ORIGINAL REPORT
Dana SHALMON1#, Efrat BAR-ILAN1#, Alon PELED1, Shamir GELLER1,2, Jonathan BAR1, Naama SCHWARTZ3, Eli SPRECHER1,2 and Mor PAVLOVSKY1
#These authors contributed equally.
1Division of Dermatology, Tel Aviv Sourasky Medical Center, 2School of Medicine, Tel Aviv University, Tel Aviv, and 3School of Public Health, University of Haifa, Haifa, Israel
Drug-associated bullous pemphigoid has been shown to follow long-term gliptin (dipeptidyl-peptidase 4 inhibitors) intake. This study aimed at identifying risk factors for gliptin-associated bullous pemphigoid among patients with type 2 diabetes. A retrospective study was conducted in a tertiary centre among diabetic patients exposed to gliptins between the years 2008–2021. Data including demographics, comorbidities, medications, and laboratory results were collected using the MDClone platform. Seventy-six patients with type 2 diabetes treated with dipeptidyl-peptidase 4 inhibitors who subsequently developed bullous pemphigoid were compared with a cohort of 8,060 diabetic patients exposed to dipeptidyl-peptidase 4 inhibitors who did not develop bullous pemphigoid. Based on a multivariable analysis adjusted for age and other covariates, Alzheimer’s disease and other dementias were significantly more prevalent in patients with bullous pemphigoid (p = 0.0013). Concomitant use of either thiazide or loop diuretics and gliptin therapy was associated with drug-associated bullous pemphigoid (p < 0.0001 for both). While compared with sitagliptin, exposure to linagliptin and vildagliptin were associated with bullous pemphigoid with an odds ratio of 5.68 and 6.61 (p < 0.0001 for both), respectively. These results suggest gliptins should be prescribed with caution to patients with type 2 diabetes with coexisting Alzheimer’s and other dementias, or patients receiving long-term use of thiazides and loop diuretics. The use of sitagliptin over linagliptin and vildagliptin should be preferred in these patients.
Key words: autoimmune blistering disease; bullous pemphigoid; bullous drug reaction; drug reaction; dipeptidyl peptidase 4 inhibitor; gliptin.
Bullous pemphigoid has been shown to follow intake of dipeptidyl-peptidase 4 inhibitors. This study identifies additional risk factors for drug-associated bullous pemphigoid among patients with type 2 diabetes receiving gliptins: Alzheimer’s disease and other dementias or chronic treatment with thiazide or loop diuretics.
Citation: Acta Derm Venereol 2024; 104: adv26663. DOI https://doi.org/10.2340/actadv.v104.26663.
Copyright: © Published by Medical Journals Sweden, on behalf of the Society for Publication of Acta Dermato-Venereologica. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/)
Submitted: Nov 17, 2023; Accepted: Mar 4, 2024; Published: Apr 4, 2024
Corr: Dr Mor Pavlovsky, Division of Dermatology, Tel Aviv Sourasky Medical Center, 6 Weizmann Street, Tel Aviv 64239, Israel. E-mail: mashapavl@gmail.com
Competing interests and funding: The authors have no conflicts of interest to declare.
Bullous pemphigoid (BP) is the most prevalent blistering autoimmune disease among elderly individuals (1), typically characterized by pruritus followed by tense blisters involving the skin (2). Mucosal membrane involvement is observed in 20% of patients (3). BP is defined immunologically by the presence of autoantibodies against collagen XVII (BP180, BPAg2) and dystonin (BP230, BPAg1), which serve as hemi-desmosomal structural proteins, located at the dermal–epidermal junction (4).
Dipeptidyl-peptidase 4 inhibitors (DPP-4i), also known as gliptins, are widely used drugs in the treatment of type 2 diabetes mellitus (DM2). As gliptins increase the level of the hormone incretin, they enhance glycaemic control by increasing insulin secretion and decreasing glucagon secretion (5). The first FDA-approved gliptin was sitagliptin in 2006. Currently, more than 10 gliptins have been endorsed and are available for clinical use (6).
