It is quite easy to produce a modified pdf of a letter.
Scan the letter, now you have an image file e.g. a JFIF(Jpeg).
Modify the image to your hearts content, then turn your finished product into a pdf.
But why go to the extra trouble to turn a scan into a image file, then into a pdf, when simply putting the document up on your web site as a Jpeg would do? Maybe to make it seem authentic?
The pdf document, in Russian, appears to have been created from a JFIF(Jpeg) source file.