Study Identifies Cardiac Phenotypes in Patients with Sickle Cell Disease

Mortality rates differed significantly across the 3 clusters.

New findings grouped patients with homozygous sickle cell disease into specific cardiac phenotypes. These phenotypes were linked to different clinical characteristics and outcomes.

A team led by Thomas d’Humières, MD, of Henri Mondor Hospital, France, performed a cluster analysis of sickle cell patients in the French Etendard Cohort. They sought to understand the onset and progression of cardiopulmonary involvement in this patient population through comprehensive echocardiographic analysis.

“Using unsupervised analysis based upon the integration of cardiac remodeling, systolic and diastolic function together with [tricuspid regurgitation velocity] TRV measurement, we aimed to improve cardiac phenotyping and provide new insights on heart involvement in SCD,” d’Humières and colleagues wrote.

The Study

The team evaluated a total of 379 patients, all of whom were ≥18 years old, had sickle cell disease, and were in steady-state condition. As such, no patient had blood transfusions or acute episodes 3 months prior to study inclusion.

They excluded anyone with end-stage renal disease, chronic restrictive lung disease, or severe liver disease with an international normalized ratio >1.7

All participants were interviewed and underwent a physical examination as well as transthoracic echocardiography.

Further, patients with TRV ≥2.5m/s underwent systematic right heart catheterization (RHC) and were thus classified using PH classifications of 2015 and 2019.

Cluster analysis was conducted using a K-mean clustering algorithm.

“The goal was to identify echocardiographic patterns that would define subsets of patients using the following parameters: LV ejection fraction (LVEF), cardiac output (CO), LV mass indexed to body surface area (LVMind), LV end-diastolic volume indexed to body surface area (LVEDVind), E′ lateral wave, E/E′ mean ratio, LA volume indexed to body surface (LAVind), TRV and Tricuspid Annular Plane Systolic Excursion (TAPSE),” the investigators wrote.

Phenotype Clusters

Overall, the echocardiographic variables identified 3 clusters:

In Cluster 1 (n = 123), patients had the lowest CO and LV volumes. Of all the clusters, they demonstrated the lowest E’. However, E/E’ ratio values, LV mass, and LA volume were slightly increased.

Although values were in normal range, TRV values were similar to cluster 2 and higher than Cluster 1.

This cluster included mostly female patients who had the lowest performance on the 6-minute walking test, had moderate hemolytic profile, and the lowest white blood cell count.

Cluster 2 (n = 102) had the highest CO, lowest LVEF, and the highest LVM volume and LAV when compared with the other 2 clusters.

Compared to Cluster 3, patients had lower E’ values and higher E/E’ ratio and TRV values. These values were similar to patients in Cluster 1.

Additionally, patients generally had higher systolic blood press pressure values, albeit in normal range, and had the highest prevalence of history of systemic hypertension, leg ulcers, and renal dysfunction. These patients had the most severe hemolytic profile.

And finally, Cluster 3 (n = 154) demonstrated the highest E’ values and the lowest E/E’ ratio and TRV values of all the clusters. The LV and LA values were significantly higher than Cluster 1 but lower than Cluster 2.

Patients tended to be younger with lower systolic blood pressure, the least severe hemolytic profile, and the best performance on the 6-minute walking test.


D’Humières and colleagues also assessed outcomes, noting that the mortality rate was higher in Cluster 1 (11%) and Cluster 2 (19%) compared with Cluster 3 (5%).

Survival analysis revealed that there were significant differences across all 3 clusters in terms of all-cause mortality (P = .003).

The team noted that death from left or global heart failure affected only patients from Cluster 2 (n = 5). Death from right heart failure complicating PH occurred in 3 patients from Cluster 1 and 1 from Cluster 3.

Out-of-hospital sudden swath occurred in patients in Clusters 1 and 2 (n = 7).

“These findings underline the weight of comprehensive echocardiographic evaluation of SCD patients, with an integrative approach based on simultaneous evaluation of TRV along with left cavities remodeling and diastolic parameters,” the investigators wrote.

“The knowledge gained in this study should alert on the current application of 2019 guidelines, which are not suited to the optimal assessment of cardiovascular risk in SCD patients,” they concluded.

The study, “Cardiovascular Phenotypes Predict Clinical Outcomes in Sickle Cell Disease: An echocardiography-based cluster analysis,” was published online in American Journal of Hematology.