Innovation in ophthalmology is often measured by what reaches the clinic—but the path to those breakthroughs begins much earlier, in how clinical trials are designed and who they include. As 2025 highlighted both the promise and the limitations of current trial ecosystems, interest has grown in leveraging global data to build more robust and representative evidence.

In this Q&A article with the Eye Care Network, Daniela Ferrara, MD, PhD, FASRS, FARVO, discusses the role of low- and middle-income country (LMIC) data in informing US ophthalmology clinical trials, the operational and ethical considerations involved, and how artificial intelligence (AI) may help harmonize global datasets in the years ahead. Ferrara is assistant professor of ophthalmology at Tufts University School of Medicine, and is involved with the optical coherence tomography team at New England Eye Center in Boston, Massachusetts. Ferrera is chair of the Low-Middle-Income Countries Working Group for the Collaborative Community of Ophthalmic Innovation¹ and is chief medical officer at Topcon Healthcare.

Note: Transcript edited for clarity and length.

What inspired your focus on using data from low- and middle-income countries (LMICs) to inform US clinical trials, and what gaps in the current US trial ecosystem does this approach aim to fill?

Daniela Ferrara, MD, PhD, FASRS, FARVO: The need for proper demographic representation in clinical trials is an important and well-known matter of scientific concern. Diseases may affect specific groups of patients differently, according for example to their sex, age range, race, ethnicity, or other specific demographic characteristics. Whether these differences are related to genetic makeup, biology, environmental exposure, or other clinically relevant features, they lead to the same recommendation—to assure that any given clinical trial is evaluating subjects that properly represent the broader population that manifest the condition under study.

This recommendation is important for many reasons, such as the generalizability of trial results, and the confidence that safety and efficacy of new therapies or devices evaluated in a trial will hold their performance when made available at the point of care.

Additionally, clinical trials offer patients the opportunity to receive the highest quality of clinical care, which unfortunately they might not have access to, especially in LMICs. Although in most trials patients do not know if they are receiving standard of care therapy, a new investigational medicinal product, or placebo, the participation in a clinical study assures access to specialized medical care, according to the study design as clinically indicated.

Thus, conducting prospective clinical trials in LMICs serves 2 important aims. The first one is to assure that proper demographic representation is included in the totality of the data collected in any given trial, which informs a robust risk-benefit profile as I mentioned before. The second one, and equally important, is to assist underserved patients that otherwise may not have access to high quality clinical care.

When conducting a clinical trial in LMICs is not possible, for example due to operational challenges, one alternative is to still leverage clinical data from LMICs to inform the design of a trial conducted in the US. Analysis of real-world data or observational studies from LMIC patients can lead to meaningful insights about clinical conditions in a broader patient population, which can further inform important elements of the trial design, such as eligibility criteria or endpoint analysis, for example.

Finally, at a minimum, clinical experts and investigators from LMICs can be consulted for the design, conduct, and interpretation of US clinical trials, with the intent to consider a broader perspective that may bring benefits for LMIC patients in the longer term. In addition, having a diverse approach to trial planning and execution also fosters the enrollment of a diverse US-based population, leading to better representation of Black, Asian, or Latino patients for example.

What are some of the unique advantages and challenges of incorporating LMIC data into US trial design?

Ferrara: There are some important advantages of incorporating LMIC data into US trials. For example, in addition to the benefits of generating robust and generalizable data or caring for underserved patients that I mentioned before, a long-lasting benefit in conducting prospective or retrospective studies in LMICs is the establishment of a specialized infrastructure that continues to benefit local clinical research and clinical practice even after the trial is concluded. In order to provide data for a clinical trial, a LMIC site will receive investments of various natures, including improvements in local facilities, specialized training for the local staff, or the implementation of best practices. These benefits can remain in place and, if well managed, create a virtuous cycle of attracting additional trials that lead to additional long-lasting advantages, which over time elevate the quality of the local clinical research and practice.

