Bullous Pemphigoid and Dipeptidyl Peptidase 4 Inhibitors: A Disproportionality Analysis Based on the Japanese Adverse Drug Event Report Database

Bullous pemphigoid (BP) is a rare autoimmune blistering skin disease. Although previous case reports and two disproportionality analyses of European populations have suggested associations between dipeptideyl peptidase 4 (DPP-4) inhibitor use and BP (1–3), the involvement of race, antidiabetes drugs, sex, age, and other risk factors in BP induced by DPP-4 inhibitors has not yet been evaluated. We conducted disproportionality analysis using the Japanese Adverse Drug Event Report (JADER) database, which contains all pharmacovigilance data that have been spontaneously reported to the Pharmaceuticals and Medical Devices Agency (PMDA) since April 2004.

As of June 2017, the database contained 454,027 reports of adverse drug reactions. For this study, data were downloaded from the PMDA website (http://www.pmda.go.jp). Potential signals of DPP-4 inhibitors or other drugs with BP were assessed with categories using the reporting odds ratio (ROR), which is an established parameter in pharmacovigilance research. An ROR was calculated using a two-by-two contingency table (4). We examined DPP-4 inhibitors (alogliptin, anagliptin, linagliptin, omarigliptin, saxagliptin, sitagliptin, trelagliptin, teneligliptin, and vildagliptin), other antihyperglycemia drug classes (sulfonylureas, biguanides, α-glucosidase inhibitors [α-GIs], glinides, glucagon-like peptide 1 receptor agonists [GLP-1 RAs], sodium–glucose cotransporter 2 inhibitors [SGLT2is], thiazolidines, and any kind of insulins), other drugs that are reported as causal reagents for BP, and negative controls (2,5). An ROR is considered significant when the lower bound of the two-sided 95% CI for the risk of BP was larger than 1. The data were analyzed using the R software (version 3.4.0; The R Foundation for Statistical Computing, Vienna, Austria).

In the database, 546 BP (referred to as pemphigoid in the database) case subjects were found. The total number of reports of BP associated with DPP-4 inhibitor use was 392. Although the RORs (95% CI) for acetaminophen were 0.25 (0.09–0.66), the RORs (95% CI) for the DPP-4 inhibitors were wholly significant and as follows; alogliptin, 8.02 (4.87–13.22); anagliptin, 10.84 (3.46–33.96); sitagliptin, 12.59 (9.86–16.06); trelagliptin, 13.77 (3.40–55.85); saxagliptin, 15.85 (5.87–42.79); linagliptin, 28.96 (21.38–39.23); omarigliptin, 43.79 (5.85–327.70); teneligliptin, 58.52 (42.75–80.10); and vildagliptin, 105.33 (88.54–125.30) (Table 1). Because most patients with diabetes receive combinations of medications, we excluded case subjects who received DPP-4 inhibitors. After such exclusion, significant RORs disappeared for case subjects receiving all the other antihyperglycemia drugs. Conversely, case subjects receiving DPP-4 inhibitors showed significant RORs even after exclusion of subjects who were receiving any other antihyperglycemia or candidate drugs, indicating that co-prescription bias was excluded. We stratified data by time trends of report and confirmed that the RORs of DPP-4 inhibitors were consistently significant over the period of 2012–2016. An analysis of ROR by therapeutic area with subjects who used antihyperglycemia drugs showed the RORs (95% CI) as follows: linagliptin, 2.67 (1.96–3.64); teneligliptin, 5.52 (4.01–7.60); vildagliptin, 12.09 (9.88–14.79); and DPP-4 inhibitors, 69.49 (34.50–139.99). Meanwhile, statistically significant associations disappeared for sitagliptin, saxagliptin, and alogliptin, implying a drug-specific effect. Vildagliptin, teneligliptin, and linagliptin, which show higher RORs than other DPP-4 inhibitors, possess lower substrate selectivity for DPP-4 or higher volume of distribution. It is possible that the inhibition of DPP-8 or DPP-9, but not that of DPP-4, in the skin evokes immunopathogenic reactions that result in the blister formation in BP. The number of males with BP taking DPP-4 inhibitors was significantly higher than the number of females with BP (male, 8,131 out of 222,567; female, 4,911 out of 215,460; χ² = 715.4; P < 2.2e-16). There was no significant difference in the frequency of BP after adjustment for DPP-4 inhibitor use (with DPP-4 inhibitor use, 238 males out of 8,131, 147 females out of 4,911, χ² = 0.05, P = 0.83; without DPP-4 inhibitor use, 71 males out of 214,436, 76 females out of 210,549, χ² = 0.27, P = 0.60). Thus, the higher number of males with DPP-4 inhibitor–related BP was presumably because DPP-4 inhibitors were used more often in males than in females. Most of the case subjects with BP were elderly people older than 60 years old. In 2015, NDB (National Database of Health Insurance Claims and Specific Health Checkups of Japan) Open Data Japan, a comprehensive prescription database, indicated that the distribution of DPP-4 inhibitor prescription in each 10-year age-group (from <40 to ≥90 years) was consistent with the adverse events induced by DPP-4 inhibitors in the JADER database. This suggests that advanced age is a risk factor for BP induced by DPP-4 inhibitor use, which is concordant with several previous reports, mainly from Europe, that showed BP was predominantly encountered in elderly people (2,3).

As the JADER database is a spontaneous reporting system, there are several limitations, such as underreporting, overreporting, missing data, mistakes in data entry, and a lack of control data. Especially among these, strong selection bias, so-called notoriety bias, must always be taken into consideration. In this study, we did not investigate the contribution of coexisting illnesses, drug dose, or the period of exposure. Further clinical monitoring and analytical observational studies that can evaluate the association are needed.