IV tPA for stroke works the same way that IV tPA works (at least to some degree) for STEMI. The half life of tPA is short, on the order of minutes, but the fibrinogen depletion (and coagulopathy) that occurs as a result lasts many hours. So even by administering IV, it still works systemically.
IA tPA must be delivered from a proximate source, so the only way to do it is via catheter-based intervention at the level of the occlusion. We certainly do this, but it is clearly a more invasive intervention than IV tPA. As previously noted, we are not clear currently on how to identify the best patients for such therapeutics, and so IV tPA remains the therapy of choice, aside from clinical trial enrollment at some centers (DIAS-4, MR-RESCUE, etc.)
We are all sort of on hold until the IMS-3 results are formally announced.