Pseudoclonality in cutaneous pseudolymphomas: a pitfall in interpretation of rearrangement studies