I will treat most patients if their prior stroke occurred >3 weeks ago. This is not a typo. There is no evidence for the 3m cut-off that I am aware of. Like many of the "contraindications" to IV tPA use, the <3m Hx of stroke was an arbitrarily selected criteria from the initial trials that was carried forward into the FDA approval parameters. It is unlikely that after 3 weeks of healing there would still be sufficient abnormalities of vascular wall integrity, so I reason that it shouldn't raise the risk of ICH. Its a much easier decision with a prior lacune than an LVO--given they are much less likely to involve multiple larger vessels, but realistically the odds of having a pt s/p large-vessel occlusion re-present with different symptoms from a new infarct so soon after the last one are very low. Also, we re-start anti-coagulation after ICH in pts with mechanical valves on average after 10d...makes sense to me that if this is safe, why not lytics after a longer healing period?
I'm admittedly very aggressive in my approach to tPA use, but even so have yet to have a symptomatic hemorrhage after tPA (after >50 uses). As with any consideration of off-label use you have to do your best to educate the pt/family on the lack of data that apply to their particular situation. I've never regretted opting to treat a patient with tPA, but often have looked back and wished I had treated after a pt went on to worsen a "mild" stroke.
You'll never get sued for treating per the manufacturer's instructions, but I think that it is really unfortunate that so many of the so-called contraindications have no data to support them as reasons not to treat...lots of patients out there who might otherwise have a better shot at good outcomes don't get treated.