Any long running transaction that caused the explosion in TempDB has already been killed by the system error produced when TempDB ran out of space. The chores now will be to 1) determine what code caused the problem, 2) fix TempDB so that it no longer consumes all the space (if it's on a separate drive, let it have all the space and no action required) and 3) fix the code so that it doesn't happen again.

If you right click on the instance in the Object Explorer and do the built in report for "Performance - Top Queries by IO", there's a pretty good chance that the code that caused the problem will show up as the first of the 10 queries reported on.