It is recommended both in the USA and Europe that children with SCD are scanned by TCD to identify increased risk of stroke and allow primary prevention. Although there are convincing trial data to support this, there is relatively little information on how this works in practice, with concerns about false positives, false negatives, implications of transfusion/chelation regimes. 20 months ago, a routine screening service was started for children with SCD attending Kings’s College Hospital (350 children). Scans were organised at clinic appointments and performed in the vascular laboratory by staff who had attended the training course at Medical College of Georgia.

Scans followed the STOP protocol, and used TCD imaging, to visualise the vessels. 178 children without a history of stroke stroke have been scanned. 143 (80%) were normal, 12 (6.7%) had conditional results (TAMX>170cm/s, <200cm/s), 4 (2.2%) had abnormal results (TAMX>200cm/s), 11 (6.2%) were unsatisfactory (not all vessels were seen), and 8 (4.5%) showed other abnormalities.

Of the 12 with conditional scans, 8 have been repeated so far; of these, 3 were subsequently normal, and 5 persistently elevated; MRI/MRA scans were normal in 2 of these, and showed a probable right MCA stenosis in the 3rd; 2 are awaiting MRI/MRA. Of the 4 with abnormal scans, all were repeated and confirmed abnormal. On MRI/MRA two were shown to have extensive cerebrovascular disease with appearances of Moya Moya, and large, but clinically silent, cerebral infarcts. One had a normal MRI/MRA scan and decided against starting regular blood transfusions. One presented with left-hemiparesis before the repeat TCD, and MRI/MRA showed the new infarct and the stenosis also seen on TCD.

Of the 11 unsatisfactory scans, 3 are awaiting repeat, 5 were normal on repeat, 1 was persistently unsatisfactory but had a normal MRI/MRA. 2 were persistently unsatisfactory and had abnormal MRI/MRA scans with stenosed vessels; one of these children presented with an extensive, hemiplegic stroke before repeat scanning or MRI/MRA was performed. All of the non-STOP abnormalities consisted of increased velocities and tortuosity/kinking in the proximal ICA and were not associated with significant intracranial abnormalities in any case.

During the 20 months, 3 children presented with new overt strokes: two were identified on TCD, as mentioned above (1 with raised velocities and stroke prior to repeat scan, and 1 with unsatisfactroy examination). One presented with nephrotic syndrome and acute hemiplegia following normal TCD, and MRI/MRA showed normal blood vessels and an old infarct. 5 (2.8%) children started on regular blood transfusions: 3 presenting with strokes, and 2 with abnormal TCDs (who both had infarcts and Moya Moya on MRA/MRI).

TCD is useful in routine service screening for cerebrovascular disease, in SCD. We found no patients who fulfilled typical STOP criteria for primary prevention, and our rate of abnormal scans was lower than predicted (2.2%); absent TCD traces/vessel images proved as significant as increased velocity in our patients. TCD also identified 2 pateints with clinically silent but extensive cerebrovascular disease, facilitating secondary prevention.

Author notes

Corresponding author

Sign in via your Institution