FLORetina 2025: EU-ROP Registry bridges borders for real-world insight

The management of retinopathy of prematurity (ROP) continues to evolve, driven by real-world data that extend beyond the scope of traditional trials. At the center of this evolution is the EU-ROP registry, a pan-European initiative expanding the original German Retina.net ROP registry into a multinational, harmonized platform capturing clinical information from more than 60 centers across 16 countries. With data now collected from nearly 950 treated patients, the registry offers a rare lens into variations in ROP presentation, treatment strategies, and outcomes across diverse healthcare systems.

At the FLORetina 2025 Congress, the 13th International Congress on OCT and OCT Angiography (ICOOR), held from December 4-7, in Florence, Italy, Andreas Stahl, Dr Med, the chair and head of the Department of Ophthalmology at Greifswald University Hospital, presented an overview of the registry. Stahl shared his perspective with Ophthalmology Times Europe, detailing the operational hurdles of multilingual consent translation, regulatory variability, and international collaboration—and how these were overcome through coordinated efforts. The conversation also highlighted insights into treatment modality trends, data linkage with international registries, and the integration of AI-driven vascular severity scoring for imaging analysis.

Editor's note: Transcript edited lightly for clarity and length.

What were the biggest hurdles in harmonising data collection across the countries included in the registry?

Stahl: To begin on a very positive note, we found exceptionally motivated and highly skilled collaborators at all participating centers. Without these individuals, it would not have been possible to establish such a large-scale registry in such a short period of time. Thanks to their dedication, expanding the existing German Retina.net ROP registry to a European level became a remarkably smooth and rewarding process.

Within just four years, we succeeded in activating more than 60 sites across 16 European countries, collectively gathering data from now 950 treated ROP patients. One of the challenges along the way was the translation of patient information and consent forms into 21 different languages. In addition, contracts had to be established with centers operating under diverse languages and legal frameworks.

All of this was coordinated and managed by our highly committed project lead, Johanna Pfeil, who maintained continuous communication with certified medical translation agencies, legal departments, and – most importantly – the clinicians and researchers at the participating centers.

How do you ensure interoperability with existing national registries?

Stahl: In many countries, we found no existing national registries for treated ROP. Sweden is a notable exception, with its well-established and long-running SWEDROP registry, which collects data not only on treated ROP but also on ROP screening. Here, we chose to establish an approach where we aim to combine the collected data for comparison and publication purposes but keep the databases separate.

When setting up the EU-ROP registries, we also looked beyond Europe and considered existing international ROP registries. An important example is the Fight Childhood Blindness! ROP registry, established by Caroline Catt in Australia. From the outset, we held joint expert sessions with neonatologists and ophthalmologists to ensure close alignment between the two registries while still accommodating regional differences and nuances.

The combined dataset of the EU-ROP and Fight Childhood Blindness! ROP registries have already been published (1), and we are currently planning the first joint analysis comparing data from Europe with data from Australia and New Zealand.

What data points are most critical for improving ROP management?

Stahl: We observe major differences in ROP demographics across Europe; the prevalence and severity of ROP vary substantially from one country to another. To me, this represents an important initial finding: even within a relatively well-developed region of the world, there are significant differences in which infants develop ROP and which stages are most common when treatment becomes necessary. This insight is crucial for accurately interpreting treatment data.

A second key finding is that ROP treatment patterns differ markedly between countries. By linking treatment data with the long-term outcomes, we will be able to identify associations not only between ROP severity at the time of treatment but also between the chosen treatment modality and long-term results. Importantly, all of this will be captured in a real-world clinical setting that more accurately reflects everyday practice than controlled clinical trials—while also providing far larger sample sizes than those available in existing clinical studies.

Are there plans to integrate AI analytics for predictive modeling?

Stahl: We have already applied AI-based analytics to imaging data from the CARE-ROP study (2). CARE-ROP was an investigator-initiated pilot trial evaluating the use of ranibizumab in ROP and served as the foundation for the later RAINBOW clinical trial. In this relatively small dataset, we found that AI-derived vascular severity scores correlated closely with clinician-assessed ROP severity. A key advantage of AI-based scoring is its ability to quantify ROP severity more objectively than clinicians, particularly with respect to plus disease, which is known for its high subjectivity in grading.

