Assay-Specific Decision Limits for Two New Automated Parathyroid Hormone and 25-Hydroxyvitamin D Assays

Abstract

The recent development of nonradioactive automated assays for serum parathyroid hormone (PTH) and 25-hydroxyvitamin D (25OHD) has made measurement of these two hormones possible in many laboratories. In this study, we compared two new assays for PTH and 25OHD adapted on an automated analyzer, the LIAISON, with two manual immunoassays used worldwide. We studied 228 osteoporotic patients, 927 healthy individuals, 38 patients with primary hyperparathyroidism, and 167 hemodialyzed patients. Serum PTH was measured with the Allegro and the LIAISON assays, and 25OHD was measured with DiaSorin RIA and the LIAISON assay. Regression analysis was used to calculate decision thresholds for the LIAISON assays that were equivalent to those of the Allegro PTH and DiaSorin 25OHD assays. The 25OHD concentrations obtained with the LIAISON assay and the RIA in osteoporotic patients were well correlated (r = 0.83; P <0.001). Regression and Bland-Altman analyses suggested that the LIAISON 25OHD assay reads lower than the DiaSorin RIA at low concentrations but higher at high concentrations. However, the cutoff (50 nmol/L) used in our laboratories to define vitamin D insufficiency with the DiaSorin RIA is applicable to the LIAISON 25OHD assay. In 927 healthy individuals, the 3rd-97th percentile intervals were 3-80 ng/L and 13-151 nmol/L for the LIAISON PTH and 25OHD concentrations, respectively. However, 506 individuals (54.6%) were vitamin D-insufficient; we therefore considered only the 421 individuals with a LIAISON 25OHD >50 nmol/L as eligible for the reference population for the LIAISON PTH assay. In this group, the 3rd-97th percentile interval for LIAISON PTH was 3-51 ng/L. Considering upper reference limits of 46 and 51 ng/L for the Allegro and LIAISON assays, respectively, the frequency of above-normal PTH concentrations in patients with primary hyperparathyroidism was similar in both assays. Regression analysis between serum PTH measured by the Allegro and LIAISON assays in 167 hemodialyzed patients and the corresponding Bland-Altman analysis of these data suggest that the LIAISON PTH assay tends to read higher than the Allegro assay at low concentrations but lower at high concentrations (>300 ng/L). Because clinical decision limits for both PTH and 25OHD should be assay specific, we propose equivalences between these assays and two manual assays used worldwide. These assay-specific decision limits should help potential users of the LIAISON PTH and 25OHD assays.