Introduction Tumor cells use immune checkpoint proteins such as programmed cell death 1 to escape immunological defenses. Immune checkpoint inhibitors (ICIs) that target these proteins have been used to treat several malignancies. ICI-induced diabetes is a rare but life-threatening adverse event. Our previous study evaluated pancreatic β-cell activity using the homeostasis model assessment of β-cell (HOMA-β) function before ICI treatment and demonstrated its association with treatment outcomes. In addition, DRB1*04:05 and DRB1*09:01 reportedly increase the risk of type 1 diabetes in Japanese patients, whereas DRB1*15:01 protected against type 1 diabetes. However, the association between human leukocyte antigen (HLA) class II alleles and treatment outcomes of ICI therapy remains unclear. Methods We included 96 patients who were diagnosed with advanced cancer. The HLA genotypes that cause type 1 diabetes in patients with ICI-treated cancers were evaluated to determine their association with the cancer prognosis. Results The median progression-free survival (PFS) in the DRB1*04:05-positive group (2 months, 95% confidence interval CI: 1.214–2.786; 31 events; 1 censored event) was significantly shorter than that in the DRB1*04:05-negative group (3 months; 95% CI: 1.933–4.067; 56 events; 8 censored events) (log rank p=0.045). Additionally, a multivariable Cox proportional hazards regression revealed that DRB1*04:05 was independently associated with shorter PFS for patients treated with ICIs. Conclusions HLA class II alleles were associated with shorter PFS in patients treated with ICIs. In particular, DRB1*04:05 positivity was associated with worse survival outcomes, suggesting a potential immunogenetic contribution to treatment outcomes.
Watanabe et al. (Wed,) studied this question.