Transcranial Doppler Alternative Stroke-Risk Threshold in Sickle Cell Disease: A Basra Cohort Study

Abstract

The Stroke Prevention in Sickle Cell Anemia (STOP) trial established time-averaged mean of maximum (TAMMx) velocity > 200 cm/s in the middle cerebral artery (MCA) as the threshold for instituting chronic transfusion in children with sickle cell disease (SCD). However, accumulating evidence from Eastern Mediterranean and Gulf cohorts — including a previously published Basra TCD analysis — suggests that intracranial velocities in these populations are systematically lower than in African and African-American populations, making the 200 cm/s threshold poorly suited to local risk stratification. To evaluate the diagnostic performance of an MCA velocity threshold of > 140 cm/s as an alternative stroke-risk indicator and compare it with the conventional STOP criteria (> 170 cm/s conditional and > 200 cm/s abnormal) in the Basra SCD cohort. Cross-sectional retrospective analysis of all transcranial Doppler (TCD) records performed at the Basra Centre for Hereditary Blood Disorders between January 2014 and May 2026. After de-duplication and merging of two source registries, the cohort comprised 1,781 patients with 2,211 TCD encounters. Sixteen patients with confirmed clinical or radiological stroke were identified; Per-patient peak (maximum) MCA velocity was used as the discriminator. Sensitivity, specificity, predictive values, odds ratio, Fisher exact test, and area under the receiver-operating-characteristic curve (AUC) were computed for the > 140 cm/s, > 170 cm/s and > 200 cm/s cut-offs. Mean peak MCA velocity was 145.6 ± 13.6 cm/s in stroke patients (n = 16) versus 134.2 ± 18.5 cm/s in non-stroke patients (n = 1765). The proportion exceeding 140 cm/s was 75.0% in stroke patients versus 31.4% in non-stroke patients. The MCA > 140 cm/s threshold yielded sensitivity 75.0%, specificity 68.6%, NPV 99.7%, odds ratio 6.56 (Fisher p < 0.001), and AUC = 0.706. No patient — stroke or non-stroke — reached the STOP abnormal threshold of > 200 cm/s, rendering that cut-off uninformative in this population. The > 170 cm/s threshold identified only the highest-velocity stroke patient (sensitivity = 6 %). In the Basra SCD cohort, the conventional STOP threshold of > 200 cm/s failed to flag any stroke patient and is therefore inappropriate as a stroke-prevention trigger locally. An MCA velocity threshold of > 140 cm/s offers substantially better discrimination (OR: 6.6, AUC ≈ 0.71) while remaining feasible in routine paediatric haematology practice. We propose > 140 cm/s as a regionally adapted, actionable threshold for initiating closer surveillance and considering hydroxyurea optimization or transfusion in Iraqi SCD patients, pending prospective validation. Recommendation considers the MCA velocity as added risk factor for Stroke screening.