Hernias and Ewing's sarcoma family of tumours: a pooled analysis and meta-analysis

Background Ewing's sarcoma family of tumours has been associated with a history of hernia and with a parental occupation of farming. However, the causes of these tumours remain unknown. We therefore aimed to investigate the association between hernia and Ewing's sarcoma family of tumours. Methods We did a pooled analysis of two case-control studies and a meta-analysis of three case-control studies of Ewing's sarcoma family of tumours that had adequate information on history of hernia. The primary endpoint was development of a tumour from the Ewing's sarcoma family. 138 patients with such a tumour and 574 controls were included in the pooled analysis, and 357 patients with these tumours and 745 controls were included in the Meta-analysis. Risk was assessed by an odds ratio (OR) and 95% CI by use of multivariate analysis with unconditional logistic regression for the pooled analysis and random effects model for the meta-analysis. Findings Pooled analysis showed that children with Ewing's sarcoma family of tumours were more likely to have had an umbilical hernia than were controls (odds ratio [OR] 3.3 [95% Cl 1 . 3-8 . 0]). Meta-analysis showed that children with Ewing's sarcoma family of tumours were more likely to have had a hernia (3.2 [1.9-5.7]). Interpretation Ewing's sarcoma family of tumours and hernias (particularly inguinal hernias) have common embryological pathways of neuroectodermal origin, and environmental factors, such as farming, might link the two entities.