Evaluation of two automated immunoassays for measurement of free deoxypyridinoline in urine using analytical goals derived from biological variation.
Free deoxypyridinoline (fDPD) is increasingly used as a specific marker of bone resorption (1). To date, the concentration of fDPD in urine has been measured using cumbersome HPLC or manual microtiter-based ELISA procedures, which require ~4 h to perform (2). Recently, two rapid, fully automated chemiluminescent immunoassays were developed: by Chiron Diagnostics for the AC5:180[R] analyzer and by DPC[R] for the Immulite analyzer. Both immunoassays use a competitive format involving the same monoclonal anti-fDPD antibody from Metra Biosystems (3, 4). If sufficiently reliable, these assays can accommodate increasing testing demands with both controlled operating costs and dramatically reduced turnaround time. This study assessed their analytical performance for the routine measurement of fDPD in our clinical laboratory. In particular, as suggested previously (5), goals for precision and accuracy were based on the biological variation of fDPD excretion in urine from healthy premenopausal women, and the results obtained during the evaluation were compared with these to assess acceptability (6).
All measurements on the two instruments were performed according to the recommendations of the manufacturers by the same trained technician. Linearity in the working ranges of the tests (AC5:180, 2-350 nmol/L; Immulite, 7-300 nmol/L) was good (r = 0.999 for the AC5:180; r = 0.998 for the Immulite). The data for the imprecision study, which used two pooled human urines, are summarized in Table 1. Eighty-one urine samples (first morning void) with fDPD concentrations ranging from 13 to 184 nmol/L were assayed, using the Pyrilinks[degrees][degrees]-D (Metra Biosystems) method as the reference (x) and the two automated procedures (J) in a correlation study. The following results were obtained: AC5:180 = 0.96 ([+ or -]0.03)x - 0.8 ([+ or -]3.0); [S.sub.y|x] = 12.9 nmol/L; r = 0.9552; and Immulite = 0.99 ([+ or -] 0.03)x + 9.8 ([+ or -]2.8); [S.sub.y|x] = 12.1 nmol/L; r = 0.9629.
When biology is used to set analytical goals, desirable imprecision (CV) is less than or equal to one-half of the average within-subject biological variation (i.e., for urinary fDPD, total CV [less than or equal to] 6.7%), and desirable inaccuracy (bias) is less than or equal to one-quarter of the group (within-subject plus between-subject) biological variation (i.e., for urinary fDPD, average bias [less than or equal to] 5.5%) (6). From our experimental results, we conclude that the ACS: 180 assay is probably accurate but is too imprecise for the between-day evaluation and that the Immulite assay shows good precision but also has a significant, constant positive bias. The final considerations depend on medical needs (7). Because the bone markers are useful adjuncts in monitoring patients and not in screening for bone disorders, low imprecision (at least as good as the goal) is required, whereas some degree of inaccuracy is probably less important (8).
We acknowledge the expert technical assistance of Cristina Serena. We also thank Chiron Diagnostics (Cassina de' Pecchi, Milano, Italy) and Medical Systems (Genova, Italy) for the generous loan of instruments and reagents to carry out the study.
(1.) Knott L, Bailey AJ. Collagen cross-links in mineralizing tissues: a review of their chemistry, function, and clinical relevance [Review]. Bone 1998;22:181-7.
(2.) James IT, Walne AJ, Perrett D. The measurement of pyridinium crosslinks: a methodological overview [Review]. Ann Clin Biochem 1996; 33:397-420.
(3.) Chen SY, Sickel M, Kline S, Corkery J, Berard-Aubin J, Hesley R. An automated chemiluminescent immunoassay for deoxypyridinoline: a specific urinary marker for bone resorption [Abstract]. Clin Chem 1995;41:S40.
(4.) Barkley J, Lei JD, El Shami AS. Immulite Pyrilinks-D: a random-access immunoassay for deoxypyridinoline in the monitoring of bone resorption [Abstract]. Clin Chem 1997;43: S276.
(5.) Fraser CG. Data on biological variation: essential prerequisites for introducing new procedures? [Editorial]. Clin Chem 1994;40: 1671-3.
(6.) Panteghini M, Pagani F. Biological variation in urinary excretion of deoxypyridinoline [Abstract]. Clin Chem 1995;41:S200.
(7.) Fraser CG, Hyltoft Petersen P. Desirable standards for laboratory tests if they are to fulfill medical needs. Clin Chem 1993;39:1447-55.
(8.) Beck Jensen JE, Kollerup G, Sorensen HA, Pors Nielsen S, Sorensen OH. A single measurement of biochemical markers of bone turnover has limited utility in the individual person. Scand J Clin Lab Investig 1997;57:351-60.
Mauro Panteghini * Franca Pagani
1st Laboratorio Analisi
25123 Brescia, Italy
* Author for correspondence. Fax 39 030 3995369; e-mail email@example.com.
Table 1. Imprecision of the evaluated immunoassays. Within-run CV Between-day Mean fDPD (n = 20) CV (n = 10) ACS:180 82 nmol/L 4.1% 9.0% 304 nmol/L 2.6% 4.2% Immulite 78 nmol/L 5.6% 3.9% 289 nmol/L 4.7% 3.2%
|Printer friendly Cite/link Email Feedback|
|Author:||Panteghini, Mauro; Pagani, Franca|
|Article Type:||Letter to the editor|
|Date:||Dec 1, 1998|
|Previous Article:||Anti-thyrotropin antibody interference in thyrotropin assays.|
|Next Article:||Clinical Investigation and Statistics in Laboratory Medicine.|