If t and k are ligatures (which is quite possible), then perhaps the VMS q symbol is just a downshifted copy of the left part of the t ligature.
In which case, it may not be related to q at all, the similarity may be coincidental (or the down-shifting might be deliberate in order to make it look like q and obscure the fact that it's related to t).
In which case, it may not be related to q at all, the similarity may be coincidental (or the down-shifting might be deliberate in order to make it look like q and obscure the fact that it's related to t).