Cumulative incidence of second primary cancers in a large nationwide cohort of Danish cancer survivors: a population-based retrospective cohort study
Trille Kristina Kjaer,Elisabeth Anne Wreford Andersen,Giske Ursin,Signe Benzon Larsen,Pernille Envold Bidstrup,Jeanette Falck Winther,Michael Borre,Christoffer Johansen,Susanne Oksbjerg Dalton
DOI: https://doi.org/10.1016/S1470-2045(23)00538-7
Abstract:Background: A new primary cancer is a serious late effect of a pre-existing cancer diagnosis, and can be attributed to hereditary cancer syndromes, immune or hormonal factors, cancer treatment, or modifiable lifestyle or environmental factors. We investigated the absolute and relative incidence of second primary cancers in a large cohort of Danish cancer survivors. Furthermore, we examined the association between alcohol-related, smoking-related, virus-related, and hormone-related first and second primary cancers. Methods: In this retrospective cohort study, we identified a cohort of Danish adults (aged ≥40 years) diagnosed with cancer from Jan 1, 1997, to Dec 31, 2014 and alive 1 year after diagnosis. Follow-up was from date of first cancer diagnosis and lasted up to 24 years, ending on Dec 31, 2020. Cohort identification and information on second primary cancers was obtained from the Danish Cancer Registry, and comorbidity and sociodemographic information was obtained from Danish population-based registries. Overall, and for 27 cancer types, cumulative incidence functions and Cox proportional hazard regression models were used to estimate the incidence of second primary cancer and death, and hazard ratios (HRs) and 95% CIs of second primary cancer adjusted for sex, age and year of diagnosis, cohabitation status, income, and comorbidity. Findings: 457 334 Danish adults were included in our study (230 150 [50·3%] male individuals and 227 184 [49·7%] female individuals; median age at diagnosis 68·3 years, IQR 59·7-76·6; median follow-up 3·6 years, IQR 0·6-9·3). The cumulative incidence of second primary cancer increased over time from 6·3% (95% CI 6·2-6·4) 5 years after diagnosis to 10·5% (10·4-10·6) 10 years after diagnosis and to 13·5% (13·4-13·7) 15 years after diagnosis. The highest cumulative incidence of second primary cancer 10 years after diagnosis was observed in survivors of cancers in the larynx (21·8%, 20·5-23·1), oropharynx and oral cavity (19·5%, 18·7-20·3), and bladder and urinary tract (18·5%, 18·0-19·0). Survivors of cancers related to alcohol (HR 1·09, 95% CI 1·06-1·13), smoking (1·73, 1·68-1·78), diet high in red or processed meat (1·32, 1·24-1·39), or virus (1·23, 1·13-1·35) were at increased risk of developing a second cancer with the same aetiology, whereas having had a hormone-related first cancer was associated with lower risk of a second hormone-related cancer (0·77, 0·73-0·81). Interpretation: Our results could help optimise prevention efforts targeting modifiable risk factors to reduce risk of developing a second primary cancer. Funding: Nordic Cancer Union and The Health Foundation (Helsefonden).