Oral Agent for Diabetic Macular Edema Misses Mark but Still Shows Promise

DENVER — An investigational oral therapy for diabetic macular edema (DME) did not significantly affect retinal thickening in a placebo-controlled randomized trial.

After 6 months of treatment, mean central subfield thickness (CST) decreased by 15 µm with the oral connexin 43 (Cx43) hemichannels inhibitor tonabersat as compared with a 5-µm increase in the placebo group. An adjusted analysis resulted in a 16-µm difference, which did not achieve statistical significance (P=0.08).

After excluding patients who had received anti-VEGF therapy, a post-hoc analysis showed an adjusted mean difference of 21 µm, which was statistically significant (P=0.02). The investigators also found that patients with higher baseline CST appeared to derive greater benefit from tonabersat.

Whether the secondary and post-hoc findings will lead to additional studies of tonabersat remains to be seen, said Jennifer K. Sun, MD, MPH, of the National Eye Institute in Bethesda, Maryland, at the Association for Research in Vision and Ophthalmology meeting.

“There was no statistically significant difference in central subfield thickness at 24 weeks between tonabersat- and placebo-treated patients,” said Sun. “However, there were consistent modest anatomic trends that favored tonabersat across additional analyses.”

“Our subgroup analyses showed a dramatic difference in terms of greater CST reductions in thicker eyes compared to thinner eyes,” she added. “We had originally planned to enroll equal numbers of thinner or moderately thick eyes and the thickest eyes, but we had difficulty enrolling patients who had thicker maculas and good vision.”

During a discussion that followed, co-moderator Arshad Khanani, MD, of Sierra Eye Associates in Reno, Nevada, noted that “it seems like there is some sort of an efficacy signal here. Obviously, phase II studies are always done to kind of learn and then enrich our patient population. Do you think the effect is real?”

Khanani pointed out the use of the last-observation-carried-forward (LOCF) method to account for patients who were excluded from the analysis, an appreciable dropout among patients who did not benefit, and the impact of side effects.

“Those little things,” he said. “Did you look at LOCF from their dropout, or was it just bad data?”

Sun responded, “Although this was not a home run and we did not make the primary outcome, I think there is enough signal here that we can conclude there probably is good evidence for some sort of biological effect. I would love to see additional studies, kind of parsing out what it all means and what the role might be in our overall population of DME patients.”

Co-moderator Aude Couturier, MD, PhD, of the French Myopia Institute in Paris, asked whether future studies might use fluorescein angiography or other techniques to aid in selecting patients who might be more likely to benefit from tonabersat in terms of blood flow, leakage, and drug penetration.

“If there was interest in doing a second, more focused phase II study, we might target it specifically for patients with thicker retinas,” said Sun. “We might be able to go through the fluorescein and take a look post hoc at whether or not there was more leakage or potential worse disruptions.”

By way of introduction to the study, Sun noted that current treatments for DME are effective but a substantial proportion of patients do not benefit. Treatment can be costly, invasive, and require frequent office or clinic visits. A need exists for novel, VEGF-independent, well-tolerated, noninvasive treatments with the potential to reduce the burden on patients.

Cx43 hemichannels are overexpressed in diabetic retinopathy and associated with activation of proinflammatory molecules. By targeting Cx43 hemichannels, tonabersat reduces cytokine activation and recycling. Prior studies in migraine showed tonabersat was well tolerated and had no significant safety issues, said Sun.

In preclinical models of diabetic retinopathy, tonabersat blocked the release of key proinflammatory cytokines, including VEGF. The drug also reduced vascular leakage and edema of ischemic retinal injury, Sun added.

For the current phase II study, the investigators enrolled patients with type 1 or 2 diabetes, center-involved DME associated with above sex-based thresholds for CST, and visual acuity of 20/32 or better. Eligible patients had received no treatment for DME in the past year and no more than four injections in the prior 3 years. Patients were randomized to tonabersat or placebo, and the trial had a primary endpoint of change in baseline CST after 6 months.

The study included 129 patients, 15 of whom had bilateral study eyes. Most patients (96%) had type 2 diabetes. Patients in the tonabersat arm had a longer duration of diabetes (19 vs 16 years), higher rate of insulin use (69% vs 54%), more renal disease (23% vs 9%), a lower mean estimated glomerular filtration rate (74 vs 82 mL/min/1.73 m²), and more heart failure (11% vs 3%).

Baseline CST averaged 359 µm in the tonabersat arm and 362 µm in the placebo group. Over 60% of patients had CST <75 µm above the threshold and 30-36% had CST ≥75 µm above the threshold. Both groups had baseline visual acuity of 20/25.

The adjusted mean difference in CST at 6 months showed that tonabersat did not achieve statistical significance (95% CI -34 to 2). The results included LOCF data for four eyes in the tonabersat arm and five in the placebo group from patients who exceeded limitations for prior treatment of DME. In the post-hoc analysis, two additional eyes were switched to LOCF because of anti-VEGF therapy, one of which had a 200-µm improvement in CST.

More patients in the tonabersat arm discontinued treatment before 6 months (13 vs 6). Serious adverse events occurred more often with tonabersat (19% vs 8%), and more patients in the tonabersat arm required hospitalization (16% vs 3%).

Please enable JavaScript to view the comments powered by Disqus.