Abstract
Studies show CYFRA 21-1 fragments of cytokeratin 19 (CK19) to be promising biomarkers for non-small cell lung cancer (NSCLC). Although previous literature identifies specific CYFRA 21-1 antibody binding epitopes, the exact molecular weight of the CK19 fragment being detected by current assays is not well-documented. Serum samples from 58 patients (lung cancer (N = 36), control (N = 22)) were used to measure CYFRA 21-1 across four different quantification assays: enzyme-linked immunosorbent assay (ELISA), chemiluminescent assay (ChLIA), electrochemiluminescence immunoassay (ECLIA), and compensated interferometric reader (CIR). In the cancer group, correlation between ECLIA and ELISA was high (R(Pearson) = 0.948, r(Spearman) = 0.868) while correlation between ECLIA vs ChLIA and ECLIA vs CIR was low (R= 0.005, r = -0.0593), (R = 0.0275, r = 0.167), respectively. In the control group, correlation between ECLIA and ELISA was high (R = 0.861, r = 0.927) while correlation between ECLIA vs ChLIA and ECLIA vs CIR was low (R = 0.0079, r = -0.0593), (R = 0.0244, r = -0.102), respectively. Compared to ECLIA, concordance coefficients (p (c) ) were poor (p (c) < 0.90) across all assays except for cancers group in ELISA (p (c) = 0.913). ECLIA was the only assay to report control ranges above 1 ng/mL CYFRA 21-1 (ECLIA, 1.14-21.59 ng/mL; ELISA, 0.79-24.26 ng/mL; ChLIA, 0.062-0.691 ng/mL; 0.08-7.68 ng/mL). Differing sizes of the protein being measured by each assay may have a role in the discrepancies observed. Given the different CYFRA 21-1 concentration estimates among assays, further characterization of the fragment and its release during epithelial malignancies, such as NSCLC, is imperative to developing effective biomarker assays.