We also observed indications that the magnitude of change in AI-based vascular severity scores between baseline and week 1 post-treatment may predict the risk of later ROP reactivation. We now aim to validate this finding in a larger dataset. For this purpose, we included questions on whether ROP imaging was performed in the EU-ROP database. This enables us to identify centers that routinely capture pre- and post-treatment ROP images and to unite them into a sufficiently large imaging cohort for validating AI-based severity grading. A major strength of this approach is that these centers contribute not only imaging data but also structured clinical information to the EU-ROP registry – an invaluable combination for validating AI-driven analytics.

Can you please share a brief overview of the 4 abstracts based on data from the EU-ROP registry that were presented at the World ROP Congress?

Stahl: We are honored to present 4 abstracts from the EU-ROP collaborative group at the World ROP Congress. Two of these will focus on in-depth country-level comparisons. First, Fabienne Wähler from our Greifswald team compared ROP treatment patterns in Poland and Germany—2 countries with broadly comparable standards of neonatal care, a key factor influencing ROP development. In the same session, Baran Özdemir from Ankara presented a comparison between Turkey and Germany, currently the 2 largest contributors to the registry.

On Friday, Johanna Pfeil, coordinator of the EU-ROP registry, presented an overall analysis of the first 651 enrolled patients. This was followed by a presentation from Teresa Barth, from Regensburg, examining anti-VEGF treatment patterns, exploring how the use of anti-VEGF varies across countries, and how these differences may relate to variations in neonatal care.

Together, these 4 presentations highlight the data currently available in the registry and illustrate the types of in-depth analyses the dataset enables. We hope to gather valuable ideas for additional analyses from discussions with the experts attending the World ROP Congress—and ideally encourage even more centers to join the registry.

The first in-person international study group meeting was held in Florence—what outcomes of that meetup do you look forward to?

Stahl: The EU-ROP study group meeting is a key opportunity to connect with all the individuals behind the registry—the clinicians, researchers, and examiners who screen and treat ROP and enter data into the system. These people are the registry; without them, none of this would be possible.

We [were looking forward] to a lively discussion about what works well within the registry and where it can be improved or expanded. We also encourage participants to share their own ideas for analyses and to suggest national or regional meetings where EU-ROP data could be presented. Ultimately, this is the essence of the EU-ROP initiative: bringing together experts in ROP to collaboratively analyze, present, and publish real-world data on ROP demographics and treatment.

Which collaborative discussions between FLORetina and World ROP Congress are you most excited about?

Stahl: It is a tremendous opportunity to have the World ROP Congress held jointly with FLORetina, as it allows both communities to interact, exchange ideas, and stimulate new insights. These 2 groups already overlap to some extent, as many retina specialists—including myself—treat ROP infants alongside adult retinal patients. For such clinicians, this joint event offers the unique advantage of attending two major conferences with a single trip.

For others who might typically attend only one meeting—either because their focus is primarily ROP or primarily on adult retinal disease—this joint format provides an opportunity to broaden their expertise and engage with perspectives outside their usual scope. ROP is, after all, a retinal disease, and integrating these two conferences is both logical and mutually beneficial.

How do you hope this year’s event will advance global strategies for retinal disease prevention and treatment?

Stahl: The World ROP Congress provides a unique platform for ROP experts from around the globe to share their experiences and perspectives. This international exchange – transcending geographic boundaries – offers an invaluable opportunity to understand ROP in all its variations and presentations. This year’s joint event with FLORetina further enhances that strength by adding a highly dedicated global community of retina specialists to the conference. Their perspectives can enrich and constructively challenge how we approach ROP care, stimulating new ideas and potentially reshaping strategies for ROP prevention and treatment worldwide.

REFERENCES

1. Catt C, Pfeil JM, Barthelmes D, Gole GA, Krohne TU, Wu WC, u. a. Development of a joint set of database parameters for the EU-ROP and Fight Childhood Blindness! ROP Registries. Br J Ophthalmol. 13. September 2023;bjo-2023-323915.

2. Eilts SK, Pfeil JM, Poschkamp B, Krohne TU, Eter N, Barth T, u. a. Assessment of Retinopathy of Prematurity Regression and Reactivation Using an Artificial Intelligence–Based Vascular Severity Score. JAMA Netw Open. 19. Januar 2023;6(1):e2251512.