In sickle cell disease (SCD), pulmonary hypertension (assessed by tricuspid regurgitant jet [TRJ] velocity ≥ 2.5 m/sec) is associated with increased mortality. The relationships among TRJ velocity and left ventricular (LV) and right ventricular (RV) systolic and diastolic function (i.e., relaxation and compliance) have not been well characterized in SCD.
A prospective study was conducted in 53 ambulatory adults with SCD (mean age, 34 years; range, 21–65 years) and 33 African American controls to define the relationship between LV and RV function and TRJ velocity using echocardiography.
Subjects with SCD had larger left and right atrial volumes and increased LV mass compared with controls. When patients with SCD were compared with controls, LV and RV relaxation (i.e., E′) were similar. Among subjects with SCD, pulmonary hypertension (TRJ ≥ 2.5 m/sec) was present in 40%. Higher TRJ velocity was correlated with larger left atrial volumes in patients with SCD. Additionally, some measures of LV (peak A, lateral and septal annular E/E′ ratio) and RV (tricuspid valve E/E′ ratio) compliance were correlated with TRJ velocity. No other measures of LV and RV systolic function or LV diastolic function (i.e., relaxation and compliance) were associated with TRJ velocity.
Ambulatory adults with SCD exhibited structural (i.e., LV and RV chamber enlargement) and functional (i.e., higher surrogate measures of LV and RV filling pressure) abnormalities compared with the control group. In subjects with SCD, few abnormalities of LV and RV structure and function were associated with TRJ velocity.
In adult patients with sickle cell disease (SCD), a tricuspid regurgitant jet (TRJ) velocity ≥ 2.5 m/sec is associated with increased mortality (risk ratio, 10.1). The reported prevalence of increased TRJ velocity (≥2.5 m/sec) is approximately 30% of adults with SCD. It is known that reduced nitric oxide bioavailability in SCD results in endothelial dysfunction and pulmonary vasoconstriction. Reduced nitric oxide availability due to increased arginase and scavenging of nitric oxide by cell-free hemoglobin released after hemolysis of sickle erythrocytes has been postulated as a mechanism for the development of pulmonary hypertension. In a subset of adults with SCD, chronic thromboembolic disease and interstitial lung disease may also contribute to pulmonary hypertension.
Abnormal left ventricular (LV) diastolic function is a well-established cause of pulmonary venous hypertension in the general population, particularly in subjects with cardiovascular disease. One study reported that abnormal LV diastolic function was present in adults with SCD, despite normal LV systolic function, and was an independent predictor of mortality. The hypotheses of the present study were that in adults with SCD, (1) indices of LV and right ventricular (RV) systolic function, relaxation, and compliance are abnormal compared with those in patients without SCD, and (2) abnormal indices of LV and RV systolic function, relaxation, and compliance are associated with TRJ velocity. We prospectively assessed LV and RV structure and function by echocardiography (two-dimensional [2D], pulsed-wave Doppler, and Doppler tissue imaging [DTI]) in adults with SCD and in normal controls.
Subjects aged > 18 years with SCD, confirmed by hemoglobin analysis, were recruited from the Adult SCD Clinic at Washington University. A cohort of African American controls without SCD was identified from an ongoing research study at Washington University. Exclusion criteria for the control group were (1) history of hypertension and/or use of antihypertensive medications, (2) type 2 diabetes and/or fasting plasma glucose > 125 mg/dL, (3) history of infection with the human immunodeficiency virus, (4) history of coronary artery disease, (5) echocardiographic abnormality (i.e., regurgitation or stenosis greater than mild, hypertrophic cardiomyopathy, pericardial disease, ventricular wall motion abnormalities, and depressed global systolic function), (6) body mass index > 35 kg/m 2 , and (7) serum creatinine > 2.5 mg/dL. Subjects with SCD with hypertension, those taking antihypertensive medications, those with type 2 diabetes, those with human immunodeficiency virus infection, and those with coronary artery disease were also excluded to maintain similar eligibility criteria between cases and controls. This study was approved by the institutional review board at Washington University School of Medicine in St. Louis, and informed consent was obtained from all participants.
