A genomic approach to study down syndrome and cancer inverse comorbidity: untangling the chromosome 21

The sequencing of chromosome 21 and the use of models of Down's syndrome in mice have allowed us to relate genes and sets of genes to the neuropathogenesis of this syndrome, and to better understand its phenotype. Research in prenatal screening and diagnosis aims to find methods to identify fetuses with Down's syndrome, and reduce or eliminate the need for amniocentesis. Other areas of active research and clinical interest include the association of Down's syndrome with coeliac disease and Alzheimer's disease, and improved median age of death. Medical management of the syndrome requires an organised approach of assessment, monitoring, prevention, and vigilance. Improvements in quality of life of individuals with Down's syndrome have resulted from improvements in medical care, identification and treatment of psychiatric disorders (such as depression, disruptive behaviour disorders, and autism), and early educational interventions with support in typical educational settings. Approaches and outcomes differ throughout the world.

The sequences of the human chromosomes 21 and 22 indicate that there are approximately 770 well-characterized and predicted genes. In this study, empirically derived maps identifying active areas of RNA transcription on these chromosomes have been constructed with the use of cytosolic polyadenylated RNA obtained from 11 human cell lines. Oligonucleotide arrays containing probes spaced on average every 35 base pairs along these chromosomes were used. When compared with the sequence annotations available for these chromosomes, it is noted that as much as an order of magnitude more of the genomic sequence is transcribed than accounted for by the predicted and characterized exons.

Individuals with Down's syndrome have a greater risk of leukaemia than the general population, but reliable estimates of the age-specific risk are lacking and little is known about the risk of solid tumours. We identified 2814 individuals with Down's syndrome from the Danish Cytogenetic Register, and linked the data to the Danish Cancer Registry. The number of person-years at risk was 48453. Standardised incidence ratio (SIR) and 95% CI were calculated of the basis of cancer rates specific for age and sex in the general population. 60 cases of cancer were found, with 49.8 expected (SIR 1.20 [95% CI 0.92-1.55]). Leukaemia constituted 60% of cases of malignant disease overall and 97% of cases in children. The SIR for leukaemia varied with age, being 56 (38-81) at age 0-4 years and 10 (4-20) at 5-29 years. No cases of leukaemia were seen after the age of 29 years. The SIR for acute myeloid leukaemia was 3.8 (1.7-8.4) times higher than that for acute lymphoblastic leukaemia in children aged 0-4 years. The cumulative risk for leukaemia by the age of 5 years was 2.1% and that by 30 years was 2.7%. Only 24 solid tumours were seen, with 47.8 expected (0.50 [0.32-0.75]). No cases of breast cancer were found, with 7.3 expected (p=0.0007). Higher than expected numbers of testicular cancers, ovarian cancers, and retinoblastomas were seen but were not significant. INTREPRETATION: The occurrence of cancer in Down's syndrome is unique with a high risk of leukaemia in children and a decreased risk of solid tumours in all age-groups. The distinctive pattern of malignant diseases may provide clues in the search for leukaemogenic genes and tumour-suppressor genes on chromosome 21.