As George stated, you OCR the scanned image of text.
If you use Searchable Image or Searchable Image (Exact) you can hack/slash at making the hidden OCR text visible and then attempt to edit. Most times the PDF will end up unusable. After all, the *intent* of either Searchable Image OCR is to support some measure of Find / Search - no more, no less.
If "edit" is a must then use ClearScan and export to Word/RTF.
Cleanup with MS Word and make a new PDF.
If minor touchups are needed they may be done, in Acrobat, to the ClearScan output.
There are some related tutorials available here within the Learning Centers.