Paper-based ELISA for fast CA 15–3 detection in point-of-care

• Colorimetric test strip on a paper substrate for CA15-3 monitoring at the point-of-care. • Modification of cellulose by reaction with silane or periodate. • Enzymatic detection based on oxidation of tetramethylbenzidine by HRP. • Specific colour coordinates show linear correlation with log concentration of CA 15–3. • Test strips within clinical values of interest and results can be collected using a smartphone. The paper-based Enzyme-Linked Immunosorbent Assay (P-ELISA) is a promising tool for diagnostic applications because the paper matrix is characterised by low cost, short analysis time, portability, and a high surface-to-volume ratio that allows the use of a small sample and reagent volume of a few microliters. In addition, colorimetric assays are suitable for low-resource areas due to their simplicity and naked-eye detectability. Although several works have been reported using paper-based colorimetric sensors for cancer biomarkers, P-ELISA for cancer antigen 15–3 (CA 15–3) has never been reported. Thus, this work reports the development of a rapid, simple and relatively inexpensive paper-based colorimetric assay for the detection of CA 15–3 as cancer biomarker. The assay was developed on a filter paper that was previously washed and chemically oxidized with periodate to generate aldehyde functional groups on the cellulose surface. After covalent binding of the first antibody, detection was performed by a colorimetric reaction based on the oxidation of the substrate 3,3′,5,5′-tetramethylbenzidine (TMB) by a peroxidase enzyme in the presence of H 2 O 2 . The colorimetric measurement was based on digital image acquisition analysed using ImageJ software. The linear range of the assay in buffer ranged 2 to 1100 U/mL. The performance of the assay was also successfully tested in human serum from Cormay®, offering a linear range from 2 to 200 U/mL. Thus, this P-ELISA sensor is suitable for the analysis of serum samples, since the physiological value in cancer patients is 30 U/mL. We believe that this proof-of-concept has the potential to be extended to be applied to other protein biomarkers and to be a suitable tool for cancer screening in developing countries.