Six Cases of Pemphigoid Nodularis during Treatment with Dipeptidyl Peptidase-4 Inhibitors

Abstract

Introduction

We investigated 6 cases of pemphigoid nodularis under treatment with dipeptidyl peptidase-4 (DPP-4) inhibitors.

Case Representation

In 3 cases, only nodules were observed at the initial presentation, but bullous lesions appeared during the course of the disease. All patients had diabetes mellitus and were taking DPP-4 inhibitors from 1 to 1.5 years. Anti-BP180Nc16a antibodies were positive in 3 cases, and low or negative in 3 cases. Full-length BP180 antibodies were analyzed in 2 of the 6 cases and were positive in both. Since patients with diabetes mellitus occasionally develop prurigo nodularis, it is important not to misdiagnose pemphigoid nodularis as prurigo nodularis.

Conclusion

There are few reports of pemphigoid nodularis under DPP-4 inhibitors. It should be noted that not only bullous pemphigoid but also pemphigoid nodularis can be induced by DPP-4 inhibitors.

Introduction

Pemphigoid nodularis (PN) has been reported to be a rare subtype of bullous pemphigoid (BP) [1]. PN is characterized by generalized pruritic nodules and may appear alone or in combination with bullous, erosive, and erythematous lesions of typical BP. Some reports have suggested that, compared with typical prurigo nodularis, PN is more often accompanied by exudates, crusts, and relatively small blisters up to the size of the head of the thumb [2]. Some cases present only with nodules and no blisters, whereas others show a delayed appearance of blisters, which may depend on the timing of diagnosis. Furthermore, in some cases, blisters precede nodules, whereas in others, the 2 lesion types are present concurrently at the initial diagnosis, making it unclear which appeared first.

Clinically resembling PN is characterized by nodular lesions with intense pruritus on the head, back, and extremities. Histopathological examination typically reveals thickening of the stratum corneum and epidermis as well as lymphocytes and eosinophil infiltrations in the superficial dermis. Direct immunofluorescence (DIF) findings in PN typically show linear deposition of IgG and C3 in the basement membrane, which is consistent with BP.

BP has been frequently reported in association with dipeptidyl peptidase-4 (DPP-4) inhibitors [3]. By contrast, association of PN and DPP-4 inhibitors is less frequent. In the present study, we report 6 cases that developed during treatment with DPP-4 inhibitors for diabetes mellitus.

Case Presentation

In total, there were 6 cases (4 males and 2 females) that developed PN under treatment with DPP-4 inhibitors. Patients’ ages ranged 53–90 years, with a median age of 63.5 years. The characteristics of the 6 patients are shown in Table 1. All patients had been taking DPP-4 inhibitors from 1 to 1.5 years, when they were diagnosed as BP or PN.

Table 1.

Summary of our six cases of PN

​	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Age	53	75	60	54	90	67
Sex	M	F	M	F	M	M
Comorbidity	Diabetes, hypertension, heart failure	Diabetes, hypertension	Diabetes, hypertension, renal failure	Diabetes, dyslipidemia	Diabetes, lacunar infarction, Alzheimer-type dementia	Diabetes, hypertension, renal failure, psoriasis
Clinical findings at first visit	Nodules, bullae	Nodules, erythema	Nodules, bullae	Nodules, erythema	Nodules	Nodules, erythema, bullae
Preceding lesion	Nodules	Nodules	Bullae	Nodules	Nodules	Bullae
Interval between preceding lesion and new eruption	1 month	About 3 year	Unknown	1.5 year	Unknown	Unknown
DPP-4 inhibitor	Teneligliptin	Teneligliptin	Linagliptin	Sitagliptin	Vildagliptin	Linagliptin
Duration from DPP-4 inhibitor initiation to eruption	Unknown	1.5 year	1 year	1 year	Unknown (over 5 years)	Unknown
Anti-BP180 NC16a antibody, U/mL	<3.0	6.3	>10,000	59.4	5.5	641
Full-length BP180 antibody, U/mL	74.61	61.36	ND	ND	ND	ND
Anti-BP230 antibody, U/mL	<5.0	<5.0	54	79	19	67
Eosinophil count, /μL	375	845	4,653	738	189.2	1,584
IgE, IU/mL	573	418	10,100	72.1	11,150	10,200
BPDAI	Bullae/erosion: 15	Bullae/erosion: 24	Bullae/erosion: 27	Bullae/erosion: 21	Bullae/erosion: 31	Bullae/erosion: 44
	Erythema/urticaria: 0	Erythema/urticaria: 27	Erythema/urticaria: 30	Erythema/urticaria: 30	Erythema/urticaria: 9	Erythema/urticaria: 23
	Mucosa: 0	Mucosa: 0	Mucosa: 2	Mucosa: 0	Mucosa: 0	Mucosa: 0
DIF	Linear deposition of IgG and C3 along the basement membrane	Linear deposition of IgG and C3 along the basement membrane	Linear deposition of IgG and C3 along the basement membrane	Linear deposition of IgG and C3 along the basement membrane	Linear deposition of IgG and C3 along the basement membrane	Linear deposition of IgG and C3 along the basement membrane
Treatment	PSL	PSL	PSL, AZP	PSL	PSL	PSL

