You can't really do this. I'm not sure it makes much sense either - the issue is that the operating point you have at time 2ms will have a set of consistent voltages - if you sweep the temperature, you'd be expecting to see those voltages change - so what do the values at that time point mean any more? You'd end up with some kind of discontinuous behaviour at that time point.
So it's not possible and I don't think it even makes sense. You could vary the temperature (gradually) over time using dynamic parameters - but not do a dc at a time point. You can do an ac sweep at a time point (using acnames/actimes), but not a dc sweep.