Hypereosinophilia following initiation of glucagon-like peptide 1 receptor agonists: a sequence symmetry analysis using the Danish health registries

Background: Glucagon-like peptide 1 receptor agonists (GLP1-RA) are increasingly used in the treatment of type 2 diabetes and obesity. Cases of hypereosinophilic syndrome (HES) following initiation of GLP1-RA have been reported. HES is a rare but potentially dangerous condition defined by hypereosinophilia (HE) and related end-organ damage.

Objectives: To investigate the association between GLP1-RA initiation and incident hypereosinophilia.

Methods: Using the nationwide Danish prescription registry and laboratory database, we conducted a sequence symmetry analysis (SSA) investigating the occurrence of HE following initiation of GLP1-RA.
We included all individuals who initiated a GLP1-RA during the period June 1st 2015 to March 1st 2023, and who had an eosinophile measurement performed during the period six months before to six months after drug initiation. Hypereosinophilia was defined as an eosinophile count above 1.5×10^9/l. To account for temporal trends in the outcome, we calculated trend-adjusted SRs with 95% confidence intervals (CI). For comparison, we repeated analyses using initiation of sodium-glucose co-transporter 2 inhibitors (SGLT2i) and dipeptidyl peptidase 4 inhibitors (DPP4i). Further, we asserted the robustness of the results in sensitivity analyses with massive eosinophilia (≥5×10^9/l), mild eosinophilia (0.5–1.5×10^9/l), window lengths of 90 and 365 days and a negative control outcome of an incident thrombocyte count < 50×10^9/l.

Results: We identified 161,567 individuals who initiated a GLP1-RA, of which 136 initiated GLP1-RA prior to occurence of hypereosinophilia and 41 experienced hypereosinophila prior to GLP1-RA initiation, yielding a SR of 3.39 (CI 2.41–4.85). For SGLT2i and DPP4i we found SRs of 0.78 (N=91/118, CI 0.60–1.03) and 0.91 (N= 67/73, CI 0.65–1.27), respectively. In the supplementary analyses with shorter and longer observation windows we found SRs of 4.25 (CI 2.74–6.81) and 2.30 (CI 1.71–3.11), respectively. Sequence ratios for non-semaglutide GLP1-RAs and semaglutide both yielded sequence ratios above 2.5. For the negative-control outcome of thrombocytopenia, we found SRs of 1.25 (0.77–2.04) and 1.33 (1.00–1.77) for GLP1-RA and SGLT2i respectively.

Conclusions: In this nationwide study, we found that hypereosinophilia is three times more likely to occur after initiation of GLP1-RAs compared to the period prior to initiation. This finding was specific to GLP1-RAs and not observed for other comparable antidiabetic drugs.