Perhaps you already know this, but there are 2 key reasons why you might be seeing "hugely" varying reported outcomes...
First, the big-picture landscape has shifted dramatically
over time. Just a generation ago, the typical newly diagnosed case was already advanced (no longer local), and so the overall survival of all men newly diagnosed with PC was much worse. The "PSA-era" which was ushered in during the period of that last generation has shifted the profile of the "typical" newly diagnosed man to a much earlier stage...and correspondingly overall survival of all men newly diagnosed in the current decade is dramatically different than 2 decades ago.
But you don't really care—for the purposes of your question—about
the "typical" newly diagnosed man, and that brings us to the second point. Prostate cancer is really two, or three, or more different diseases along a spectrum. You probably know this, but long term survival at one end of the spectrum is so high that it makes no difference whether a man seeks treatment or not. And the thing is that there are SO MANY of the lower risk cases newly diagnosed in comparison to the much smaller numbers of high risk cases (today, only about
4% of newly diagnosed cases are distant mets) that these don't-bother-treating cases skew the combined statistics of ALL newly diagnosed men towards the "good side." Those skewed overall numbers today are 99+% survival for 5 years.
For the much smaller number of contemporary
newly diagnosed advanced cases, the numbers are obviously not as good as the don't-treat cases the the heavily skewed overall set of cases. So the data you seek is the survivability of men initially diagnosed with advanced disease. I have that info, published by the American Cancer Society (previously provided in another thread by fellow HW/PC member George_). What's NOT available is the variations on these numbers if you do this, that or the other treatment modes—and that is generally because those variation tend to make little general difference, once the disease is established as distant.
(Sorry to walk you through all that background...you probably already knew that, but let's recognize that some others may read this now or in the future without the background, and we don't want anyone to misunderstand the basis.)
So ACS reports that using contemporary data (2005-2011; I wouldn't expect today to be much different), the 5-year
relative survival rates are as follows for those initially diagnosed with Localized, Regionalized and Distant disease:
- Localized = >99%
- Regional = >99%
- Distant = 28%
Here's a summary chart which George_ originally posted: LINK
And here's the whole ACS document from which he got that chart: LINK
Is this what you were looking for?