Parsing the wrong content stream in RadPdfViewer results in handled exception and this way some of the PDF content is missing when rendering the page.
Weirdly, Adobe Reader works fine with this.