Retinal and Retinal Pigment Epithelium (RPE) Autoimmunity in Age-related Macular Degeneration (AMD)

Up to 10% of patients with neovascular AMD treated with ranibizumab respond poorly or worsen despite therapy. The reason for this lack of response is unclear. We have preliminary data that suggests abnormal autoimmune activity is apparent in these patients. Previous studies have shown evidence of retinal autoimmunity in AMD patients, but there is very little data describing any specific immunologic commonality that correlates with disease and/or poor response. (8,9) Perhaps just as significantly there is little data regarding the immunologic activity of age-matched normals, making published data hard to evaluate especially in this age group in which autoimmunity is known to increase. (8,9) While there are many known retinal antigens in autoimmune retinal disease, the role of these antigens is not well established in AMD and not all the antigens have been identified. (24) Moreover, RPE-reactivities are only beginning to be understood in ocular disease. (25-27) We intend to address humoral responses in AMD by making a systematic comparison of the immunologic activity of ranibizumab responders, ranibizumab initial non-responders, and a comparable population of age-sex-race matched normal controls. Data suggests that 5 groups of patients are evident after 3 treatments with ranibizumab: 1) rapid responders, 2) delayed responders, 3) gradual responders, 4) acutenon-responders and 5) chronic non-responders. We hypothesize that non-responders and gradual responders may in fact be patients with complicating underlying autoimmune activity involving retinal and RPE antigens, which are exposed secondary to the breakdown of the blood-retinal barrier during CNV development. We will study this humoral response (antibody production) over the treatment period, as it likely is changing at different rates in the patients with different responses. In addition we will correlate underlying genetic phenotype in these patients.

For this study, we plan to look at 2 treatment groups and 2 control groups:

• Group 1: patients with neovascular AMD who respond to ranibizumab after 4 consecutive injections with ranibizumab
• Group 2: age-sex-race matched normal population controls (without AMD)
• Group 3: patients with neovascular AMD who are acute non-responders to anti-VEGF treatment after 4 or more consecutive injections
• Group 4: age-sex-race matched dry AMD patients (AREDS category 2/3 ou) controls
• Group 5: patients with neovascular AMD who are chronic non-responders to anti-VEGF treatment after 4 or more consecutive injections

This is an open-label study assessing antibody formation (anti-RPE and anti-retinal) in 5 groups. Group 1 (n=40) will include neovascular AMD patients treated with ranibizumab. Patients will be included and receive 4 ranibizumab 0.5mg intravitreally at 4-6 week intervals and then twice more "as needed" (PRN) at 4-6 week intervals. After the 4th ranibizumab injection, if a Group 1 patient has not responded (persistent fluid on OCT), they will be moved into Group 3 (anti-VEGF acute non-responders) or Group 5 (anti-VEGF chronic non-responders). This will reduce the eventual number of subjects enrolled in Group 1 to approximately 36, as we anticipate approximately 4 subjects to have to move to either Group 3 or Group 5 as a non-responder. Group 2 (n=40) will be an age-sex-race matched normal subjects from the population that does not have AMD. Group 3 (n=8) and Group 5 (n=7) (for a combined total of 15 subjects, approximately 4 of whom transferred from Group 1) include patients treated with 4 or more injections of anti-VEGF treatment at 4-8 week intervals without an initial response (Initial non-response is defined as < 100 microns of improved [decreased] retinal thickening by OCT). Group 3 patients, the acute non-responders, will be included after the 4th injection and followed for 2 more visits at 4-8 week intervals during which time they can receive "as needed" anti-VEGF treatment(s) at the investigator's discretion for any fluid on OCT. Group 4 (n=40) will be age-sex-race matched patients with Dry AMD as controls for immune response before there is a neovascular response. Group 5 patients, the chronic non-responders, will be included after the 4th injection and followed for one (1) visit at Month 4 during which time they can receive an "as needed" anti-VEGF treatment at the investigator's discretion for any fluid on OCT.

NOTE: Only 10% of Group 1 (approximately 4 patients) are expected to be non-responders, therefore, 11 of the Group 3 and Group 5 subjects will be patients treated outside the study who are found to be non-responders by chart review. These patients will then be enrolled at the Month 4 visit to supplement the subjects transferred from Group 1 for a total of 15 patients in Groups 3 and 5.

We will use Western blotting for global assessment of all autoantibodies against the full complement of retinal proteins in both normal individuals (Group 2) and those treated for exudative AMD (Group 1), those initial non-responders to ranibizumab (Group 3), and patients with "dry" AMD (Group 4).

Genotyping (CFH and HTRA1) will be performed on all Groups. Approximately 25 ml (2 tablespoons) of blood will be sent to Dr. Khang Zhang of the Shiley Eye Center at the University of California, San Diego and he will perform the genetics analysis.

This study will investigate if antibody production differs between nv AMD patients (Groups 1, 3 and 5) and the normal population (Group 2), if it differs between ranibizumab responders (Group 1) and non-responders to any anti-VEGF treatment (Groups 3 and 5), and we will also see how patients with dry AMD (Group 4) compare with the nv AMD groups (Groups 1, 3, and 5).