Several reasons have been credited for the rise in the incidence of BP over the past decades, one being the increasing use of medications associated with the eruption of this cutaneous disease (6, 7). The most common drugs that induce BP are gliptins, PD-1/PD-L1 inhibitors, loop diuretics, and antibiotics, especially penicillin (6). In recent years, the emergence of DPP-4i as a significant contributor to drug-induced BP has been reported (7–12). The association between DPP-4i intake and development of BP was initially described in 2012 (13). A recent meta-analysis by Kridin and Cohen (14) reported a 3.2-fold increased risk of developing BP following DPP-4i administration, of which vildagliptin was found to have the highest association with BP among the DPP-4i.
Several common comorbidities (15–17) and medications have been shown to be associated with BP. It has yet to be determined whether gliptin intake alone might induce BP or whether additional factors such as concomitant drug therapy and comorbidities are required. The aim of this study is to define the patient population at risk of developing BP secondary to DPP-4i treatment.
We conducted a single-centre retrospective study among DM2 patients who were treated with gliptins from 1 January 2008 to 31 March 2021 at the Tel Aviv Sourasky Medical Center. Identification of cases and data collection were performed using MDClone software version 6.1 (https://www.mdclone.com/), which enables obtaining big data using our hospital’s digital patient database.
This study protocol was approved by the institutional Helsinki research ethics committee in Tel Aviv Sourasky Medical Center, approval number 09-36-20-TLV. The study group included patients with a new diagnosis of BP and history of diabetes and gliptin therapy. BP patients included in the study were all diagnosed by a dermatologist based on clinical, histological, and direct immunofluorescence (DIF) parameters as previously described by Kershenovich et al. (18). All BP patients included in the study had a DIF microscopy of perilesional, uninvolved skin with linear, continuous deposits of IgG and/or C3 along the epidermal basement membrane. Manual revision of all BP patients’ medical files was performed. The control group consisted of DM2 patients with a history of gliptin intake, but without BP, admitted to our hospital. The specific gliptin used for each patient was recorded, and if a patient consumed more than one gliptin only the last medication was analysed.
Categorical variables were presented with frequencies and percentages, continuous variables were presented with mean (SD) (median, IQR). Association between categorical variables and the binary endpoint (i.e. BP yes/no) was performed using a χ2 test (or Fisher’s exact test). For continuous variables, t-test (or Wilcoxon two-sample test) was implemented. Univariate and multivariable logistic regression was performed presenting odds ratios (OR) and adjusted OR (aOR) with 95% confidence intervals (95% CI). The statistical analysis and data management were performed using SAS 9.4 software (SAS Institute Inc, Cary, NC, USA). P-value < 0.05 was considered statistically significant.
The study group included 76 BP patients with a history of gliptin intake (research group) and 8,060 DM2 patients who received gliptins and did not develop BP (control group). The mean age at the time of data extraction for BP patients and control group was 81.9 and 77.2 years, respectively. Additional demographic characteristics, HbA1c values, and target organ damage are summarized in Table I.
