AI and gene therapy: The next frontier in uveitis treatment

Artificial intelligence and a ‘cross-sciences’ research model could be key to transforming how uveitis is diagnosed and treated, a leading international expert in the field told the Irish College of Ophthalmologists (ICO) 2025 Annual Conference in Kilkenny.

Delivering the annual Mooney Lecture on the evolution of treatment regimens for uveitis, Prof Andrew Dick, Director of the Institute of Ophthalmology at University College London and Professor of Ophthalmology at the University of Bristol, UK, said uveitis remains underdiagnosed and undertreated.

However, treatment is now entering a new era of sustained disease control, with fewer systemic side-effects and greater personalisation.

This progress, he said, is coming from using more targeted therapies, detecting disease earlier, and, eventually, will incorporate gene therapy into treatment protocols.

“Despite being a leading cause of preventable blindness, uveitis remains underdiagnosed and undertreated. However, we have made significant inroads treatment-wise in the last two decades,” Prof Dick said.

Historically, corticosteroids were the mainstay of treatment. However, they carry considerable short- and long-term risks, and Prof Dick stressed the need to use them as sparingly as possible.

“We know steroids are fantastic at inducing remission of acute inflammation. But you shouldn’t have patients on them for longer than three months as they start accruing side-effects, short-term with change of mood, and looks. Furthermore, the longer-term data is deeply worrying on cardiovascular, diabetes, and osteoporosis risks, as well as increased glaucoma and cataracts.”

Improved understanding of inflammatory pathways – particularly those involving cytokines – in uveitis has supported the shift towards more targeted therapies.

The introduction of disease-modifying anti-rheumatic drugs and biologics, such as adalimumab, has significantly improved outcomes, especially for severe or recurring cases of uveitis.

However, many challenges remain, including optimising biologic treatment access globally, improving timely diagnosis, and predicting response.

“There has certainly been a lot of progress [but] there is a lot more to do. Among the key things we need to do are to identify patients that are going to fail current very successful treatments, particularly anti-TNFs,” he commented. “We know about 30-to-40 per cent of patients do not respond adequately to anti-TNF therapies, but why is this? We need to know from the get-go who are those patients, and have the necessary biomarkers to predict treatment response.”

Prof Dick and colleagues are working to identify the molecular and cellular signatures that drive an individual’s uveitis. “If we can find that, then we could offer them earlier better treatment, so they don’t fail their first treatment.”

He stressed the need for treatment to be guided by the underlying cause of inflammation. Non-infectious uveitis is typically managed with immunosuppressive drugs, while infectious types require antimicrobial therapies.

“By tailoring treatment approaches, we can significantly decrease the likelihood of vision loss, which is the most serious potential outcome of untreated eye inflammation,” Prof Dick said.

Because uveitis is often linked to systemic diseases, Prof Dick emphasised the importance of a multidisciplinary care model, involving ophthalmologists, rheumatologists, and infectious disease specialists. Such collaboration, he said, improves early detection and enables more targeted treatment.

He also advocated for a cross-sciences approach to research, bringing together diverse fields to address the remaining treatment challenges in uveitis.

Prof Dick is a co-investigator on CLUSTER, a UK-based consortium linking immunology, clinical and genomic medicine, and computer science to analyse disease behaviour in JIA-associated uveitis.

“Now we are engaging much more with computational biologists, mathematicians, and we are bringing in deep learning and artificial intelligence, because we need to unravel a huge amount of data – patient or biological data – that we are getting from the lab or patient.”

Looking ahead, Prof Dick pointed to gene therapy as the next major development in the field of uveitis. Research is now focusing on refining vector design, delivery methods, and patient selection.

“We are pushing really hard. We are now in the process of developing a human-ready gene therapy and will see where we get to. There have been some setbacks, but we are close.”