Risk of a second cancer in Canadians diagnosed with a first cancer in childhood or adolescence

BACKGROUND: Second cancers are an adverse outcome experienced by childhood cancer survivors. We quantify the risk and correlates of a second cancer in Canadians diagnosed with a first cancer prior to age 20 years. METHODS: Using death-linked Canadian Cancer Registry data, a population-based cohort diagnosed with a first cancer between 1992 and 2014, prior to age 20 years, were followed for occurrence of a second cancer to the end of 2014. We estimate standardized incidence ratios (SIR), absolute excess risks (AER), cumulative probabilities, and hazard ratios (HR). FINDINGS: 22,635 people contributed 204,309•1 person-years of follow-up. Overall risk of a second cancer was 6•5 (95% CI: 5•8–7•1) times greater than expected resulting in an AER of 16•5 (14•4–18•5) cancers per 10,000 person-years and a 4•8% (3•8%–6•0%) cumulative probability of a second cancer at 22•6 years of follow-up. SIRs decreased with increasing age at diagnosis and time since diagnosis; were larger in more recent calendar periods of diagnosis; and varied by type of first cancer. Large SIRs in the first year after diagnosis and in those diagnosed in 2010–2014 were partly associated with changing registry practices. For the whole cohort, factors associated with the hazard of a second cancer included: being female vs. male [HR = 1•439 (95%CI: 1•179–1•760)]; being diagnosed in 2005–2014 vs. 1992–2004 [2•084 (1•598–2•719)]; having synchronous first cancers [4•814 (2•042–9•509)]; and being diagnosed with certain types of cancer. Factors varied, however, by type of first cancer. INTERPRETATION: Risks of a second cancer are not equally distributed and can be impacted by changes in registry practice and the methods used to define second cancers.