We described an infant with severe HI resulting from a paternally-inherited
ABCC8 mutation in conjunction with mosaic segmental pUPD11p15 demonstrated in the pancreatic tissue from the second resection but not in peripheral blood leucocytes, suggestive of BWS/BW-spectrum HI. With pUPD11p15, the loss of maternal allele resulted in a loss of
H19 and
CDKN1C expression, which usually negatively regulates cell proliferation; whereas the biallelic
IGF-II expression promotes cell growth [
9]. Therefore, pancreatic adenomatous hyperplasia and hyperinsulinism were attributed to the combination of the K
ATP defect along with the pUPD11 and the imbalance of imprinted genes at 11p15 region. In contrast to the classical histological findings in focal HI related to a paternally-inherited
ABCC8 mutation with lesion confines to a small localized area, our patient had multiple foci of adenomatous hyperplasia throughout the pancreas. Furthermore, the level of mosaicism of UPD cells in the pancreas correlated with the shifted allele frequency of the
ABCC8 mutation. To our understanding, this is the first report using the accurate and sensitive assays to demonstrate the direct correlation of the paternally inherited ABCC8 c.1792 T level with mosaic level of pUPD.
Other than being macrosomic, our patient had no other somatic features of BWS. The consideration of testing for BWS was triggered by the atypical histological findings. This distinct pancreatic histology had been described in children with Beckwith-Wiedemann Spectrum [
5,
10,
11]. In a large series of 28 patients with BWS and persistent HI, their phenotypes were reported to range from isolated, subtle hemihypertrophy or umbilical hernia to frank BWS phenotype with multiple somatic features [
6]. Only four of them had concomitant K
ATP mutations. Therefore, It was suggested that, even in the absence of somatic features of BWS, testing should be considered in HI cases with large ‘focal’ pancreatic lesions with or without a K
ATP mutation [
5]. The diagnosis of BWS is important due to their inherent vulnerability to embryonal tumours, affecting up to 8% of BWS patients [
4,
12]. Calton et al. reported a similar case of large/multifocal focal HI resulting from a paternally inherited recessive
ABCC8 mutation [
11]. That patient, like our patient, had no clinical features of BWS. BWS testing was only performed at the age of 20 months when he developed hepatoblastoma. Again, similar to our patient, pUPD11p was identified in the affected tissue (hepatoblastoma tissue and the stored pancreatic tissue), but not in peripheral blood or buccal DNA [
11]. This highlights that infants with HI related to mosaic BWS could also develop BWS-associated tumours due to mosaic UPD, and that tumour surveillance is indicated. It has been suggested that the tumour risk could be associated with the level of mosaicism for UPD within specific organs [
13]. Since tissues from other organs were not available for testing in our patient, it is unclear whether other organs are affected by pUPD11p. Therefore, tumour surveillance during early childhood is warranted.
With the variability of mosaicism between tissues in patients with BWS, the source of DNA for molecular analysis is extremely important. In our patient, absence of mosaicism in the peripheral blood leukocytes would have wrongly concluded as ‘normal study’ if the pancreatic tissues were not sent for further analysis. Therefore, similar to other mosaic conditions, affected tissue should always be sent for further molecular analysis if possible [
4].