Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are severe cutaneous adverse reactions (SCARs) characterized by extensive epidermal necrosis, blistering, and mucosal involvement. These disorders often have an abrupt onset and rapid progression, with high morbidity and mortality, underscoring the importance of early recognition and multidisciplinary management (1). Epidemiological studies indicate that the mortality rate is 4.8% for Stevens-Johnson syndrome (SJS), 19.4% for SJS-TEN overlap, and 14.8% for toxic epidermal necrolysis (TEN) (2). Since the approval of the first immune checkpoint inhibitor (ICI), the cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) antibody ipilimumab, in 2011, ICIs targeting PD-1, PD-L1, and CTLA-4 have revolutionized cancer therapy and significantly improved survival outcomes across multiple tumor types (3). However, their potent immune activation may also trigger immune-related adverse events (irAEs), among which cutaneous manifestations are the most frequent. While most dermatologic irAEs are mild and reversible, a small subset may progress to life-threatening SCARs such as SJS/TEN (4–6). With the expanding use of PD-1/PD-L1 inhibitors, increasing reports of ICI-related SJS/TEN have emerged in the literature. These cases may differ from classical drug-induced SJS/TEN in latency, clinical features, histopathology, and treatment response. Nevertheless, most published data are limited to individual case reports, and comprehensive analyses remain scarce. Therefore, our study retrospectively analyzed 21 patients with ICI-related SJS/TEN to summarize their clinical characteristics, laboratory findings, therapeutic approaches, and outcomes. By providing real-world evidence from a relatively larger cohort, this study aims not only to enhance clinical recognition and management of this rare but potentially fatal complication but also to contribute to broader awareness among clinicians and stakeholders and to support the improvement of post-marketing surveillance strategies for immune checkpoint inhibitors.

In normal immune responses, T-cell activation and inhibition are tightly regulated by multiple immune checkpoints to maintain immune homeostasis and prevent autoimmunity. Tumor cells frequently exploit these inhibitory pathways to escape immune surveillance, among which the cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) and programmed cell death protein 1 (PD-1) pathways are the most representative. Both CTLA-4 and PD-1 act as negative regulators of T-cell immunity but differ in their sites and timing of action (9, 10). CTLA-4 primarily functions during the early stage of T-cell activation within lymphoid organs by competitively inhibiting CD28-CD80/CD86 interactions, thereby suppressing T-cell proliferation and clonal expansion. In contrast, PD-1 mainly operates during the effector phase in peripheral tissues, where its ligand PD-L1–widely expressed on tumor cells and interferon-γ-stimulated immune cells—transmits inhibitory signals that lead to T-cell exhaustion, reduced cytokine production, and weakened antitumor activity (11–13). Immune checkpoint inhibitors (ICIs) block these inhibitory signaling pathways, effectively releasing the “brakes” on T cells and restoring antitumor immunity. Anti–CTLA-4 antibodies (such as ipilimumab) primarily enhance early T-cell priming, while anti–PD-1/PD-L1 antibodies (such as nivolumab and pembrolizumab) sustain effector T-cell responses within peripheral tissues. These therapies have achieved remarkable and durable efficacy in several malignancies, including melanoma, non–small cell lung cancer, and hepatocellular carcinoma (14, 15).

The precise pathogenesis of ICI-induced SJS/TEN has not been fully elucidated. Goldinger et al. (16) performed immunohistochemical analysis on six patients who developed immune-related cutaneous adverse events (ircAEs) following PD-1 inhibitor therapy. They observed CD8+ T-cell infiltration at the dermo-epidermal junction in all rash tissue samples, along with upregulated expression of PD-1 on lymphocytes and PD-L1 on keratinocytes. The authors proposed that PD-1 inhibitors disrupt the homeostatic PD-1/PD-L1 interaction, thereby inducing severe cutaneous adverse reactions. Research published in Nature further delineates a key pathogenic axis in ICI-induced SJS/TEN, centered on macrophage-derived CXCL10 and TNF signaling. This mechanistic framework also explains why TNF-α inhibitors may demonstrate superior therapeutic effects compared with glucocorticoids–primarily through the disruption of CXCL10-driven CD8+ T-cell activation, thereby preventing blister formation and accelerating epidermal repair (17). Another recent research has established that a shared Th2-dominant immune mechanism underlies the pathogenesis of both lichenoid and eczematous eruptions associated with immune checkpoint inhibitors (18). Ellis et al. (19) further identify that ICI-induced SJS/TEN is characterized by a dominant CD8+ T-cell infiltrate, compensatory PD-L1 upregulation, and a unique cytotoxic/chemokine gene signature, directly linking its pathology to the mechanism of immune checkpoint inhibition.