However, conducting studies in LMICs is not short of challenges, mostly operational in nature. The initial effort and investment to initiate a new study site may be higher, in comparison to a similar task in the US, because of local limitations – lack of expertise or appropriate resources. In addition, country-specific features must be observed, from an understanding of local policies and regulations to an appreciation of the local language and cultural nuances. These challenges may be substantial in prospective interventional clinical trials but are often minor in retrospective clinical data analysis from LMICs.

In what ways can collaboration between researchers in LMICs and the US reshape global clinical research equity and innovation?

Ferrara: As discussed before, investments to enable prospective clinical trials or retrospective clinical research in LMICs bring short- and long-term benefits to the local community, as well as meaningful scientific contributions to any given clinical trial. The critical component to achieve the maximum positive impact is indeed the establishment of a true “collaboration” between researchers from LMICs and the US, preserving an ethical governance and data sovereignty.

From the LMIC perspective, a true collaboration leads to long-lasting benefits for international clinical education and healthcare at large, which then fosters more equitable access to healthcare resources in a previously depleted community. From the US perspective, in addition to the obvious benefits of getting access to diverse clinical data for drug and device development, the establishment of an international research network enables innovative technological exchange, derisks future trials, and may bring additional benefits to global brands present in various markets.

Looking ahead, what role do you see AI or machine learning playing in harmonizing and validating LMIC data for US regulatory and clinical applications?

Ferrara: The broad adoption of AI-based solutions in virtually all sectors of the economy is regarded as the new industrial revolution, leading to fundamental changes in how experts perform their core tasks.

Thus, the implementation of AI-based solutions is fundamentally changing the design, conduct, and interpretation of clinical trials. One immediate benefit that is already observed in many studies is related to the automatic processing or analysis of very large and complex datasets, leading to substantial increase in efficiency and reliability in comparison to manual analysis. Such methodological improvements may lead to scientific breakthroughs, which would not be possible based solely on manual data analysis.

These powerful AI-based solutions enable a more comprehensive characterization of clinical datasets, collected in clinical research or real-world settings, and increase the scientific relevance of investigating diverse datasets. In this context, AI-based solutions can be applied to analyze clinical data collected from LMICs in a prospective or retrospective fashion, assisting in the harmonization and validation of such datasets, among other scientific endeavors.

In addition, the utilization of the same AI-based solutions in both clinical research and clinical practice settings will empower health care professionals (HCPs) to truly practice “evidence-based medicine.” By having the same tool for clinical data analysis in both settings, HCPs will be able to better translate scientific insights, from the trial to the clinic. In this context, again, the risk-benefit profile of new therapies can be better characterized in diverse populations.

What’s the most important takeaway for clinicians and industry leaders to remember about equity in ophthalmology research?

Ferrara: Fostering diversity, equity, and inclusion in clinical trials is critical to the betterment science and society at large. However, most clinical trials are based on data from high-income countries, and this poor representation may lead to poor generalizability of clinical trial results and may impact the global access to therapies and medical interventions.

Patients from LMICs have been historically marginalized and often still don’t have access to medical advances, in both clinical research and clinical practice settings. In the medical field of ophthalmology, this must be addressed. Retinal disease burden is most significant in LMICs, and the main causes of blindness worldwide are preventable or treatable, including diabetic retinopathy, myopia, cataract, and glaucoma.

A globalized world, now empowered by AI-based technological advances, can change this concerning scenario and help improve care for people in LMICs, the US, and everywhere around the world.

Daniela Ferrara, MD, PhD, FASRS, FARVO
E: [email protected]
Ferrara is assistant professor of ophthalmology at Tufts University School of Medicine, and is involved with the optical coherence tomography team at New England Eye Center in Boston, Massachusetts. Ferrera is chair of the Low-Middle-Income Countries Working Group for the Collaborative Community of Ophthalmic Innovation. She is chief medical officer at Topcon Healthcare.

Reference

1. Ferrara D. Leveraging LMIC data to power US trials. Presented at: CCOI–Stanford Summit; July 23, 2025; Byers Eye Institute, Stanford University; Palo Alto, California.