LIQUID BIOPSY IN UVEAL MELANOMA: PROTEOMICS OF AQUEOUS HUMOR

Posterior uveal melanoma (UM) is the most common primary intraocular malignancy in adults. UM occurs in a closed microenvironment characterized by a direct contact with ocular fluids. Therefore, the analysis of these fluids may allow to search and identify specific biomarkers providing information about the different pathways involved in UM pathogenesis and in the retinal complications related to its conservative treatment. To evaluate, in vivo, the presence of specific AH biomarkers in eyes affected by UM before and after the conservative treatment. Seventy-two eyes affected by primary UM underwent full ophthalmic examination. During brachytherapy (Iodine-125) surgical procedure, AH sample collection and 25-gauge transscleral fine needle aspiration biopsy of the primary tumor were performed. AH samples were analyzed to detect the concentration of selected proteins. Cytologic material underwent fluorescence in situ hybridization for chromosome 3. Tumor thickness and largest basal diameter (LBD) were quantified. The same seventy-two UM subjects underwent AH sampling at the time of a planned para surgical procedure 20 months after brachytherapy. The AH of thirty-six normal eyes, scheduled for cataract surgery, was used as control. Compared with the control group, significantly higher levels of SSTR1 (p=0.028), HMB45 (p=0.018), IL-6 (p=0.047), IL-8 (p=0.008), RANTES (p=0.008), PEDF (p=0.048), Osteopontin (p=0.048), EGF (p=0.041), bFGF (p=0.019), MIF (p=0.009), MCP-1 (p=0.023). VEGF concentration between the two groups was statistically borderline (p=0.058). Comparison between clinical characteristics and cytokine concentrations showed a positive correlation between tumor thickness and IL-8 (p=0.032), and degree of serous retinal detachment and IL-6 (p=0.021). Monosomy 3 was detected in 20 cases (57%) and disomy 3 in the remaining 15 cases (43%). No correlation was found between chromosome 3 status and tumor dimensions (thickness: p=0.301; LBD: p=0.321), and between monosomy 3 and tumor location (ciliary body: p=0.871, choroid: p=0.931). Compared to UM with disomy 3, significant higher levels of IL-6 (p=0.022) IL-8 (p=0.003), RANTES (p=0.005), EGF (p=0.018), bFGF (p=0.012), VEGF (p=0.026), MIF (p=0.005) and MCP-1 (p=0.005) were detected in UM with monosomy 3. Comparison between clinical characteristics of UM with monosomy 3 and cytokine concentration showed a positive correlation between: tumor thickness and IL-8 (p=0.022), tumor thickness and VEGF (p=0.031), LBD and RANTES (p=0.015), and degree of retinal detachment and IL-6 (p=0.021). Compared with the control group, significantly higher protein levels of: GNAQ (p=0.022), BAP1 (p=0.013) and SF3B1 (p=0.023) were detected in eyes with UM. Cluster analysis of UM group revealed 2 clusters, one showing higher expression of GNAQ and BAP1 protein and one of EIF1AX protein. Moreover, the 2 clusters corresponded with the chromosome 3 status of UM. the AH of the UM group 20 months after the treatment showed that 22 subjects (30.5%) developed RME. All RME eyes at OCT analysis had significant increase in thickness of inner and middle retinal layers (p=0.002) mostly determined by intraretinal fluids, in number of hyperreflective retinal foci (p=0.003) and in areas of disrupted external limiting membrane (p=0.004). GFAP, Kir 4.1 and all inflammatory biomarkers were dramatically increased in RME eyes compared to those without RME (p=0.035, p=0.021 and p=0.032, respectively). Specific and selected proteins may be detected in the AH of eyes affected by UM. These novel findings confirm the possibilities provided by AH analysis in UM before and after its treatment.