The clinical presentation was mixed erythema and nodules in some cases and nodules alone in others (Fig. 1a, b). Only nodular lesions were present at the initial presentation in one case (case 5). Mixed nodules, blisters, and erythema were present in 5 cases (cases 1, 2, 3, 4, and 6). There were no cases with blisters alone. Case 5 had multiple nodules on the trunk and extremities; cases 1 and 3 had multiple pruritic nodules on the trunk and extremities with scattered erosions, tense blisters, and nodules; and cases 2, 4, and 6 also had erythematous edema. Of the 3 cases with no blisters at the time of initial presentation, blisters appeared during the course of the disease. The interval between the initial lesions and blister development ranged from 1 month to 3 years, although the interval was unknown in 3 cases. Among the 3 patients who presented with mixed nodules and blisters at the initial presentation, one had a skin rash that developed from nodules to blisters before presentation, whereas in the other 2 cases, the sequence of lesion development was unknown.

Fig. 1.

Representative clinical features. a Pruritic erythema is observed on the chest and abdomen, and multiple red papules and nodules are observed on the anterior trunk and upper arms (case 2). b Multiple prurigo nodules are present on the buttocks (case 3).

The Bullous Pemphigoid Disease Area Index (BPDAI) at the first visit ranged from 15 to 44 (mean 27) for blisters/erosions and from 0 to 30 (mean, 19.8) for wheals/erythema. Only 1 case was classified as severe, while 5 cases were classified as moderate. Mucocutaneous rash was present in only 1 case (case 3).

Skin biopsies from nodules were performed in 2 cases, while the remaining 4 cases underwent biopsy from blisters. Histopathological examination of the nodular lesions revealed liquefaction degeneration in the epidermal basal layer in 2 cases, lacunae immediately beneath the epidermal basal layer in one case, and inflammatory cell infiltration, mainly lymphocytes, in the upper dermis, with eosinophils present in 2 cases (Fig. 2a). Skin biopsy tissue from blisters showed subepidermal blisters and infiltration of lymphocytes and eosinophils within the blisters and in upper dermis. DIF showed linear deposition of IgG and C3 in the epidermal basement membrane in all cases (Fig. 2b). In case 5, linear IgE deposition was also observed along the epidermal basement membrane zone (Fig. 2c).

Fig. 2.

Representative histopathological findings. a The epidermis is irregularly thickened, and the stratum corneum is also thickened. Inflammatory cell infiltration, predominantly lymphocytes, is observed in the upper dermis (case 3) (HE. stain ×40). b DIF showing linear deposition of IgG along the epidermal basement membrane (case 6). c DIF showing linear deposition of IgE along the epidermal basement membrane (case 6).

Complications included hypertension in 4 cases, with one case each of heart failure, renal failure, and psoriasis vulgaris. The orally administered DPP-4 inhibitors were linagliptin in 2 cases, teneligliptin in 2 cases, sitagliptin in 1 case, and vildagliptin in 1 case. The time from initiation of DPP-4 inhibitor treatment to the appearance of skin rashes ranged from 1 to >5 years, excluding 2 cases with unknown durations.

Peripheral blood eosinophil counts ranged 189.2–4,653/µL (mean, 791.5/µL), and 4 of the 6 patients had hyper-eosinophilia of 500/µL or more. Serum IgE levels ranged 72.1–11,150 IU/mL (mean 5,336.5 IU/mL), and 5 out of 6 patients had elevated IgE levels above the reference value (170 IU/mL).

Three patients had positive antibody titers, with anti-BP180Nc16a antibody levels ranging 3.0–10.0 U/mL; titers below 3.0 U/mL were considered negative. The remaining 3 patients had low or negative anti-BP180Nc16a titers; however, 2 of these patients were tested for full-length BP180 antibodies, and both were positive. Serum anti-BP230 antibodies were positive in 4 cases. All patients discontinued taking the DPP-4 inhibitors and were simultaneously treated with glucocorticoids (prednisone 15–60 mg/day); five received oral steroids alone, while one received additional azathioprine.

