Russel Dinh, MD
Vanderbilt University, Nashville, TN
Charlie Zhang, MD
Bascom Palmer Eye Institute, Miami, FL
Uveitic Macular Edema Treatment: Suprachoroidal Triamcinolone Acetonide versus Intravitreal Triamcinolone Acetonide versus Dexamethasone Intravitreal Implant 
Dr. Kareem Moussa opened the debate by advocating for the use of suprachoroidal triamcinolone acetonide for uveitic macular edema, citing favorable outcomes from both randomized controlled trials (such as the PEACHTREE trial) and real-world studies. He emphasized the targeted delivery to the suprachoroidal space as a key advantage, allowing for significant anatomical improvement while minimizing steroid-related IOP elevations. The PEACHTREE trial showed a reduction of central subfield thickness by approximately 150 microns and a relatively low IOP elevation rate (11.5%) despite patients receiving two injections. Dr. Moussa argued that this route combines efficacy, safety, and innovation, positioning it as the future of posterior segment steroid therapy for uveitic macular edema.
Dr. Akshay Thomas countered by highlighting the consistency, generalizability, and ease of use of intravitreal triamcinolone. He noted that while suprachoroidal delivery offers theoretical advantages, its durability and real-world performance remain variable, with many patients requiring rescue therapy within a few months. He also pointed out the technical difficulty of accessing the suprachoroidal space, especially in eyes with surgical hardware or hypotony – common among uveitic patients – making intravitreal triamcinolone a more practical and broadly applicable option across diverse clinical scenarios.
Dr. Arthi Venkat presented a compelling argument in favor of the intravitreal dexamethasone implant (Ozurdex), citing its predictable delivery, prolonged duration of effect, and manageable side effect profile. She challenged the notion that newer delivery methods like the suprachoroidal route offer substantial clinical advantages, arguing instead that Ozurdex provides consistent control of inflammation with a well-established injection technique with which retina surgeons are familiar. While acknowledging the shared risks of cataract formation and IOP elevation among all steroid therapies, she emphasized the user-friendliness, reliability, and patient tolerability of the dexamethasone implant as reasons it remains her preferred treatment.
Following the debate, an informal audience poll revealed that approximately 75% of attendees favored the dexamethasone implant as their preferred treatment modality for uveitic macular edema.
How to Manage Foveal-Threatening Geographic Atrophy
Dr. Diana Laura opened the debate with a strong argument in favor of initiating treatment for foveal-threatening geographic atrophy using complement inhibition. She emphasized that once GA encroaches upon the fovea, patients experience rapid vision loss, resulting in reduced independence, increased depression, and significantly impaired quality of life. Dr. Laura cited data from the AREDS study, which found the median time to foveal involvement is approximately 2.5 years, arguing that early intervention may delay this debilitating progression. She presented compelling evidence from the GALE extension study, where monthly pegcetacoplan slowed GA progression by over 35% from months 24 to 48, preserving approximately 3.16 mm² of retinal tissue—equivalent to 1.5 optic disk areas. Additionally, she noted that earlier and sustained treatment preserved nearly three times more retinal area than delayed therapy.
Dr. Laura also addressed safety concerns, including the rare but serious risk of retinal vasculitis, stating that such events were not seen in clinical trials and are estimated to occur in approximately 1 in 4,000 injections, primarily as a first dose phenomenon. She acknowledged the increased risk of exudative AMD but emphasized that anti-VEGF treatments are highly effective and widely available. She concluded by urging clinicians to reframe their thinking—highlighting that preserving functional retina, even if modest in area, may significantly extend patients’ quality of life.
Countering this view, Dr. Yasha Modi raised concerns regarding the real-world impact, cost, and risk of complement inhibition therapy. He began by questioning whether slowing GA growth by 3.16 mm² over four years—at a cost of over $140,000 and requiring 48 injections—translates into meaningful functional benefit for patients. He noted that there is no published data showing these therapies significantly slow progression toward the fovea, nor is there clear improvement in visual acuity or even microperimetry across the broader patient population.
Dr. Modi also outlined an important cumulative risk associated with long-term treatment: a >1% chance of infectious endophthalmitis. He walked through real-world patient scenarios, including those who might lose foveal fixation before the drug takes effect, patients who die shortly after initiating treatment, or those who suffer treatment-related complications. His main message was that treatment should be individualized, taking into account patient age, comorbidities, functional goals, and overall life expectancy.
In the end, the audience was polled, and 56% voted in favor of not treating foveal-threatening GA with complement inhibition, reflecting the importance of patient selection for these new therapeutic agents.