In this study, all patients with immune checkpoint inhibitor (ICI)–related SJS/TEN discontinued immunotherapy immediately after diagnosis and received comprehensive systemic and multidisciplinary management. The main therapeutic measures included systemic corticosteroids, intravenous immunoglobulin (IVIG), and supportive and symptomatic treatments, such as anti-infective therapy, nutritional support, and maintenance of fluid and electrolyte balance. In addition, some patients were treated with topical corticosteroids or antihistamines to alleviate cutaneous symptoms.

In terms of therapeutic strategies, corticosteroids remain the cornerstone of SJS/TEN management. Although some studies have suggested that high-dose corticosteroid therapy may increase the risk of secondary infections and therefore should be used with caution, an increasing body of evidence supports that early and adequate corticosteroid administration can effectively suppress immune-mediated inflammation, promote epidermal healing, and reduce mortality, without significantly prolonging recovery time (20). Most patients shown marked improvement in cutaneous and mucosal lesions after systemic corticosteroid therapy; for those with insufficient response, adjunctive agents such as cyclosporine, TNF-α inhibitors, or intravenous immunoglobulin (IVIG) can be added to enhance therapeutic outcomes.

Once SJS/TEN is confirmed, immune checkpoint inhibitors should be permanently discontinued, and corticosteroid therapy (prednisone or methylprednisolone, 1–2 mg/kg·d) should be initiated as early as possible. Patients are advised to be managed in an intensive care unit or burn unit under multidisciplinary collaboration involving dermatology, ophthalmology, urology, and otorhinolaryngology. For patients at risk of gastrointestinal bleeding, corticosteroid dosage should be carefully evaluated in close consultation with gastroenterology specialists to balance anti-inflammatory efficacy and hemostatic safety (21, 22).

In addition to corticosteroids, intravenous immunoglobulin (IVIG) represents an important therapeutic option. IVIG can inhibit Fas/FasL-mediated keratinocyte apoptosis, suppress the release of inflammatory cytokines, and exert immunomodulatory effects through toxin neutralization. It is recommended as an adjunctive therapy at a dosage of 0.4 g/kg·d (21, 23, 24). In our cohort, most patients received combined therapy with corticosteroids and IVIG, and several severe cases achieved favorable recovery under multidisciplinary team (MDT) management. Taken together, these observations suggest that early and coordinated intervention with corticosteroids plus IVIG may contribute meaningfully to improved clinical outcomes in patients with ICI-related SJS/TEN (25).

In recent years, targeted therapies aimed at key inflammatory pathways in severe cutaneous adverse reactions have advanced rapidly, providing more controllable immunomodulatory strategies for patients with inadequate responses to systemic corticosteroids or those with underlying malignancies who develop ICI-related SJS/TEN. Cyclosporine, which rapidly suppresses drug-specific T-cell activation, has been shown to shorten re-epithelialization time and reduce mortality, and is now recommended as a steroid-sparing first-line option in several international guidelines (26–29). TNF-αinhibitors represent the most well-supported class of biologics; TNF-α plays a central role in keratinocyte apoptosis and inflammatory amplification, and multiple studies have demonstrated that etanercept or infliximab combined with systemic corticosteroids significantly shortens re-epithelialization time, reduces gastrointestinal bleeding, and does not increase infectious risk, with similarly favorable outcomes reported in ICI-related SJS/TEN (30–32). Furthermore, JAK inhibitors (e.g., selective JAK1/3 or broader JAK1/2 agents) have shown promising preclinical and early clinical activity in aborting IFN-γ/IL-15–driven cytotoxic programs and accelerating skin recovery in severe cases, although careful monitoring for infectious and thromboembolic risks is required (33–35). Overall, the integration of targeted biologics, JAK inhibitors, and conventional immunosuppressive therapy may allow more precise modulation of pathogenic pathways, accelerated skin repair, and improved outcomes, although their long-term safety and optimal sequencing in the setting of ICI-related SJS/TEN require further validation (36).

