Anatomic Outcomes in Clinical Trials for Retina

OCT has changed retina practice, but how are clinical trials being affected?

Anatomic Outcomes in Clinical Trials for Retina

OCT has changed retina practice, but how are clinical trials being affected?

Andrew E. Mathis, PhD, Medical Editor

An FDA briefing document published on June 29 is significant not only for its overall topic of cellular and gene-based therapies for retinal diseases, but also for its third chapter, in which it discusses the use of efficacy endpoints in clinical trials for these therapies. The document notes that it is important, when establishing endpoints for retinal studies, that improvement in patient quality of life be considered, but it concedes the inherent problems in measuring such outcomes accurately.

Among the less qualitative outcomes that the FDA discusses here are anatomic measures, including fundus photography, fluorescein angiography and optical coherence tomography. Regarding this latter modality, the FDA's Center for Drug Evaluation and Research has already accepted photoreceptor loss as measured on OCT as a valid outcome measurement. However, in addressing the use of OCT in neovascular AMD, the FDA writes, “changes in retinal thickness, as measured by OCT, may be due to changes in retinal tissue cellularity, retinal cell volume, or interstitial fluid. Therefore, the clinical meaningfulness of changes in retinal thickness or macular volume is unclear.”

This pronouncement may come as something of a surprise to retinal physicians, who are becoming more and more used to measuring reduction of subretinal fluid using spectral-domain OCT. To assess the way that OCT is revolutionizing the design of clinical trials in retina, Retinal Physician spoke to several physicians involved in the clinical trials process, either in private practice or at academic medical centers, both here in the United States and abroad.


Optical coherence tomography has been an endpoint in retinal trials since it was used in a trial of foscarnet for cytomegalovirus retinitis over two decades ago. Today, with the advent of anti-VEGF drugs that demonstrably reduce central retinal thickness, OCT is commonly used to measure the efficacy of these drugs. By measuring an anatomic outcome, trial investigators are offered a more objective way of measuring the effects of treatment.

Among the ways OCT has changed the clinical trials process is in having been added to trials as a secondary endpoint. “OCT has only recently been added as an endpoint,” says Wiley A. Chambers II, MD, deputy director of the FDA's Division of Transplant and Ophthalmology Prod ucts, “primarily because there is not a direct correlation be tween central retinal thickness and visual acuity.”

Dr. Chambers' comments point up an ongoing issue in retina practice, ie, the lack of strict correlation between decreases in macular thickness and improvements in visual acuity. In practices where clinical trials are undertaken, this lack of strict correlation has had varied impacts. According to Sebastian Wolf, MD, of the University Clinic for Ophthalmology in Bern, Switzerland, “We use OCT endpoints for many studies, but we stopped using thickness measurements as an endpoint. Instead, we look for structural changes and sub- or intraretinal fluid.”

There has also been the suggestion that using OCT should expand the number of endpoints used, rather than being used as a single endpoint. For instance, Anat Loewenstein, MD, head of the department of ophthalmology at the Tel-Aviv Sourasky Medical Center and vice dean of the Sackler Faculty of Medicine at Tel-Aviv University, stresses the need to retain visual acuity as an endpoint. “I think the benefit of every treatment is in preserving or improving visual acuity,” she says. “Therefore, I don't think BCVA should be replaced.”

Figure 1. OCT is used increasingly as a secondary endpoint.

However, she continues, “We need to look at the morphology of the retinal layers, including the outer retina layers—loss of photoreceptors, hyper-reflective dots, regularity of RPE—and the inner retinal layers. These should be added to the endpoints. Also, when we recruit patients to studies, we need to know the duration of edema prior to entry and mention this in the baseline characteristics.” Limiting enrollment to patients with shorter disease duration, she explains, may increase the correlation between BCVA and anatomical endpoints.

However, such choices may not always be available, particularly in early phase trials, which tend to recruit patients with refractory cases. OCT thickness changes are still important measurements in these trials, as BCVA may not be expected to improve. “OCT endpoints have been most important in early-phase trials,” says Jeffrey S. Heier, MD, of Ophthalmic Consultants of Boston, “This is true in phase 1 especially, but even some phase 2 trials, in which patients with more advanced disease may be enrolled. Anatomic endpoints allow some measure or evidence of biologic activity that the measurement of visual acuity would not necessarily pick up. In fact, retinal thickness, or improvement in retinal fluid, is often a much better gauge of biologic activity in early-phase trials, especially as many of these trials are unmasked. They provide objective evidence of change.”

