This case report describes a patient with a remote history of uveitis who developed irAE–associated uveitis following ICI-based therapy for BC, occurring 25 years after the initial episode.

The reported incidence of irAE-associated uveitis ranges from 0.3% to 6% [4], with melanoma representing the most common underlying malignancy, followed by lung cancer and urological cancers [5, 10]. A retrospective cohort study demonstrated that the 1-year incidence of uveitis among patients treated with ICIs was significantly higher among individuals with melanoma compared to non-melanoma malignancies (1.2% vs. 0.2%), with an adjusted odds ratio of 6.45 [6]. Similarly, a recent systematic review including 52 patients reported melanoma in 69.2% of irAE uveitis cases, whereas renal cell carcinoma and lung cancer each accounted for 11.5% [8]. One proposed mechanism underlying the high incidence of irAE uveitis in melanoma involves immune cross-reactivity between melanoma antigens and melanocytes within the retina and choroid, potentially leading to VKH-like pathology [11]. Additionally, VKH disease has been reported more frequently among individuals with specific human leukocyte antigen (HLA) genotypes [11, 12]. Regarding ICI regimens, 63.5% of patients who developed irAE uveitis received programmed death-1 (PD-1) inhibitor monotherapy, whereas 19.2% received combination therapy that included a cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) inhibitor [8]. Dual PD-1 and CTLA-4 blockade has been used more frequently in melanoma, renal cell carcinoma, and lung cancer, a pattern that may partially explain the higher incidence of irAE uveitis reported in these malignancies.

In contrast, irAE-associated uveitis is rarely reported among patients receiving ICIs for BC. In the KEYNOTE-522 trial, which employed the regimen administered in the present case, uveitis occurred in only two of 784 patients (0.3%) [9]. Similarly, the KEYNOTE-355 trial involving patients with metastatic TNBC treated with pembrolizumab plus chemotherapy reported a uveitis incidence of 0.2% [13]. Other pivotal clinical trials evaluating ICI-based regimens for early-stage or advanced BC either did not report uveitis or reported very low frequencies [14,15,16,17,18,19,20] (Table 1). The relatively low frequency of irAE uveitis in breast cancer may be explained, at least in part, by the absence of melanocytes in normal breast tissue. However, because melanocytes are likewise rarely identified in most other non-melanoma tumors treated with ICIs, the greater number of reported cases in lung cancer and renal cell carcinoma may be better explained by differences in ICI responsiveness, the intensity of immune activation, earlier and more widespread clinical implementation of ICIs, and the use of anti–PD-1/anti–CTLA-4 combination regimens, rather than by organ-specific biological features. Given the rarity of this toxicity, however, such interpretations should be considered hypothesis-generating.

Patients with pre-existing autoimmune diseases have an increased risk of developing irAEs, including flares of underlying conditions. A meta-analysis including 23,897 patients with cancer and pre-existing autoimmune diseases reported an overall irAE incidence of 61% after ICI-based therapy, with 36% representing flares of the underlying autoimmune disease [3]. Another study demonstrated an association between pre-existing autoimmune disease and irAE development, with an adjusted odds ratio of 2.52 [2]. These findings emphasize the importance of obtaining a detailed medical history and performing careful risk assessment before initiating ICI-based therapy.

In the present case, the distinction between a flare of VKH disease and ICI-associated VKH-like uveitis is inherently challenging. However, several features favored ICI-associated uveitis in our patient: the onset occurred shortly after pembrolizumab initiation, the patient had remained free of recurrence for more than 25 years, and the ocular inflammation responded promptly to discontinuation of pembrolizumab and relatively low-dose systemic corticosteroid therapy. In contrast, recurrent VKH generally requires more intensive and prolonged immunosuppressive treatment to achieve sustained disease control. Therefore, although recurrence of VKH cannot be completely excluded, the clinical course in this case was considered to be more consistent with ICI-associated uveitis.