Furthermore, special attention should be paid to one patient in our study who developed TEN and ultimately died after receiving the PD-1/CTLA-4 bispecific antibody (Ipilimumab-Tovoralimab, as presented in Table 1). This case highlights that simultaneous blockade of the PD-1 and CTLA-4 immune checkpoint pathways may not only potentiate antitumor immunity but also markedly amplify autoreactive immune responses, leading to severe or even fatal toxicities. Previous studies have shown that combination therapy with PD-1/PD-L1 and CTLA-4 inhibitors can induce deeper tumor responses in some patients, but it is also associated with substantially higher rates of severe adverse events, immune-related toxicities, and treatment discontinuation compared with PD-1/PD-L1 monotherapy (37, 38). Extending this therapeutic concept, bispecific antibodies enable dual checkpoint inhibition within a single molecular framework, thereby providing even stronger immune stimulation. However, such intensified immune activation may simultaneously erode self-tolerance, increasing susceptibility to fulminant immune reactions such as SJS/TEN (39).

This fatal case suggests that the toxicity profile of dual immune checkpoint blockade may differ substantially from that of monotherapy, with potential survival benefits accompanied by unprecedented risks. For patients receiving such potent immune combination therapies (including bispecific antibodies or combination regimens), clinicians should perform comprehensive risk assessments before treatment, maintain heightened vigilance throughout therapy, and promptly recognize and manage even mild cutaneous or mucosal symptoms to improve clinical outcomes.

Overall, the clinical outcomes in our cohort–including a 14.3% mortality rate and generally favorable responses to corticosteroids combined with IVIG–align with previous reports of ICI-related SJS/TEN, which similarly describe high SCORTEN scores, substantial morbidity, and rapid disease escalation (40, 41). Our findings further reinforce emerging evidence that dual checkpoint blockade or bispecific antibodies may confer heightened severity, likely due to amplified T-cell activation that lowers immune tolerance and predisposes to fulminant cutaneous injury. Nevertheless, several limitations should be acknowledged. The rarity of ICI-related SJS/TEN resulted in a relatively small sample size, which may restrict generalizability. Skin biopsies were not performed in all patients because of critical illness and concerns regarding infection risk, potentially affecting diagnostic certainty. In addition, the retrospective design and heterogeneity in cancer types, treatment history, and supportive care introduce unavoidable bias and may cloud interpretation of therapeutic efficacy. Despite these constraints, our study provides clinically relevant insights that extend beyond individual patient management. The findings highlight the need to strengthen clinician awareness of severe cutaneous irAEs during ICI therapy and emphasize the importance of robust post-marketing pharmacovigilance systems to promote early recognition, standardized reporting, and timely intervention for rare but catastrophic toxicities. Looking ahead, future research should prioritize the establishment of multicenter prospective registries to improve case capture and refine estimates of incidence, risk factors, and outcomes. Mechanistic studies–such as immunogenomic profiling and microenvironmental characterization of lesional skin–will be critical to elucidate the pathways linking checkpoint inhibition to epidermal necrosis. At the policy level, strategies such as mandatory reporting frameworks, harmonized dermatologic monitoring protocols, and structured guidance for high-risk individuals could further enhance patient safety. Ultimately, a deeper understanding of the clinical trade-off between antitumor efficacy and severe immune-related toxicity is essential. While combination ICIs and bispecific antibodies may yield superior tumor control, they also appear to heighten vulnerability to life-threatening dermatologic reactions. Balancing these competing considerations will require thoughtful patient selection, individualized treatment intensity, and vigilant surveillance–particularly for those receiving intensified checkpoint inhibition.

This study has several limitations. Its single-center, retrospective nature may limit the generalizability of the findings. The small cohort size (n=21) restricts statistical power and precludes meaningful subgroup analyses. In addition, key immunologic biomarkers—including cytokines and cytotoxic mediators implicated in SJS/TEN pathogenesis (42)—were not systematically assessed across all patients. Due to heterogeneous institutional testing practices and the severe condition of many patients at presentation, the available biomarker data were incomplete and insufficient for cohort-level analysis, limiting our ability to investigate mechanistic associations. Furthermore, heterogeneity in treatment regimens and potential biases inherent to retrospective data collection could confound outcomes interpretation. Finally, the relatively short follow-up period may not fully capture long-term prognosis. Future multicenter, prospective studies incorporating standardized biomarker profiling are warranted to validate these findings and improve mechanistic understanding.

Our findings highlight that early recognition, immediate discontinuation of immunotherapy, and prompt initiation of systemic corticosteroids—often in combination with intravenous immunoglobulin—are critical for improving prognosis. Multidisciplinary management plays a pivotal role, particularly in severe cases, to minimize complications and mortality. Given the increasing use of immune checkpoint inhibitors, clinicians should maintain high vigilance for severe cutaneous adverse reactions and differentiate them from other immune-related events at an early stage. Future studies are warranted to elucidate the immunopathogenic mechanisms of ICI-induced SJS/TEN and to identify predictive biomarkers that can guide personalized prevention and management strategies.