We can`t say that there is an incubation period of cancer. All we can say is that there is a period where cancer is not big enough to produce symptoms, but the disease is already there. Certain tumors, like pancreas adenocarcinoma produces change in life even before becoming really symptomatic Depression which may precede for many months other symptoms of pancreas cancer is perhaps the result of some paraneoplasic phenomenon to which we may find an answer in the biochemistry field in the future. The real changes in a person's life start when he gets to know that he has a cancer. Is the knowledge of what he (or she) has what elicits the changes.
i have the same caveat as Tomas. Incubation period is used in regard to infections. Clearly it covers the time between the initial exposition to e.g. a virus and the first symptoms, or responses that can be used to tell wether the infection has indeed followed the exposition. E.g. in HIV secondary prevention programs, blood tests can be positive before first symptoms show and thus further spreading can be prevented.
So in cancer there is no infection that precedes the disease as the main aetiological cause. There are often early stages that can be used for screening and early therapy. E.g. cervix carcinoma.
You see in that example that early stages do not affect the patient at all.
In anthropology, at least, how you analytically deal with the period before noticeable symptoms appear, should come out of your data. If you do research about cancer patients, you can only talk to them once they have learnt what they are suffering from. And then, they will construct their past (i.e. the period you are interested in) in relation to their knowledge of suffering from cancer. Therefore, naming that period in a positivist manner prior to research does not really make sense if your analytical stance is an actor-centered one.