Naturally, I was curious, so I took a look. First, some background... PDF files are binary files (just like JPEG or MP3 files). Certain software will look at at the first few bytes of a file and attempt to determine whether it's binary or plain ASCII text. This is important when transmitting a file: in some contexts (like email), binary files need to be encoded so they're not damaged in transit. Despite being a binary file, the first few bytes of a PDF file might be all 7-bit ASCII, so Adobe strongly recommends adding a comment at the top, with a few non-7-bit characters, just to be sure that the file gets identified as binary. It doesn't really matter what those characters are, so long as their high bits are set. It appears that whoever wrote the PDF generation software built into MacOSX took the string "Derek's PDF", set all the high bits, and used that as the comment. I have no idea who wrote that software, but I'm guessing his name just might be Derek.
From my quick testing, this applies to any PDF file generated on MacOSX (using the built-in PDF generator). The White House PDF file was generated on OSX 10.6.7. I have an old Mac running 10.4.11, and the string is there too. (If you want to verify this, go find a Mac, load up a random web page or whatever, hit print, and then save to PDF.)
Alas, no conspiracy involved. It's just a cute little easter egg.
Anyway, the reporter asked if I could help decipher the remaining header info in that PDF file. I'm happy to say that I have a tool which can extract hidden headers from PDF files, which you can download here. You'll need a Windows PC (any recent version of Windows should do), and it runs from the command line, like this: "analyzepdfheaders c:\whatever.pdf". Let me know if you find anything interesting.