Are solid tumours different in children with Down's syndrome?

Abstract

Listen

Translate article

A review of the literature showed a peculiar distribution of neoplasms in persons having Down's syndrome (DS). It revealed a well-known increase in leukaemia combined with a less often described global decrease in solid neoplasms. It also found a difference in organ site and histology distribution of solid neoplasms.1 In adults this was confirmed in 3 subsequent epidemiologic studies on cancer incidence and on causes of death in DS conducted in Denmark,2 Israel3 and the United States,4 showing a very low incidence of carcinomas while germ cell tumours were more frequent than in the general population. However, for children <15 years of age, these studies could not provide sufficient data. The largest epidemiologic study reported only 7 cases of solid tumours in children <15 years in Great Britain.5 To better know the tumour profile of children with DS, we conducted a retrospective study in France using the network of the Société Française d'Oncologie Pédiatrique (SFOP), which covers the whole country. Our study was conducted through the centres of Paediatric Oncology covering, among the French metropolitan population of 58 million inhabitants, the 10.5 million <15 years of age. It aimed at gathering all cases of solid malignant tumours diagnosed from 1980 onward in children with DS up to 14 years of age. We included teratomas, since these tumours can be either benign or malignant and their evolution is unpredictable, as well as all types of brain tumours. Each tumour was histologically documented. A questionnaire requesting a listing of any encounter during clinical practice with patients having both DS and a solid tumour, as defined previously, was sent twice at 5-year intervals to every physician including radiologists and pathologists in each Paediatric Oncology centre. When a centre did not reply to our inquiry, physicians were contacted by phone. In case of a positive answer, a detailed clinical history was requested. This provided the basis for the computation of a number of observed cases. The number of expected cases of solid tumours in children with DS for the whole French population was calculated on the basis of population rates of solid tumours in the total paediatric population obtained by the French Childhood Cancer Registry of Lorraine (CCRL) and extrapolated to the whole DS French paediatric population, taking into account the cumulative incidence of DS at birth (10.4 cases/10 000 births) and the survival proportion of DS at age 15 (76.6%). During the 22-year period of 1980–2001, we could identify 13 cases of solid tumours in children <15 years with DS (Table I), whereas 27.2 were expected (p = 0.007 based on Poisson distribution) based on the rates observed in CCRL and extrapolated to the whole French paediatric population with solid tumours (23,244 estimated cases). As a comparison, from 1990–99, the annual number of incident cases of leukaemia in children with DS was about 10 new cases, according to the French National Registry of Childhood Leukaemia and Lymphomas, which covers the whole French paediatric population. These data confirm that solid tumours are less frequent in children with DS than in children of the general population and are much less frequent than leukaemia in DS. It also strengthens the idea of a peculiar distribution of solid tumours in DS. There was a complete lack of central nervous system (CNS) tumours, which usually account for one-third of solid neoplasms in children, and a lack of embryonal neoplasms, particularly neuroblastoma, nephroblastoma, hepatoblastoma and medulloblastoma, which altogether provide the largest solid tumour burden in childhood. On the contrary, there was an excess of both gonadal and extra-gonadal germ cell tumours of lymphomas and of retinoblastoma. It is noteworthy that retinoblastoma does not follow the pattern of other embryonal tumours such as neuroblastoma, nephroblastoma, hepatoblastoma and medulloblastoma since they are not increased in incidence in conditions that favour the onset of other embryonal tumours such as Beckwith-Weidemann syndrome.6 Our study indicates that, from an oncologic point of view, the diagnostic procedure to follow in a child with DS and a suspected tumour may be different from that of a child in the general population. It may also suggest that among the 225 identified genes on chromosome 21,7 some could protect against embryonal and CNS neoplasms, while others may favour germ cell tumours and lymphomas. Interestingly, the excess of germ cell tumours, particularly of the testis, is also observed in adults with DS.4, 8 Yours sincerely, Daniel SATGÉ*, Annie J. SASCO and Brigitte LACOUR We thank Dr. J. Clavel for the information she kindly provided based on the French National Registry of Childhood Leukaemia and Lymphomas.