Probably not: the coax shield, inside the tube, makes another coaxial cable itself, and thus supports the TEM00 mode with propagation down to DC. The coupling might not be all that strong, say if the jacket is PVC (poor dielectric at microwave frequencies, strong absorption), and the impedance match might be random, but the coupling will definitely be nonzero.
It will be up to you, to calculate or measure the attenuation of this geometry, or to determine what attenuation or isolation is required for the application. Or what leakage is tolerable on the other side of this tube.
To prevent this, a bulkhead connector is needed, so that the coax shield becomes bonded to the chamber. If this isn't practical (e.g. isolation is required?), a common mode filter can be used. Which in this case, is just a normal mode filter, involving the coax cable's shield. Without breaking the cable open, its internal signal remains referenced to the inner shield. Put another way, the signal wire is perfectly coupled (to the extent the coax isn't leaky) to the shield, so to the extent the shield runs through a CLC(LC..) filter, the signal remains perfectly coupled through the chain of L's in that filter.
Capacitors aren't particularly easy to connect to an unbroken coax shield, while preserving good performance at microwave frequencies. Bonding is preferred.