Increased Cerebral Blood Flow Velocity in Children With Sickle Cell Disease: Adenotonsillectomy or Transfusion Regimens?: In Reply

Abstract

We appreciate the possible implications for children with sickle cell anemia (SCA) in the response by Bader-Meunier et al to our article on increased cerebral blood flow velocity in children with SCA with mild sleep-disordered breathing.1 Although there seems to be a link between sleep-disordered breathing and changes to the cerebral vasculature, the mechanism has not been elucidated. Candidates include the mechanical effects of snoring,2 associated hypoxemia,3 and infection/inflammation.4 Tonsillectomy is sometimes performed for recurrent tonsillitis or relief of upper airway obstruction in children with SCA,5 although mean overnight oxyhemoglobin saturation may not improve,5 and there are few systematically collected data that have addressed a relationship with SCA complications such as transcranial Doppler (TCD) abnormality.TCD testing has been undertaken at least once in 384 children with homozygous SCA in the London cohort, and adenotonsillectomy has been performed on clinical grounds. In 183 (47.7%) patients, TCD data are available on 2 occasions (mean age at follow-up: 11.2 years [SD: 4.5 years]). Of these, 40 (21.9%) had undergone adenotonsillectomy between the 2 TCD measurements (mean age at surgery: 7.7 years [SD: 3.9 years]), 28 had adenotonsillectomy before the first TCD (not included in this analysis), and 115 had not undergone adenotonsillectomy. We applied hierarchical regression analysis retrospectively to examine the effect of 3 predictor variables on middle cerebral artery velocity (MCAV): age at follow-up TCD (step 1); MCAV at baseline (preoperative in surgical cases: step 2); and surgical status (removal of tonsils or not; 0, 1: step 3). Age at follow-up TCD significantly predicted follow-up MCAV (F1,153 = 28.14 [P < .001]; R2 = .16). When baseline MCAV was added (step 2), the explanatory power of the model significantly increased (F2,152 = 57.87 [P < .001]; ΔR2 = .28 [P < .001]). However, at step 3, surgical status did not add to the power of the model (F3,151 = 38.45 [P < .001]; ΔR2 = .00 [P = .656]). As can be seen from the summary in Table 1, older age was predictive of lower MCAV at follow-up, whereas higher baseline MCAV was predictive of higher follow-up MCAV. Surgical status did not affect final MCAV levels. Of 18 children with MCAV of ≥200 cm/second, 15 had a repeat TCD measurement, of whom 2 had had adenotonsillectomy between the measurements; MCAV decreased from 262 to 250 cm/second in one patient and from 212 to 205 cm/second in the other; both still fell into a range that, according to Stroke Prevention in Sickle Cell Anemia (STOP) criteria, would make transfusion required. Of 17 children with MCAV of ≥170 but <200 cm/second (conditional), all had repeat TCD measurements, of whom 2 had had adenotonsillectomy between the measurements; MCAV decreased from 179 to 170 cm/second in one and from 194 to 14 cm/second in the other.Although these data suggest that adenotonsillectomy has limited influence on MCAV in children with SCA, surgery may be performed for different reasons, including obstructive sleep-disordered breathing, chronic or intermittent oxyhemoglobin desaturation, or, alternatively, chronic infection. In addition, adenotonsillectomy for patients with SCA is often undertaken some time after the onset of symptoms because of operative risk factors. Postoperative MCAV, therefore, may be differentially affected by the original etiology and the duration and severity of exposure to infection, inflammation, and hypoxemia as well as the treatment per se. Perioperative carotid dissection4 is also a risk; this was a possible cause of the very low postoperative velocity in 1 of our conditional patients. Once initiated, cerebrovascular disease in SCA is rarely completely reversible,6 but it is possible that in the cases reported by Bader-Meunier et al, a detrimental effect of hypoxemia and chronic infection on the endothelium was reversed before secondary vascular changes commenced.Polysomnography would be informative for future study of young children with SCA to clarify the associations between adenotonsillar hypertrophy, objectively rated obstructive sleep-disordered breathing, and increased MCAV. This would be comparable with the design of our original study in children without SCA.