Annual evaluation by Trans-Cranial Doppler (TCD) of cerebral artery blood flow is useful for determining stroke risk in children with sickle cell disease. Since the publication of the Stroke Prevention Trial in Sickle Cell Anaemia (STOP) results in 2006 it has been standard practice at our centre to perform annual TCD imaging scans on all children with sickle cell disease to identify individuals with high risk features. Data from the STOP trial demonstrated that the incidence of stroke was reduced by 90% in children with high risk TCDs when they were managed on a regular transfusion programme.

In more recent years, the use of hydroxyurea as a disease modifying therapy has increasingly become our practice, particularly so since the publication of the BABY-HUG trial data in 2014. Hydroxyurea has significantly altered the clinical course of sickle cell disease in children and prior to this era, 11% of patients would experience a stroke before the age of 20 years.

In light of the changing management strategies for children with sickle cell disease, we reviewed the past 10 years TCD results from our patients.

Annual TCD evaluation of cerebral artery blood flow was performed by radiologists and radiographers accredited to the Kings College national training programme. Scans were performed using Philips iU22 scanners with dedicated F5-1 transducers as per the UK National Screening protocol. Velocities were measured in the middle (MCA), anterior (ACA) and posterior cerebral arteries (PCA), internal carotid artery (ICA) and bifurcation on right and left sides (10 measurements per scan). The TCD service at our centre has had a positive external peer review.

Over the past 10 years, 303 TCD imaging scans have been performed on 53 children aged 2-18 years. The number of abnormal, high-risk results (cerebral blood flow velocity (CBF-V) >200 cm/s) has been zero. 137 scans were performed in children prior to those individuals commencing hydroxyurea therapy, and the results of 3 scans were conditional, based on an increased CBF-V (171-200 cm/s) detected in 1 or more arteries. The other 168 scans were performed on children who were receiving hydroxyurea, of which 4 results were conditional. 49 of the 53 children have continuously had a normal TCD with no conditional or abnormal results. 4 children developed transient conditional TCD.

The 3 conditional results recorded prior to commencing hydroxyurea therapy relate to 2 patients. One case in which a child with severe phenotype already on regular blood transfusions developed a conditional TCD which subsequently normalised without adjustment to their regular transfusion programme. Transfusions were then stopped and hydroxyurea commenced and this patient has since maintained normal TCD features (8 years follow up). The other 2 conditional scans relate to one patient where a conditional TCD developed in a young child not receiving any therapy. This feature normalised without treatment by the time of the following annual scan but subsequently reoccurred the following year. Hydroxyurea was commenced because of the second conditional TCD result, which then normalised and has remained normal (7 years follow up).

The 4 conditional results that occurred in individuals receiving hydroxyurea relate to 2 patients. Both children developed conditional TCDs and the increased CBF-V persisted until their next annual scan. In both cases hydroxyurea was escalated to maximum tolerated dose following the second conditional TCD result. TCDs in both patients normalised by the time of their next annual scan and have remained normal since (6 years and 2 years follow up).

Our hydroxyurea practice has resulted in not a single child being started on transfusion therapy from our TCD programme. Importantly, at the end of this observation period all 53 children have normal TCD features and no child is managed on a transfusion programme. In the era of high dose hydroxyurea therapy, our experience is that we have not seen any abnormal TCD results. We need to consider whether TCD remains relevant or whether alternative strategies to predict and prevent stroke need to be developed.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract

Sign in via your Institution