@1.6GHz you should be able to contrive a simple shorted 1/4 wave stub as the DC injection choke, then cap, GDT and transorb the hell out of the DC end of the choke, that way the RF short circuit at the end of the stub is transformed to an open at the feedline at the operating frequency so the capacitance of the transorb on the DC side does not matter.
In fact a few such stubs along the line are an interesting possibility, take a 1/4 wave stub, short one end to the ground plane with a largeish cap, and place a transorb across the cap, the stub appears as a short circuit down near DC where the discharge energy is, so the transorb can work and appears as an open at resonance.
Low frequency energy picked up by the feedline will go straight thru the choke and be absorbed by the protection on the DC side.
1.6GHz is high enough that printing the 1/4 wave choke straight onto the PCB is practical.
Regards, Dan.