When Finnish geneticist Leena Peltonen-Palotie was diagnosed with rare sarcoma in 2008, her colleagues started big project to save her life. Her cancer became the "worlds most studied cancer" for a short period. They sequenced her genes and used large screening robot to test tens of thousands combinations of different drugs against tissue samples taken from her.
They actually fond a cure for the sarcoma she was diagnosed with, but the cancer had already mutated into a form that did not respond to the treatment and she died two years later 2002. I believe this might have been the first ever for this kind of large scale tailored cancer research. Weirdly enough I can't find any mention of this research effort in English speaking magazines.
One could also cite Steve Jobs. Despite heroic measures, his delay doomed him.
(To expand a little bit on this: the Gompertz curve means that even if a 85yo billionaire contracts some cancer and is able to buy a cure no one else can which isn't useless or iatrogenic, he is going to die very soon anyway as the annual mortality risk increases exponentially. This is what is behind those surprising observations like 'curing all cancer would only add a few years to the average life expectancy' - curing cancer just means that you die of dementia, Alzheimers, a heart attack or something else a year or two later.)
I collected all Wikipedia articles on American billionaires who died in 2008 - 2017: http://tools.wmflabs.org/dschwenbot/intersection/index.php?l...
That got me 74 results. I wgot the articles and manually removed the women based on related Wikipedia categories (turns out https://en.wikipedia.org/wiki/Aubrey_McClendon was a man, I would have removed him just based on the name).
Then I grepped the files for '"[0-9]\{4\} births"' to get the year of birth (this string marks the corresponding category). Two men didn't have a known birthdate, so I threw out their articles, too. This left me with the years of birth for 65 American billionaires. I got their years of death analogously.
After that I calculated their ages as the difference (yes, this could be off by +-1 year, but my method is not exactly rigorous anyway). The mean of these values is ~82.6. 29 of them are >= 86. The mean of those is ~91.3, or 5.3 years past 86. Considering the huge error bars, this is well in line with the general population average.
