Macular edema consists of intra- or subretinal fluid accumulation in the macular region. It occurs during the course of numerous retinal disorders and can cause severe impairment of central vision. Major causes of macular edema include diabetes, branch and central retinal vein occlusion, choroidal neovascularization, posterior uveitis, postoperative inflammation and central serous chorioretinopathy. The healthy retina is maintained in a relatively dehydrated, transparent state compatible with optimal light transmission by multiple active and passive systems. Fluid accumulation results from an imbalance between processes governing fluid entry and exit, and is driven by Starling equation when inner or outer blood-retinal barriers are disrupted. The multiple and intricate mechanisms involved in retinal hydro-ionic homeostasis, their molecular and cellular basis, and how their deregulation lead to retinal edema, are addressed in this review. Analyzing the distribution of junction proteins and water channels in the human macula, several hypotheses are raised to explain why edema forms specifically in the macular region. "Pure" clinical phenotypes of macular edema, that result presumably from a single causative mechanism, are detailed. Finally, diabetic macular edema is investigated, as a complex multifactorial pathogenic example. This comprehensive review on the current understanding of macular edema and its mechanisms opens perspectives to identify new preventive and therapeutic strategies for this sight-threatening condition.

The anti-CD20 monoclonal antibody (mAb) rituximab has been used successfully for lymphoma therapy for more than 10 years. Although several direct mechanisms by which anti-CD20 mAbs act have been characterized in vitro, their specific role in clinical efficacy is still debated. Little is known about the possible antitumor immune response that they may induce in patients, despite clinical data suggesting a "vaccinal" effect. We show here that an initial treatment with anti-CD20 induces protection against human CD20-expressing tumor cells and allows immunocompetent mice to survive tumor challenge. This long-lasting protection requires the presence of the Fc portion of the anti-CD20 mAb and is achieved through the induction of a cellular immune response. Only CD4 ؉ cells were needed at the beginning of the treatment, but both CD4 ؉ and CD8 ؉ cells were required after tumor challenge to achieve protection. Finally, we show that interleukin-2 treatment, given after tumor challenge, improves the overall survival rate, compared with that obtained by anti-CD20 treatment alone. These findings demonstrate that anti-CD20 mAbs exert therapeutic effects through the induction of an adaptive cellular immune response, aside from any direct mechanisms involving effectors from innate immunity. (Blood. 2010;116(6):926-934)