There's a little discussion on that here
replace string in PDF document (ITextSharp or PdfSharp) - Stack Overflow[
^] - the code shown may or may not work ..
My approach would be 'different', and depends on how many document formats you have - note, under no circumstances just redact text by drawing/stamping a 'black box' over the text to be redacted, because the pdf document itself still holds the data, and a binary inspection could reveal the details.
I would parse all the text from a document
Read Text from a PDF in C# with iTextSharp – Chris Schiffhauer[
^] and build the redacted document from scratch - ok, its easy for me to say that, it depends how complicated your documents are