5] | 76.5 ± 9.1 [72.4–80.6] | 0.005 | 0.026 | 0.009 | 0.051 |
| RV ESVI (ml/m2) | 31.1 ± 5.8 [28.9–33.2] | 27.7 ± 5.3 [25.3–30.1] | 0.037 | 0.108 | ||
| RV SVI (ml/m2) | 53.6 ± 6.8 [51.1–56.1] | 48.8 ± 5.1 [46.6–51.1] | 0.006 | 0.047 | ||
| RV EF (%) | 63.6 ± 4.6 [61.9–65.2] | 64.4 ± 3.7 [62.7–66.1] | 0.477 |
LV, left ventricular; RV, right ventricular; EDVI, end-diastolic volume index; ESVI, end-systolic volume index; SVI, stroke volume index; EF, ejection fraction; MI, mass index; CI, cardiac index; BP, blood pressure.
All variables are expressed as mean ± SD with 95% confidence intervals in square brackets.
On univariate analysis, serum ferritin levels were not significantly associated to LV parameters so they were not used as covariates. The systolic blood pressure was significantly associated with all LV volumes and with the LV mass for both sexes without interaction with respect to disease state. As a result, SBP was used as covariate in the ANCOVA, but the adjustment for the systolic blood pressure did not change the systematic volumetric differences between the diseases. SBP was associated with the CI in males but it showed an interaction with the disease (P = 0.044), so ANCOVA correction was not performed.
For males, cardiac R2* values were significantly associated with all LV volume indexes and with the LV mass index. Disease-specific differences remained after ANCOVA correction. For females, cardiac R2* values were only associated with the LV end-diastolic and end-systolic volume indexes; ANCOVA correction did not remove the disease-specific differences in these variables. Simultaneous adjustment for both SBP and cardiac R2* did not qualitatively change the overall results.
RV parameters: SCD versus TM
Table II shows the comparison of RV parameters between SCD and TM, by sex; overall findings are similar to those from the LV RV volumes were significantly higher in SCD than in patients with TM, while the RV EF was comparable between them.
Right ventricular dimensions and function had similar association with serum ferritin, SBP, and cardiac R2* as for the left ventricle. All RV volumes remained significantly higher in SCD also after ANCOVA adjustments, except RV ESVI. Simultaneous adjustment for both SBP and cardiac R2* did not qualitatively change the overall results.
Discussion
In this study, we characterized the LV and the RV of regularly transfused, pediatric patients with SCD and TM having no evidence of cardiac iron overload. These data represent cardiovascular compensation for chronic anemia and a baseline from which to interpret toxic changes produced by cardiovascular iron deposition.
As expected, both patient groups demonstrated dilated ventricles, relative to published population norms [24]. The anemia is accompanied by elevation of blood volume (increased preload) and by decrease in systemic vascular resistance (decreased afterload) [3]. Both conditions enhance the ventricular pump performance and the anatomical-functional expression of this hemodynamic state is the enlargement of cardiac cavities [25,26]. In patients with SCD, the mean LV and RV EFs were slightly larger than those reported for healthy pediatric patients [24], with a greater difference for the RV. In patients with TM, the LV EF for males was quite close to the mean value detected in healthy subjects while the females exhibited a lower LV EF than the healthy females (61.7 vs. 63.6%). The RV EFs of both sexes were slightly higher than those reported for healthy pediatric patients.
Sex differences were large in both TM and SCD, becoming evident as patients entered puberty. Sex differences in biventricular parameters have been demonstrated in normal adults [27–30] and in patients with TM [13,14]. However, for normal pediatric populations there is not full consensus: sex differences have been found in some studies [24,31] but not in others [32,33], indicating the need for targeted studies in both normal subjects and patients with anemia. Our study clearly demonstrated that males have larger left and right ventricular dimensions in both pediatric patients with SCD and with TM. Sex-differences for ventricular volumes were more marked in SCD than in patients with TM, particularly for the right ventricle. The percentage difference for the RV was larger than that one seen in normal subjects (from 8 to 14%). One could speculate that males may have larger, preclinical pulmonary vascular changes but cardiac catheterization would be required to prove this hypothesis. Cardiac index was increased in males with TM but not with SCD because females with SCD were more tachycardic.
