Effect of intravitreal methotrexate and rituximab on interleukin-10 levels in aqueous humor of treated eyes with vitreoretinal lymphoma.

Intraocular cytokines are promising diagnostic biomarkers of vitreoretinal lymphoma. Here, we evaluate the utility of IL-10, IL-6 and IL-10/IL-6 for discriminating lymphoma from uveitis and report the effects of intraocular methotrexate and rituximab on aqueous cytokine levels in eyes with lymphoma. This is a retrospective case series including 10 patients with lymphoma and 7 patients with uveitis. Non-parametric Mann-Whitney analysis was performed to determine statistical significance of difference in interleukin levels between lymphoma and uveitis. Compared to eyes with uveitis, eyes with lymphoma had higher levels of IL-10 (U = 7.0; two-tailed p = 0.004) and IL-10/IL-6 (U = 6.0; two-tailed p = 0.003), whereas IL-6 levels were more elevated, although insignificant, in those patients with uveitis than in lymphoma (U = 15.0; two-tailed p = ns). Using a receiver operating characteristic analysis to identify threshold values diagnostic for lymphoma, optimal sensitivity and specificity improved to 80.0% and 100%, respectively, for IL-10>7.025 pg/ml and 90.0% and 100.0%, respectively, for IL-10/IL-6>0.02718. In patients in whom serial interleukin levels were available, regular intravitreal treatment with methotrexate and rituximab was associated with reduction in IL-10 levels over time. In conclusion, optimal IL-10 and IL-10/IL-6 threshold values are associated with a diagnostic sensitivity ≥80% and specificity of 100%. Therefore, these cytokines may serve as a useful adjunct in the diagnosis of lymphoma. While negative IL-10 and IL-10/IL-6 values do not exclude a diagnosis of lymphoma, elevated levels do appear to be consistent with lymphoma clinically. Moreover, elevated levels of IL-10 in the setting of a clinically quiet eye may point to impending disease recurrence. Lastly, once lymphoma is diagnosed, IL-10 levels can be monitored over time to assess disease activity and therapeutic response.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.