Should Nurse Practitioners and/or Physician Assistants Perform Intravitreal Injections?
Dr. Charles Calvo argued in favor of expanding the role of NPs and PAs to include performing intravitreal injections, highlighting both safety and practicality. He cited robust real-world data from Moorfields Eye Hospital in the UK, where nurses safely performed over 25,000 injections annually with extremely low complication rates—comparable to those reported in landmark clinical trials such as MARINA and ANCHOR. Dr. Calvo emphasized that the growing burden of retinal diseases, such as neovascular AMD and diabetic retinopathy, outpaces the number of retina specialists trained each year, creating an urgent need for increased treatment capacity.
He further argued that patients are already accustomed to receiving care from advanced practice providers in other medical specialties and would likely adapt well to a similar model in ophthalmology. Dr. Calvo also addressed concerns about complications, noting that retina specialists routinely manage adverse outcomes from procedures performed by others, such as cataract surgeons. He concluded by emphasizing the potential for improved clinic flow, physician work-life balance, and patient access—especially in high-volume practices—through appropriate training and delegation to NPs and PAs.
Opposing the motion, Dr. Rebecca Soares delivered a passionate and data-driven rebuttal, cautioning against the expansion of non-physician injectors in ophthalmology. While acknowledging the access-to-care argument, she pointed out that intravitreal injection visits often uncover other retinal pathology, and it remains unclear who would be responsible for interpreting OCTs or making real-time clinical decisions in NP/PA-led encounters. Dr. Soares raised additional concerns about the variability in training, regulation, and oversight of NPs and PAs, particularly in states where they can practice independently without direct physician supervision.
She warned that expanding NP/PA roles might not meaningfully improve access, referencing studies from primary care that show non-physician providers tend to cluster in the same urban areas as physicians. From a broader policy perspective, Dr. Soares noted that increasing non-physician billing under Medicare contributes to declining reimbursement rates under the current budget-neutral system, potentially harming ophthalmologists financially in the long run. She closed with a call to protect the scope and integrity of the specialty, arguing that quality of care, professional respect, and long-term sustainability may all be compromised by such delegation.
The audience overwhelmingly sided with Dr. Soares, with 94% voting against allowing NPs and PAs to perform intravitreal injections.
Will Artificial Intelligence Be Good for the Field of Retina?
Dr. Matthew Starr opened the debate in favor of AI, arguing that AI has the potential to augment retina care, expand access, and alleviate workforce shortages. He emphasized that AI encompasses machine learning and deep learning models capable of image interpretation, risk prediction, and clinical decision support. Dr. Starr cited the example of autonomous AI screening for diabetic retinopathy, which became the first FDA-approved AI system in any field of medicine. He also highlighted data suggesting a looming mismatch between patient volume and retina specialists by 2035, noting that AI could help mitigate disparities in access, especially in underserved regions.
Dr. Starr went on to provide examples of the benefits of AI included AI-assisted ROP diagnosis, OCT-based fluid segmentation, and decision-support tools for treatment timing in neovascular AMD. He referenced applications of AI in surgical robotics, predictive modeling for complications like PVR, and automated imaging interpretation. Dr. Starr suggested that AI could help retina specialists streamline documentation, increase clinic efficiency, and potentially reduce burnout. He closed by likening AI adoption to other technological leaps in ophthalmology, predicting that it will eventually become as integral to care as OCT is today.
In contrast, Dr. Esther Kim delivered a compelling argument against unchecked adoption of AI in retina. She cautioned that AI threatens to erode the unique expertise of retina specialists by democratizing the role of the retina specialist as a diagnostician. This could lead to scope-of-practice encroachment by non-retina providers or even optometrists and PCPs who feel empowered by AI tools. She cited real-world examples, such as nurse-administered injections in the UK, to underscore the risk of retina specialists being bypassed in care delivery.
Dr. Kim warned that AI is not infallible—algorithms may miss systemic signs or comorbid conditions, and their narrow focus could give providers and patients a false sense of security. She questioned who bears responsibility when AI makes a clinical error, and raised concerns about the financial burden of acquiring, maintaining, and integrating AI tools—especially amid declining reimbursements. Additionally, she emphasized the lack of regulatory guardrails and the potential for insurers to use AI as a gatekeeping tool to reduce specialist referrals. Ultimately, she argued that retina specialists must define the terms of AI integration or risk ceding control of the field.
The audience sided with Dr. Starr, with 67% voting in favor of AI being good for the field of retina.
*Of note, the debate topics were assigned, and may not reflect the personal preferences of the speakers.*