Although impaired oxygen carrying capacity drives increased cardiac output in patients with anemia, there was no significant inverse correlation between hemoglobin levels and LV and RV parameters. The most likely explanation is that pretransfusion hemoglobin level is tightly controlled in our chronically transfused populations, making it statistically challenging to detect a correlation against the background of high inter-subject variability. Furthermore, our MR observations were sampled randomly across the transfusion cycle. While LVEF does not change prior to and following transfusion [34,35], and LV mass is unlikely dependent on timing within the transfusion cycle, absolute ventricular volumes and cardiac index decrease immediately following blood transfusion [35]. Similar findings were observed in adult patients with SCD by Johnson et al. [36], and in adult patients with TM by Westwood and Carpenter [13,14].
During puberty, ventricular dimensions, stroke volume and mass increased faster than changes in body surface area; this effect was much stronger in males. This observation most likely represents the anabolic effects of steroid hormones causing disproportionate increases in lean body mass [37] and this is in agreement with studies involving healthy subjects [24,33]. While patients with TM often have iron-mediated hypogonadism, elimination of patients having significant cardiac iron and the availability of iron chelation since birth in this cohort eliminated patients with known hypogonadism from this cohort.
Although sex differences in TM and SCD mirrored changes observed in normal subjects and were not unexpected, the larger ventricular dimensions, mass, and cardiac index observed in SCD remain unexplained and could not be explained by anemia severity. Westwood et al. [38] noted a similar disparity in adult patients, but hemoglobin levels in that cohort were not comparable, in contrast to our study. We previously observed increased cardiac mass and index in SCD but the TM cohort had significant cardiac iron, confounding the comparison [17]. While hemoglobin levels were matched near the time of scan, we do not have longitudinal hemoglogin data. Patients with SCD usually start transfusions later in life than TM subjects and thus have greater lifetime exposure to more severe anemia and cardiovascular remodeling. Another possibility is that vascular iron deposition is greater in TM despite a cardiac R2* < 50 Hz (undetectable cardiac iron); two lines of evidence support this hypothesis. First, cardiac benefits of iron chelation has been previously demonstrated in patient with TM cohorts without significant myocardial iron overload [39]. Second, cardiac dimensions were significantly correlated with cardiac R2* in our study, even with R2* in the “normal” range. However, disease-specific differences remained after ANCOVA correction, suggesting that this could not be the sole cause of the disparity. Patients with SCD showed higher SBP, which would tend to increase ventricular dimensions and mass [40]; however, ANCOVA correction for SBP failed to eliminate disease-specific differences.
Unfortunately, we were not able to determine if differences in oxygen saturation could contribute to the disparity. Decreased asleep and awake oxygen saturations are associated with LV hypertrophy in children with SCD [36].
Another important limitation of this study is the limited availability of clinical data on patients who were referred for CMR from outside institutions. Furthermore, it is impossible to simultaneously match patients with SCD and TM for age, duration, and intensity of blood transfusions, given the different clinical indications for chronic transfusion therapy in the two populations.
In summary, pediatric patients with SCD and TM exhibit significant sex differences in biventricular volumes, LV mass, and cardiac index, beginning at the onset of puberty; these differences parallel changes described in normal subjects. Compared to patients with TM, patients with SCD showed significantly increased biventricular dilation and LV hypertrophy. This difference could not be explained by different hemoglobin levels, cardiac iron overload, and systolic blood pressure. These data represent important baseline findings that place changes introduced by iron overload as well as systemic and pulmonary vasculopathy in proper context.
Supplementary Material
Acknowledgments
Contract grant sponsor: NHLBI; contract grand number: 1 RO1 HL075592-01A1; contract grant sponsor: General Clinical Research Center at the Children’s Hospital Los Angeles; contract grand number: RR000043-43; contract grant sponsor: Center for Disease Control (Thalassemia Center Grant); contract grand number: U27/CCU922106; contract grant sponsors: Novartis Pharma, Department of Pediatrics.
Footnotes
Conflict of interest: Nothing to report.
Additional Supporting Information may be found in the online version of this article.
