Aims: To find genetic predictors of reocclusion after successful fibrinolytic therapy during the acute phase of ischemic stroke. Patients & methods: This was a case-case prospective study analyzing 236 polymorphisms in a cohort of 222 patients treated with tissue plasminogen activator, from which 16 patients suffered a reocclusion event (7.2%). A predictive scale was generated using independent polymorphisms with a dominant/recessive model and tandem occlusion, weighted by their-coefficients in logistic regression. Results: Using a dominant/recessive model, the rs1800801 SNP from the MGP gene (odds ratio [OR]: 15.25; 95% CI: 2.23-104.46; adjusted p = 0.006) and the rs1883832 SNP from CD40 gene (OR: 0.077; 95% CI: 0.009-0.66; adjusted p = 0.019) were independently associated with reocclusion after logistic regression adjustment by clinical predictors. In an additive model, only the rs1883832 SNP (OR: 4.43; 95% CI: 1.62-12.15; adjusted p = 0.004) was related to reocclusion occurrence. The predictive model that was generated stratified the reocclusion risk from less than 1% to more than 70%. Reocclusions were associated with neurological worsening at 24 h (patients with reocclusion: 26.7%, versus patients without reocclusion: 4.9%; p = 0.002), as it was seen for MGP-7A>G (AA: 17.2% vs AG+GG: 4.5%; p = 0.027), but not for CD40 1C>T (CC: 4.5% vs CT+TT: 7.7%; p = 0.565). There was an association between CD40-1C>T genotype and CD40 transcriptional activity in peripheral blood mononuclear cells (median expression values TT: 65.75%, CT: 70.80%, CC: 96.00%; p = 0.023). However, CD40 soluble fraction was not a useful biomarker of reocclusion status. Conclusion: An association was found between MGP-7A>G and CD40-1C>T polymorphisms, and reocclusion risk. The predictive scale that was generated permits the stratification of patients by their reocclusion risk with higher accuracy than clinical parameters alone. © 2010 Future Medicine Ltd.
- Matrix-carboxyglutamic acid protein
- Tissue plasminogen activator