Strange encoding of non-latin characters in Microsoft Word metadata embedded in a PDF file

I have a PDF file created from a Microsoft Word document that contains a block:

20 0 obj
<<
/Author (^P=4@59 ^S025=:>)
/Producer (Microsoft: Print To PDF)
/Title (Microsoft Word - ^T>:C<5=B1)
>>
endobj

I suppose this is Microsoft Word metadata. This is not a part of the info dictionary that can be displayed with pdfinfo or any PDF-reader. I wonder how to decode these fields. I assume that these are Cyrillic letters.

Update: I could find a couple of word documents on the web with the same text in the Title field, so this might be just a signature of some pirated version of Microsoft Office.