I am assuming the chromatography solvent is keeping the MS clean so no problem there (if someone knows whether this is wrong, please correct this assumption). Then it is a question of protecting the LC column. Two ways, filter all samples as Doc Oc suggested or guard column or per-filter. I lean more for speed and toss guard column or filter. If you take an LC aliquot, how much palladium will be in your sample?
If this is an open access machine, then follow the rules of the person that maintains it. If it is a group instrument, then it depends on whose time is most valuable, yours for filtering every sample or whether you can afford to toss a guard column every X injections.