A phase 3, double-masked, multinational trial found that low-dose sirolimus (rapamycin) injected every 2 months can reduce vitreous haze in non-infectious posterior segment-involving uveitis, with low rates of ocular side effects.
Sirolimus is a bacteria-derived immunosuppressant that inhibits the activation of T cells and B cells and reduces cytokine production.
Investigators randomized 347 subjects with a spectrum of intermediate, posterior and panuveitis to 2 different doses of intravitreal sirolimus (880 μg or 440 μg) or active control (44 μg). At 5 months, more patients receiving 440-μg injections of sirolimus every other month achieved complete resolution of inflammation than those treated with an active control dose of 44 μg or 880-μg (22.8%, 10.3% and 16.4%, respectively).
Mean BCVA was maintained throughout the study in each dose group, and most patients in the 440-μg group who received corticosteroids at baseline successfully tapered off corticosteroids (76.9% vs. 63.6%) by month 5.
There were no significant differences in number of subjects from each group who required rescue therapy or who had ocular or systemic adverse events, with the most common complication being iridocyclitis (18.8-19.7%) likely due to the requirement for a topical taper before day 1 of the study. There was a dose-dependent trend for inflammation-related adverse events such as uveitis (9.4-22.2%) and cataracts (2.6-7.7%), though it did not reach significance. IOP elevation occurred at a similar rate across the groups, yet patients were given IOP-lowering medications at the investigators discretion during the study period.
Due to the greater efficacy and numerically lower risk of adverse events, the authors suggest that the risk-to-benefit ratio of the 440 µg dose is superior to the 880 µg.
The study does have several limitations. As active control was used rather than a placebo, there is potential that the treatment effect was underestimated, though the authors argue that consistent results from subgroup analysis demonstrate clear benefit of the 440 µg dose over control. Because the intent-to-taper population was small (most subjects were already corticosteroid-free at randomization), the power of the statistical analysis of the steroid-sparing ability of each dose was also limited. Furthermore, the use of vitreous haze as a primary outcome endpoint is prone to confounders due to subjectivity of the masked grader, potentially explaining why the active control group also had major improvements in VH.
In conclusion, low dose (440 μg) intravitreal sirolimus injected every two months can reduce vitreous haze in non-infectious posterior segment-involving uveitis patients, with low rates of ocular side effects. Yet the utility of this agent as an adjunctive vs. a single therapy and its long-term efficacy for the treatment of uveitis requires further study.