References
- 1.Bunn HF Pathogenesis and treatment of sickle cell disease. N Engl J Med 1997;337:762–769. [DOI] [PubMed] [Google Scholar]
- 2.Weatherall DJ. The thalassaemias. BMJ 1997;314:1675–1678. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Varat MA, Adolph RJ, Fowler NO. Cardiovascular effects of anemia. Am Heart J 1972;83:415–426. [DOI] [PubMed] [Google Scholar]
- 4.Graettinger JS, Parsons RL, Campbell JA. A correlation of clinical and hemodynamic studies in patients with mild and severe anemia with and without congestive failure. Ann Intern Med 1963;58:617–626. [DOI] [PubMed] [Google Scholar]
- 5.Ware R Principles and indications of chronic transfusion therapy for children with sickle cell disease. Clin Adv Hematol Oncol 2007;5:686–688. [PubMed] [Google Scholar]
- 6.Gordeuk VR, Bacon BR, Brittenham GM. Iron overload: Causes and consequences. Annu Rev Nutr 1987;7:485–508. [DOI] [PubMed] [Google Scholar]
- 7.Li W, Coates T, Wood JC. Atrial dysfunction as a marker of iron cardiotoxicity in thalassemia major. Haematologica 2008;93:311–312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cheung YF, Chan GC, Ha SY Arterial stiffness and endothelial function in patients with beta-thalassemia major. Circulation 2002;106:2561–2566. [DOI] [PubMed] [Google Scholar]
- 9.Cheung YF, Chan GC, Ha SY Effect of deferasirox (ICL670) on arterial function in patients with beta-thalassaemia major. Br J Haematol 2008;141:728–733. [DOI] [PubMed] [Google Scholar]
- 10.Wood JC, Enriquez C, Ghugre N, et al. Physiology and pathophysiology of iron cardiomyopathy in thalassemia. Ann N Y Acad Sci 2005;1054:386–395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Anderson LJ, Holden S, Davis B, et al. Cardiovascular T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload. Eur Heart J 2001;22:2171–2179. [DOI] [PubMed] [Google Scholar]
- 12.Bellenger NG, Davies LC, Francis JM, et al. Reduction in sample size for studies of remodeling in heart failure by the use of cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2000;2:271–278. [DOI] [PubMed] [Google Scholar]
- 13.Westwood MA, Anderson LJ, Maceira AM, et al. Normalized left ventricular volumes and function in thalassemia major patients with normal myocardial iron. J Magn Reson Imaging 2007;25:1147–1151. [DOI] [PubMed] [Google Scholar]
- 14.Carpenter JP, Alpendurada F, Deac M, et al. Right ventricular volumes and function in thalassemia major patients in the absence of myocardial iron overload. J Cardiovasc Magn Reson 2010;12:24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wood JC, Enriquez C, Ghugre N, et al. MRI R2 and R2* mapping accurately estimates hepatic iron concentration in transfusion-dependent thalassemia and sickle cell disease patients. Blood 2005;106:1460–1465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Pepe A, Positano V, Santarelli F, et al. Multislice multiecho T2* cardiovascular magnetic resonance for detection of the heterogeneous distribution of myocardial iron overload. J Magn Reson Imaging 2006;23:662–668. [DOI] [PubMed] [Google Scholar]
- 17.Wood JC, Tyszka JM, Carson S, et al. Myocardial iron loading in transfusion-dependent thalassemia and sickle cell disease. Blood 2004;103:1934–1936. [DOI] [PubMed] [Google Scholar]
- 18.Kaushik N, Eckrich MJ, Parra D, et al. Chronically transfused pediatric sickle cell patients are protected from cardiac iron overload. Pediatr Hematol Oncol 2012;29:254–260. [DOI] [PubMed] [Google Scholar]
- 19.Ghugre NR, Enriquez CM, Coates TD, et al. Improved R2* measurements in myocardial iron overload. J Magn Reson Imaging 2006;23:9–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ghugre NR, Enriquez CM, Gonzalez I, et al. MRI detects myocardial iron in the human heart. Magn Reson Med 2006;56:681–686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Carpenter JP, He T, Kirk P, et al. On T2* magnetic resonance and cardiac iron. Circulation 2011;123:1519–1528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wang Y, Moss J, Thisted R. Predictors of body surface area. J Clin Anesth 1992;4:4–10. [DOI] [PubMed] [Google Scholar]
