tPA treatment for ischemic stroke

This review from the Journal of Emergency Medicine (full text via Medscape) may be the best evidence synthesis available on the topic. It deals with all the studies and focuses both on the use of tPA in general as well as the new extended window. Concerning the latter question here's the author's bottom line:

The publication of the ECASS III trial has produced much excitement and discussion in the Stroke community. The Heart and Stroke Foundation of Canada and the European Stroke Organisation have both recommended treatment with IV tPA up to 4.5 h from symptom onset for appropriate patients. At this time, treatment with IV tPA for acute stroke beyond 3 h from symptom onset remains without FDA approval, but it has been endorsed by a Scientific Advisory from the American Heart Association Stroke Council.

The paper notes that many tPA associated head bleeds may be of little clinical importance, partly because those who suffer them tend to have large infarcts already destined for bad outcome. This was illustrated in a 2007 paper cited by the author:

Background and Purpose—A clinically relevant number needed to harm for tissue plasminogen activator (tPA)-related symptomatic intracerebral hemorrhage (SICH) would greatly assist therapeuticdecision-making.

Methods—A 15-variable prognostic model was derived from a placebo group enrolled in National Institute of Neurological Disorders and Stroke tPA Trials 1 and 2 and used to predict final global disability outcome for patients with tPA-related SICH had they been treated with placebo, rather than tPA, and not experienced SICH.

Conclusions—Most patients who experience SICH have severe baseline infarcts and already are destined for poor outcomes. For every 100 patients treated with tPA, approximately 1 will experience a severely disabled or fatal final outcome as a result of tPA-related SICH.

In other words the clinically relevant number need to harm is higher than one might expect.

But there's more than one way to interpret the NINDS data as illustrated by this recent analysis:

Methods
We used the original data from the NINDS trials to create graphs showing the effect of treatment on neurologic function in all 624 individual patients in the trial. Our goal was to show detailed graphics of the 90-day outcomes, stratified on relevant confounders and effect modifiers.

Results
Final outcomes were highly dependent on stroke severity. In many graphs, the small difference between groups favored tissue plasminogen activator, particularly when baseline NIHSS score was between roughly 5 and 22. These differences diminish or disappear when 90-day change in NIHSS is graphed. Our graphs fail to support the time-is-brain hypothesis.