Once I knew that my cancer was invariably fatal, the most pressing question was “How long?”. I got some averages from the urologist, but those numbers were quite unsatisfying, so I researched a bit and created this graph.
(click the image to enlarge)
This graph is not the whole story, but let’s start with what it does tell us.
The red line labled “still alive (actual)” is what percentage of people are still alive after some amount of time.[note 1] “still alive (interpolated)” used the average 6 month survival rate to project into the gap between the 5 year data point and the 10 year data point. It fits with the actual survival curve. The difference at the 10 year data point is 8.1% vs 10%.
The green line is the probability of being alive in another 6 months, assuming you are still alive at the time where the dot is located. I computed this from the “still alive (actual)” numbers. I also plotted a line of the average 6 month survival, which I used to interpolate the valuse from year 5.5 to 10.
The blue lines are survival for 5 years and 10 years, assuming you are still alive at the time where the dot is located. The 10 year line has a discontinuity because the first part is based on the actual data and the second part is based on the interpolated data.
So, starting at zero, I had about a 34% chance of seeing the next US President, and about a 25% chance of seeing the launch of the James Webb Space Telescope.
Of course, these are just probabilities derived from a whole population of cancer patients. To understand your own survival from this graph, you need to remember two important things:
1. The average survival for the next 6 months is about 88% no matter which 6 month interval you are in. Another way to look at the 6 month survival is to convert it to some other interval: 77% for 1 year, 60% for 2 years, 50% for 2.5 years.
2. This graph does not consider the two phases of metastatic prostate cancer: At first the most effective treatment is to suppress the hormone testosterone; after a time, the cancer becomes “castrate resistant”, and that is the beginning of the end, but the end is not necessarily instantaneous. The survival graph for after the cancer becomes castrate resistant (which I have not included, but you can see in footnote [note 2]) shows a probability of several more months of survival. If your cancer is not castrate resistant yet, it is pretty unlikely to kill you tomorrow.
For me, 0 is about February 2013. As of this writing, I’ve made it through the first year, and the cancer is not yet castrate resistant. That’s good news so far.
 The first 5 years of this data came from Cancer Treatment Centers of America, which, at the time I was collecting this data, had a web page comparing survival of their patients to the the US NCI SEER data for metastatic prostate cancer. The last data point (at 10 years) came from the UK National Health Service. If I were to publish this graph in a peer-reviewed journal, this is where I would give you actual references, but I did this online and did not keep track of the URLs.
 there is an interesting graph on page 5 of this paper: http://clincancerres.aacrjournals.org/content/16/1/203.full.pdf
 I have the raw numbers I used for the graph, but not in a format that WordPress will allow me to upload. Leave a comment if you are interested.