Discussion

The precise pathogenesis of PN remains unclear; however, PN may be triggered by drugs or ultraviolet (UV) radiation, as well as in association with several comorbidities such as rheumatoid arthritis and diabetes mellitus [4]. External stimuli such as scratching can cause capillary dilation and increased vascular permeability, leading to increased deposition of immunoglobulins and complement in the epidermal basement membrane [1]. Alternatively, antigenic denaturation of the epidermal basement membrane is induced in the course of remodeling process of repetitive scratching, leading to the production of anti-basement membrane antibodies [1]. Shimauchi et al. [5] speculated that low titers of anti-basement membrane antibodies may not cause blister formation, but instead promote epidermal proliferation and dermal remodeling, resulting in the development of pruritic nodules. In the present study, blister formation was observed after the initiation of phototherapy in one case (case 2), whereas in other cases, the triggering factors for blistering were unclear. There have been scattered reports of PN in which blisters appeared following UV therapy, and the proposed mechanism is that UV irradiation damages the epidermal basement membrane, resulting in antigenic alteration or antigen exposure, and subsequent autoantibody production [6].

Since the launch of DPP-4 inhibitors in Japan in 2009 and the first report of DPP-4 inhibitor-induced BP in 2011 [7], numerous cases have been reported both in Japan and internationally [3]. By contrast, there are only several reports of PN occurring while taking DPP-4 inhibitors. To date, 13 cases of DPP-4 inhibitor-associated PN have been reported in Japan and internationally, including our 6 cases (Table 2) [8–13]. The cohort comprised 7 males and 6 females, with a mean age of 69.1 years. Six patients were positive for anti-BP180Nc16a antibodies, whereas 7 had low or negative titers, including 4 who were positive for full-length BP180 antibodies. The DPP-4 inhibitors involved were vildagliptin in 4 cases, teneligliptin in 4 cases, sitagliptin in 2 cases, sitagliptin in 2 cases, and alogliptin in 1 case. The duration of medication ranged 12–42 months (mean 23.6 months), with a tendency for disease onset after long-term use; duration was unknown in 5 cases. Clinically, 5 patients initially presented with nodules alone, whereas 6 presented with a combination of nodules and blisters or erythema. In 10 of the 13 patients, nodules preceded the appearance of blisters, and the interval nodule and blister onsets ranged 1–36 months (mean, 16.6 months); the duration was unknown in 5 cases.

Table 2.

Previous reports of PN that developed during DPP-4 inhibitor therapy

Case	Author	Age	Sex	DPP-4 inhibitor	Duration of intake	Clinical symptoms at first visit	Course (duration)	BP180Nc16a antibody, U/mL	Full-length BP180 antibody, U/mL
1	Tsuchiyama et al. [11]	55	F	Alogliptin	30 months	Nodules, bullae	Nodules → bullae (unknown)	Positive	Not examined
2	Sugimoto et al. [9]	74	F	Sitagliptin	15 months	Nodules, bullae	Nodules → bullae (unknown)	<3.0	Positive
3	Okudaira et al. [12]	67	F	Teneligliptin	42 months	Nodules, bullae	Nodules → bullae (34 months)	<3.0	26.31
4	Murashima et al. [13]	83	M	Vildagliptin	Unknown	Nodules	Nodules → bullae (unknown)	173	Not examined
5		74	M	Vildagliptin	Unknown	Nodules	Nodules → bullae (8 months)	<3.0	Not examined
6	Tahara et al. [10]	76	M	Vildagliptin	32 months	Nodules	Nodules → bullae (32 months)	4.5	Not examined
7	Ichimura et al. [8]	70	M	Teneligliptin	28 months	Nodules	Nodules → bullae (2 months)	67.8	Not examined
8	Present case (case 1)	53	M	Teneligliptin	Unknown	Nodules, bullae	Nodules → bullae (1 month)	<3.0	74.61
9	Present case (case 2)	75	F	Teneligliptin	18 months	Nodules, erythema	Nodules → bullae (36 months)	6.3	61.36
10	Present case (case 3)	60	F	Linagliptin	12 months	Nodules, bullae	Unknown	>10,000	Not examined
11	Present case (case 4)	54	F	Sitagliptin	12 months	Nodules, erythema	Nodules and erythema simultaneously	59.4	Not examined
12	Present case (case 5)	90	M	Vildagliptin	Unknown (>5 years)	Nodules	Nodules → bullae (3 months)	5.5	Not examined
13	Present case (case 6)	67	M	Linagliptin	Unknown	Nodules, bullae, erythema	Unknown	641	Not examined

In our cases, 4 of 6 patients (66.7%) had hyper-eosinophilia (≥500/µL) and 5 of 6 patients (83.3%) had elevated IgE levels, including 4 patients who exhibited both findings. Ngan et al. [14] reported elevated IgE levels in 60% of patients with BP and a correlation between serum IgE levels and eosinophil count. However, no such correlation was observed in our cases.

