Proteomic biomarkers in metastatic UM (explanations below).
Marker
Who/year/method
Results
Function
Number
Materiel
S-100-β
Missotten et al. [56]/2003/ Immuneluminometric assay
No significant correlation with any of the investigated factors, including metastasis = 0,96
Unknown
64 UM patients, 20 metastasis
Serum
S-100-β
Missotten et al. [57]/2007/ Immunoluminometric assay
Not related to any prognostic parameter, significantly elevated in metastatic UM (median 0,7 to 0,23 μg/L, < 0,001), along with LD better than MIA and γ-GT
Calcium-binding protein in the calmodulin/troponin C superfamily
Increased in metastatic UM (median of 6,6 to 26,28 ng/mL, < 0,001) and metastasis development (median of 6,6 to 29,2 ng/mL, < 0,001) Not a prognostic parameter for size and survival (NB metastasis results more scattered than nonmetastatic)
Induces cell-matrix detachment because it binds to fibronectin and laminin
139 UM patients, 8 metastasis (3 developed), 61 followed over time
Increased in metastatic UM (median of 6 versus 13,03 ng/mL, < 0,001) and in their development (5,92 to 12,21 ng/mL, = 0,005) Not a predictive marker for tumor height and treatment (NB metastasis results more scattered than nonmetastatic)
Induces cell-matrix detachment because it binds to fibronectin and laminin
Increased in metastatic UM (median 6,97 to 11,69, < 0,001), ROC analysis with AUC = 0.84 giving a suggested MIA threshold of 8,3 ng/mL (NB: related to ELISA kit) with a sensitivity of 82% and a specificity of 77%, positive predictive value 0,30 and a negative predictive value of 0,97. Metastasis development showed an increase (median of 6,8 to 19,6 ng/mL)
Induces cell-matrix detachment because it binds to fibronectin and laminin
503 UM patients, 54 metastasis, 28 developed metastasis during followup
Serum
OPN
Kadkol et al. [59]/2006/OPN mRNA, IC and serum OPN ELISA
OPN mRNA was increased in highly invasive primary and metastatic UM (6- and 250-fold) Histologically it was related to vasculogenic mimicry, not macrophage presence. Serologically it was significantly increased in metastatic UM versus 10-Y DF (median 17,62 versus 7,15 ng/mL, = 0,0001) and pre- versus postmetastasis (6,19 versus 19,66 ng/mL, = 0,0004) with a cutoff of 10 ng/mL sensitivity and specificity of 83,7% and ROC with AUC being 96% (the probability of correct diagnosis of metastasis)
Component of the noncollagenous bone matrix, actively promotes tumoregenic phenotype and contributes to metastasis
3 UM cell lines, serum from 37 10-Y DF, 15 metastatic (8 with pre- and postmetastatic sampling)
Increased metastatic versus 10-Y DF in OPN (8-9 to 14–18 ng/mL, = 0,0037), S-100β (4–6 to 12–14 = 0,0111), and MIA (0,05–0,1 to 0,325–0,45 ug/L, = 0,0005). Development of metastasis OPN (4-5 to 18–23 ng/mL, = 0,002), S-100β (6-7 to 17–22 ng/mL, = 0,046), and MIA (0,03–0,06 to 0,12–0,18 ug/L, = 0,045) AUC = 91%
To wide interpersonal variability to show any significant difference between the groups. Ratios showed a significant increase from after treatment to 3 years after diagnose of 53%, but this disappeared after excluding two outliners
Barak et al. [61]/2007/Direct injection of UM cells (MUM2B) into mice (ELISA)
TSP was markedly elevated in MUM2B injected mice versus noninjected (84,7 U/L versus 601 μg/L) In humans TSP was significantly elevated in metastatic versus 10-Y DF and controls (139,63 versus 69,29 and 54,23 U/L, < 0,01).