Endocrine Abstracts

Introduction: Immune checkpoint inhibitors have revolutionized the treatment of advanced malignancies. These monoclonal antibodies target cytotoxic T-lymphocyte antigen-4 (CTLA-4), programmed cell death 1 (PD-1) or its ligand (PD-L1). Endocrine side effects are common, and while hypophysitis and thyroid disorders are predominant, checkpoint-blockade associated diabetes mellitus deserves further notice.

Methods: We present the case of a 61-year old male with metastatic non-small cell lung carcinoma, who was treated with pembrolizumab (anti-PD-1). He was admitted one week before his third cycle with a clinical manifestation of diabetic ketoacidosis (DKA). Blood analysis confirmed marked hyperglycemia (1194 mg/dl), capillary ketones and metabolic acidosis. Glutamic acid decarboxylase autoantibodies (GAD) were detectable, along with a low c-peptide. We subsequently reviewed the literature to identify similar cases. PubMed was searched through August 2018, by two reviewers, working independently (J.d.F. and A.V.K.).

Results: Aside from our patient, we identified 64 additional cases (56% male; aged 22–84 y/o). Half of these were treated for advanced melanoma (52%). Most patients developed diabetes mellitus while being treated with anti-PD-1 in monotherapy (45/64, 70%). On average, patients were diagnosed after 5.7 cycles (range 1–42), while this appears to be earlier for the combination of anti-CTLA-4 and PD-1 blockade (2.9 cycles). Three-quarter of patients presented with DKA, with an average glycaemia of 601 mg/dl and glycated haemoglobin of 7.6%. C-peptide levels at diagnosis were low in 85% of cases. Pancreatic autoantibodies were found positive in 50% of cases, with GAD being the most reported (47%, 28/59 analysed patients). Other endocrine adverse events were mentioned in 25% of cases, of which the thyroid was clearly the most frequently affected.

Conclusions: Checkpoint-blockade associated diabetes mellitus is a rare, but potentially lethal complication, as DKA is often the first presentation. Health-care workers should be aware and patients need to be educated. This is of utmost importance as immune checkpoint therapy is increasingly used, affecting patients in earlier stages and for a longer period of time. Routine measurement of blood glucose and glycated haemoglobin seems to be the most practical and feasible options. Predisposing factors, such as HLA haplotype, may explain why some individuals are at greater risk, though further exploration (e.g. identifying those individuals carrying an increased risk) is required.