Major vascular events after transient ischaemic attack and minor ischaemic stroke: Post hoc modelling of incidence dynamics

Background: Only few follow-up studies have compared the long-term risk of such major vascular events (MVE) as myocardial infarction (MI) and/or stroke following transient ischaemic attack (TIA) or minor ischaemic stroke (MIS). Estimates of relative risk and cumulative long-term occurrence of MVE may provide better information and contribute to the optimization of treatment decisions. Methods: In the current post hoc modelling study with unique data from Bulgaria, we analysed 183 consecutive patients with TIA (n = 89) or MIS ( n = 94), aged 1 40 years, who were prospectively followed over 36 months for non-fatal or fatal MVE. The cumulative survival, hazard and risks (with 95% confidence intervals) for MVE ( combined or by stratification) were calculated by Kaplan-Meier analysis and adjusted ( age, sex) by multivariate Cox proportional hazard models. A set of regression models was then applied to MVE incidence ( per 100 patients; 4-month intervals). Results: Median follow-up was 36 months (interquartile range 30.8 - 36.0); no differences by age or sex were found (p > 0.05). The risk of non-fatal or fatal MVE was approximately 28% ( stroke 19.7%, MI 8.2%). The adjusted cumulative risk of stroke was 0.21 versus 0.10 for MI. The odds ratio of TIA versus MIS was 0.75 ( 95% CI 0.43 - 1.32), i. e. lower for stroke ( 0.63, 0.31 - 1.25) than for MI ( 1.12, 0.40 - 3.14). The risk of non-fatal MVE was higher in MIS than in TIA (p(Breslow) = 0.0497), especially for non-fatal stroke ( p = 0.0325). Time series regression models provided best estimates of the different outcome dynamics in TIA versus MIS (R-TIA(2) = 0.969 with b(power) = 1.04 vs. R-MIS(2) = 0.989 with b(linear) = 0.84; p(1-tailed) = 0.04) over the study period. Conclusions: The age-and sex- adjusted cumulative 36-month hazard of MVE is higher after MIS than after TIA, but MVE fatality was higher after TIA than after MIS. Although stroke incidence was higher (up to 3 times that of MI), with the highest difference between months 8 and 18, MI fatality was always higher in absolute, relative or adjusted terms. The above alarming patterns and increasing, diverging tendencies for MVE indicate a higher long-term cumulative risk after MIS compared with TIA. These results confirm our hypothesis of a differential risk of TIA versus MIS and, at least, point toward equal importance of therapies aimed at preventing MVE in both types of preceding conditions (TIA or MIS) and the increased fatality after MI, in particular in patients with TIA. Copyright (c) 2008 S. Karger AG, Basel.