Supplementary Material for: Head positioning for stroke blood flow augmentation assisting reperfusion therapies (HEAD-START) study

Background
It is uncertain whether lowered head position meaningfully improves cerebral perfusion in ischemic stroke.
Aims
We performed a prospective, single-arm, single-centre, self-controlled, non-randomised, pre-post intervention study, testing whether 20-degree head-down (Trendelenburg) positioning in patients with acute stroke improves perfusion of ischaemic brain tissue, as measured by automated quantitative computed tomography perfusion (CTP).
Methods
We enrolled patients aged ≥60, 0-24h after acute stroke onset, with ≥30mL anterior circulation CTP lesion volume (delay time [DT]>3seconds, MISTAR software). CTP was acutely repeated after 5 minutes of on-table 20-degree Trendelenburg positioning (achieved by a custom-designed foam wedge). Clinical severity (National Institutes of Health Stroke Scale [NIHSS]) and blood pressure were recorded in routine (30 degree up) and Trendelenburg position. Trendelenburg positioning was maintained for 24h if lesion volume significantly decreased (≥5mL) and stroke reperfusion was suboptimal or undetermined.
Results
We enrolled 25 patients (14 [56%] male, age 76 (interquartile range [IQR]70-85), baseline modified Rankin scale score 0 [IQR0-0], median pre-CT NIHSS 20 [IQR 13-25]). All patients had anterior circulation large vessel occlusion (LVO); 15/25 [60%] M1 middle cerebral artery (MCA) occlusion, 6 (24%) proximal M2 MCA and 4 (16%) ICA). Stroke etiology was predominantly cardioembolic (15/25 [60%]). Median DT>3 lesion volume was reduced by 18mL [2-48] following Trendelenburg compared with conventional horizontal CT positioning (114mL [94-204] vs 149mL [76-153] p=0.0027)). Systolic blood pressure was unaltered (mean 148mmHg (+/- standard deviation 29) vs 143 (+/-27); p=0.129). Head position did not alter clinical severity (post-CT NIHSS 13 [IQR 9-28]) in both positions). A significant lesion volume reduction with Trendelenburg positioning was seen in 15/25 patients (60%); 7 received continued Trendelenburg positioning (6 due to incomplete reperfusion following thrombectomy). Head down positioning caused no serious adverse events and was mostly well tolerated (6/7 [86%]).
Conclusion
Head-down (Trendelenburg) positioning appears to modestly improve penumbral perfusion in acute LVO ischemic stroke and is generally well-tolerated. Clinical benefits of this approach may be best tested in patients for whom reperfusion is delayed or not achieved.
Registered ANZCTR.org.au ACTRN12618000698279