Impact of TCD Screening Protocol on the Incidence of Hemorrhagic Stroke in Children and Young Adults with Sickle Cell Disease

Background: Primary hemorrhagic stroke is a rare complication of sickle cell disease (SCD) that usually occurs in adults. The Stroke Prevention Trial in Sickle Cell Anemia (STOP) and Optimizing Primary Stroke Prevention in Sickle Cell Anemia (STOP 2) established routine transcranial Doppler ultrasound (TCD) screening with indefinite chronic red cell transfusions (CRCT) for patients with abnormal TCD as standard of care.  Despite a notable improvement in the incidence of infarctive stroke in children with SCD after the introduction of TCD screening protocols, it is unclear how this protocol will affect the rate of hemorrhagic stroke.  Presumably, early TCD screening and subsequent initiation of CRCT in high risk patients will prevent the progression of cerebral vasculopathy, which should decrease the risk of hemorrhagic stroke; however this has not been proven. Using the large multicenter cohort of children who participated in STOP and/or STOP 2 trials, we sought to assess whether the rate of hemorrhagic stroke was impacted by the use of TCD screening and/or CRCT.

Subjects and Methods: Between 1995 and 2005, STOP and STOP 2 (STOP/2) were conducted at 26 sites in the US and Canada.  These studies included 3835 children, ages 2 to 16 y with SCD type SS or S-beta-0-thalassemia.  Participation in STOP/2 required at least a single screening TCD for randomization. Patients on STOP 2 also had an observational arm for children started on CRCT who had an abnormal TCD.  The Post-STOP study was designed to follow-up the outcomes of children who participated in one or both of trials.  For all participants the date of their last encounter in STOP/2 was defined as the start of their Post-STOP period. 19 of the 26 original study sites participated in Post-STOP, contributing a total of 3539 (92%) of the STOP/2 subjects.   After exit from STOP/2, these children received TCD screening and treatment according to local practices.  Data abstractors visited each clinical site and obtained retrospective data from STOP/2 study exit to 2012-2014 including follow-up TCD and brain imaging results, clinical information, and laboratory results. Two separate neurologists, blinded to STOP/2 status and prior TCD and neuroimaging results adjudicated all suspected strokes. 

Results: Follow-up data were available for 2850 of the 3539 subjects (81%). Twelve children who had a stroke during the STOP study period were further excluded from this analysis resulting 2838 subjects. The mean age at the start of Post-STOP was 10.5 years and mean duration of follow-up after exiting STOP/2 until time of last medical encounter was 9.1 years. A total of 31 patients had a primary hemorrhagic stroke during the Post-STOP observation period (incidence 0.12 per 100 pt years). The mean age at time of stroke was 16.2+5.6 (median 15.3 range (4.8-30.2) years of age.   Of those 31 patients, only 52% had a TCD during Post-STOP prior to the event. Seven of those children who underwent screening had documentation of an abnormal TCD prior to the event (5 during STOP era and/or 3 Post-STOP). However, only 1/7 patients (14%) were documented on CRCT at the time of the stroke (4 patients were receiving HU and 2 patients had unknown treatment).

Discussion:  Although less common than infarctive stroke, patients with SCD are at increased risk for hemorrhagic stroke.  There is an increased risk of mortality for patients who suffer from hemorrhagic stroke (up to 26% in some reports in the 2 weeks after the event).  It is unclear if TCD screening and subsequent initiation of CRCT will impact the rate of hemorrhagic stroke in the long term.  In our results, a similar incidence of primary hemorrhagic stroke was noted although the patients were overall younger than previously reported (16.5+/- 5.5 years versus 20-29 years in Cooperative Study of Sickle Cell Disease).  Many patients who had a hemorrhagic stroke (48%) had not undergone TCD screening during the Post-Stop period. Additionally, although a safe stopping point for CRCT has not been established in patients who have had an abnormal TCD, only 1 patient was documented on CRCT at the time of the event (14%).  Thus, it is unclear at this time whether TCD screening and subsequent, lifelong continuation of CRCT could have prevented these other events. Clearly, these results demonstrate that improved implementation of STOP protocol is needed as well as further evaluation of the impact of this protocol on the incidence of hemorrhagic stroke.