That being said, it is extremely unlikely that OCT will ever replace visual acuity as a primary endpoint in later phase or registration trials, provided the hope of recovering visual acuity remains in the patient group being treated. Abdhish Bhavsar, MD, of the Retina Center in Minneapolis, currently uses OCT in trials ranging from wet AMD to documenting cystoid macular edema and pigment epithelium detachments.

Nevertheless, Dr. Bhavsar remains convinced that OCT as an endpoint is an addition and not a replacement. He says, “We are a long way off from using anatomic endpoints alone for primary endpoints in clinical trials. Best-corrected visual acuity is still a very reliable and reproducible endpoint, and it is also clinically meaningful.” The FDA's position is the same, as enunciated by Dr. Chambers: “OCT provides useful information; it's just not a surrogate for visual acuity. We have a noninvasive tool as an additional tool to evaluate the retina, so if it replaces anything, it replaces subjective assessment of retinal thickness.”


For clinical trials examining geographic atrophy, anatomic endpoints other than OCT may play important roles. As Frank G. Holz, MD, of the department of ophthalmology at the Uni versity of Bonn in Germany, explains, “The enlargement of geo graphic atrophy is now used as the main outcome parameter in various interventional clinical trials. These area measurements are based on fundus autofluorescence images using confocal scanning laser ophthalmoscopy. They are highly re pro ducible and represent objective measurements, as opposed to psychophysical functional testing, including visual acuity.”

Dr. Holz notes that visual acuity has particular limitations in the late phenotype of dry AMD because the foveal area may be spared until late in the disease process despite marked enlargement of the atrophic area. As a result, he says, BCVA has been partly replaced as an endpoint in dry AMD trials, although he is quick to point out that BCVA is “highly important for outcome assessments, as there is also a high correlation with patient-reported outcomes.”

Dr. Heier agrees that it is important in assessing dry AMD treatments to use anatomic endpoints in addition to BCVA. “The enlargement of geographic atrophy, as measured by fundus autofluorescence, has also been used to help gauge activity in agents attacking dry AMD,” Dr. Heier says, “and this is another instance in which change in VA may not actually reflect disease activity, especially if the center vision is already damaged.”

As noted by Dr. Chambers above, the FDA agrees that an eye that anatomically has an area that lacks photoreceptors is an eye that has a visual defect. Prevention of that loss of photoreceptors is a useful clinical endpoint.

Beyond geographic atrophy measurements, a few other anatomic endpoints have been incorporated into clinical trials. For example, Dr. Bhavsar reports that his practice has used fluorescein angiography to identify leakage and to distinguish between types of CME and choroidal neovascular membranes. In addition, Dr. Heier suggests that as ophthalmologists learn more about the relationship between wet and dry AMD and as treatments are further developed for the dry version of the disease, drusen volume may also become a useful anatomic endpoint for clinical trials.

Figure 2. Snellen visual acuity is being augmented by objective measures.


As noted above, the FDA and retinal physicians are in agreement that, for the majority of clinical trials and conditions, visual acuity will remain the key endpoint used. Outside the United States, the regulatory environment varies from place to place. Dr. Holz, who practices in Europe, reports that the European Medicines Agency, which is based in London and evaluates drugs and medical devices for approval, is currently considering anatomic endpoints for clinical trials in retinal disease.

Meanwhile in Israel, where Dr. Loewenstein practices, she reports, “Anatomic endpoints are looked at seriously when dealing with regulatory agencies and institutional review boards regarding trials, but needless to say, visual acuity seems to the regulators an extremely important parameter, which is easier to understand and follow up.”

It is also important to understand why variation exists between anatomic measurements and visual acuity assessment. Dr. Heier explains, “When larger populations are studied, there are clearly relationships between the two. For instance, anti-VEGF trials typically show VA improvements and retinal thickness decreases that are mirror images of each other. Retinal thickness still remains an unequivocal marker of biologic activity. However, there are a number of reasons that VA and retinal thickness may not correlate directly—underlying retinal damage, duration of disease, ischemia, lag between anatomic and functional recovery, and so on.”

Until those differences can be accounted for in some way, anatomic endpoints will only supplement more subjective, patient-oriented outcomes. The FDA's position reflects that point of view, as Dr. Chambers elaborates: “Ultimately, prior to a new drug application, we expect to see adequate and well-controlled trials that demonstrate safety and efficacy. Regarding how they get there, as long as they bear safety in mind as they proceed, we try to give freedom to the investigators.” RP