My technique, developed for an Agilent 6890 with an OI concentrator, is to cool the line, concentrator 6 port valve and GC inlet. Disconnect on both ends (or include the 6 port valve) and flush with methanol followed by water. Heat the transfer line and flush with hot water. Reassemble and clean the sparge tube, the headspace above, and the inline filter in the concentrator. Do not flush a hot transfer line with a flammable solvent.
This, plus a new trap, will remove the carryover naphthalene from the nastiest creosote laden samples.
I have done the same cleaning procedure. Sometimes I will heat the line before doing the water, to do a "steam" cleaning on it.
Honestly though I usually just replace the transfer line with a new Silcosteel line. If you can use vespel ferrules in the fittings it is really easy if you use a section of the 0.53ID Silcosteel Hydroguard guard tubing as it is smaller outer diameter and slides through the transfer line heater much more easily than the standard 1/16 inch tubing does, and yet will carry the flow without much extra backpressure. The higher linear velocity also can reduce peak tailing sometimes.