Editor,—We greatly enjoyed reading the article by Cole on overgrowth disorders in childhood.1While the article alluded to the increased risk of malignant disease among children with Beckwith–Wiedeman syndrome (BWS) no mention was made of the increased risk of cancer in Sotos syndrome.2

We recently treated a 3 year old boy with Sotos syndrome for abdominal B cell lymphoblastic lymphoma (Murphy stage III) in accordance with United Kingdom children’s cancer study group protocol 9002. Although his lymphoma exhibited only a partial response to standard chemotherapy, he remains disease free two years later following second look surgery and subsequent high dose treatment with autologous bone marrow rescue.

Non-Hodgkin’s lymphoma has also been reported in three other patients with Sotos syndrome.2 3 Malignant disease in this condition is not restricted to lymphoproliferative disorders and several other types of cancer including Wilms’s tumour, neuroblastoma, hepatocellular carcinoma, and acute lymphoblastic leukaemia have been described. The risk of malignant disease in Sotos syndrome has been estimated as 1 in 41 under the age of four years.2

The literature suggests that the incidence of specific tumour types may differ between individual overgrowth syndromes. In Sotos syndrome, non-Hodgkin’s lymphoma constitutes almost one quarter of reported cases; a very different pattern has been observed in children with BWS in whom Wilms’s tumour, hepatoblastoma, and neuroblastoma are responsible for the vast majority of observed cancers.4Although subject to reporting bias these figures suggest that specific overgrowth syndromes are associated with differing, albeit overlapping, patterns of tumour formation. Such an observation may result from differences in the aetiology of the cancers involved. Dr Cole’s annotation details the relation between overexpression of insulin-like growth factor II and BWS.1 This discovery has stimulated considerable interest in the role of this peptide and its binding proteins in the aetiology of many human cancers including Wilms’s tumour, hepatoblastoma, and neuroblastoma. The idea that overexpression of insulin-like growth factor II is important in the aetiology of these malignancies is an attractive hypothesis that may partially explain the observed differences in the incidence of individual tumour types between BWS and other overgrowth disorders.