Introduction: Plasminogen activator inhibitor type 1 (PAI-1) is the most important regulator of fibrinolysis. Elevated PAI-1 levels are reported to be prothrombotic and have been associated with cardiovascular disease, obesity and cancer. PAI-1 synthesis is regulated by insulin, glucocorticoids and cytokines like tumor necrosis factor-α. The single base pair guanine deletion 4G PAI-1 genetic polymorphism has been linked to elevation of PAI-1 antigen and/or activity levels. Genotype frequencies in Caucasian populations have been estimated at 0.25 (5G/5G), 0.49 (4G/5G) and 0.26 (4G/4G). Increased thrombosis risk has been reported in patients with the combination of PAI-1 4G and factor V Leiden (FVL) genetic polymorphisms. Our pediatric specialty laboratory offers PAI-1 genotyping, PAI-1 antigen and PAI-1 activity assays as part of thrombophilia testing. The laboratory at the children's hospital is also a reference lab for the community. We retrospectively reviewed the results of PAI-1 testing over the last 10 years.
We hypothesized that the PAI-1 4G polymorphism would be more prevalent in young patients with thrombosis. We also hypothesized that there would be an additive thrombophilia risk from the combination of PAI-1 4G and FVL in young patients with thrombosis.
Methods: After IRB approval, using information from Cerner and EPIC, we created a deidentified thrombophilia database of 11,459 unique patients who had thrombophilia testing between March 8, 2007 and June 6, 2017. We identified 1,896 who had PAI-1 genotyping and 2,573 who had antigen and activity testing. Removing those <3 months of age for whom normal ranges for PAI-1 have not been established, we identified 1,418 subjects who had genotype, antigen and activity levels. We also evaluated a subset of 510 pediatric patients with thrombosis, of whom 338 had all three PAI-1 tests and were 3 months or older, and 292 who had both PAI-1 and FVL genotyping. Local genotype frequency was established by the clinical laboratory on 100 controls and was in agreement with published estimates. All comparisons were made using two-sided Fisher's Exact tests. Significance was assessed at the 0.05 level.
Results: Ages for subjects tested for PAI-1 ranged from 0.1 - 89.7 years with a mean of 14.3 years, and 41% were male. Compared to the local genotype frequency, PAI-1 4G was not significantly altered in either the thrombophilia cohort or in the subset with pediatric thrombosis (Table 1). This was true even when we re-evaluated by gender or by selecting for those with thrombophilia testing <22 years of age. Among 53 pediatric thrombosis patients with cancer, genotype frequencies for PAI-1 4G did not differ strongly from control group (p=0.14), but numbers were very small. Of the 1,418 thrombophilia testing subjects who had PAI-1 genotype, antigen and activity, correlation between abnormal antigen (p=0.07) or activity (p=0.28) levels and the 4G polymorphism did not reach statistical significance, though it did for those with both abnormal antigen and activity (p=0.05). Among 292 pediatric thrombosis patients with testing (Table 2), there was a suggestion of an effect from the combination of FVL and PAI-1 4G polymorphisms and thrombosis (p=0.06).
Conclusions: Abnormal PAI-1 antigen and activity levels are not solely dependent upon the 4G genotype and are likely influenced by other congenital and acquired factors. Among pediatric patients with thrombosis and particularly those with cancer and thrombosis, a possible effect from the PAI-1 4G polymorphism alone or in combination of FVL should be validated in a larger cohort.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.