Cascode?
2V BVceo is damn wimpy. Must be a high-zoot SiGe or
sump'n.
Maybe you just want to ensure that the base impedance
is adequately low, and push it toward BVces. But this is
probably outside reliability ratings and if you have data
to go on, best be sure it covers high temp (worse) and
process outliers. We used to respect the BVceo as the
"no worries" limit and very rarely tried to push it, in the
high-rel business. If it's just gotta last until the lab grade
comes back, then maybe you just go for it.
But I think you ought to look at cascode topologies. Just
make real sure your Zbase is very low on the upper
device because it'll act quasi-inductive if not.