DOZ047.32: Infantile hypertrophic pyloric stenosis in patients with esophageal atresia: proposal for a causative seesaw model

Aim of the Study Patients born with esophageal atresia (EA) appear to have a 30 times higher prevalence of infantile hypertrophic pyloric stenosis (IHPS). This makes sense from a developmental perspective as both the esophagus and the pyloric sphincter are foregut derived structures. We hypothesized that genetic defects, disturbing foregut morphogenesis, are responsible for the specific combination of EA and IHPS. Methods Patients with both EA and IHPS born between 1970 and 2017 and where possible their parents were included. This study was approved by the internal review board of the Erasmus MC. After parental written informed consent, we determined genetic profiles with whole exome sequencing and SNP-array-based copy number variation analysis. We focused on (1) genetic variation in known EA and IHPS disease genes, (2) pathways important for foregut morphogenesis, (3) shared rare genetic variation, and (4) ultrarare variants in variant-intolerant genes, which have a high chance of being de novo. Segregation analysis of possible candidate variants was performed. Results Twenty-seven out of 664 patients (4.1%) born with EA during the study period developed IHPS, of which 15 cases were analyzed. As none of the parents were affected, we considered dominant (de novo) or recessive and X-linked inheritance models. Unfortunately, we could neither identify rare de novo mutations, nor variants fitting a recessive model. We did however identify inherited putative deleterious heterozygous variants in genes either known to be involved in EA or IHPS (e.g., COL7A1, TNXB, WDR11, GDF6) or important in foregut morphogenesis (e.g., GLI3, NKX2.1) in all patients. Moreover, seven genes were affected by rare variation in ≥2 patients (ADAMTSL4, ANKRD26, CNTN2, HSPG2, KCNN3, LDB3, SEC16B) and expressed in the developing foregut, esophagus or pyloric sphincter in mice between E8.25 and E18.5. However, burden analysis did not show a significant difference with unaffected controls. Conclusions Although the presence of genetic variation in likely candidate genes suggests a genetic component, genetic factors alone could not explain the abnormalities seen in these patients. Therefore, we propose a multifactorial model in which the combination of high impact genetic, mechanical, and environmental factors together can shift the balance from normal to abnormal development.