Novel Drug Fails to Improve Diabetic Retinopathy, but Slows Progression

SAN DIEGO -- A novel oral angiogenesis inhibitor failed to improve diabetic retinopathy but significantly slowed the rate of progression in a randomized placebo-controlled trial.

Numerically, more study eyes treated with the Ref-1 inhibitor APX3330 met the primary endpoint of ≥2-step improvement in the Diabetic Retinopathy Severity Scale (DRSS) at 12 and 24 weeks (4% vs 0% and 8% vs 0%, respectively), but neither difference achieved statistical significance. In an analysis that included the fellow eye, more than three times as many eyes met the primary endpoint, but the difference from placebo still did not reach statistical significance.

Significantly fewer patients in the APX3330 arm had ≥3-step worsening in the DRSS, which could be an approvable endpoint, said Christina Y. Weng, MD, of Baylor College of Medicine in Houston, during the American Society of Cataract and Refractive Surgery annual meeting.

"While the phase II study primary endpoint was not met, a lot of lessons were learned, and potentially a path forward, looking at binocular ≥3-step worsening of the DRSS as a potentially acceptable endpoint by the FDA," said Weng. "Phase III trials are being planned. Finally, good total tolerability in diabetic patients."

Anti-VEGF therapy to prevent diabetic retinopathy has a checkered history. A large randomized placebo-controlled trial of aflibercept (Eylea) showed that early intervention led to a slowing of disease progression but no visual acuity benefit. The results did not change with longer follow-up, leading the NIH to release a statement noting that "treating eyes only as needed is the best approach."

A separate trial showed that a year of more intensive treatment with aflibercept led to improvement in the DRSS and a reduced risk of complications as compared with sham intravitreal injections.

APX3330 targets reduction-oxidation effector factor-1 (Ref-1), which activates downstream molecules such as HIF-1α and NF-κB, leading to abnormal angiogenesis and inflammation, Weng said during her introduction to the study. APX3330 has an extended history that includes investigation in hepatic inflammation and solid tumors and evidence of safety and efficacy in multiple preclinical studies. ZETA-1 is the first clinical trial evaluating APX3330 for an ophthalmic indication.

Investigators enrolled and randomized 103 patients with moderate or severe nonproliferative diabetic retinopathy or mild proliferative diabetic retinopathy (DRSS 47, 53, 61). Patients with anti-VEGF therapy in the previous 6 months and those with HbA1c ≥12% were excluded. Patients received APX3330 twice daily or matching placebo for 24 weeks. The primary endpoint was the proportion of patients with ≥2-step improvement in baseline DRSS.

At study entry, 40 patients had a DRSS of 47, 53 had a DRSS of 53, and 10 had a DRSS of 61. Scores for the fellow eye were <43 in 26 patients, 47 in 32 patients, 53 in 21 patients, 61 in five patients, and ≥65 in 10 (moderate or severe proliferative diabetic retinopathy). Best corrected visual acuity in the study eye was 80, and 77 in the fellow eye.

DRSS assessment at 12 and 24 weeks showed no significant difference in the proportion of study eyes achieving ≥2-step improvement, even though no patient in the placebo arm met the primary endpoint. Assessment of the fellow eye at the same time points showed that 14% of patients allocated to APX3330 had at least a 2-step improvement in DRSS versus 4% in the placebo group. At 24 weeks, the rates were 27% and 5%, respectively. Still, the differences did not reach statistical significance.

According to Weng, the sponsors had preliminary discussions with the FDA about allowing binocular ≥3-step DRSS worsening (sum of DRSS change in both eyes) as an acceptable endpoint for registration. For systemic drugs, the FDA accepts improvement or worsening (prevention of progression) in diabetic retinopathy as endpoints.

In the ZETA-1 trial, use of binocular improvement or worsening in DRSS at 24 weeks would have resulted in a statistically significant difference in favor of APX3330, as 16% of patients in the placebo arm would have met that outcome versus none in the APX3330 arm (P=0.04). Additionally, 8% of patients in the placebo arm versus none in the APX3330 group had ≥4-step worsening at 24 weeks.

Patients allocated to APX3330 had a trend toward preservation of binocular visual acuity. The proportion of patients who had ≥5-letter loss in best corrected visual acuity at 24 weeks was 19% with placebo and 5% with APX3330 (P=0.07).

Adverse events (AEs) were generally mild. No serious AEs related to study medication occurred, and no ocular AEs occurred except as expected with progression of diabetic retinopathy. No effects on major organ systems were observed, said Weng.

Comment