I dont have much experience with the CAT reader, but, according to the documentation it does not seems to be possible....
After reading the file, can you then use a TextDecoder or TextEncoder to properly convert to UTF-8?
Another option woule be to use the Textfile reader instead of the CAT - reader, and specify the Character encoding. And afterwards one SubstringExtractor for each column to get. Does not work with Generic Readers - however if you always have the same formatted input-file it will work.
Also you can request safe to include the Character Encoding option on ALL NON-BINARY files READ and WRITE, not sure if that is already a PR.