OBJECTIVE: To define risk factors associated with thrombosis in patients with antiphospholipid antibodies (aPL). METHODS: Ninety-nine patients with aPL, most of whom had prior thrombosis, were evaluated for the presence of acquired and inherited thrombophilic states. Genomic testing was performed for factor V(R506Q), 3' prothrombin (PTG) and methylene tetrahydrofolate reductase (MTHFR) polymorphisms. Clinical records were reviewed for the presence of acquired risk factors (RF) for thrombosis and events associated with aPL. Univariate statistical analysis was performed using Fisher's exact testing. A neural network statistical model was also used to identify which thrombophilic risk factors were most important in development of arterial and venous thrombosis. RESULTS: For arterial thrombosis, hypertension, tobacco use, hyperlipidemia, and diabetes mellitus were the most important predictors of thrombosis. By contrast, tobacco use, the 3' PTG and factor V(R506Q) polymorphisms, and previous cardiac surgery were the most important predictors of venous thrombosis. CONCLUSION: In this hypothesis-generating retrospective study, acquired risk factors were most important in arterial thrombosis, while the presence of factor V(R506Q) and 3' PTG polymorphisms were more important in the development of venous thrombosis. These findings are being validated in an ongoing, prospective study.
OBJECTIVE: To define risk factors associated with thrombosis in patients with antiphospholipid antibodies (aPL). METHODS: Ninety-nine patients with aPL, most of whom had prior thrombosis, were evaluated for the presence of acquired and inherited thrombophilic states. Genomic testing was performed for factor V(R506Q), 3' prothrombin (PTG) and methylene tetrahydrofolate reductase (MTHFR) polymorphisms. Clinical records were reviewed for the presence of acquired risk factors (RF) for thrombosis and events associated with aPL. Univariate statistical analysis was performed using Fisher's exact testing. A neural network statistical model was also used to identify which thrombophilic risk factors were most important in development of arterial and venous thrombosis. RESULTS: For arterial thrombosis, hypertension, tobacco use, hyperlipidemia, and diabetes mellitus were the most important predictors of thrombosis. By contrast, tobacco use, the 3' PTG and factor V(R506Q) polymorphisms, and previous cardiac surgery were the most important predictors of venous thrombosis. CONCLUSION: In this hypothesis-generating retrospective study, acquired risk factors were most important in arterial thrombosis, while the presence of factor V(R506Q) and 3' PTG polymorphisms were more important in the development of venous thrombosis. These findings are being validated in an ongoing, prospective study.