| BP (n = 76) | No BP (n = 8,060) | p-value | |
| Demographic characteristics | |||
| Age, years, mean (SD) Median [IQR] |
81.9 (8.7) 83.1 [75.3–88.7] |
77.2 (11.3) [77, 69.9–85.7] |
< 0.0001 |
| Female gender, n (%) | 37 (48.68) | 3285 (40.76) | 0.1617 |
| BMI, kg/m2,* mean (SD) Median [IQR] |
28.7 (5.1) [27.4, 25.4–30.5] |
30.7 (50.2) [27.6, 24.6–31.2] |
0.7605 |
| Weight, kg,* mean (SD) Median [IQR] |
77.3 (15.8) [72.5, 66–85] |
80.1 (18) [79, 68–90] |
0.1792 |
| Height,* mean (SD) Median [IQR] |
163.1 (8.3) [164, 160–167] |
166.4 (15.8) [168, 160–174] |
0.0038 |
| Smoking (current/past),* n (%) | 11 (18.64) | 748 (27.38) | 0.1357 |
| Smoking years,* mean (SD) Median [IQR] |
27.5 (16) [25, 20–40] |
34.6 (14.4) [37, 25–45] |
0.2710 |
| Mortality, n (%) | 23 (30.26) | 2588 (32.11) | 0.7315 |
| Diabetes characteristics | |||
| Retinopathy, n (%) | 3 (3.95) | 338 (4.19) | |
| Nephropathy, n (%) | 1 (1.32) | 133 (1.65) | |
| Neuropathy, n (%) | 3 (3.95) | 224 (2.78) | 0.4717 |
| HbA1C,* mean (SD) Median [IQR] |
7.4 (1.3) [7.3, 6.4–8.1] |
7.7 (1.8) [7.3, 6.6–8.3] |
0.5382 |
| Chronic medications | |||
| Antihypertensives, n (%) | 22 (28.95) | 2569 (31.87) | 0.5858 |
| Thiazides, n (%) | 21 (27.63) | 694 (8.61) | < 0.0001 |
| PD1/PD-L1, n (%) | 0 (0) | 1 (0.01) | |
| Loop diuretics, n (%) | 18 (23.68) | 627 (7.78) | < 0.0001 |
| NSAIDS, n (%) | 0 (0) | 23 (0.29) | |
| Penicillins, n (%) | 0 (0) | 16 (0.2) | |
| *Missing data for percentage of study participants: body mass index (BMI) – 25%, weight – 22%, height – 27%, smoking status – 66%, smoking years – 47%, HbA1C–63%. Statistically significant values are shown in bold. | |||
Table II depicts the prevalence of comorbid conditions in the study and control groups (19). A statistically significant higher rate of Alzheimer’s disease and other dementias, essential hypertension, and kidney failure was observed in BP patients who received DPP-4i, compared with DM2 patients treated with gliptins but who did not develop BP. In a multivariable analysis after adjusting for age and other covariates, only Alzheimer’s disease and other dementias remained statistically significant for BP development (Table III). We did not observe a significant difference for other conditions, including autoimmune diseases, between the two groups.
Additionally, our results showed that the chronic intake of loop diuretics and thiazides was significantly more prevalent among research group when compared with the control group (Table I). Upon performing a multivariable analysis adjusted for covariates (Table III), the association of these drugs with BP was still significant (p < 0.0001 for thiazides and p = 0.0003 for loop diuretics).
Twenty-two patients (32.35%) developed BP within one year of gliptin intake, 26 patients (38.24%) within one year to 2 years, and 20 patients above 2 years (29.41%). For the remaining 8 patients, there was no information regarding the time period between intake of gliptin and the development of BP in their medical records. The relative proportion of specific gliptin medication varied between the study and control groups (Table IV). Linagliptin and vildagliptin were more prevalent among BP patients than in the control group (32.89% vs 14.31% and 38.16% vs 14.24%, respectively), while sitagliptin had a reduced relative proportion in BP patients (28.95% to 71.45%); these results were all statistically significant.
Using sitagliptin as a reference for comparison, the three different gliptins were also compared using univariate and multivariable logistic regression models (Table III). For the BP group, linagliptin and vildagliptin had statistically significant higher odds than sitagliptin with aORs of 5.25 (p < 0.0001, 95% CI 2.83–9.73) and 6.55 (p < 0.0001, 95% CI 3.64–11.81) respectively.
BP is a severe blistering autoimmune disease that mainly affects elderly patients with significant morbidity and potential mortality. The rise in BP incidence worldwide has been attributed to the increase in life expectancy, as well as improved diagnostic tests and increased usage of specific medications (20, 21). More than 89 drugs have been reported to induce BP, yet robust evidence suggests that DPP-4i carries the greatest risk for BP induction (22–24).Our study examined comorbidities and concomitant medications that might predispose DM2 patients to develop gliptin-associated BP. Delineating the specific risk factors for gliptin-associated BP in DM2 patients can guide the physician when choosing gliptin therapy in susceptible patients and thus help to decrease gliptin-associated BP incidence.
