Context: Congenital hyperinsulinism (HI) is characterized by hypoglycemia related to inappropriate insulin secretion. Focal and diffuse forms of hyperinsulinism share a similar clinical presentation, but their treatment is dramatically different. Preoperative differential diagnosis was based on pancreatic venous sampling, a technically demanding technique.
Objective: Positron emission tomography (PET) after injection of [18F]fluoro-L-DOPA (L-dihydroxyphenylalanine) has been evaluated for the preoperative differentiation between focal and diffuse HI, by imaging uptake of radiotracer and the conversion of [18F]fluoro-L-dopa into dopamine by DOPA decarboxylase. We propose to validate this test by immunohistochemical approach.
Patients and methods: Pancreatic surgical specimens of four focal and three diffuse HI were studied, using anti-DOPA decarboxylase and proinsulin antibodies. The effect of an inhibitor of DOPA decarboxylase (carbidopa) on insulin secretion was evaluated in vivo and in cultured INS-1 cells.
Results: Immunohistochemical detection of DOPA decarboxylase showed diffuse staining of Langerhans islets in the whole pancreas in all diffuse cases, in contrast with dense focal staining in all focal cases. Staining of Langerhans islets outside the focal lesion was diffusely but weakly positive. We correlated the localization of DOPA decarboxylase and proinsulin in normal pancreas and in both diffuse and focal HI tissues. The diffuse PET uptake found before treatment in one child with diffuse HI disappeared completely after carbidopa administration, suggesting in vivo that pancreatic cells can take up amine precursors and contain DOPA decarboxylase. The insulin secretion measured in the supernatant was the same whether INS-1 cells were treated by dopamine or Lodosyn or untreated.
Conclusion: We validate PET with as a consistent test to differentiate diffuse and focal HI.