Sickle cell disease is a hereditary haemolytic anaemia characterised by the synthesis of the haemoglobin S (HbS, sickle cell haemoglobin). Homozygous sickle cell disease (SS) results from the inheritance of two sickle cell genes (β^s). Coinheritance of a β^s gene and a β thalassaemia gene (β^thal) results in different types of sickle β thalassaemia (Sβ) depending on the mutation, which is carried by the β thalassaemia gene, and the amount of HbA synthesised. If the β^s gene is inherited with a β⁺ thalassaemia gene, in which some β chains are present, the genotype is Sβ⁺ and if the β thalassaemia gene is β⁰, in which β chains are absent, the genotype is Sβ⁰. It is well known that whether the interacting β thalassaemia gene is βo the clinical picture is similar to sickle cell anaemia whereas, when the interacting β thalassaemia gene is of β+ type it has milder clinical signs and symptoms.¹ Cardiac abnormalities in sickle cell disease have been well described but mainly in sickle cell anaemia.² However, little is known about heart involvement in sickle β thalassaemia.

The purpose of this study was to investigate the changes in systolic and diastolic function of the left and right ventricle (LV and RV respectively) in middle aged patients with sickle β thalassaemia.

METHODS

Study population

Forty three consecutive patients between 45 and 66 years (15 men, 28 women, aged mean (SD) 54 (10) years) with sickle β thalassaemia and no underlying heart disease and 55 (18 men, 37 women, aged 53 (11) years) healthy volunteers were recruited in our study. Patients with sickle β thalassaemia were recruited from the haemoglobinopathies unit of Laiko General Hospital, Athens, Greece. Inclusion criteria were, diagnosis of sickle β thalassaemia, made by solubility screening test for sickling, haemoglobin electrophoresis, and α/β chain synthesis to confirm the thalassaemic genotype. All patients were Sβ+ type, with haemoglobin concentrations higher than 70 g/l; they had painful crises and red cell transfusions rarely (less than five per year) while none was receiving drugs such as digoxin, diuretics, or vasodilators. The study protocol was approved by the institutional committee on human research and all subjects gave informed consent.

Echocardiographic evaluation

Comprehensive echocardiographic examination with pulsed, continuous, and colour Doppler was performed with a Hewlet Packard Sonos 1000 ultrasound System, using a 2.5 MHz transducer by the same cardiologist and was recorded on tape. The measurements were performed, independently, by two cardiologists unaware of the subject’s status. From the two dimensional guided M-mode echocardiogram, LV end systolic and end diastolic diameters as well as interventricular septum and posterior wall thickness at end diastole were measured for the calculation of fractional shortening (FS) and LV mass with the Penn convention formula.^3,4 Measurements of LV mass were divided by body surface area to obtain LV mass index. The LV ejection fraction (EF) was determined by the biplane Simpson’s method. The LV stroke volume was also calculated as the product of the cross sectional area of the LV outflow tract and velocity time integral of the LV outflow velocity.⁵ The LV cardiac output was derived as the product of the stroke volume and heart rate and, divided by body surface area; the LV cardiac index was calculated. Furthermore, we measured several parmaeters of LV systolic function such as the circumferential fibre shortening velocity (Vcf) as the ratio FS to ejection period, the meridional end systolic stress (ESS), the ratio end systolic stress/end systolic volume index (ESS/ESVI), and the mitral annular plane systolic excursion (MAPSE).^6–9 RV end diastolic diameter and thickness of the free wall was measured and the RV systolic function was evaluated from the tricuspid annular plane systolic excursion (TAPSE).^10,11 We also evaluated the myocardial performance index (Tei index) of LV and RV.^12,13 Using pulsed Doppler from the mitral and tricuspid inflow velocity curves the following parameters were calculated: peak early velocity (E-wave), peak velocity at the time of atrial contraction (A-wave), E/A ratio, deceleration time (DT) of the peak early velocity, and the isovolumic relaxation time (IVRT). For the calculation of the RV diastolic variables at least three beats from the end inspiration and three beats from the end expiration were recorded and their values were averaged.^14–16 Flow velocities of the hepatic veins and upper pulmonary veins were also recorded for systolic to diastolic forward flow velocity ratio (S/D) calculation. The intraobserver and interobserver mean percentage error (absolute difference between two measurements divided by the mean and expressed in percentage) of Doppler measurements were respectively (for the LV: 3.1% and 3.3% for E-wave, 3.3% and 3.6% for A-wave, 3.5% and 4.2% for DT, 3.5% and 4.3% for IVRT whereas for the RV: 6.3% and 6.9% for E-wave, 6.5% and 7% for A-wave, 5.6% and 5.9% for DT, 5.2% and 5.5% for IVRT). Continuous wave Doppler echocardiogram recorded in the apical four chambers or parasternal short axis view was used to estimate the peak systolic pressure gradient across the tricuspid valve. A tricuspid gradient greater than 30 mm Hg was considered indicative of pulmonary hypertension.^17,18

Statistical analysis

Data are expressed as mean (SD). Differences between groups were compared using the unpaired Student’s t test as appropriate. Linear correlation and regression were used to test the correlations between serum ferritin concentrations with the echocardiographic findings. A probability (p) value of less than 0.05 was considered significant.

RESULTS

As table 1 shows, the patients with sickle β thalassaemia had lower levels of diastolic blood pressure, packed cell volume, and haemoglobin and increased concentrations of serum ferritin compared with controls.

Chamber enlargement (left atrium, LV, and RV), greater LV mass index (109 (8) compared with 86 (19) g/m², p<0.01), cardiac index (4.80 (1.80) compared with 3.70 (1.20) l/min/m², p<0.01) and RV wall thickness (0.32 (0.07) compared with 0.29 (0.06) cm, p<0.05) were noted in the anaemic group (table 2).

