Elucidating The Thrombogenic Risk Of Non-O Blood Groups In Children With Acute Lymphoblastic Leukemia (All): Is Von Willebrand Factor A Culprit

Background: Individuals with non-O blood group are shown to have increased risk of thromboembolism (TE). The exact pathogenesis of the prothrombotic effect of non-O blood group, is however not known. Because individuals with O-blood group have low levels of von Willebrand factor (vWF) compared to those with non-O blood group, vWF has been contemplated as a pathogenetic mechanism in ABO blood group-related prothrombotic risk. However, the available data regarding the role of vWF in the thrombotic risk of non-O blood group are inconclusive. Children with acute lymphoblastic leukemia (ALL) are at increased risk of TE. Several factors such as older age, leukemia phenotype and asparaginase have been shown to impact the risk of TE in children with ALL. We have recently shown that non-O blood group and circulating blasts were significant risk factors for TE in children with ALL. We have also shown that at diagnosis of ALL patients with circulating blasts have significantly higher levels of vWF compared to those without circulating blasts.  Within the context of a larger study aimed to define risk factors for symptomatic TE (sTE) in children with de novo ALL, we undertook a sub-study to evaluate the relationship of ABO blood groups and vWF level at diagnosis of ALL, and to evaluate the impact of circulating blasts on the vWF levels in children with O and non-O blood groups. We hypothesized that compared to patients with O-blood group, those with non-O blood group will have significantly higher levels of vWF and that circulating blasts will have additive effect on the vWF levels in patients with non-O blood group.  Methods : The multicenter, prospective, analytical cohort study included consenting patients (1-≤18 yrs. of age) with de novo ALL enrolled on the Dana-Farber Cancer Institute  05-001 therapeutic trial. Details of patient demography including ABO blood group, ALL diagnosis, therapy and symptomatic TE (sTE) were collected. Samples collected prior to starting ALL-therapy were analyzed centrally for prothrombotic defects (PD) including [protein C, S, antithrombin, Factor VIII:C, vWF, anticardiolipin antibodies and gene polymorphisms of methylene tetrahydrofolate reductase C677T, prothrombin G20210A, Factor V Leiden]. Age-adjusted standardized laboratory data defined PD. Regression analyses evaluated relationship between risk factors and sTE. Thrombosis-free survival was estimated using Kaplan-Meier method.  Results : Of 131 enrolled patients [mean age (range) 6.4 (1-17) yrs.; 70 boys], 21 (16%) developed sTE. ABO blood group information was available for 127 patients; 51 patients had blood group O and 76 non-O (57 with blood group A, 15 with B, and 4 with AB). There was no impact of PD including vWF on the risk of sTE. Older age compared to age ≤ 5 yr. [Odds Ratio (OR) 1.9, p=0.029] and non-O blood-group (OR 4.27, p=0.028) compared to O group were identified as independent predictors for development of sTE. Patients with peripheral blasts had higher odds of developing sTE (OR 7.79; p=0.059).The sTE-free survival was affected by older age (Hazard ratio (HR) 1.1, p 0.03), ALL risk type (HR 3.0, p 0.025) and blood group (O blood group vs non-O blood group, HR 0.23, p 0.03). Table 1 compares the vWF levels in patients with O and non-O blood group and those with and without circulating blasts. Overall, there was no difference in the vWF level at ALL diagnosis between patients with O vs. Non-O blood group. Patients with circulating blasts had higher levels of vWF at ALL diagnosis compared to those without circulating blasts; this comparison was statistically significant for non-O blood group. However, there was no interaction between ABO blood group and circulating blasts on vWF levels (p=0.723)  Conclusion : There was no effect of blood group type on the vWF level at diagnosis of ALL. Patients with circulating blasts had significantly higher levels of vWF at ALL diagnosis and the vWF levels were significantly higher in patients with non-O blood group and circulating blasts. Although it is likely that the relationship between blood group and vWF may be affected by effect of circulating blasts on vWF, we showed no interaction between ABO blood groups and circulating blasts on the vWF levels at diagnosis of ALL in children. Small sample size is a limitation of current study. Further studies with larger sample size are needed to elaborate the relationship between vWF, ABO blood groups, and circulating blasts. Disclosures No relevant conflicts of interest to declare.