Subjects with SCD underwent echocardiography during scheduled well visits to the outpatient clinic, when they reported no illness or increased pain and were not within 2 weeks of a hospitalization or emergency room visit for any reason. Comprehensive echocardiography, including 2D, M-mode, pulsed-wave Doppler, and DTI, was performed in all study participants using commercially available ultrasound equipment (Sequoia C512; Siemens Medical Solutions USA, Inc., Mountain View, CA). Two-dimensional echocardiographic measurements included LV end-diastolic and end-systolic volumes and LV ejection fraction calculated using the method of disks (modified Simpson’s method). Left atrial (LA) volume was measured at end-systole in the apical four-chamber view. Stroke volume was calculated by the LV outflow tract time-velocity integral determined by pulsed-wave Doppler multiplied by the LV outflow tract cross-sectional area. LV mass was measured by the M-mode-derived cubed method and indexed to height 2.7 . LV diastolic function was assessed using pulsed-wave Doppler–derived peak early diastolic mitral inflow (E-wave) velocity, late diastolic mitral inflow (A-wave) velocity, the ratio of E-wave velocity to A-wave velocity (E/A), E-wave deceleration time (DT), and isovolumic relaxation time (IVRT). DTI-derived myocardial early diastolic (E′) tissue velocities were obtained from the apical four-chamber view at the lateral and septal annulus; mitral E/E′ ratios were determined as a surrogate for LV filling pressures. LV septal and LV lateral tissue Doppler velocities were averaged from three cardiac cycles at end-expiration. Diastolic dysfunction was graded as (1) mild (E/A < 1.0 and/or DT > 240 msec and E/E′ ≤ 10), (2) moderate (E/A ≥ 1.0 and/or E/E′ > 10), or (3) severe (E/A > 95% for age or DT < 140 msec and E/E′ > 10).
Right atrial (RA) volume was measured at end-systole in the apical four-chamber view. RV areas at end-diastole and end-systole were determined in the apical four-chamber view. RV fractional area change was calculated as (RV end-diastolic area − RV end-systolic area)/RV end-diastolic area × 100. RV systolic and diastolic function was assessed by DTI-derived peak systolic (S′) and early diastolic (E′) velocities at the lateral tricuspid annulus in the apical four-chamber view at end-expiration. The peak tricuspid E-wave velocity was determined by pulsed-wave Doppler at the tricuspid leaflet tips; RV filling pressure was calculated as tricuspid valve (TV) E/E′.
The myocardial performance index was derived by pulsed-wave Doppler. The time intervals from valve closure to valve opening for the mitral and tricuspid values were used to determine total systolic duration for the left and right ventricles, respectively. The LV and RV ejection times were measured from the onset to the end of LV and pulmonary outflow, respectively. The myocardial performance index was calculated as (total systolic duration − ejection time)/ejection time for each ventricle.
The TRJ velocity was examined in the apical four-chamber (standard or modified) and parasternal views; the view yielding good-quality Doppler envelopes with the highest velocities was reported ( Figure 1 ). The average of three TRJ velocities was used to report the peak TRJ.
All measurements were performed in accordance with published guidelines or prior studies and represent the average of three consecutive cardiac cycles obtained by a single observer blinded to all clinical parameters.
Clinical and Laboratory Data Collection
Standardized data extraction forms were used to perform a review of individual medical records. Laboratory data were collected at scheduled well visits to an outpatient clinic. No laboratory studies were performed within 2 weeks of a subject with SCD reporting illness or increased pain or hospitalization for any reason.
Clinical characteristics were compared between adults with SCD and the control population using Student’s t tests. To compare echocardiographic parameters between subjects with SCD and controls, a general linear model was used. Differences in echocardiographic parameters (dependent variables) between cases and controls were adjusted for age, body mass index, gender, and gender–by–case/control interaction (independent variables). Echocardiographic parameters were normally distributed, with the exception of TRJ velocity, which was subsequently log-transformed for all analyses. Correlations of echocardiographic parameters and TRJ velocity among subjects with SCD were assessed using Pearson’s correlation. P values were adjusted for multiple testing using the false discovery rate method. The SAS Multtest procedure (SAS version 9.1.3; SAS Institute Inc., Cary, NC) was used for the adjustment.
A power analysis was performed to determine whether the available SCD subject and control sample sizes were sufficient to detect medium to large effect sizes between measures of LV diastolic function and TRJ velocity ( r ≥ 0.40). Although our sample size may not have detected small associations between measures of LV diastolic dysfunction and TRJ velocity, our analyses were powered to identify associations with effect sizes likely to be clinically meaningful. To detect a correlation coefficient ≥ 0.40 with β = 0.80 and α = 0.05, sample sizes of ≥50 cases and ≥30 controls are required.
Clinical Characteristics of the Study Population
The study population included 53 adults with SCD (mean age, 34 ± 11 years; range, 21–65 years) and 33 healthy African American controls (mean age, 39 ± 10 years; range, 18–53 years). The SCD cohort was composed of adults with HbSS (68%), HbSC (19%), HbSβ-thalassemia 0 (7%), and HbSβ-thalassemia + (6%). Compared with controls, subjects with SCD had higher systolic blood pressures, lower diastolic blood pressures, and higher heart rates. Subjects with SCD also had lower body weights ( Table 1 ). Among subjects with SCD, the mean hemoglobin was 9.0 g/dL (range, 5.8–13.3 g/dL). Compared with historical data from the Cooperative Study of Sickle Cell Disease, the current SCD group had higher rates of hospitalization for pain (1.2 vs 0.8 per year) and acute chest syndrome (0.2 vs 0.13 per year).
