Ah ok, already out.
In that case, I'd check for continuity map out continuity of the pins against the PCB footprint. If it's a 10.7MHz filter then it's probably an LC configuration.
Bypassing in-out pads on the PCB might give some clue, otherwise try a 10.7MHz ceramic filter as PA0PBZ suggests (not sure how lossy they are in comparison).
[EDIT] Hang on, checking the IC datasheets, aren't we past the IF strip by now - that's all handled in the LA1266, the LA3401 is the AM / FM switching and stereo decode. I don't think FL5 is 10.7MHz. Also leaky or reversed C56 would screw up the internal biassing of LA3401 pin 2. Worth a try.
[EDIT1] Have you checked that the AM/FM select pin (10) of the LA3401 is being driven correctly?