Abstract

Background: Sickle cell anemia (SCA) is one of the commonest monogenic disorders, with 90% of the world’s population living in sub-Saharan Africa. Cerebrovascular accident (CVA) is a major cause of morbidity, but its clinical prediction in resource rich countries has allowed effective primary and secondary prevention. Measurements of time-averaged maximum of the mean (TAMM) cerebral blood flow velocity (CBFv) in the internal carotid/middle cerebral (ICA/MCA) arteries by Transcranial Doppler (TCD) ultrasonography and of mean overnight oxyhemoglobin saturation (SpO2) have been useful in predicting CVA. The criteria used in Western populations may not be appropriate to children living in Africa.

Aims: The aims of this study were to evaluate the TAMM CBFv in patients with SCA in Kilifi district hospital, Kenya, to assess risk factors associated with high ICA/MCA TAMM CBFv and to examine any association with neurological complications.

Study design: This was a cross sectional descriptive study, where TCD ultrasonography was performed on all SCA patients attending the outpatient clinic at CGMR-C, Kilifi, Kenya in 2002. Previous data from 1990 and follow-up data from 2004 were included.

Results: In 140 patients with SCA, aged 3 months to 16 years, the median ICA/MCA TAMM CBFv was 116cm/sec (SD 38, range 0–219 cms/s) compared with 97 (SD 24, range 46–190) cm/sec in 142 controls aged 2 months to 14 years (p=0.0001). 28 SCA patients (20%) had TAMM CBFv greater than and 16 (11%) had TAMM CBFv less than 2 standard deviations from the mean for controls in one or both ICA/MCA’s, but only seven (5%) had a velocity above 170 cm/sec (one >200cm/sec), with the highest proportion of patients aged between 5–9 years (p=0.02). In only two of the patients with low velocities, both with previous CVA, was there no ultrasound signal from either side. 45 (32%) SCA patients had a second TCD after 2 years (two after 14 years). Of the 21 restudied who had high TAMM CBFv at baseline, 14 remained high and 2 became low. Of the 15 restudied who had low TAMM CBFv at baseline, 14 remained low and none became high. Patients with abnormal TCD had lower daytime SpO2 oxygen saturation (p=0.01) and hematocrit (p=0.05). Abnormal TCD was also associated with lower haemoglobin level, red blood cell count and higher white cell count, but not significantly. Neurological abnormalities included history of convulsions in 25 (18%) and history of CVA in 5 (4%). Of those with CVA, maximum TAMM CBFv on either side were 157, 156, 108, 0 and 0; the last patient subsequently died. Three patients who had convulsions in the interim attended for follow-up TCD; compared with those without seizures there was a trend for a greater increase in TAMM CBFv in these patients (p=0.06).

Conclusion: Compared with the developed world, in Africa a smaller proportion of patients with SCA have conditional or abnormal TCDs or CVA, although convulsions are common. The proportion of those with low velocities, perhaps due to ICA/MCA occlusion with moyamoya, may increase with time. Further population-based studies in a birth cohort will determine whether cerebrovascular disease is rare or lethal and, together with imaging and neuropsychology, will establish whether abnormal TCD predicts neurological events in Africa.

Author notes

Corresponding author