Without cherry picking data. Compare cancer rates in non vax to vax would be a start, controlling for all other cancer causing factors (diet, environment, age, other diseases etc). This would be a start, I am sure an epidemiologist would have additional dimensions to consider.