Disease activity criteria shapes dosing interval outcomes in clinical trials

The definition of active disease in clinical trials affects the estimates of the durability of the agents that are tested in those trials. Image credit: ©tippapatt – stock.adobe.com

Earlier this year, Marco Zarbin, MD, PhD, FACS, delivered his talk titled “An Assessment of the Impact of Disease Activity Criteria on Dosing Interval Assignment in Clinical Trial Patients with Neovascular AMD” at the American Society of Retina Specialists (ASRS) meeting 27 July to 1 August in Seattle, Washington, United States. He presented the data on behalf of coauthors Christina Weng, Audrey Souverain, Ivaylo Stoilov and Philippe Margaron.

The emphasis on neovascular age-related macular degeneration (nAMD) allowed Dr Zarbin to focus on differing patient outcomes based on dosing intervals. “The take-home message of this talk is that the definition of active disease in clinical trials affects the estimates of the durability of the agents that are tested in those trials,” Dr Zarbin explained.

In clinical trials, he said, it is necessary for disease activity to drive treatment decisions. In cases of nAMD, for example, there are a set of common activity indicators that physicians consider when giving a diagnosis. “I think for most of us in practice, we judge the presence of active nAMD to exist if there’s a loss of vision, or if there’s increased disease activity on the optical coherence tomography [OCT], or if there’s a presence of new macular hemorrhage,” Dr Zarbin said.

But due to a lack of consistent disease activity criteria across clinical trials, it is challenging to compare the durability of different agents, which is a vital consideration when dosing patients in a clinical setting. Dr Zarbin explained that criteria which "reflect real-world clinical practice" increase the likelihood of trial results being reproduced in a clinical setting.

To illustrate this principle, Dr Zarbin discussed the phase II CANDELA trial (NCT04126317), in which 2 mg aflibercept was compared with 8 mg aflibercept. Investigators performed disease activity assessments at weeks 24, 28, 36 and 40. Dr Zarbin highlighted two criteria that reflected active disease: anatomic findings that were considered to be vision threatening, or a loss of vision of five or more letters.

Meanwhile, in the phase III PULSAR trial (NCT04423718), investigators used different indicators of active disease. In this trial, active disease was judged to be present if there was both a loss of vision of more than five letters, a change in the OCT and/or new macular hemorrhage. Researchers compared dosages of 2 mg aflibercept every 8 weeks, 8 mg aflibercept every 12 weeks and 8 mg aflibercept every 16 weeks. Disease activity assessments were completed at weeks 16, 20 and 24.

These differences in active disease criteria were reflected in the study outcomes. “It may not be surprising that the durability estimates were not the same,” Dr Zarbin said. “In CANDELA, where either a change in vision or a change in anatomy mandated an injection, about 53% of the patients could be maintained with injections every 12 weeks, whereas in the PULSAR trial, where both a change in vision and a change in anatomy were required to mandate an injection, durability increased such that 83% of the patients could be maintained with injections every 12 weeks or longer.”

Dr Zarbin continued, discussing the TENAYA [NCT03823287] and LUCERNE [NCT03823300] studies, which were the registration trials for faricimab. In these trials, he said, 2 mg aflibercept every 8 weeks was compared with faricimab 6 mg, and disease activity assessments were done at weeks 20 and 24. Active disease was judged to be present if there was either a loss of vision or a change in the OCT or a new macular hemorrhage. “Using these criteria, by year 2, about 80% of the patients [receiving] faricimab could be maintained with dosing intervals of every 12 to 16 weeks,” Dr Zarbin said. “The change that most frequently mandated injection was a change in the OCT, which was the case for about 70% of the time that injections were administered.”

Around 25% of the time, a change in vision indicated the presence of active disease and around 10% of the time, it was a presence of new macular hemorrhage.

“My point is that if any one of the above events occurred at weeks 20 or 24, clinicians administered faricimab treatment as each event independently indicated the presence of active neovascular AMD,” Dr Zarbin said. “We wondered, what would the durability of faricimab be if we changed the disease activity criteria such that both a change in vision and a change in OCT were required to mandate an injection using the criteria employed in the PULSAR trial.”

In this post hoc analysis of the TENAYA and LUCERNE data, investigators recalled that in those trials, either a change in vision or OCT or new macular hemorrhage resulted in 78% of the patients being maintained at injection intervals of 12 to 16 weeks. When investigators required both a change in vision and a change in OCT, the durability seemed to increase; with these criteria, 96% of the patients could be maintained with injections every 12 to 16 weeks.

“By changing the disease activity criteria, we changed the estimate of the durability of faricimab,” Dr Zarbin said. He stated that the disease activity criteria in TENAYA and LUCERNE, in which either a change in vision or anatomy triggered requirement for an injection, “more accurately reflect routine clinical practice.”

“And as we found in the trial, and as I think is the case in practice, the majority of the time, the change in disease activity criteria that mandates the injection is a change in the OCT,” he added. Dr Zarbin said this consideration was vital to understanding durability estimates in both TENAYA and LUCERNE, as well as in the CANDELA and PULSAR trials. “When we apply the requirement that both a change in vision and a change in anatomy are needed in order to maintain to mandate an injection, we increase the assignment to longer dosing intervals,” he summarised.