Late metabolic consequences of childhood cancer

Over the past decades, cure of children with cancer has become increasingly feasible. However, excess multimorbidity and mortality due to treatment are observed throughout adulthood. Metabolic syndrome (MetS), a risk factor for cardiovascular disease and diabetes, is more prevalent among childhood cancer survivors (CCS). Therefore, survivors need to be closely monitored, but MetS and the separate components can be underdiagnosed with the current criteria. Additionally, knowledge on national prevalence and risk factors remains incomplete. This thesis aimed to identify better methods to increase MetS diagnosis, and to describe prevalence and clinical and genetic determinants of MetS and the components overweight and dyslipidemia. In our single-center study among 103 survivors of nephroblastoma and neuroblastoma, we observed that MetS was present in 14% of survivors, and that 33% had at least two components. Abdominal radiotherapy was significantly associated with MetS and the components high triglycerides, low high-density-lipoprotein cholesterol and hypertension. We observed that four serum biomarkers may be useful for enhancing MetS diagnosis: adiponectin, LDL cholesterol, apolipoprotein B (apoB) and uric acid. We also found that vascular ultrasonography was not of additional benefit to improve diagnosis. In our systematic review on biomarkers we identified five biomarkers as useful for MetS diagnosis and prediction: uric acid, leptin, hsCRP, adiponectin and apoB. The level of evidence was high or moderate, mainly based on general population studies. Studies among survivors are scarce, so assessment of the use of these biomarkers in CCS is required. We described the methodology of the DCCSS-LATER Metabolic Syndrome study, among the Dutch cohort of 2,338 first treated (1963-2002) long-term CCS. In this thesis we presented the first results of the overweight component. Overweight occurred in almost half of all CCS. Compared to general population data, there was a significantly higher prevalence for overweight among women aged 50+ and for morbid obesity among men aged 50+. Overweight at diagnosis, cranial radiotherapy and growth hormone deficiency were associated with long-term overweight. Historic use of corticosteroids was not. Waist-circumference-based methods classified a substantial number of survivors with normal BMI as overweight. DXA scan identified overweight in an additional 30% of survivors, particularly those treated with abdominal radiotherapy, total body irradiation, anthracyclines and platinum chemotherapy. Adiponectin did not add diagnostic value. In the first genome-wide association study (GWAS) on the MetS component dyslipidemia in three large cohorts of CCS – CCSS Original, St Jude Lifetime (SJLIFE) and CCSS Expansion – none of nine SNPs that were identified in discovery analysis of the entire cohort, replicated. Next, we identified rs114017774 on chromosome 2 as a potential genetic variant for dyslipidemia, specifically in cranially irradiated CCS. Meta-analysis of the CCSS Original and SJLIFE cohorts revealed a pooled odds ratio of 11.30 (95%CI=5.03-25.40, p=4.5x10-9), but this locus did not individually replicate in the CCSS Expansion or SJLIFE cohorts. Additional replication and functional validation are needed. Furthermore, we replicated the protective variant rs676210 in the APOB gene, which was identified in a previous study among survivors of acute lymphoblastic leukemia, in the CCSS Original cohort