The purpose of this study was to establish the incidence of carcinomas
in children, changes in incidence over a 30-year period, and to ident
ify features of possible aetiological significance. A total of 173 cas
es were identified, but after review of the histopathology, 30 patient
s were excluded because they were considered to have benign epithelial
tumours or malignant tumours of nonepithelial origin. Seven other cas
es were excluded because pathology material was not available. Overall
, in 28% of cases, the diagnoses were changed by pathology review. Thu
s, 136 children in the West Midlands Region diagnosed 1957-1986 were i
ncluded, with carcinoid tumours (44) and tumours of skin (22), nasopha
rynx (14), salivary gland (13), adrenal cortex (13), thyroid (9), larg
e bowel (5), other (16). Excluding carcinoids, the age-standardised in
cidence rate was 2.4 x 10(6) per year. Male:female ratio was 0.7:1 and
66% were aged >10 years. Incidence increased from 1.5 to 3.3 x 10(6)
per year. Genetic factors predisposing to carcinoma included tyrosinos
is, MEN II and III, congenital adrenal hyperplasia and basal cell naev
us syndrome. There was a case of Li-Fraumeni syndrome and several othe
r patients had relevant family histories. Probable ''environmental'' c
auses included antenatal exposure to stilboestrol or hydroxyprogestero
ne hexanoate, stilboestrol given for premature menarche, neonatal hepa
titis and prior radiotherapy. The aetiology of carcinomas in children
is multifactorial, both genetic and environmental factors being import
ant. The incidence is increasing. (C) 1993 Wiley-Liss, Inc.