Early Recurrent Stroke or Neurological Deterioration?
To the Editor:
The article by Coull and Rothwell1 calls for a standard definition of recurrent stroke. They suggest the definition “any stroke ≥24 hours after the incident event excluding early deterioration not caused by a stroke” as they emphasize the importance of using a definition that does not miss early recurrences. I heartily agree with the authors about the need of such definition. However, I do not think that their suggestion meets with what we need from the definition.
It is not clear if it is any stroke ≥24 hours after the onset of the incident event or ≥24 hours after neurological stability.
The new neurological deficit should in my view be quantified by a stroke scale.
Neurological deterioration/stroke in progression must be separated from stroke recurrence in a way that makes both biological and clinical sense.
Studies on neurological deterioration have described this phenomenon with different time frames, some up to 7 days. The causes of neurological deterioration are not well defined, but the clinical course has been linked to radiological findings including large infarcts, mass effect, and hemorrhagic infarction. 2 However, the classic interpretation that neurological deterioration is caused by continuous thrombosis in the same vascular territory still stands,3 even though it has not actually been demonstrated.
A practical approach to this problem could be a distinction based on time from stroke onset. A reasonable distinction could be 72 hours because at this time brain swelling as well as secondary hemorrhage in infarctions based on late reperfusion is most likely to have occurred.
In my experience, it is probably not possible to make a watertight distinction between neurological deterioration and recurrent stroke, which calls for some caution in working with this field.
I am grateful to Christensen for these comments. However, I do not agree that 72 hours should be required before a “recurrent stroke” can be said to have occurred. There is a very high risk of recurrent stroke during the first 72 hours after an initial stroke.1 Such strokes might well be preventable and it is essential that they are not excluded from risk estimates in epidemiological studies or from outcome events in randomized controlled trials.
I agree with Christensen that neurological deterioration caused by complications resulting from the initial stroke pathology should be excluded. I said in the discussion of our article, “it is, of course, important that progression of the initial stroke, hemorrhagic transformation of infarction, systemic disturbances, or edema and mass effect resulting in fluctuations in cerebral perfusion, are not misclassified as recurrent strokes.” However, the distinction between deterioration caused by such processes and that caused by a further episode of embolism or thrombosis should be left to clinical judgment rather than being based on an arbitrary time period. Moreover, it should be noted that early recurrence is most common after minor ischemic stroke,1,2 in which hemorrhagic transformation, edema, mass effect, and systemic disturbances are least likely.
A consensus is clearly needed before a standard definition of recurrent stroke can be determined. The aim of our article was simply to point out the problems inherent in the use of the current standard epidemiological definitions, which exclude events within either 21 or 30 days of the incident stroke, irrespective of the severity of the presenting event. I illustrated the differences in risk obtained using the definition: any recurrent stroke occurring >24 hours after the onset of the incident stroke, irrespective of vascular territory, that is considered to be a new episode of embolism or thrombosis rather than a deterioration caused by hemorrhagic transformation of infarction, systemic disturbances, or edema and mass effect resulting from the initial stroke. Similar definitions have been used previously.2–4
I also disagree with Christensen’s view that the definition should be based on a change in a particular stroke scale. This is unlikely to improve reliability and would introduce potential artifacts caused by intra-observer and inter-observer variability in administering such scales.
I did not address the important issue of recurrent strokes within the first 24 hours after a transient ischemic attack because epidemiological studies frequently have insufficient detail to determine the course of events reliably in the acute phase. However, this is an issue that also needs to addressed given the high frequency of such events5 and the potential for prevention by emergency intervention.
Coull A, Lovett JK, Rothwell PM, on behalf of the Oxford Vascular Study. Early risk of stroke after a TIA or minor stroke in a population-based incidence study. BMJ. 2004; 328: 326–328.
Lovett JK, Coull A, Rothwell PM, on behalf of the Oxford Vascular Study. Early risk of recurrent stroke by aetiological subtype: implications for stroke prevention. Neurology. 2004; 62: 569–574.
Petty GW, Brown RD, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers DO. Ischaemic stroke subtypes. A population-based study of functional outcome, survival and recurrence. Stroke. 2000; 31: 1062.
Kolominski-Rabas PL, Weber M, Gefeller O, Neundoerfer B, Heuschmann PU. Epidemiology of ischaemic stroke subtypes according to TOAST criteria. Incidence, recurrence and long-term survival in ischaemic stroke subtypes: a population-based study. Stroke. 2001; 32: 2735–2740.
Rothwell PM, Warlow CP. Timing of transient ischaemic attacks preceding ischaemic stroke. Neurology. 2004; In press.