Study Population

Cases were recruited prospectively from patients enrolled in the MSC and controls from children who had attended the clinic for sickle screening and were determined not to have SCD. Inclusion criteria included age between 9 and 19 years and, for patients with SCD, sickle status confirmed by hemoglobin (Hb) electrophoresis and high-performance liquid chromatography, while controls without SCD had the HbAS or HbAA phenotype by electrophoresis. Blood pressure was measured twice after 5 min of rest using an automated digital device (PRO 400 V2 Dinamap; GE Healthcare, Little Chalfont, United Kingdom) and the mean calculated. Axial temperature, height, and weight were recorded, and a clinical examination was performed to exclude acute complications. Exclusion criteria included history of blood transfusion within the previous 2 months, symptoms of sickle crisis within the previous 2 weeks, febrile illness within the 7 days before echocardiography, fever or signs of acute illness on the day of echocardiography, and known or echocardiographically identified hemodynamically significant congenital heart disease.

The study was approved in the institutional review board of Children’s Hospital Colorado (Reference No. COMIRB protocol #10-0030) and by the Muhimbili University of Health & Allied Sciences Research and Publications Committee (Reference No. MU/RP/AEC/Vol.XIIV/01). Written informed consent was obtained from all study participants.

Definitions

Rotation was defined as circumferential myocardial motion around the centroid, or long axis, of the ventricle. By convention, rotation is said to be positive in the counterclockwise direction when viewed from the apex of the left ventricle and vice versa. As depicted in Figure 1 , a basal short-axis plane used to determine twist was obtained just below the mitral valve annulus, where muscle appears in the plane of insonation throughout the cardiac cycle. The apical short-axis plane used was defined as a short-axis plane just above the apex, where the LV lumen can be seen throughout the cardiac cycle. Net twist angle was defined as the instantaneous difference in rotation between the apex and the base at the time of aortic valve closure, or end-systole. Peak differential twist is the maximum difference in rotation between the apex and base, independent of time. Torsion was defined as net twist divided by the length of the left ventricle at end-diastole to account for heart size. Untwisting is defined as the difference between apical recoil and basal recoil. The timing of these events is described both in milliseconds and as a percentage of systole, with the onset of systole defined as the initial deflection of the Q wave on the surface electrocardiogram and end-systole defined as aortic valve closure.

Assessment of rotational mechanics using two-dimensional speckle-tracking echocardiography. Representative rotation-time plots for a subject with SCD. Basal and apical short-axis planes were obtained as shown, and speckle-tracking analysis was performed. Peak global rotation on each image was measured ( white dots ) with reference to the electrocardiogram below (the Q wave is marked by a yellow dot ) and aortic valve closure (indicated by a vertical green line ). SCD , Sickle cell disease.

Echocardiography

For each patient, measurements of LV chamber size and fractional shortening were made from short-axis basal images. From the apical window, LV peak systolic mitral annular velocity (LV S′), LV peak early diastolic mitral annular velocity (LV E′), early diastolic mitral inflow (LV E), duration of late diastolic mitral inflow (mitral valve a-wave duration), and duration of the late diastolic a wave in pulmonary vein inflow (pulmonary vein a-wave duration) were made. The LV myocardial performance index was also calculated from mitral inflow and aortic outflow spectral Doppler patterns.

Short-axis images of the base and apex of the left ventricle were obtained using the VividQ platform (GE Healthcare) and an M4S, 5S, or 7S probe, depending on patient characteristics and image quality. Frame rates of >75 and <125 frames/sec were accepted. An acquisition of five consecutive beats was obtained. An apical view was also obtained for the purposes of measuring end-diastolic LV length and performing spectral Doppler measurement of mitral inflow and aortic outflow. End-diastolic LV length was measured from the apical four-chamber view as the distance from the mitral valve to the apex of the left ventricle.

Commercially available software (EchoPAC version 113, revision 0.4; GE Healthcare) was used to perform speckle-tracking analysis. A single cardiac cycle was isolated for the base, and a region of interest was identified. The magnitude and timing of basal rotation was calculated. The process was repeated for the short-axis apical image. Net twist was calculated as the difference between apical and basal rotation. Torsion was calculated by dividing the net twist by LV end-diastolic length. This process was repeated two additional times and an average over three cardiac cycles was calculated.

Laboratory Data

Hb phenotype was determined by alkaline Hb electrophoresis (Helena; Sunderland, United Kingdom) and by quantification of Hb fractions, including HbF, by high-performance liquid chromatography (β-Thalassemia Short Program, VARIANT analyzer; BioRad, Hercules, CA). In subjects with SCD, the most recent steady-state Hb, total bilirubin, conjugated bilirubin, and lactate dehydrogenase levels were obtained from MSC records (steady state defined as absence of pain or fever, malaria rapid test negative, and with no recorded hospital admission or blood transfusion within 30 days). Full blood counts were performed using an automated cell counter (Pentra 60; Horiba ABX, Kyoto, Japan). Biochemical tests were performed using an automated chemistry analyzer (Cobas Mira [Roche, New York, NY] or Abbott Architect [Abbott Diagnostics, New York, NY]).

Statistical Analysis

SAS version 9.2 (SAS Institute Inc, Cary, NC) was used to perform statistical analyses. Continuous variables are presented as mean ± SD or median (interquartile range), and categorical variables are presented as proportions. The Student independent two-sample t test was used to compare mean values between two patient groups. Simple linear regression was used to evaluate associations between independent variables and a continuous, dependent variable. Pearson correlation was used to evaluate associations between measures of rotational mechanics and Hb or laboratory measures of hemolysis in the children with SCD. A P value < .05 was determined to be the threshold for statistical significance. Analyses were compared when the five children with HbSB+ were included or excluded from the patients with SCD and no material differences were observed, and therefore all reported analyses include both the HbSS and HbSB+ cases.

Interobserver and Intraobserver Variability

Eleven consecutive patients were evaluated separately by two independent observers (H.H.H. and G.A.), who evaluated net twist, torsion, twist rate, and peak untwisting rate in a blinded fashion. To reflect the methodology used in this study, each measurement was made on what the investigator felt was the best cardiac cycle among the five that were acquired. A single observer (H.H.H.) performed serial analysis of the same 11 patients 2 months later using the same methodology to determine intraobserver variability. Analysis included calculation of intraclass correlation coefficients of variation.