The first says it's basically holding for everyone but the very old who weren't well represented 100 years ago, and the second seems to show a slight decline in the past ~20 or so years with a peak in 1975 and holding almost steady after that. Basically what you would expect with improved diagnosis.
Has it? All I can find is
http://www.skepticalraptor.com/skepticalraptorblog.php/cance...
and
http://seer.cancer.gov/statfacts/html/all.html
The first says it's basically holding for everyone but the very old who weren't well represented 100 years ago, and the second seems to show a slight decline in the past ~20 or so years with a peak in 1975 and holding almost steady after that. Basically what you would expect with improved diagnosis.