International Glaucoma Symposium 2026 recap: From AI to MIGS in practice

The second International Glaucoma Symposium—with the theme of Future of Glaucoma Diagnosis and Management—took place on 31 January 2026 at the University Eye Clinic in Mainz, Germany. The one-day meeting brought together leading researchers and clinicians to examine the latest advances in glaucoma diagnosis, management, and emerging technologies.

Through interviews with key presenters, the Eye Care Network explored some of the symposium’s innovations including discussions on population‑based artificial intelligence (AI) data sets, the evolving role of microinvasive glaucoma surgery (MIGS), emerging insights into optic disc drusen, and strategies for identifying pre‑perimetric glaucoma.

OCT and neuro‑ophthalmology insights

Prof. Dr. med. Wolf A. Lagrèze, University Hospital Mannheim, Germany, highlighted the nuanced clinical role of optical coherence tomography (OCT) in both glaucoma and neuro‑ophthalmic disease. He noted that OCT “has to be regarded as one diagnostic tool, which is part of a spectrum” alongside fundoscopy, fundus photography, perimetry, visual acuity testing, pupillary assessment, and careful history taking, emphasising that it should not be relied on in isolation. In the glaucoma setting, he underscored OCT’s value particularly for monitoring disease progression, which he described as “the mainstay in glaucoma.” However, he also noted that experienced clinicians can often distinguish glaucomatous optic atrophy based on characteristic optic disc features without OCT: “Honestly, you don’t need an OCT for that.”

Lagrèze explained that OCT can still play a helpful role in specific clinical scenarios, such as differentiating glaucoma from optic disc drusen by using enhanced depth imaging to visualise deeper optic nerve head structures. Nonetheless, he stressed caution when interpreting OCT in the broader context of neuro‑ophthalmic diseases, pointing out that there is “no specific fingerprint” on OCT for common neurodegenerative conditions like Alzheimer disease or multiple sclerosis. Despite this limitation, OCT was described as “highly useful” for detecting subtle papilloedema and objectively assessing disc swelling linked to raised intracranial pressure.

A central theme of Lagrèze’s presentation was the risk of misinterpretation when OCT findings are evaluated outside a comprehensive clinical framework. He cautioned that over-reliance on single parameters—for example, inter-eye differences in retinal nerve fibre layer (RNFL) or ganglion cell layer thickness—may lead to false-positive diagnoses, particularly in light of revised multiple sclerosis criteria. He noted that thorough interpretation must consider refraction, coexisting ocular pathologies such as glaucoma, and the patient’s full clinical history to avoid diagnostic errors.

Lagrèze described OCT as “a super helpful tool,” especially when integrated with AI techniques capable of analysing large datasets and supporting earlier disease detection. He underscored the importance of specificity in AI models to prevent harmful false diagnoses and concluded by advocating for strong collaboration between neurologists and ophthalmologists to ensure accurate interpretation and safe clinical application of OCT in neuro‑ophthalmic practice.

Population-based normative data and AI in glaucoma

Prof. Dr. Luís Abegão Pinto, Department of Ophthalmology, Hospital de Santa Maria, Lisbon, Portugal, highlighted the transformative potential of population-based normative data for advancing AI in glaucoma care. He explained that “the performance of AI in glaucoma is fundamentally limited by the reference data we use,” noting that many current AI systems are trained on hospital-based, referral-biased datasets rather than broad, real-world populations. Anchoring AI to robust normative datasets that reflect real-world biology is essential if these tools are to become reliable aids in clinical decision-making at scale.

Pinto detailed that population-based data collection, coupled with AI, enables automated analysis, standardised feature extraction, and consistent quality control across very large datasets. He noted that although AI does not inherently eliminate bias, it allows bias “to be identified and handled systematically,” rather than being implicitly embedded in smaller, highly selected cohorts. This systematic approach is key to building trustworthy AI tools for glaucoma screening, diagnosis, and monitoring across diverse populations.

Discussing real-world deployment, Pinto noted that defining what AI should detect—such as “glaucoma” or “referable glaucoma”—depends on the goals of the screening programme and the healthcare context. His work addresses these challenges by combining population-based screening with clinic-level validation, including OCT, visual field testing, and detailed phenotyping, to stratify disease severity and establish clinically meaningful targets for AI systems.

Pinto envisioned that AI-enhanced normative databases will become part of routine clinical infrastructure rather than niche research tools. He anticipates that larger and more credible datasets will further improve AI models, deepen understanding of glaucoma progression across different populations, and ultimately support more informed, equitable, and effective care.

Expanding role of AI

Priv.-Doz. Dr. med. Jan H. Terheyden, FEBO, Department of Ophthalmology, University Hospital Münster, Germany, highlighted the rapidly expanding role of AI in ophthalmology, focusing on how today’s AI tools are already reshaping clinical practice. He noted that AI “is already supporting ophthalmologists in everyday clinical care,” particularly through retinal imaging applications where algorithms for diseases like diabetic retinopathy are “approaching or matching expert performance.” Terheyden also noted the emerging impact of AI on workflow, including tools like ambient clinical scribes and automated documentation, which are “beginning to have a meaningful impact by reducing administrative burden and allowing ophthalmologists to focus more on patient care.”

