Performance assessment of a fully automated electro-chemiluminescence immunoassay system for serum S100B protein.

The altered expression levels of S100 proteins can lead to four different categories of diseases: diseases of the heart and of the central nervous system, inflammatory disorders and cancer. Various studies have shown the lack of harmonization of the results obtained with different methods, mainly due to different performances and measurements of S100B. The purpose of this work was to compare quantitatively the fully automated Elecsys® immunoassay with the reference immunoenzimatic method CanAg® EIA for serum S100B protein. In the study serum samples were analyzed of 161 patients: 85 females (aged 22-83 years) and 76 males (aged 16-90 years), affected by oncological and non-oncological pathologies. Passing–Bablok regression was used to analyze the comparison between the assays; it showed a strong interassay correlation: r = 0.9350 (95% CI =0.9122 – 0.9520), with an intercept of 0.02063 (95% CI=-0.02850 – 0.01400) and a slope of 1.1125 (95% CI=1.0200 – 1.2417). Elecsys® S100 assay should be preferred to CanAg® S100 for better standardization, good reliability and precision but also with the aim to reduce costs and obtain results in a shorter time.

CLEIA CA125 evidences: good analytical performance avoiding "Hook effect".

Cancer antigen 125 (CA125) is a coelomic epithelium-related antigen carried by a high molecular weight glycoprotein complex. It is commonly used as a tumor marker for ovarian cancer to monitor disease progression and response to therapy and as an early detection for recurrence after treatment. The aim of this study was to test the reliability of two different assay methods, a radioimmunometric assay (RIA) and an automated chemiluminescent enzyme immunoassay (CLEIA) system, by measuring CA125 serum levels using both methods in 357 patients and comparing the results. Patients were recruited from Oncologic Unit A, Policlinico Umberto I, Roma. Eighty-six were healthy donors, while 271 were oncologic patients representing a variety of diagnoses. Within this group, 76 patients were diagnosed with an ovarian related pathology (28 cancerous and 48 benign). The evaluation of CA125 marker blood levels showed a high agreement in healthy donors group (R (2) = 0.9003). Interesting results emerged when sera collected from oncologic patients were assessed: significant differences between the two assays were found in nine samples. When assayed again with RIA after a dilution, new values agreed with undiluted CLEIA values (R (2) = 0.9847). Our data suggest an overall good comparison between the two methods. However, some artifacts were obtained with RIA and indicate an underlying presence of "hook effect". CLEIA automated assay showed a good reliability and should be preferred to one-step radioimmunoassays in order to minimize errors.