Diffuse myocardial fibrosis occurs in young patients with sickle cell anemia despite early disease-modifying therapy

TO THE EDITOR:

We read with great interest the recent report from Niss et al.¹ Our group has recently reviewed data from a similar population at our institution, which does not support the conclusion of Niss et al that diffuse myocardial fibrosis is prevented by early initiation of disease-modifying therapy (DMT).

Cardiovascular complications are the leading cause of early mortality in patients with sickle cell anemia (SCA).² Yet, the pathobiology of sickle cell cardiomyopathy is incompletely understood and likely multifactorial.³ Chronic anemia leads to global cardiac enlargement and myocardial hypertrophy. Global cardiac chamber dilation is a common feature in patients with SCA, and enlarged left atrial (LA) volume is a marker of diastolic dysfunction in SCA.⁴ Repetitive microvascular ischemic insults and associated reperfusion injury may contribute to diffuse myocardial fibrosis and myocardial remodeling in individuals with SCA.^5-7 

Cardiovascular magnetic resonance (CMR) imaging is the only noninvasive method capable of accurately and reproducibly assessing heart size and function as well as characterizing the myocardium. Emerging data suggest that patients with SCA develop diffuse myocardial fibrosis,^1,5-7 detected by quantifying the myocardial extracellular volume (ECV). Increased ECV fraction correlates strongly with histologically quantified myocardial fibrosis.⁸ Our goal was to estimate the prevalence of diffuse myocardial fibrosis in a cohort of young individuals with SCA.

At our institution, all patients with SCA undergo cardiac surveillance starting at 16 years of age, which comprises a clinical assessment by a cardiologist and an echocardiogram. The presence of LA dilation (LA volume > 34 mL/m²), left ventricular dilation, or prominent trabeculations seen by echocardiogram led to patients being referred for CMR (detailed echocardiographic data in supplemental Table 1, on the Blood website). We performed a retrospective chart review of pediatric and young adult patients with SCA who underwent CMR at our institution between 2020 and 2021. Demographic and clinical information was obtained from reviewing the electronic medical record. A waiver of the requirement for informed consent was granted by the institutional review board for use of deidentified data.

CMR was performed on a 1.5T scanner (Avanto Fit; Siemens Medical Solutions, Inc, Erlangen, Germany). The protocol included ventricular long- and short-axis planes for chamber sizes, ejection fraction, and myocardial mass. ECV was measured from T1 maps acquired with a modified Look-Locker inversion recovery sequence in short-axis pre- and postcontrast.⁹ T2 and T2∗ maps were obtained before contrast administration. Regions of interest for analysis were drawn in the interventricular septum at the midmyocardial level (supplemental Figure 1). All planimetric and region-of-interest analyses were performed with cvi42 (Circle Imaging; Alberta, Canada) by a single technologist (C.G.) and reviewed for accuracy by a cardiac imaging specialist (C.E.M., A.M., or J.N.J.). Myocardial MRI-T2∗ values were obtained from clinical reports and were measured as previously described.¹⁰ Hematocrit was collected within 24 hours of the patients undergoing a cardiac MRI.

Thirty-one patients with SCA (SS genotype) underwent CMR during the study period (Table 1). The median age at the time of CMR was 16.7 years. At the time of CMR, 11 (36%) patients were receiving hydroxyurea for a median treatment duration of 3.3 years (interquartile range [IQR] 0.1-9.3 years), and 14 (45%) patients were receiving chronic monthly transfusions (CTXN) for a median treatment duration of 1.1 years (IQR 0.0-3.5 years). Some patients switched from hydroxyurea to CTXN or vice versa, therefore the collective median duration of any DMT was 9.6 years (IQR 2.2-11.8 years). In the 2 years preceding the CMR, the frequency of SCA acute complications was as expected for a DMT-exposed young population: 5 (16.1%) patients had 1 to 2 episodes of acute chest syndrome and 10 (32.3%) patients had 1 or more episodes of vaso-occlusive pain crises. The remaining participants had no vaso-occlusive events in the 2 years before undergoing CMR.

