Geographic Atrophy Growth Rate Reduced by Half With Anti-Inflammasome Implant

NEW ORLEANS — A new therapeutic approach to geographic atrophy (GA) showed promise for slowing disease progression, according to a preliminary clinical trial.

The rate of GA lesion growth decreased by 54% after 6 months with a bioerodible implant that continually releases the dual inflammasome inhibitor kamuvudine K8 (K8). The effect on GA growth rate increased over time, from 43% during the first 3 months to 70% during months 3-6. The GA growth rate was slower in the treated versus untreated eye in 25 of 30 patients.

Of the three dose levels evaluated, the mid-level dose proved most effective. No drug-related adverse events (AEs) occurred at any dose, reported Liangbo Shen, MD, of Duke University Medical Center in Durham, North Carolina, at the American Academy of Ophthalmology meeting.

Enumerating convenience advantages K8 offers over existing GA treatment options, Shen said, “K8 is a quarterly injection, requires no refrigeration, and comes in a preloaded injector.”

On the basis of the results, a phase III pivotal clinical program has been planned, he added.

During a discussion, Jay M. Stewart, MD, of the University of California San Francisco, noted that investigators pooled results from the untreated eyes, resulting in “a lot of complexity that we’re not used to seeing. To what extent do you think an outlier or two in one of those groups could sway the results and make us think that one of the groups is having better efficacy than the other?”

Shen agreed and said investigators also performed an analysis comparing treated and untreated eyes in the same patients, and the results were similar after adjustment for sample size and other factors.

“We need a longer, larger study to investigate the treatment efficacy,” said Shen. “But it is encouraging to see a phase II trial with a significant reduction in GA growth rate.”

Although the trial was not powered for visual acuity, investigators compared best-corrected visual acuity (BCVA) at baseline and at the end of the study and found an improvement favoring the K8-treated eyes. Noting that vision was not masked in the study, Stewart asked how the results should be interpreted.

Shen acknowledged the limitation and also the fact that investigators treated the eye with the more severe GA, which might have skewed the results.

“Obviously, we all want something that can slow GA progressing in terms of visual function, but I think we have to have a much larger study to answer that question,” Shen added.

In response to another question, Shen agreed that the 6-month follow-up might have been too short to allow for emergence of a true dose effect, noting no difference from control with the lowest dose and no significant difference between the middle and high doses.

In contrast to the two approved treatments for GA, both of which are complement inhibitors, K8 is a nucleoside reverse transcriptase inhibitor that blocks activation of two inflammasomes, NLRP3 and NLRC4. Nucleoside reverse transcriptase inhibitors have intrinsic anti-inflammatory and anti-viral activity. K8 preserves the anti-inflammatory activity but loses the antiviral activity and associated mitochondrial toxicity, said Shen.

In preclinical models, NLRP3 and NLRC4 both contribute to degeneration of retinal pigmented epithelium (RPE), he continued. Three different NLRP3-selective inhibitors failed to prevent RPE loss, but simultaneous inhibition of both inflammasomes preserved RPE.

Shen reported findings from a phase II trial of K8 in 30 patients with bilateral GA. Ten patients each received implants with one of three dose levels in the eye with the worse vision. The fellow eye served as a control. The primary endpoint was GA growth rate at 6 months.

Eligible patients had a GA lesion area of 2.5-20 mm², presence of hyperautofluorescence in the junctional zone, no prior treatment for GA, BCVA of 24+ letters (≥20/320 Snellen equivalent), and no history of exudative age-related macular degeneration.

At 6 months, the GA lesions in the control eyes had progressed by an average of 1.18 mm². By comparison, the eyes treated with the mid-level dose of K8 had a 54% reduction in growth rate by the FDA-preferred slope model (P=0.016), and a 50% reduction versus control after adjustment for baseline lesion size (P=0.005). The GA slope reduction by dose group (versus pooled control eyes) achieved statistical significance for the mid-level and high doses of K8 (P=0.005, P=0.029).

A prespecified analysis of BCVA in 14 treated eyes and 16 control eyes showed a difference of +4 letters in favor of K8 (P=0.004).