- 23.Quinn GP, Keough MJ. Chapter 12: Analyses of Covariance In: Experimental Design and Data Analysis for Biologists. United Kingdom: Cambridge University Press; 2002. p. 339–358. [Google Scholar]
- 24.Sarikouch S, Peters B, Gutberlet M, et al. Sex-specific pediatric percentiles for ventricular size and mass as reference values for cardiac MRI: Assessment by steady-state free-precession and phase-contrast MRI flow. Circ Cardiovasc Imaging 2010;3:65–76. [DOI] [PubMed] [Google Scholar]
- 25.Lindsay J Jr, Meshel JC, Patterson RH. The cardiovascular manifestations of sickle cell disease. Arch Intern Med 1974;133:643–651. [PubMed] [Google Scholar]
- 26.Kremastinos DT, Tsiapras DP, Tsetsos GA, et al. Left ventricular diastolic Doppler characteristics in beta-thalassemia major. Circulation 1993;88:1127–1135. [DOI] [PubMed] [Google Scholar]
- 27.Bella JN, Palmieri V, Kitzman DW, et al. Gender difference in diastolic function in hypertension (the HyperGEN study). Am J Cardiol 2002;89:1052–1056. [DOI] [PubMed] [Google Scholar]
- 28.Salton CJ, Chuang ML, O’Donnell CJ, et al. Gender differences and normal left ventricular anatomy in an adult population free of hypertension. A cardiovascular magnetic resonance study of the Framingham Heart Study Offspring cohort. J Am Coll Cardiol 2002;39:1055–1060. [DOI] [PubMed] [Google Scholar]
- 29.Maceira AM, Prasad SK, Khan M, et al. Normalized left ventricular systolic and diastolic function by steady state free precession cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2006;8:417–426. [DOI] [PubMed] [Google Scholar]
- 30.Maceira AM, Prasad SK, Khan M, et al. Reference right ventricular systolic and diastolic function normalized to age, gender and body surface area from steady-state free precession cardiovascular magnetic resonance. Eur Heart J 2006;27:2879–2888. [DOI] [PubMed] [Google Scholar]
- 31.Buechel EV, Kaiser T, Jackson C, et al. Normal right- and left ventricular volumes and myocardial mass in children measured by steady state free precession cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2009;11:19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lytrivi ID, Bhatla P, Ko HH, et al. Normal values for left ventricular volume in infants and young children by the echocardiographic subxiphoid five-sixth area by length (bullet) method. J Am Soc Echocardiogr 2011;24:214–218. [DOI] [PubMed] [Google Scholar]
- 33.Cain PA, Ahl R, Hedstrom E, et al. Age and gender specific normal values of left ventricular mass, volume and function for gradient echo magnetic resonance imaging: A cross sectional study. BMC Med Imaging 2009;9:2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Davis BA, O’Sullivan C, Jarritt PH, et al. Value of sequential monitoring of left ventricular ejection fraction in the management of thalassemia major. Blood 2004;104:263–269. [DOI] [PubMed] [Google Scholar]
- 35.Detterich J, Noetzli L, Dorey F, et al. Electrocardiographic consequences of cardiac iron overload in thalassemia major. Am J Hematol 2012;87:139–144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Johnson MC, Kirkham FJ, Redline S, et al. Left ventricular hypertrophy and diastolic dysfunction in children with sickle cell disease are related to asleep and waking oxygen desaturation. Blood 2010;116:16–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.de Simone G, Devereux RB, Daniels SR, et al. Gender differences in left ventricular growth. Hypertension 1995;26:979–983. [DOI] [PubMed] [Google Scholar]
- 38.Westwood MA, Shah F, Anderson LJ, et al. Myocardial tissue characterization and the role of chronic anemia in sickle cell cardiomyopathy. J Magn Reson Imaging 2007;26:564–568. [DOI] [PubMed] [Google Scholar]
- 39.Pennell DJ, Porter JB, Cappellini MD, et al. Efficacy of deferasirox in reducing and preventing cardiac iron overload in beta-thalassemia. Blood 2010;115: 2364–2371. [DOI] [PubMed] [Google Scholar]
- 40.Haji SA, Ulusoy RE, Patel DA, et al. Predictors of left ventricular dilatation in young adults (from the Bogalusa Heart Study). Am J Cardiol 2006;98:1234–1237. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.