Retina World Congress 2026: Uveitis risk rises as immune checkpoint inhibitor use expands

Immune checkpoint inhibitors (ICIs) have transformed oncology, enabling patients with previously untreatable malignancies to achieve durable responses. However, as the use of agents such as pembrolizumab and nivolumab continues to expand, so does the spectrum of immune-related adverse effects they can cause—including uveitis. Although the per-patient risk remains low, the growing number of patients on these therapies means retina specialists are increasingly encountering ICI-associated uveitis, particularly at tertiary referral centers.

In this Q&A conversation with the Eye Care Network, Charlie Zhang, MD—a retina fellow at the Bascom Palmer Eye Institute at the University of Miami Miller School of Medicine in Miami, Florida—discusses the mechanisms, diagnostic challenges, and management of uveitis associated with cancer therapy at Retina World Congress 2026, May 14 to 17, in Fort Lauderdale, Florida.

Note: Transcript edited for clarity and length.

What prompted your focus on uveitis associated with cancer therapies, and how frequently are retina specialists encountering these cases in current practice?

Charlie Zhang, MD: Uveitis associated with cancer therapies is relatively uncommon and reported to be less than 1% in the landmark clinical trials. What has changed recently is the scale of exposure as more patients are now receiving these life-saving therapies. Immune checkpoint inhibitors (ICI) have fundamentally reshaped oncology with previously untreatable metastatic disease patients living longer while receiving these therapies. As a result, drugs such as pembrolizumab and nivolumab have seen an increase in their use and have even surpassed many of the anti-VEGF drugs in Medicare Part B expenditure. [Although] the per-patient risk remains low, the absolute number of patients is increasing. In practice, most of these referrals are still for ocular surface disease, but there has been an increase in ICI-associated uveitis, particularly in tertiary referral centers such as ours.

What are the underlying mechanisms driving uveitis in patients undergoing cancer treatment, and are there particular therapies or patient populations at higher risk?

Zhang: ICIs function by removing the natural brakes on T-cell activation, which is something that many tumors often exploit to evade the immune system. By inhibiting these pathways with ICI therapy, there is enhanced T-cell activation but at the cost of loss of immune tolerance which can lead to autoreactive T-cells and inflammation complications such as uveitis. From a mechanistic standpoint, the PD-1/PD-L1 inhibition primarily affects peripheral tolerance, whereas CTLA-4 inhibition acts more centrally at the level of T-cell/antigen presenting cells in the lymph nodes. By combining both these therapies, both layers of regulation are impaired, which is why these patients often have the highest risk of ICI-associated uveitis. In a cohort study of 157k patients, we found nearly a 5 fold hazard of ICI-uveitis in patients receiving combination therapy as compared [with] either drug class as monotherapy.

In terms of patient populations, melanoma patients appear to be at higher risk compared to those with other malignancies. One compelling hypothesis is antigenic overlap between melanoma-associated antigens and the melanocytic structures found in the immunoprivileged milieu of the eye. The activation of T-cells directed against the melanoma antigens may predispose to higher rates of intraocular inflammation. Clinically, we and others have also observed that melanoma-associated cases tend to have a higher proportion of posterior segment involvement and in the same cohort study we found a 2.5 fold higher hazard of ICI-uveitis as compared [with] non-melanoma associated cases.

What are the key challenges in distinguishing treatment-related uveitis from other inflammatory or infectious causes, and how should clinicians approach management in these complex cases?

Zhang: One of the challenges regarding ICI uveitis is that there is nothing pathognomonic about the disease and the presentation varies from a smoldering mild anterior uveitis to a panuveitis with exudative detachments mimicking VKH. That said, unlike traditional cytotoxic chemotherapy, ICIs do not typically cause profound immunosuppression meaning that we are not seeing any signal that there is an increased rate of atypical opportunistic infections.

In practice, we approach these patients much like any other uveitis workup by performing a thorough review of systems, medical and social history, and a careful exam with multimodal imaging to look for risk factors or patterns that might suggest a specific uveitic entity or infection that guides our targeted testing. The most important additional consideration in this population is masquerade syndromes, particularly in patients with metastatic melanoma. There are now multiple reports, including one from our group, of patients on ICI presenting with what appears to be panuveitis or posterior uveitis but ended up being vitreoretinal metastatic melanoma. Importantly, up to half of these cases were found to be amelanotic, which makes the diagnosis especially challenging. Often in cases with profound vitreous involvement, if there is a lack of response to steroids, or if there is pigment on exam, it's critical to have a low threshold for diagnostic vitrectomy to establish the diagnosis and guide management.

How important is collaboration between oncologists and ophthalmologists in managing these patients, and what future research or guidelines are needed to improve outcomes?

Zhang: Collaboration with our oncology colleagues is absolutely central in managing these patients. The ocular findings must be framed within the patient's oncologic context. That includes understanding whether there are alternative systemic options, whether dose modification is feasible, if there are other immune-related adverse effects present and how well the tumor is responding to ICI therapy. In many cases, ICI therapy may represent the patient's best or only effective oncologic option, which helps signal to us to be more aggressive with local or system ocular steroid therapy in an effort to preserve treatment. In other situations, if the tumor is not responding to ICI therapy, the development of uveitis may be the signal to consider an alternative regimen altogether. Because of this complexity, these cases are best approached in a highly individualized, multidisciplinary fashion. The existing oncology guidelines, such as the CTCAE-based recommendations, are helpful as a framework, but in practice function as a starting point. From the ophthalmic side, we're fortunate to have a wide range of local steroid options that can provide effective control of inflammation and often allow for continuation of ICI therapy even in cases of higher grade inflammation.

In our experience at Bascom Palmer Eye Institute, we found that despite the majority of patients meeting criteria for CTCAE grade 3 or higher toxicity, where guidelines typically recommend holding or discontinuing ICI therapy, only half of these patients ultimately required cessation with a combination of topical, periocular, intravitreal, and systemic corticosteroid use. Furthermore, most of these patients were able to successfully restart ICIs after control of inflammation. Going forward, it will be critical that we continue to report these cases from a multidisciplinary approach to better define when ICIs may be safely continued and the optimal role of local and systemic immunosuppression to provide improved ophthalmology-specific guidelines.

Charlie Zhang, MD

E: cxz679@med.miami.edu

Zhang is a retina fellow with Bascom Palmer Eye Institute at the University of Miami Miller School of Medicine in Miami, Florida.