ESPE Abstracts

Background: New-onset diabetes in youth can classically fall into three categories: type 1 diabetes (T1D), type 2 diabetes (T2D), or monogenic diabetes (MD). This classification is based on the underlying disease mechanisms, dictates treatment, and predicts outcome. Type B insulin resistance syndrome (TBIR) and ketosis-prone diabetes (KPD) are described in the African population, but their incidence in the pediatric population is very low.

Case Presentation: We present the case of an adolescent girl of African descent recently diagnosed with systemic lupus erythematosus (SLE). She was admitted to the emergency room with ketoacidosis, positive anti-insulin antibodies, a hemoglobin A1c (HbA1c) of 12.1% (109 mmol/mol) and was diagnosed with T1D and managed accordingly with intravenous insulin. Metformin was added due to obesity and signs of insulin resistance. She also received treatment with hydroxychloroquine, mycophenolate mofetil, and prednisone in the context of SLE. After discharge from the hospital, the family slowly discontinued insulin injections and metformin without informing the treating physicians. Five months after the diabetes diagnosis, the HbA1c values and glycemia had normalized without any anti-diabetic treatment, despite continuous glucocorticoid therapy accompanied by an increase in body weight. Exome analysis did not reveal any variants involved in MD. Type 1 genetic risk score (GRS), to assess type 1 diabetes susceptibility, was below the 1st centile.

Conclusion: We illustrate a complex form of a reversible diabetes subtype. Several features, such as the anti-insulin antibodies, an additional autoimmune disease, SLE, point towards an autoimmune origin. The overlapping HLA risk haplotypes for T1D and SLE supports this idea. Further, we hypothesize that a decrease in SLE disease activity, supported by the drop of SLE antibodies, could have favorably influenced the remission of diabetes in this case. In addition, the current immunomodulating treatment for SLE could have helped to abort the developing T1D. The GRS does not support this hypothesis, but does not rule out T1D entirely, because it was not developed for people of African ancestry. Ketosis-prone diabetes, classified as idiopathic but not an HLA-associated form of T1D, could be another differential diagnosis. We rather rule out T2D because remission should not have occurred during weight gain and while on glucocorticoid therapy. The clinical phenotype does not quite fit with the diagnosis of TBIR. In the near future, the recently developed ancestry-tailored GRS will improve the prediction of T1D across ancestries and lead to a better understanding of the different underlying disease mechanisms.