There was no difference in the LV systolic function variables such as the EF, ESS, ESS/ESVI, Vcf, and MAPSE between patients and controls. Similarly, the RV systolic function was not impaired assessed by the TAPSE. Conversely the Tei index expressing combined systolic and diastolic performance was increased for both ventricles in the anaemic group (0.39 (0.07) compared with 0.36 (0.06), p<0.05 for LV and 0.37 (0.11) compared with 0.32 (0.09), p<0.05 for RV) distinguishing global LV and RV dysfunction (table 3).

The diastolic function was also impaired in anaemic patients compared with controls. Specifically, the LV and RV A-wave was increased (53 (12) v 44 (15) cm, p<0.01 and 47 (9) v 40 (8), p<0.01 respectively), the LV and RV IVRT was longer (83 (17) v 76 (11) ms, p<0.05 and 52 (13) v 46 (11) ms, p<0.05 respectively) and the RV DT was also longer (198 (39) v175 (34) ms, p<0.01) in the group of patients. Furthermore, the ratio S/D in pulmonary and hepatic veins was increased (1.43 (0.50) v 1.22 (0.30), p<0.05 and 2.20 (1.08) v1.64 (0.90), p<0.01 respectively) in the anaemic group (table 4). Additionally pulmonary hypertension was also present in nine (21%) patients and was mild.

Serum ferritin concentrations, for the whole group of 43 patients, were correlated significantly with RV Tei index (r = 0.55, p<0.01). Correlation with LV mass index and RV IVRT was also less exact but still significant (r = 0.43, p<0.05 and r = 0.39, p<0.05 respectively).

DISCUSSION

It is well known that the clinical course for patients with sickle β thalassaemia is similar to that seen in patients with homozygous sickle cell disease.¹⁹ Thus it is therefore logical to state that there is a good number of studies on sickle cell anaemia and heart involvement but only a few concerning sickle β thalassaemia.^2,20–22

Acquired chronic anaemia, in young or older patients, was found causing systolic hyperfunction attributable to the hyperdynamic state and no impairment of diastolic function.^23,24 Similarly cardiac abnormalities in sickle β thalassaemia are thought to be secondary to anaemia related volume overload. However, more factors such as transfusion therapy with iron overload and toxicity, renal, pulmonary and heart injury attributable to vaso-oclusive effect of sickle haemoglobin on the circulation are involved in the sickle syndrome.²⁵

In our study the patients with sickle β thalassaemia had increased dimensions of the left atrium, and left and right ventricle than the controls. Moreover they had a greater LV mass index, a thicker RV wall, and a higher cardiac index.

We found that the LV EF and the other traditional contractility parameters—Vcf, ESS, and MAPSE—were not different between patients and controls. It is possible that the above variables are insensitive because they are load dependent and in sickle β thalassaemic patients high preload and low afterload are present. However, the ratio ESS/ESVI, which is comparatively load independent, was similar in patients and controls. Likewise the RV systolic function was not impaired as assessed by the TAPSE. Conversely the LV and RV Tei index expressing global myocardial performance significantly differs between patients and controls, distinguishing global dysfunction. The fact that the Tei index decreases in increased preload and/or decreased afterload and we found opposite results supports the notion of an abnormality mainly in diastolic function in such patients.^26,27

Significant differences as far as the parameters of ventricular filling and pulmonary and hepatic vein flow were found between patients and controls. These diastolic variables however are load dependent. Specifically the high preload and low afterload in anaemic patients would be expected to cause physiologically an increase of transtricuspid and transmitral inflow velocities and shortened IVRT and DT, which was not the case.^28,29 Thus the RV and LV filling pattern show impaired relaxation. Additionally the pulmonary and hepatic vein flow characteristics are also compatible with abnormal relaxation as the increased preload decreases the S/D ratio and our findings are opposite.^28,30

Our findings are in accordance with the findings of Braden et al who found abnormal diastolic filling pattern in patients with sickle cell anaemia and coexisting α thalassaemia-2.³¹ As in our study, they found significantly greater wall thickness but not chamber enlargement. It is possible that this happens because their population was younger but mainly because of the sickle cell anaemia and coexisting α thalassaemia-2 is less severe than sickle β thalassaemia.³² Previous studies have also established left sided chamber enlargement and biventricular dysfunction in patients with sickle cell anaemia. However, there is controversy about the frequency of the influenced ventricle.^2,31,33–35 Other investigators have found the LV ventricular performance to be normal in these patients, studied by the ejection phase parameters, and only the index ESS/ESVI ratio discriminated the above population from controls.^36–39 In our study the preserved LV and RV systolic function may be associated with the Sβ+ type of sickle cell thalassaemia, which is less severe.¹

In this study, impairment of diastolic parameters of RV function was more pronounced than impairment of the respective LV function. The explanation for this may be multifactorial. This dissociation possibly reflects a preferential involvement or a higher susceptibility of the RV. The fact that the RV diastolic variables are correlated positively with the serum ferritin concentrations, support the hypothesis that lung and heart haemochromatosis affects predominantly the RV function. Hyperkinetic circulation and vaso-oclusive effects may have also more effect on the RV, which has substantially a smaller mass than the LV, resulting in faster functional derangement.

In conclusion, our findings show that in middle aged patients with sickle β thalassaemia, Sβ⁺ type, the diastolic function is abnormal in both ventricles but more noticeable in the RV. The systolic function remains unchanged. However, these abnormalities seem insignificant and well tolerated, because of sufficient cardiovascular adaptation, suggesting longer living in this population.