Characteristics of insulinopenic and non insulinopenic diabetes related to immune checkpoint inhibitors: A French pharmacovigilance study - 07/04/21

Doi : 10.1016/j.therap.2021.03.004

Marie Bastin ^a, Marion Allouchery ^b, Marion Sassier ^c, Franck Rouby ^d, Pirayeh Eftekhari ^e, Bénédicte Lebrun-Vignes ^f, Fabrizio Andreelli ^a, Kevin Bihan ^f,⁎
^a Pitié Salpêtrière Hospital, Diabetology Dpt, Sorbonne University, APHP, 75013 Paris, France
^b Regional Pharmacovigilance Center, Department of clinical Pharmacology, University of Poitiers, Poitiers University Hospital, 86073 Poitiers, France
^c Regional Pharmacovigilance Center, Department of Pharmacology, Caen University Hospital, 14000 Caen, France
^d Regional Pharmacovigilance Center, Department of Pharmacology, Aix-Marseille University Hospital, 13015 Marseille, France
^e Regional Pharmacovigilance Center, Department of Pharmacology, Fernand-Widal Hospital, 75010 Paris, France
^f Regional Pharmacovigilance Center, Department of Pharmacology, Pitié-Salpêtrière Hospital, 47-83, boulevard de l’Hôpital, 75013 Paris, France

^⁎Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Wednesday 07 April 2021
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Summary

Introduction

Immune checkpoint inhibitor-induced diabetes mellitus (ICI-DM) is an immune-related adverse drug reaction (irADR). Hyperglycemia can be linked to endogenous insulin deficiency with ketoacidosis or associated with preserved beta-cell function.

Objectives

We aimed to identify the characteristics of both types of ICI-DM (type 1 and type 2 DM), to improve our understanding of this irADR and its management.

Methods

Data for ICI-DM recorded in the French Pharmacovigilance Database from 2015 to October 2019 were analyzed according to the French causality assessment.

Results

In total, 60 subjects were included. Anti-PD1/PDL1 pathway blockade therapy (nivolumab: 61.7%+3.3% in association with ipilimumab pembrolizumab: 28.3%) was most frequently implicated in ICI-DM, but some reports involved anti-CTLA4 drug (ipilimumab: 6.7%+3.3% in association with nivolumab). One third of reports occurred within one month of the initiation of immunotherapy. Decreased insulin secretion (defined by the presence of ketone bodies) were confirmed in 80% of reports. Among them, 54% of patients met the diagnostic criteria for fulminant diabetes. Overall, 17.7% of the reports had pre-existing type 2 diabetes T2D. Four deaths due to hyperglycemia were declared, with altered insulin secretion in only two of these reports. BMI was lower in the insulinopenic group (23.4±0.7 vs. 27.9±1.6, P=0.004) and other irADRs were more frequently observed in patients with persistent insulin secretion (66.7 vs. 18.8%, P=0.02). We found no difference in age, indication or cumulative ICI dose between the two groups (with and without insulinopenia). The presence of GAD antibodies was associated with a shorter time to diabetes onset (42.6±6.1 vs. 208.1±41.6 days, P=0.029).

Conclusions

ICI-DM is a rare but serious irADR triggered by all classes of immunotherapy. The observation period for ICI-DM can be shortened for patients positive for anti-GAD antibodies. Endogenous insulin deficiency did not appear to be the only mechanism involved in ICI-DM, as beta-cell function was preserved in 20% of reports. Improvements in our understanding of this complication will be required for its prevention.

Le texte complet de cet article est disponible en PDF.

Keywords : Immunotherapy, Diabetes, Ketoacidosis, Insulin secretion, Autoimmunity, Cancer therapy