With reference to TEXTUTIL(1) manual page it seems pdf is not between formats managed by this utility:
fmt is one of: txt, html, rtf, rtfd, doc, docx, wordml, odt, or webarchive
On Linux/Unix installing a scriptable product like XPDF/pdftotext may be a valid solution like some comment already suggested.
For those on OS X it's possible extract text from PDF by a native OS X automator action ( ..see this answer or last 4' of this tutorial) then consider that automator's workflow can be "scripted" via CLI automator command