A multiplex immunoassay of serum biomarkers for the detection of uveal melanoma.
Abstract
Approximately 50% of uveal melanoma (UM) patients develop metastases preferentially in the liver leading to death within 15 months. Currently, there is no effective treatment for metastatic UM, in part because the tumor burden is typically high when liver metastases are detected through abnormal liver function tests (LFTs) or imaging studies. The use of LFTs results followed by diagnostic tests has high specificity and predictive values but low sensitivity, and better tests are needed for early diagnosis of the primary tumor as well as its metastatic spread. To evaluate serum biomarkers for the early detection of UM, multiplex immunoassays were developed.
Magnetic bead-based multiplex immunoassays were developed for the selected serum biomarkers using a Bio-Plex 200 system. The dynamic ranges, lower limits of detection and quantification, cross-reactivity, and intra- and inter-assay precision were assessed. All proteins were analyzed in sera of 48 patients diagnosed with UM (14 metastatic, 9 disease-free (DF) ≥ 5 years, 25 unknown) and 36 healthy controls. The performance of the biomarkers was evaluated individually and in combination for their ability to detect UM.
A 7-plex immunoassay of OPN, MIA, CEACAM-1, MIC-1, SPON1, POSTN and HSP27 was developed with negligible cross-reactivity, recovery of 84-105%, and intra-assay and inter-assay precision of 2.3-7.5% or 2.8-20.8%, respectively. Logistic regression identified a two-marker panel of HSP27 and OPN that significantly improved the individual biomarker performance in discriminating UM from healthy controls. The improved discrimination of a two-marker panel of MIA and MIC-1 was also observed between metastatic UM and DF, however not statistically significant due to the small sample size.
The multiplex immunoassay provides sufficient analytical performance to evaluate serum biomarkers that complement each other in detection of UM, and warrants further validation with a larger number of patient samples.
EDRN PI Authors
Medline Author List
- Chan DW
- Merbs SL
- Sokoll LJ
- Song J
- Zhang Z