Analysis of the clinical characteristics of pembrolizumab-induced bullous pemphigoid.

Background: Pembrolizumab, a programmed cell death protein 1 checkpoint inhibitor, is a novel drug used to treat a variety of advanced malignancies. However, it can also result in many immune-related adverse events, with cutaneous toxicities being the most frequent. Regarding pembrolizumab-induced skin adverse reactions, bullous pemphigoid (BP) has the worst effects on quality of life. Recently, there have been more and more reports of BP incidents resulting from pembrolizumab therapy in patients with cancer. This study aimed to define the clinical characteristics, diagnosis and management of pembrolizumab-induced BP and identify potential differences between classical BP and pembrolizumab-induced BP. Methods: Case reports, case series, and case analyses of pembrolizumab-induced BP up to 10 December 2022 were collected for retrospective analysis. Results: Our study included 47 patients (33 males and 14 females) from 40 studies. The median age was 72 years (range 42-86 years). The median time to cutaneous toxicity was 4 months (range 0.7-28 months), and the median time to bullae formation was 7.35 months (range 0.7-32 months). The most common clinical features were tense bullae and blisters (85.11%), pruritus (72.34%), and erythema (63.83%) on the limbs and trunk. In 20 of the 22 cases tested, the serum anti-BP180 autoantibodies were positive. However, in 10 cases (91.90%, 10/11) the circulating autoantibodies of anti-BP230 were negative. 40 patients had skin biopsies and the skin biopsy revealed subepidermal bullae or blister eosinophil infiltration in 75.00% of patients with pembrolizumab-induced BP, 10.00% of patients with lymphocyte infiltration and 20.00% of patients with neutrophil infiltration. There were 20 patients (50%) with eosinophilic infiltration around the superficial dermis vessels, 8 patients (20.00%) with lymphocyte infiltration around the superficial dermis vessels, and 4 patients (10.00%) with neutrophil infiltration around the superficial dermis vessels. Direct immunofluorescence detected linear immunoglobulin G (IgG) IgG and/or complement C3 along the dermo-epidermal junction in 36 patients (94.74%) with BP. IgG positivity was detected by indirect immunofluorescence in 81.82% of patients with BP. All patients were in complete remission (95.65%,44/46) or partial remission (4.35%, 2/46) of BP, whereas 9/46 patients had a relapse or refractory. The majority of patients achieved BP remission after discontinuation of pembrolizumab with a combination of topically and systemically administered steroid treatments, or other medications. The median duration of BP remission was 2 months (range 0.3-15 months). Conclusion: A thorough diagnosis of pembrolizumab-induced BP should be made using clinical signs, biochemical markers, histopathological and immunopathological tests. Pembrolizumab-induced BP had similar clinical characteristics to classic BP. Temporary or permanent discontinuation of pembrolizumab therapy may be required in patients with perbolizumab-induced BP depending on the severity of BP and the response to medication. Pembrolizumab-induced BP may be effectively treated using topical and systemic steroid treatments in combination with other medications (e.g., doxycycline, niacinamide, dapsone, rituximab, intravenous immunoglobulins, dupilumab, cyclophosphamide, methotrexate, mycophenolate mofetil, and infliximab). Clinicians should provide better management to patients with BP receiving pembrolizumab to prevent progression and ensure continuous cancer treatment.