Large cohort studies have demonstrated that patients with a prior history of uveitis are at increased risk of ophthalmic irAE, including uveitis, after ICI-based therapy. One study reported recurrence rates of 44% (44/100) after pembrolizumab administration and 50% (8/16) after ipilimumab administration [7]. Another report demonstrated 1-year recurrence rates of 28.6% among patients with melanoma and 12.6% among patients with non-melanoma malignancies who had a history of uveitis after ICI initiation, supporting pre-existing uveitis as a risk factor for ophthalmic irAEs regardless of tumor type [6] (Table 2). Notably, uveitis developed in the present patient despite a prolonged remission period of 25 years, indicating that even a very remote history of uveitis may confer an increased risk of irAE-associated ocular inflammation. We did not initially recognize her remote history of uveitis as a potential risk factor and therefore did not perform a baseline ophthalmologic evaluation before treatment. Although mild ocular symptoms developed several days before the second dose of pembrolizumab, ophthalmologic evaluation was delayed until after the second dose, partly because we hesitated to interrupt neoadjuvant treatment given with curative intent. This delay may have contributed to treatment interruption, whereas earlier recognition of her high-risk status might have enabled earlier diagnosis.

Current irAE management guidelines recommend topical corticosteroids for mild uveitis, discontinuation of ICIs with local and/or systemic corticosteroids for moderate uveitis, and high-dose systemic corticosteroids for severe cases [21,22,23]. Regarding outcomes, complete resolution occurred in 58% of patients, partial resolution in 25%, and recurrence despite treatment in 8% [8]. In the present case, uveitis responded promptly to low-dose systemic corticosteroids and resolved completely after steroid pulse therapy and intravenous immunoglobulin administered for incidentally developed SJS.

From an ophthalmologic perspective, immune-related uveitis appears to differ from conventional non-infectious uveitis in several important aspects. Previous studies have shown that ICI-associated uveitis often responds favorably to relatively low-dose corticosteroid therapy compared to typical autoimmune uveitis, suggesting a distinct inflammatory profile and generally good steroid sensitivity [24]. Additionally, uveitis may occur in the absence of overt structural retinal abnormalities on OCT, highlighting the importance of functional and clinical assessment beyond structural imaging alone [25]. Accordingly, functional testing, such as ERG, may aid in detecting subclinical retinal involvement and complement OCT in selected patients, as illustrated by bilateral ERG abnormalities in the present case despite unilateral structural changes on OCT. These findings indicate that ICI-associated uveitis may present with subtle or subclinical retinal involvement and may be effectively controlled with prompt corticosteroid therapy. Nevertheless, patients with high-risk features, including a prior history of uveitis or melanoma, require close ophthalmologic surveillance during ICI treatment, as early detection and intervention remain critical for preventing vision-threatening complications.

In clinical practice, TMP-SMX prophylaxis for Pneumocystis jirovecii pneumonia is recommended for patients receiving ≥ 20 mg/day of prednisolone for at least 4 weeks, particularly when combined with other immunosuppressive agents [26]. Although TMP-SMX is associated with a relatively high incidence of cutaneous adverse reactions, recent reports suggest that ICI-based therapy increases the risk of severe drug eruptions by lowering the threshold for drug hypersensitivity through immune activation [27, 28]. In retrospect, given the favorable response of uveitis to low-dose corticosteroids, prolonged steroid therapy and prophylactic TMP-SMX may have been unnecessary, and SJS might have been avoided. These findings highlight the importance of recognizing the potential risks associated with TMP-SMX use in patients receiving ICIs.

In the present case, pembrolizumab was discontinued after two cycles due to immune-related uveitis and SJS, leading to treatment interruption and reduced treatment intensity. Although marked tumor regression was observed, pCR was not achieved; therefore, postoperative treatment was completed with capecitabine based on the CREATE-X trial, which reported a 42% reduction in recurrence risk and a 48% improvement in OS among patients with residual disease [29]. In the event of disease recurrence, non-ICI-based therapies remain viable treatment options.

In conclusion, including the present case, previous reports have suggested that patients with a history of uveitis may be at increased risk of developing uveitis after ICI-based therapy. Careful pre-treatment evaluation, thorough patient counseling, and close multidisciplinary monitoring, with early involvement of ophthalmology specialists, are essential for the safe administration of ICIs in patients witha history of uveitis.