Influence of PTH assay methodology on differential diagnosis of renal bone disease.

Abstract

Determination of plasma parathyroid hormone (PTH) is routinely performed to diagnose and monitor renal bone disease. Recently, a new PTH assay ('whole PTH') using an antibody directed specifically against PTH(1-4) has been introduced. It was the aim of the current study to evaluate whole PTH and parameters derived from whole PTH in renal bone disease. The following measurements were carried out in blood samples from 141 unselected haemodialysis patients: three intact PTH assays (Nichols, Roche Elecsys), Scantibodies total); whole PTH (Scantibodies); bone-specific alkaline phosphatase (bAP); tartrate-resistant acid-phosphatase 5b (TRAP 5b); osteocalcin, 25-hydroxyvitamin D. Parameters derived from whole PTH were: (i) non-PTH(1-84), difference between intact PTH (Scantibodies assay) and whole PTH; (ii) whole PTH/non-PTH(1-84) ratio. The values generated by the intact PTH assays were comparable. The mean whole PTH concentration was lower than mean intact PTH concentrations (16.9+/-18.1 vs 26.4+/-30.5 pmol/l, Nichols, P<0.05). The correlation coefficients between all four PTH assays were comparable and were very high (r>0.96, ns). The rank order of values generated by the whole PTH assay was statistically not significantly different from the rank order generated by the Nichols intact PTH assay. The median non-PTH(1-84) concentration was 5.2 pmol/l (range 0-49.4). All PTH assays correlated highly significantly with non-PTH(1-84) (correlation coefficients 0.83-0.92). Corrected serum calcium was also associated with non-PTH(1-84) but the correlation was weaker (r=0.28). Regression analysis indicated that the non-PTH(1-84) concentration could be predicted by 76.6-84.6% by the prevailing intact PTH concentrations. Other parameters contributed only marginally to prediction of non-PTH(1-84). In the entire patient group, there was no statistically significant correlation between the whole PTH/non-PTH(1-84) ratio and any of the PTH assays or biochemical bone markers. Eight of 141 patients had a whole PTH/non-PTH(1-84) ratio <1. TRAP 5b, bAP and osteocalcin had high correlations with intact PTH assays and the whole PTH assay (correlation coefficients 0.51-0.56, no significant difference). None of the PTH assays was superior to any other PTH assay in predicting serum concentrations of the bone markers. Therapy with active vitamin D metabolites (n=70) did not alter the results of our analyses. With respect to information about bone turnover we were not able to find differences between whole PTH and intact PTH assays. Our data also suggest that whole PTH and intact PTH assays give similar information. (i) The correlation between all PTH assays was very high. (ii) The rank order between whole PTH and Nichols intact PTH assays was comparable. (iii) The association between intact PTH assays and non-PTH(1-84) was very high. Albeit non-PTH(1-84) was mostly determined by the prevailing intact PTH concentration, diagnostic information on parathyroid activity provided by whole PTH or intact PTH, respectively, may differ in individual patients. How often this would happen cannot be answered with the currently available data. Unequivocal structural identification of the non-PTH(1-84) fraction would facilitate the answer to that question. The use of the whole PTH/non-PTH(1-84) ratio as a biochemical bone marker in renal bone disease requires further investigation.