Chimeric antigen receptor (CAR)-T-cell therapy, particularly CD19-targeted agents such as tisagenlecleucel, has revolutionized and is increasingly used for the treatment of refractory B-cell malignancies [1, 2]. However, it carries a common risk of immune-mediated adverse events called cytokine release syndrome (CRS), presenting with a variety of symptoms ranging from mild, flu-like symptoms to severe life-threatening manifestations of the overshooting inflammatory response, which onset within a few days up to several weeks after infusion [3]. Besides rash and edema as cutaneous symptoms of CRS, patients undergoing CAR-T-cell therapy develop skin lesions such as secondary cutaneous malignancies, disseminated infection, and CAR-T-cell therapy-related ‘eruptions of lymphocyte recovery’ [4]. In addition, erythematous, purpuric and vesiculobullous lesions with substantial inflammation have been reported as cutaneous adverse events (AEs) of CAR-T-cell therapy [5–10].

Here we report an uncommon case of subepidermal bullous eruption with minimal inflammation, transient drug hypersensitivity and a favorable clinical course, potentially reflecting T-cell dysregulation due to post-CAR-T immune remodeling. We also compare our case with prior reports.

A 70-year-old male with diffuse large B-cell lymphoma was administered with tisagenlecleucel, a CD19-directed CAR-T therapy. He was prophylactically administered with sulfamethoxazole-trimethoprim, acyclovir, fluconazole, esomeprazole magnesium hydrate, and magnesium oxide. On 2 months post-treatment, he was consulted to our department for multiple tense bullae on the extremities, including a 4 × 5 cm blister on the right ankle, and facial erythema (Figures 1a,b). He claimed no subjective symptoms such as pruritus associated with the skin eruptions. No history of bullous skin was noted. The patient developed a high fever up to 39 °C from day 2 after infusion, which was considered grade 1 CRS. Tocilizumab was administered on days 3 and 4, and the fever resolved gradually, without any neurological complications. The CRP peaked at 17.43 mg/dL on day 4 and normalized by day 13. The bullous eruptions appeared on day 58, approximately 1 month after the resolution of obvious CRS. Blood tests revealed eosinophilia (11.5%), elevated LDH (271 U/L), and low IgG (294 mg/dL). CRP was not elevated and CD19⁺ B cells were undetectable. Anti-BP180 antibody was negative. Skin biopsy showed subepidermal blistering with red blood cell extravasation and minimal inflammatory cell infiltration (Figure 2). No deposition of IgG, IgA, or C3 was detected by direct immunofluorescence. Immunohistochemistry showed mild perivascular infiltration of T lymphocytes, predominantly CD4⁺ cells rather than CD8⁺ cells, with no CD19⁺ B cells. Small number of CD8⁺ T cells were also detected in the epidermis close to the blister.

Drug-induced lymphocyte stimulation tests (DLSTs), also referred to internationally as the lymphocyte transformation tests (LTTs), were strongly positive for all five drugs (sulfamethoxazole-trimethoprim, acyclovir, fluconazole, esomeprazole magnesium hydrate, and magnesium oxide) initially, but gradually diminished over time after discontinuation of all drugs (Table 1). While skin symptoms relapsed at 4 months after the first visit, anti-BP180 antibody again showed negative and DLST revealed positive for only sulfamethoxazole-trimethoprim and magnesium oxide. Closed patch testing was negative for sulfamethoxazole-trimethoprim on both post-lesional and non-lesional skin. Simple coverage of bullae with adhesive plasters and occasional topical betamethasone led to resolution without scarring or milia formation. No systemic immunosuppressants were administered and the facial erythema resolved without topical intervention. At 15-month follow-up, DLSTs turned completely negative and the patient remained in remission without significant recurrence. Drug use was initially restricted strictly but relieved after improvement.

We have reported a case of subepidermal bullous eruption with minimal inflammation, who also showed transient multiple drug hypersensitivity and a favorable clinical course. Although a sulfamethoxazole-trimethoprim-induced non-scarring bullous drug eruption had been reported [11], multiple positivity on DLST and negative closed patch test reduce the possibility of the allergic drug eruption. Along with the bullous drug eruption, autoimmune bullous disease and the side effect of CAR-T-cell therapy should be considered for differential diagnosis. As autoimmune bullous disease can be denied by negative direct immunofluorescence, we consider that this case represents the side effect of CAR-T-cell therapy. However, current knowledge on the pathogenesis of dermatologic bullous change associated with CAR-T-cell therapy is limited. Our case differs from so far reported cutaneous AEs of CAR-T-cell therapy developing bullous eruptions, which all exhibit intense inflammation and/or necrosis (Table 2). Although CRS commonly occurs after CAR-T-cell therapy and can present with cutaneous manifestations, in this case, initial CRS was grade 1 and resolved within 2 weeks after infusion. As the bullous eruptions developed about 1 month later, late-onset CRS, rather than direct contribution of the initial CRS, may be involved. However, it has not been clarified by a measurement of cytokine levels. The patient initially achieved disease control, but residual FDG uptake on PET imaging suggested possible early relapse, indicating that mild CRS in this case might predict temporal remission. This may be consistent with previous studies, suggesting that the severity of CRS may correlate with the efficacy of CAR-T-cell therapy, as robust immune activation could reflect higher CAR-T expansion and anti-tumor activity [2, 3]. Further studies are needed to clarify these associations. Although the overall inflammatory infiltrate was minimal, immunohistochemistry revealed a mild predominance of CD4⁺ over CD8⁺ T cells. This suggests that limited T-cell activation might have contributed to the pathogenesis, despite the lack of massive inflammation. Given the delayed timing of eruption after CAR-T infusion, we hypothesize that localized T-cell reactivity, possibly associated with immune reconstitution, may have played a role. This may explain the initial DLST hyperresponsiveness, which normalized over time, in accordance with the change of CD8⁺ T cell ratio in blood lymphocytes (Table 1). In this context, our case is considered to be another example of CAR-T-cell therapy-related ‘eruptions of lymphocyte recovery’ [4].

While trauma and diabetic bullae were clinically excluded, the localized nature of bullae, including a solitary large blister over the medial malleolus, suggests participation of any site-specific factors. Although fixed drug eruptions (FDEs), especially non-pigmented bullous FDEs, were suspected in this case, patch testing was negative for the most suspected drug, sulfamethoxazole-trimethoprim, even on the lesional skin. Regarding sulfamethoxazole-trimethoprim, a patient was reported who rapidly developed nonscarring sublamina densa blisters within hours after the intravenous administration [11]. In our case, histological analysis revealed that some basal cells still attached to the dermis after the blister formation, suggesting that the blister is not formed sublamina densa (Figure 1b). Although we did not perform a formal drug provocation test, suspected drugs were re-administered after symptom improvement, but importantly, no recurrence of the skin lesions was observed following re-exposure. This clinical course indirectly supports the notion that the skin lesions were unlikely to be caused by the suspected drugs.

This report contributes to the understanding of atypical cutaneous AEs of CAR-T-cell therapy. Since not all bullous eruptions post-CAR-T represent drug allergies or autoimmune disease, a careful work-up including histopathology, immunostaining, and DLST interpretation in temporal context is crucial. This case highlights a rare presentation of post-CAR-T subepidermal blistering with minimal inflammation and transient DLST positivity. Lack of early immunophenotyping, cytokine measurements and inability to perform drug rechallenges are listed as limitations. We advocate for heightened clinical awareness of immune reconstitution–related cutaneous reactions.