Abstract T P377: Vitamin B12 Deficiency - A Poorly Recognised and Treatable Childhood Stroke Risk Factor
Introduction: Risk factors in childhood stroke are multiple and causality hard to determine. Current guidelines recommend prothrombotic workup, yet yield is variable, often without therapeutic consequence. Raised homocysteine is associated with vitamin B12 deficiency (B12), which is not routinely measured. Hypothesis: We hypothesised that B12 deficiency is a treatable risk factor for stroke in children not reliably identified by analysis of homocysteine only. Method: We retrospectively reviewed paediatric stroke patients admitted to a tertiary care hospital from 2010-2014. All patients with plasma homocysteine measured as part of their prothrombotic workup were selected. All clinical data closest to stroke diagnosis were reviewed. B12 deficiency was defined as low total and/or holo (functional) vitamin B12 with/without raised methylmalonic acid (MMA). Parental prothrombotic workup and B12 status was analysed where available. Results: Of 134 patients, 61 had homocysteine levels, 20 were tested for vitamin B12 (age range 0-14.8 years), 14/20 for MMA. B12 deficiency was found in 7/20 with median age of acute stroke presentation of 1 day (25th percentile= 0; 75th percentile= 51days). Median age of stroke presentation in B12 replete group was 3.5 years (25th percentile=2.0;75th percentile=7.0years). There was a statistical difference in age of stroke presentation between the B12 deficient and replete groups (Kruskal-Wallis, p=0.001). Raised homocysteine (10/20) was not sensitive in detecting B12 deficiency (sensitivity 42.8%, 95% CI 15.82-74.95). Two of seven had additional risk factors (1/7 iron deficiency, 1/7 MTHFR 667 homozygous). Seizures were the most common stroke presentation regardless of B12 status. Vitamin B12 (im) was given in 4/7 children and 3/9 parents (2 mothers, 1 father). Five of seven were breast fed. All treated parents were vegetarian. Conclusion: Our study suggests vitamin B12 deficiency is associated with early childhood stroke. Homocysteine alone is not a sensitive screen for this. Investigation with functional biomarkers such as holo B12 and MMA would allow for improved detection of a treatable risk factor in childhood stroke. Further studies may support this recommendation being added to guidelines.