Analytical validation of the LiCA (R) high-sensitivity human thyroid stimulating hormone assay

Background: We aimed to assess the analytical characteristics of a new high-sensitivity human thyroid stimulating hormone (hTSH) assay on a light-initiated chemiluminescent immunoassay system (LiCA Smart) and examine the utility of this assay in the context of profoundly low TSH levels (<0.01 mIU/L). Methods: Analytical validations included precision, linearity, reportable range, analytical sensitivity, interference, reagent lot-to-lot and between-instrument variability, and method comparisons. Additionally, a cross-sectional study was performed to evaluate the assay for the detection of profoundly low TSH levels in comparison to those of two other ultrasensitive hTSH assays. Results: Within-run and within-lab imprecisions (%CV) were < 5% at all concentrations studied. A satisfactory linearity (R = 0.998, change in recovery < 5%) was verified over the entire measuring range. Method comparisons demonstrated a reasonable agreement (R > 0.99, median bias < 5%) between LiCA and Cobas, ADVIA, UniCel or Architect. The limit of quantitation was 0.0019 mIU/L. Comparative measurements of 236 patient samples with profoundly low TSH levels (<0.01 mIU/L) by LiCA, Cobas, and Architect revealed that the detection rate observed with LiCA (67.8%) was significantly higher than that with Cobas (28.0%) or Architect (21.7%). In a further comparative follow-up of patients with overt hyperthyroidism who were receiving treatment, an earlier recovery response of TSH was observed in LiCA. Conclusions: The LiCA Smart hTSH is a precise and highly sensitive fourth-generation assay. The assay demonstrated superior detection sensitivity for profoundly low TSH levels and was acceptable for clinical use.