Because adverse effects of DPP-4i are rare, they are used extensively in elderly DM2 patients (25).
Univariate analysis showed that BP patients were significantly older than the control group and had higher prevalence of chronic renal failure, hypertension, and Alzheimer’s disease and other dementias. Upon performing a multivariable analysis with adjustment for age and other covariates only the association of Alzheimer’s disease and other dementias with gliptin-associated BP remained significant. Neurological diseases are known to be intertwined with BP onset (26). Studies have credited this to epitope spreading secondary to a cross-reaction between the neuronal and cutaneous isoforms of BPAg1 (27). The results of our study highlight the increased risk for patients with Alzheimer’s disease and other dementias to develop BP after exposure to gliptins. Despite BP being linked with other autoimmune disorders and skin diseases (15), the results of our study did not find these comorbidities to increase the risk for BP onset among gliptin-treated DM2 patients.
In previous studies, gliptin has been implicated as a predisposing factor for BP even several years after its onset (28).We confirm these findings, with more than 65% of patients developing BP more than 1 year after gliptin initiation.
Chronic intake of thiazides or loop diuretics was found to be an independent risk factor for developing BP among DM2 patients receiving gliptins. This is not surprising as diuretics have long been associated with drug-induced BP (20, 21). However, the additive risk of gliptin consumption when taken concurrently with thiazides or loop diuretics has not been examined. The pathogenesis of drug-induced BP secondary to thiazides and loop diuretics has been linked to the fact that these medications are thiol-based and may disrupt the integrity of the basement membrane through either immunologic or non-immunologic mechanisms (21). Gliptins can elicit BP in patients with underlying susceptibility by modifying the immune response and altering the antigenic properties of the epidermal basement membrane zone. Therefore, gliptins and sulphur-containing drugs may have an additive effect on BP induction (13). In a recent case-control study by Gravani et al. (29), anticoagulants, proton pump inhibitors, and selective serotonin reuptake inhibitors were associated with a significantly higher risk of drug-associated BP. In light of these results, we suggest that patients’ prescribed medications should be thoroughly reviewed. Caution should be advised when prescribing gliptins with thiazides or loop diuretics. Future prospective studies are needed to clarify whether the combination of these drugs has an additive effect on BP induction. Several epidemiological studies have confirmed the link between DPP-4i, particularly vildagliptin, and drug-associated BP risk (10). However, data comparing the risk of BP development among different DPP4i have just begun to emerge. In a study published in 2021 by Jedlowski et al. (30) anagliptin, vildagliptin, and teneligliptin had a higher reporting odds ratio than alogliptin, linagliptin, saxagliptin, and sitagliptin to develop BP. A systematic review and meta-analysis published in 2020 by Phan et al. (31) demonstrated a strong association between vildagliptin and BP. BP has also been linked to linagliptin; however, the association was weaker than with vildagliptin. There was no association between sitagliptin administration and BP. Similar results were also observed in an analysis of cases published by Sun et al. in 2022 (32). Our results are in accordance with previous studies showing linagliptin and vildagliptin had a stronger association with drug-associated BP than sitagliptin. Accordingly, sitagliptin may be a safer option for DM2 patients at risk of developing BP.
The results of our study emphasize the importance of physician vigilance in the treatment of DM2 patients. When considering gliptins for diabetics, attention to patient comorbidities and chronic medications could aid in identifying patients who may be at risk, preventing drug-associated BP and the development of additional unnecessary comorbidities associated with long-term corticosteroid treatment. Our study’s limitations arise from its design being a single-centre retrospective study. The study’s strength lies in its large cohort size of BP patients with previous exposure to gliptins and comparison of different gliptin subtypes.
In conclusion, our study highlights the need for caution when prescribing gliptins in DM2 patients receiving thiazides or loop diuretics and in those suffering from Alzheimer’s disease and other dementias. Sitagliptin should be preferred in these patients over linagliptin and vildagliptin.