Many patients had left ventricle (LV, 25 [81%]) and right ventricle (20 [65%]) enlargement. Thirteen (42%) patients also had LA enlargement. Most patients had elevated LV cardiac index (24 [77%]). Both LA volume and LV cardiac index were elevated despite DMT, whether hydroxyurea, CTXN, or a combination of both, though this was not statistically significant (supplemental Figure 2).

All patients in our study had increased ECV with a median ECV of 32% (IQR 29%-34%, reference normal 20.8% ± 2.4% in children⁹ and 25.9% ± 0.4% in adults¹⁴). This is consistent with prior studies in a similar population.^5,7 Duration of exposure to DMT was not associated with ECV (slope of ECV percent vs years on DMT: −0.07, P = .5). All patients had normal myocardial MRI-T2∗ values indicating the lack of significant iron overload of the myocardium.

Our study highlights that diffuse myocardial fibrosis is nearly universal in children and young adults with SCA, even those patients with normal cardiac size and function, confirming prior reports.^5,15 In contrast to Niss et al, we found no difference in the ECV in patients who began DMT early (less than 6 years of age) vs late in life (at least 6 years or after) despite the 2 cohorts being similar in demographics and treatment received. Although mean ECV was comparable in the early therapy groups in our study and the one described by Niss et al (32% vs 30%), ECV in the no early therapy group was lower in our study compared with Niss et al (31% vs 44%). Participants in the Niss et al study were older (median age in the early therapy and no early therapy groups was 18 and 19 years respectively, compared with 17 and 16.8 years, respectively, in our study). In the cohort described by Niss et al, the median age of initiation of any DMT in the early therapy group was 2.9 years. In the no early therapy group, hydroxyurea was prescribed at a mean age of 11.3 years (patients receiving CTXN were excluded). In comparison, in our study, the median age of initiation of any DMT in the early therapy group was 2.6 years and in the no early therapy group was 9 years. The median duration of treatment with DMT in the Niss et al study was 13.7 years in the group that received early therapy. In our study, the median duration of treatment with DMT in the group that received early therapy was 10.5 years, which is slightly shorter but still comparable to the Niss et al study. It is important to note, though, that patients who subsequently received either hydroxyurea or CTXN were included in our no early therapy group, in contrast to exclusion of individuals who ever received CTXN in the no early therapy group in the Niss et al study. Even after exclusion of the patients who had received CTXN from the no early therapy group, our results remained the same—we found no significant differences in the ECV of those patients who received early therapy vs those who did not (supplemental Figure 3). We further explored whether starting DMT at an even earlier age would have a further protective effect and dichotomized our cohort into those who started DMT before or after 3 years of age and found no difference in the ECV of the 2 groups (Figure 1B, supplemental Figure 4). Lastly, even though 40% of patients in the early therapy group were receiving beta-blockers for management of diastolic dysfunction on echocardiogram and higher resting heart rates on 24-hour Holter monitors, the echocardiographic findings between the 2 groups were not significantly different. Our cohort of patients are younger with similar cardiac index and LV end-diastolic volumes but smaller LA volumes compared with the Niss et al cohort.¹⁵ Only 8 (25.8%) patients had LA volume >50 mL/m² in our cohort, potentially reflecting less diastolic dysfunction compared with the Niss et al cohort.¹⁵ 

We cannot conclusively explain why our results differed from Niss et al. In general, increased ECV may be partially related to increased LV mass and may explain a portion of the elevated ECV seen in patients with SCA. The lower ECV observed in the no early therapy group compared with the Niss et al cohort may be related to smaller LA volumes, reflecting potentially less diastolic dysfunction. Although our no early therapy group included CTXN, excluding patients who had ever received CTXN from the analysis did not significantly change the results. Nearly all patients had abnormal ECV, suggesting that even CTXN is not completely protective. Our findings suggest that damage to the myocardium occurs before symptoms or other clinical markers of cardiac dysfunction and that our current screening protocols may require revision to include earlier cardiac assessments. Although we cannot conclude that early DMT is protective against the development of myocardial fibrosis, the presence of nearly universal diffuse myocardial fibrosis will require further longitudinal evaluation. CMR provides additional clinically relevant information that may help guide treatment in patients with SCA.