Terheyden explained that one of the central challenges in integrating AI is bridging the gap between algorithmic performance and real clinical value. He pointed to concerns around generalisability, explainability, and trust, noting that many models are trained on curated datasets that may not reflect broader populations. Discussing AI’s role in improving diagnostic accuracy and efficiency, Terheyden said it excels at automating repetitive technical tasks like image quality control, segmentation, and risk stratification, freeing clinicians to focus on more complex decision-making. He anticipates particular promise in combining longitudinal and multimodal data to enhance early detection and assessment of disease progression.

Terheyden outlined that the most significant developments in the next 5–10 years will come from clinically integrated, multimodal, and prospective AI systems rather than standalone diagnostic tools. Progress in foundation models, real-world validation, and regulatory-ready evaluation frameworks will be critical to move AI from research into routine, equitable ophthalmic care.

Modern glaucoma procedures and MIGS

Andrew Tatham, MD, MBA, FRCOphth, FEBO, Oxford University Hospitals NHS Foundation Trust, UK, outlined the evolving role of micro-invasive glaucoma surgery (MIGS) alongside established procedures, noting how modern approaches are reshaping clinical decision-making. He said the principal advantage of MIGS is safety, noting that these procedures have “a much better safety profile compared with traditional surgery” with “fewer complications” and faster visual recovery. Although trabeculectomy remains highly effective for lowering intraocular pressure (IOP), Tatham conveyed that MIGS should not be seen as a direct competitor but as an additional option for earlier-stage disease, particularly in patients with milder glaucoma.

Discussing patient selection, he highlighted the presence of a visually significant cataract as a key factor in deciding when to incorporate MIGS, as these procedures can be efficiently performed in combination with standard cataract surgery. He commented that “there is a device for most patients” and outlined other considerations such as glaucoma severity, target IOP age, comorbidities, and anticoagulant use. Despite the growth in options, many eligible patients with mild to moderate glaucoma still do not receive MIGS, partly due to technical demands of intraoperative gonioscopy and uncertainty about device selection.

Tatham addressed adoption barriers and the need for clearer clinical guidance, stressing that training and mentorship are crucial. He also highlighted advances in complementary technologies such as imaging, drug delivery, AI, and genomics. One promising development is aqueous angiography, which may enable clinicians to visualise outflow pathways with precision analogous to coronary angiography. He concluded that MIGS’ most significant benefit lies in enhancing patient quality of life by reducing medication burden, improving adherence, and delaying more invasive interventions.

Optic disc drusen: Clinical perspectives

Steffen Hamann, PhD, FEBO, FRCOphth, Department of Ophthalmology, University Hospital Hamburg-Eppendorf, Germany, provided insights into optic disc drusen, a condition that can mimic glaucoma and other optic nerve pathologies. He described drusen as “kind of odd man out” at a glaucoma meeting but stressed its clinical relevance due to overlapping features such as optic neuropathy, slow progression, RNFL loss, and visual field defects. Hamann noted that recognising these similarities and differences is essential to avoid misdiagnosis during routine evaluation.

He highlighted that optic disc drusen can cause appearances resembling true optic disc swelling, especially when peripapillary hyperreflective ovoid mass-like structures (PHOMS) are present. He clarified that while PHOMS are characteristic, they cannot alone distinguish pseudopapilloedema from true papilloedema, making comprehensive clinical and imaging assessment crucial. Advances in OCT now allow detection of smaller and less calcified drusen and enable quantitative assessment of afferent pathway damage.

Hamann also noted the potential of AI algorithms, OCT angiography, and genetic studies to improve detection, risk stratification, and eventually treatment, aiming “to get to a stage where we can safely treat optic disc drusen.”

Detecting pre-perimetric glaucoma

Marta Pazos, MD, PhD, Department of Ophthalmology, Hospital Clínic de Barcelona, Spain, highlighted clinical strategies for detecting pre-perimetric glaucoma, noting early recognition of structural damage before visual field loss. She noted that optic nerve head examination “still remains fundamental,” noting that localised neuroretinal rim thinning or notching, particularly in inferotemporal and superotemporal sectors, are among the earliest and most reliable signs. She described OCT as “indispensable” because it can reveal RNFL and ganglion cell layer thinning years before perimetric changes, providing a window to initiate treatment earlier and preserve vision.

Pazos discussed advanced imaging and AI tools as enhancements for early detection. AI algorithms analysing optic disc photographs, OCT maps, and multimodal data can predict which eyes are at greatest risk, supporting a probabilistic, risk-based approach rather than a binary diagnosis. She also noted limitations in the context of anatomical variations—including split nerve fibre bundles, high myopia, tilted discs, and peripapillary atrophy—adding that imaging must always be integrated with IOP measurements, structural assessment, and clinical history to avoid misclassification.

Looking beyond individual practice, Pazos highlighted the potential of AI-based screening and referral tools using fundus photography to improve early case detection and triage, especially in underserved populations. She described ongoing work in Barcelona to refine referral pathways from primary care, aiming to identify high-risk patients without overwhelming specialist services, underscoring how systemic approaches may enhance glaucoma diagnosis at earlier stages.