Autoimmune Diabetes Presented with Diabetic Ketoacidosis Induced by Immunotherapy in an Adult with Melanoma.

INTRODUCTION: Immunotherapy has been approved for treatment of melanoma. Autoimmune endocrinopathies have been reported in trials involving immunotherapy but autoimmune diabetes has not been definitively linked to them. Here we describe a case of autoimmune diabetes presenting with DKA after receiving combined immunotherapy with anti-CTLA4 and anti-PD1 monoclonal antibodies. CASE: A 47year old gentleman with metastatic melanoma presented to our institution with confusion, abdominal pain and decreased oral intake. The patient had a history of diabetes on metformin which was discontinued two years prior. He was started on Novilumab/Iplimumab for metastatic melanoma. He had received two cycles of immunotherapy and treatment was initially well tolerated. However, eight days after the second cycle the patient developed lethargy, confusion, vomiting and abdominal pain. CT of the head was negative for intracranial abnormalities and without evidence of brain metastasis. His laboratory results included: serum sodium 126 mmol/L, potassium 6.7 mmol/L, BUN 55 mg/dL, creatinine 3.5, bicarbonate 5 mmol/L, chloride 94 mmol/L, albumin 3.2 g/dL. Serum beta-hydroxybuterate was elevated (4.7 mmol/L, N: 0.0-0.5 mmol/L) and the calculated anion gap was 43 mmol/L. Serum lipase elevated (535 u/L, N: 4-60 u/L). The diagnosis of diabetic ketoacidosis was made and he was started on intravenous fluids and insulin therapy. Given his history of metastatic melanoma, his DKA was initially thought to be secondary to pancreatic metastasis especially considering the elevated lipase level. A non-contrast CT of the abdomen showed no evidence of pancreatic metastasis. Interestingly, further investigation identified high serum titers of anti-glutamic acid decarboxylase (anti-GAD) antibodies (0.43 nmol/L, N: less than 0.02 nmol/L), a low C-peptide level (0.2 ng/ml, N: 0.9-5.5 ng/ml), supporting an autoimmune etiology of the diabetes. Other islet autoantibodies were not elevated and his Hemoglobin A1C was 8.0 percent . DISCUSSION: There are few case reports about diabetes and immunotherapy. Autoimmune mechanism was suggested as the culprit, although not all cases reported with positive antibodies. Moreover, it is unlikely that patient developed latent autoimmune diabetes (LADA); and not related to immunotherapy due to the course of LADA is quite more gradual and our patient presented with acute DKA few days post the second cycle. Physicians and patients should be aware that autoimmune disorder such as DKA may be a rare but important immunotherapy related adverse events.