Genotype/epigenotype/phenotype correlations define Beckwith-Wiedemann syndromes. A. Mussa1, A. De Crescenzo2, S. Russo3, L. Calzari3, N. Chiesa1, C. Molinatto1, G. Baldassarre1, D. Melis4, D. Milani5, M. F. Bedeschi5, L. Tarani6, A. Selicorni7, M. Cirillo Silengo1, L. Larizza3,8, A. Riccio2, G. B. Ferrero1 1) Department of Pediatrics, University of Torino, Italy, Torino, Torino, Italy; 2) Genetics and Biophysics Institute A. Buzzati-Traverso, CNR, Naples, Italy; 3) Laboratory of Cytogenetics and Molecular Genetics, Istituto Auxologico Italiano, Milan, Italy; 4) Department of Pediatrics, University Federico II, Naples, Italy; 5) Medical Genetics Unit, IRCCS Ca' Granda Foundation, Ospedale Maggiore Policlinico, Milan, Italy; 6) Genetica Clinica, Dipartimento di Pediatria, UniversitÓ "La Sapienza" di Roma, Policlinico Umberto I, Roma; 7) Pediatric Clinical Genetics, Department of Pediatrics, University of Milano Bicocca, S. Gerardo Hospital MBBM Foundation, Monza, Italy; 8) Medical Genetics, Department of Medicine, Surgery and Odontoiatry, University of Milan, Italy.

   Background Beckwith-Wiedemann syndrome (BWS) is characterized by cancer predisposition and a variable association of overgrowth, macroglossia, abdominal wall defects, renal anomalies, nevus flammeus, ear malformations, hypoglycemia, hemihyperplasia, and organomegaly. BWS molecular bases are heterogeneous as several mechanisms lead to the disruption of the transcription of genes regulated by the two Imprinting Center (IC1 and IC2) in the 11p15 chromosomal region responsible for the syndrome. Objective To search for genotype-phenotype correlations in a large cohort of BWS patients with positive molecular tests. Methods 281 patients with BWS and proven molecular defect were included and characterized clinically. Patients characteristics were compared among BWS molecular subclasses: KvDMR1 hypomethylation (IC2, n=172), H19/IGF2 hypermethylation (IC1, n=30), chromosome 11p15 paternal uniparental disomy (UPD, n=68), and CDKN1c mutation (n=11). Results Each group showed different growth patterns: neonatal macrosomia was typical and almost constant in IC1 patients (90%), postnatal overgrowth in IC2 patients (50%), and hemihyperplasia more common in UPD patients (88%, p<0.001). Exomphalos was more common in IC2/CDKN1c patients (63% and 29%), whereas minor abdominal wall defects as diastasis recti and umbilical hernia were associated with IC1 defects (40% and 50%, p<0.001), consistent with organomegaly (20-25%) and polyhydramnios (11%). Renal defects were typical of UPD/IC1 patients (33%), and uretheral malformations of IC1 cases (23%, p<0.001). Ear anomalies and nevus flammeus were associated with IC2/CDKN1C genotype (52-63%, p<0.001). Macroglossia, almost always present (90% of cases), was less common among UPD patients (72%, p<0.001). Wilms tumor was associated with IC1 defects (20%) or UPD (4%) and never observed in IC2 patients (p<0.001). Hepatoblastoma occurred only in UPD cases (6%) and other tumors were randomly scattered among molecular subclasses that displayed a different cancer risk, lower in IC2/CDKN1c defects, intermediate in UPD, and very high in IC2 cases (p=0.050). Conclusions In BWS is definable a clear phenotype-(epi)genotype correlation is definitely present allowing to define three different syndromes with large clinical overlap. These clinical molecular correlations will likely allow a tailored follow-up and cancer screening procedures.

You may contact the first author (during and after the meeting) at