The STOP trial provided conclusive, level 1 evidence that TCD screening reduces the risk of stroke in patients with homozygous S sickle cell disease. These findings were based on data collected from a discrete population of children in the United States using standardised non-imaging TCD instrumentation and a strict protocol for TCD data acquisition and interpretation. Since this time there has been considerable energy addressed to provide a TCD screening service to all children with SCD. Initial efforts were hampered by a lack of available TCD equipment and lack of technical expertise in TCD scanning. A solution to both problems was provided by duplex colour mapping systems which have facilitated the widespread propagation of TCD surveillance programmes. There was initial caution in the substitution of imaging TCD in the sickle surveillance programme, particular in view of the recognised differences in velocities acquired by the two TCD techniques. A number of elegant studies have been undertaken to establish the variance between the techniques and adjustments to imaging TCD diagnostic thresholds have been advised. Surveillance is now being undertaken with a range of TCD imaging systems. The present study reviewed the screening program at our institution where STOP classification is based on imaging and non-imaging TCD.
Data from previous comparative studies have indicated that TCD imaging velocities, obtained with a rigorous protocol, underestimate non-imaging TCD by 10%. Based on these findings a 10% adjustment was applied to the STOP thresholds. TCD imaging time-averaged maximum (TAMX) velocity data was therefore classified as normal if velocities were <155cm/s; conditional for velocities of 155–179cm/s and abnormal for >180cm/s velocities. Children seen in a One Stop clinic were first scanned with imaging TCD (Philips IU22) and those who had velocities falling in the conditional or abnormal ranges were rescanned with non-imaging TCD (Nicolet, Companion III).
Based on imaging-TCD velocities, 3 middle cerebral arteries were classified as abnormal, 7 conditional and the remaining 4 normal. However, repeat scanning with non-imaging TCD demonstrated; no abnormal classifications, 3 conditional and 11 normal. Comparison of TAMX velocities obtained by the two techniques showed no systematic difference (mean 1±8cm/s).
The TCD imaging system used in this study did not demonstrate any systematic difference in velocity compared to non-imaging TCD. Reported differences in velocities between the two techniques have been attributed to manufacturer derived and technical expertise factors consequently, current STOP thresholds for imaging TCD are flawed. All centres should evaluate their imaging TCD velocities against expert derived non-imaging TCD values to determine the correct STOP velocity thresholds for their institution.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.