Project Summary
Recent clinical trials of platelet-inhibiting medications for preventing a second stroke following an initial minor
stroke or transient ischemic attack (TIA) have shown that recurrences of strokes over time follow a distinct time
course, with a high rate of recurrence over the first few days, slowing to a steady lower rate over subsequent
months. We have shown that this temporal course can be predicted from a model of stroke recurrence risk
based on kinetic analysis, in which rates of events are proportional to numbers of subjects in a given state
multiplied by a fixed kinetic rate for that state. A kinetic model that postulates both a transient vulnerable state
and a stabilized state following initial ischemic events predicts a specific mathematical form for the time course
of the event-free survival curve (the proportion of subjects not yet having experienced a recurrence, over time).
The predictions of this model closely match the data from the POINT, SOCRATES, and THALES trials, and
allow for estimation of the kinetic rate constants that describe the rate of ischemic stroke recurrence in the
vulnerable state (k1), the rate in the stabilized state (k2), and also the rate of transition from the vulnerable to
stabilized state (k0).These different rates that determine risks of stroke recurrence are likely to relate to
different underlying biological mechanisms, and to be distinct in their responses to different treatments. In fact,
we have shown that the added anti-platelet medication strategies used in these trials only affected only k1, not
the other rates.
Understanding the condition of patients following stroke in these kinetic terms of distinct vulnerable and
stabilized states, subject to different biologically-determined rates of stroke recurrence, allows for exploring
what patient risk factors or treatments may selectively affect the different rates. These insights have
implications for how clinical trials of treatments should be designed. For instance, if a certain type of treatment
only affects k1, the rate of stroke recurrence in the transient vulnerable state, then it is only likely to make an
impact when applied rapidly and for the short-term following the initial stroke. In contrast, interventions
affecting the recurrence rate in the stabilized state will need to be applied for longer periods to impact overall
stroke recurrence. Determining the distinct effects of risk factors or treatments on the individual rate constants
requires data from large numbers of patients to quantify the effects. By combining the data from the large
SOCRATES, POINT, and THALES trials into a unified and harmonized data set, these types of careful
analyses can be performed, testing important questions as to which kinetic rates are specifically affected by
patient risk characteristics, by treatments such as cholesterol-lowering medicines or anti-inflammatory
medicines, or by the presence of significant plaque and stenosis in intracranial arteries. These important
insights, with implications for the design of future clinical trials, are the goal of this project.
