If the voltage is kept low (below the ionization energy), a gas can be used. This is the basis of some new "vacuum tube" research, using field emission cathodes and sometimes no vacuum (as such) at all.
It can also be made smaller than the mean free path at ambient pressure, which is a necessary feature of any vacuum tube -- since the classics are macroscopic, a very good vacuum is needed (mean free path of ~mm). Microscopic ones can operate at higher pressure.
And of course, some tubes did use gas, usually for the conductivity enhancement that ionization brings. This ranges from gas rectifiers and thyratrons, to even just phototubes (where the ionization cascade has much the same effect as an avalanche photodiode).
Tim