Histopathological examination of biopsies from the nodules showed liquefaction degeneration in both cases and fissure formation at the epidermal basement membrane in 1 of the 2 cases. Since liquefaction degeneration and subepidermal bulla formation are sometimes obscure in PN, the condition is sometimes misdiagnosed as prurigo nodularis. In addition, both PN and prurigo nodularis showed similar integrin expression patterns, suggesting a shared epidermal hyperproliferative state [15]. Therefore, it is important to perform DIF in parallel with HE staining. In our case series, IgG and/or C3 deposition was observed on DIF in all cases, and linear IgE deposition in the epidermal basement membrane zone was additionally observed in 1 of the 6 cases. Yayli et al. [16] reported that IgE deposition was observed in the epidermal basement membrane zone in 41% of patients with BP. Kamata et al. [17] compared cases with and without IgE deposition in the epidermal basement membrane zone as seen by DIF and reported that cases with IgE deposition had significantly higher BPDAI values. Nishie [18] suggested that, in PN, IgE may induce cytokine release from basophils, contributing to the development of pruritic nodules, whereas IgE deposition on the basement membrane may lead to blister formation [18].

Charneux et al. [19] reported that 86% of patients with BP tested positive for BP180 ELISA, and 59% tested positive for BP230 ELISA. Mai et al. [3] also reported that the BP230 antibody positivity rate in DPP-4 inhibitor-associated BP was low at 11.1% [20]. There have been reports of BP180 antibodies being negative and BP230 antibodies being positive in nodular pemphigoid [3]. Therefore, when BP180 antibodies are negative, it is necessary to also measure BP230 antibodies. In addition, Izumi et al. [21] identified many cases that were positive for full-length BP180 antibodies. In our case series, anti-BP230 antibodies were positive in 4 out of 6 cases, and the 4 positive cases also had elevated anti-BP180 Nc16a antibody levels.

DPP-4 inhibitors have been associated with an increased risk of BP, with several epidemiological studies demonstrating a two- to four-fold increase in risk [22]. In our 6 cases studied, since discontinuation of the DPP-4 inhibitor was accompanied by the simultaneous initiation of prednisone, there were no cases in which improvement was achieved by discontinuing the DPP-4 inhibitor alone. The mechanism of DPP-4 inhibitor-induced BP remains unclear, but it is suspected that DPP-4 inhibitors modulate immune responses by inhibiting DPP-4 (CD26) activity, which is expressed on lymphocytes and macrophages [9]. DPP-4 also suppresses the activity of cytokines and chemokines such as IL-1α, IL-6, vascular endothelial growth factor, and thrombopoietin. Therefore, DPP-4 inhibitors are thought to enhance the activities of these molecules and induce inflammation. It has been suggested that DPP-4 inhibitors may promote the release of cytokines responsible for both blister formation and tissue injury [9, 23]. The mechanism of DPP-4 inhibitor-induced nodular lesion formation may involve a Th2-dominant cytokine profile and type 2 inflammation. In prurigo nodularis, recent studies have implicated that various cytokines, such as IL-4, IL-13, and IL-31, play important roles in the induction of epidermal hyperplasia and dermal fibrosis. In recent years, there has been growing recognition that BP is a Th2-dominant inflammatory disease, and targeted therapies such as dupilumab and Janus kinase inhibitors have emerged as promising treatment options, particularly for refractory or steroid-dependent cases [23, 24]. Dupilumab has minimal immunosuppressive effects and may be beneficial for elderly patients and those with compromised immune systems [24]. Janus kinase inhibitors have shown promising efficacy in BP by suppressing multiple cytokine pathways involved in type 2 inflammation, although current evidence is limited to case reports and small case series [4]. Further investigation is needed to clarify the pathogenesis of PN induced by DPP-4 inhibitors. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000552572).

Acknowledgment

We would like to thank the Department of Dermatology, Hokkaido University, for performing the full-length BP180 antibody assay.

Statement of Ethics

This study protocol was reviewed and approved by the Research Ethics Committee of Fukushima Medical University, Approval No. 2024-089. Written informed consent was obtained from all patients regarding publication of the details of their medical case and any accompanying images.

Conflict of Interest Statement

The authors have no conflicts of interest to declare. Toshiyuki Yamamoto was a member of the journal’s Editorial Board at the time of submission.

Funding Sources

The authors do not have any research relevant to our study.

Author Contributions

Misaki Kusano and Aki Honda collected the data, and Natsumi Norikawa wrote the initial manuscript. Toshiyuki Yamamoto evaluated and revised the manuscript, provided critical feedback, and contributed to the final version of the manuscript.

Funding Statement

The authors do not have any research relevant to our study.

Data Availability Statement

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.