Article Text

Download PDFPDF

1252 Immune-checkpoint inhibitor induced autoimmune diabetes is a heterogeneous disease with distinct clinical and immune phenotypes
  1. Karina Ruiz-Esteves1,
  2. Kaitlyn Shank2,
  3. Caitlin Colling1,
  4. Tianqi Ouyang1,
  5. Leyre Zubiri1,
  6. Chloe Villani1,
  7. Kerry Reynolds1,
  8. Meghan Sise1 and
  9. Michelle Rengarajan1
  1. 1Massachusetts General Hospital, Boston, MA, USA
  2. 2Brigham and Women’s Hospital, Boston, MA, USA


Background Immune checkpoint inhibitors (ICIs) can lead to immune-related adverse events (IrAEs). One such irAE, ICI-induced autoimmune diabetes (ICI-DM), affects approximately 1% of ICI recipients. Most reports suggest resemblance to fulminant type 1 diabetes, with low or absent endogenous insulin production at presentation.1,2 However, more recent reports have revealed milder cases, suggesting a heterogeneous disease entity.3 We identified distinct clinical phenotypes of ICI-DM that could allow for better optimization of insulin therapy.

Methods We performed a retrospective chart review of 16,582 patients treated with anti-PD1, anti-PDL1, anti-CTLA4, or combination between 2010 and 2022 at a multi-centered academic hospital system. Criteria for inclusion were new diagnosis of type 1 diabetes or drug-induced diabetes and new prescription for insulin following ICI therapy; cases were confirmed by two board-certified endocrinologists. To phenotype ICI-DM based on endogenous insulin production, we reviewed patients with ICI-DM and a C-peptide level, allowing us to classify 45 cases of ICI-DM based on endogenous insulin production and autoantibody status. We are characterizing ICI-DM in these patients at both initial presentation and up to 30 months of follow-up.

Results We identified three distinct clinical phenotypes of ICI-DM based on initial presentation: patients with preserved endogenous insulin production (insulin+) and patients with decreased or absent endogenous insulin production (insulin -), with the latter divided into autoantibody positive (Ab+) and negative (Ab-) patients. Insulin+ ICI-DM patients have high levels of pre-existing insulin resistance, longer time-to-diagnosis (median 175 days), higher A1c at presentation (median 9.8%) and lower likelihood of presenting in diabetic ketoacidosis (DKA). Insulin-Ab+ patients have lower levels of insulin resistance, the lowest time-to-diagnosis (median 65 days), intermediate A1c at presentation (median 8.2%), high rates of DKA and nearly universal hospitalization at presentation. Insulin-Ab- had no pre-existing insulin resistance, a longer time-to-diagnosis (median 154 days) with a relatively low A1c at presentation (media 7.6%) and moderate rates of DKA and hospitalization. As expected, insulin- patients had higher total daily insulin needs compared to insulin+ patients.

Conclusions Our study examined the largest data set to date of deep clinical phenotyping of patients with ICI-DM. We define at least two distinct subsets of ICI-DM, with antibody-positive, islet negative patients being most similar to cases of ICI-DM initially described. However, patients with a milder form of ICI-DM (islet positive patients) may require simpler insulin regimens that could profoundly improve their quality of life. Our findings may also suggest distinct underlying mechanisms for ICI-DM with implications for other IRAEs.


  1. de Filette JMK, Pen JJ, Decoster L, Vissers T, Bravenboer B, Van der Auwera BJ, Gorus FK, Roep BO, Aspeslagh S, Neyns B, Velkeniers B, Kharagjitsingh AV. Immune checkpoint inhibitors and type 1 diabetes mellitus: a case report and systematic review. Eur J Endocrinol. 2019;181(3):363-374.

  2. Tsang VHM, McGrath RT, Clifton-Bligh RJ, Scolyer RA, Jakrot V, Guminski AD, Long GV, Menzies AM. Checkpoint Inhibitor-Associated Autoimmune Diabetes Is Distinct From Type 1 Diabetes. J Clin Endocrinol Metab. 2019;104(11):5499–5506.

  3. Chen X, Affinati AH, Lee Y, Turcu AF, Henry NL, Schiopu E, Qin A, Othus M, Clauw D, Ramnath N, Zhao L. Immune checkpoint inhibitors and risk of type 1 diabetes. Diabetes Care. 2022;45(5):1170–1176.

Ethics Approval This study was approved by the Mass General Brigham Institutional Review Board, Protocol #2017P000501. There was no direct interaction with patients or their samples so individual consent was not deemed necessary by the IRB.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.