You can't backtrack a specific cancer to a specific source. Skin cancer and lung cancer have famous "causes" but there statistical in nature with some people getting skin cancer despite minimal exposure to sunlight. Expose a billion people to low doses of a well known carcinogen and you expect a statistical increase in cancers, but there only estimates and it's impossible to track down specific people affected.
The question at the end of the day though is when we are going to pass regulation that a set of chemicals must be used who has the burden of proof to show that this is a safety gain? I think that the regulator has that burden, and if you have a sort of revolving door of chemicals used there (use it for a time throw it out because of health problems, move on to the next without testing) then you know you have a problem.
I know that this is hard, and it means effectively if there is such a burden of proof patents will expire long before it is met, but it's time to stop that cycle.