LAUSR

Dipeptidyl peptidase-4 inhibitor (DPP4i) is a widely used antidiabetic agent. Emerging cases of DPP4i-associated bullous pemphigoid (DBP), whose pathogenesis remains unclear, have been reported. Thus, a retrospective study was conducted from January 2016 to June 2021 to determine the clinical, laboratory, and histopathological features of DBP and idiopathic bullous pemphigoid (IBP). We set up in vitro experiments using vildagliptin-treated HaCaT keratinocytes to validate what we found by analyzing published RNA sequencing data about the genes related to the dermal–epidermal junction. We also observed IL-6 expression by HaCaT cells treated with vildagliptin. We enrolled 20 patients with DBP and 40 patients with IBP. The total Bullous Pemphigoid Disease Area Index (BPDAI) score was similar in both groups. However, the BPDAI score of erosions and blisters in DBP was significantly higher than that in IBP (24.6 vs. 16.68, p = 0.0189), and the score for urticaria and erythema was lower in DBP (12 vs. 19.05, p = 0.0183). The pathological features showed that the mean infiltrating eosinophil number per high-power field was significantly lower in DBP than in IBP (16.7 vs. 27.08, p = 0.023). The expression of LAMA3, LAMB3, LAMC2, DST, and COL17A1 decreased significantly in vildagliptin-treated human keratinocytes. On the other hand, IL-6, the hallmark cytokine of bullous pemphigoid (BP) severity, was found to be upregulated in HaCaT cells by vildagliptin. These experimental findings imply less of a requirement for eosinophil infiltration to drive the inflammatory cascades in DBP blistering. Both immunologic and non-immunologic pathways could be employed for the development of DBP. Our findings may help explain the higher incidence of non-inflammatory BP that was observed in DBP.