I'm not quite sure as to why I have a year missing in France (maybe it was a strike year ...). Still, there does seem to be massive volatility in lung cancer over time and space. It's most likely that my code is wrong but over time my code will be right. My main probem at the moment is that ICD codes are a little opaque - see here. "Deaths" in the charts should be "Deaths per 100,000".
My aim in the future is to recreate this map using the CDC data but as this will be a little more complicted, I shall leave it until later.