( n = 33)
|Patients with SCD |
( n = 53)
|Age (y)||39 ± 10||34 ± 11||.02|
|Height (cm)||167 ± 7||170 ± 12||.09|
|Weight (kg)||78 ± 16||70 ± 15||.03|
|BMI (kg/m 2 )||28 ± 5||25 ± 6||.003|
|Heart rate (beats/min)||66 ± 9||71 ± 11||.04|
|Systolic BP (mm Hg)||112 ± 11||119 ± 15||.02|
|Diastolic BP (mm Hg)||76 ± 6||71 ± 7||.004|
|Pulse pressure (mm Hg)||36 ± 7||47 ± 12||<.001|
Echocardiographically Derived Parameters of Cardiac Structure and Function
Right and Left Cardiac Structures
Measurements of cardiac structure, including LA volumes, RA volumes, LV mass, and end-systolic volume, were significantly higher in the SCD group compared with the control group ( Table 2 ). RV end-diastolic areas were larger in the SCD group compared with controls.
( n = 33)
|Patients with SCD |
( n = 53)
|Correlation with TRJ velocity|
|LVESV (mL)||34 ± 10||41 ± 16||.02||.03||.09||.52||.66|
|LVEDV (mL)||92 ± 26||106 ± 33||.09||.12||.06||.66||.74|
|LV mass (g)||143 ± 38||219 ± 67||<.001||.002||.30||.03||.06|
|LV mass/height 2.7 (g/m 2.7 )||36 ± 8||52 ± 13||<.001||.002||.30||.03||.06|
|RVESA (cm 2 )||9.8 ± 2.3||11.8 ± 3.2||.28||.37||.26||.11||.22|
|RVEDA (cm 2 )||15.6 ± 3.1||19.4 ± 4.6||.02||.03||.24||.14||.22|
|LA volume (mL)||43 ± 14||80 ± 30||<.001||.002||.50||<.001||.005|
|LA volume/BSA (ml/m 2 )||23 ± 6||45 ± 17||<.001||.002||.49||<.001||.005|
|RA volume (mL)||42 ± 13||63 ± 22||<.001||.002||.31||.03||.06|
|LV systolic function|
|Stroke volume (mL)||62 ± 13||86 ± 19||<.001||.002||.09||.54||.66|
|Ejection fraction (%)||63 ± 4||60 ± 6||.03||.048||−.003||.98||.99|
|LV MPI||0.47 ± 0.10||0.41 ± 0.14||.07||.10||.02||.91||.98|
|RV systolic function|
|RV fractional area change||0.36 ± 0.12||0.39 ± 0.10||.04||.06||.002||.99||.99|
|RV MPI||0.28 ± 0.09||0.25 ± 0.11||.43||.52||.22||.14||.22|
|RV S′||13.2 ± 2.5||9.9 ± 2.0||<.001||.002||−.08||.58||.68|
|LV diastolic function|
|Peak E (m/sec)||0.73 ± 0.14||0.90 ± 0.24||.001||.002||.21||.14||.22|
|Peak A (m/sec)||0.48 ± 0.12||0.54 ± 0.16||.002||.004||.35||.01||.04|
|E/A ratio||1.6 ± 0.5||1.8 ± 0.7||.82||.87||−.17||.22||.32|
|DT (msec)||192 ± 30||184 ± 33||.66||.77||.13||.35||.47|
|IVRT (msec)||79 ± 10||77 ± 20||.86||.87||.21||.13||.22|
|Lateral E′ (cm/sec)||15.6 ± 4.0||16.8 ± 4.4||.87||.87||−.33||.02||.06|
|Lateral E/E′ ratio||4.9 ± 1.2||5.6 ± 2.0||.008||.02||.35||.01||.04|
|Septal E′ (cm/sec)||11.5 ± 2.5||12.0 ± 2.6||.75||.84||−.17||.23||.32|
|Septal E/E′ ratio||6.5 ± 1.3||7.6 ± 1.8||<.001||.002||.34||.01||.04|
|RV diastolic function|
|RV E′ (cm/sec)||13.2 ± 2.5||14.5 ± 3.4||.37||.47||−.30||.03||.06|
|TV E/E′ ratio||4.4 ± 0.9||6.4 ± 1.8||<.001||.002||.40||.004||.03|
|TRJ velocity (m/sec)||2.1 ± 0.2||2.5 ± 0.5||.001||.002||NA||NA|