Analytical performance of 4 automated assays for measurement of Cystatin C.
Ethics approval was obtained from the Johns Hopkins Internal Review Board. For patient comparison studies, we selected 102 specimens; approximately half of the samples with creatinine values lower than the upper limit of the reference interval [<1.3 mg/dL (0.115 mmol/L) for men, <1.2 mg/dL (0.106 mmol/L) for women, n = 58], and half above the upper limit of the reference range [>1.3 mg/dL (0.115 mmol/L) for men, >1.2 mg/dL (0.106 mmol/L) for women, n = 44]. We prepared serum pools with low, medium, and high concentrations of cystatin C by pooling patient serum with low (<1.2 mg/dL, <0.106 mmol/L), medium (3-7 mg/dL, 0.265-0.619 mmol/L), and high (>10 mg/dL, >0.884 mmol/L) concentrations of creatinine. Specimens were stored refrigerated for no more than 1 week after collection, and then stored at -20 [degrees]C until analysis.
Assays were evaluated following guidelines from the Clinical and Laboratory Standards Institute (CLSI). We evaluated limit of blank (LoB), limit of detection (LoD) and limit of quantification (LoQ) following guideline EP17-A (7) and total imprecision following a modification of guideline EP15-A2 (8). Using single lots of reagents, we analyzed 2 replicates of the 3 serum pools twice per day for 5 days, with at least 2 h separating the daily runs. We evaluated linearity following guideline EP6-A (9). We prepared serial 10% dilutions (e.g., 90:10, 80:20) of the high concentration serum pool using cystatin C-free serum (HyTest Ltd.) as diluent. We measured these samples in triplicate and plotted the measured value against the expected value (mean of 100% serum pool x dilution factor) using EP Evaluator (David G. Rhodes Associates, Inc.). We calculated percent recovery as the measured concentration divided by the expected concentration multiplied by 100% and compared results from each assay on 102 patient samples following guideline EP9-A2 (10). For data analysis, we used EP Evaluator release 8 or Microsoft Excel (LoD only).
The LoD and LoQ values demonstrated the most variability among assays (Table 2). Although the actual values for LoD and LoQ showed differences of up to 10-fold, all ewre below the quoted reference intervals for patient specimens (0.53-0.95 mg/L by Siemens, 0.47-1.09 mg/L by Roche, 0.61-1.17 mg/L by Genzyme, and 0.52-0.97 mg/L by Tosoh). Therefore we considered all assays to have sufficient analytic sensitivity for clinical use. Compared with the manufacturer's specifications, the total CVs observed by us were generally higher, the largest discrepancy being observed with the Genzyme assay (8.9% vs 2.3% for the low-concentration cystatin C pools, 3.7% vs 1.3% for the medium-concentration pools, and 9% vs 2.4% for the high-concentration pools) (Table 1). Because the serum pools used in this study were different from the pools used by the manufacturer, we did not calculate statistical significance of the observed differences.
In the linearity study (Table 1), the Roche assay tended to overrecover, especially at low cystatin C concentrations. There was a gradual increase in recovery, from 100% on the highest sample (assigned value 5.869 mg/L, mean of triplicate measurement) to 142% on the lowest sample (assigned value 0.5869 mg/L, as 10% of the high-concentration pool). Regression analysis of the recovery data suggested both proportional and constant bias by the Roche assay, with a slope of 0.945 and an intercept of 0.55. The other 3 assays demonstrated a mean recoveryclose to 100% (Siemens, 100.05%, range 96.2%-103.4%; Genzyme, 98.49%, range 97.4%-101.7%; Tosoh, 100.67%, range 98.1%-103.2%) with slopes close to 1 and intercepts close to 0.
For method comparison, we compared the results on 102 patient samples from each method with the Siemens assay, since this assay is the most widely used based on 2008 College of American Pathologists survey participants (40 of 48 laboratories submitted results using the Siemens assay) and was also the first commercial cystatin C assay to obtain FDA approval. The equations (Fig. 1) obtained using Deming regression were y = 1.184x + 0.089, [S.sub.y|x] = 0.246, R = 0.987 (Genzyme vs Siemens); y = 0.937x + 0.231, [S.sub.y|x] = 0.134, R = 0.995 (Roche vs Siemens); and y = 1.010x + 0.216, [S.sub.y|x] = 0.135, R = 0.997 (Tosoh vs Siemens). The Genzyme assay appeared to give higher results than the Siemens assay, which is consistent with a higher reference interval specified by the manufacturer (0.61-1.17 mg/L by Genzyme, 0.53-0.95 mg/L by Siemens) (Table 1). Based on a previous publication evaluating the Siemens (x) and the Genzyme (y) cystatin C assays, a linear regression of y = 0.975x + 0.197, R = 0.995 was reported ([S.sub.y|x] was not provided) (11). In this case, a large positive intercept contributed to the higher results by the Genzyme assay. We are unclear about the cause of this difference on regression analysis, but 1 possibility may be that 1 or both manufacturers has changed calibration or reagent formulation in the interval between evaluations.
Although all assays were acceptable for clinical use according to commonly accepted criteria, their diagnostic performances were not optimal. Concerns for different assays include imprecision observed for the Genzyme assay, overrecovery at low cystatin C concentrations by the Roche assay, and assay specific differences in the patient comparison studies. The latter situation means that assay-specific cystatin C-based GFR-prediction equations are required. This problem might be solved if the assays' calibration can be standardized using the international cystatin C calibrator now being developed (12). Factors such as available instrumentation and cost of reagents mayalso contribute to the suitability of each assay for a given laboratory. The Siemens N Latex cystatin C assay is a nephelometric method, and the Tosoh ST AIA-PACK cystatin C assay is a 2-site immunometric assay; both must be run on specific immunoassay instruments. The Roche Tina-quant cystatin C and the Genzyme cystatin C assays are particle-based immunoassays that may be run on automated chemistry analyzers and therefore are less restricted with respect to instrumentation.
[FIGURE 1 OMITTED]
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures of Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: L.J. Sokoll, National Academy of Clinical Biochemistry.
Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: None declared.
Research Funding: None declared.
Expert Testimony: None declared.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: The authors thank all 4 manufacturers (Siemens, Roche, Genzyme, and Tosoh) for providing reagents and test kits for this study.
(1.) Stevens LA, Coresh J, Greene T, Levey AS. Assessing kidney function: measured and estimated glomerular filtration rate. N Engl J Med 2006;354: 2473-83.
(2.) Preiss DJ, Godber IM, Lamb EJ, Dalton RN, Gunn IR. The influence of a cooked-meat meal on estimated glomerular filtration rate. Ann Clin Biochem 2007;44:35-42.
(3.) Laterza OF, Price CP, Scott MG. Cystatin C: an improved estimator of glomerular filtration rate? Clin Chem 2002;48:699-707.
(4.) Kyhse-Andersen J, Schmidt C, Nordin G, Andersson B, Nilsson-Ehle P, Lindstrom V, et al. Serum cystatin C, determined by a rapid, automated particle-enhanced turbidimetric method, is a better marker than serum creatinine for glomerular filtration rate. Clin Chem 1994;40:1921-6.
(5.) Shlipak MG, Praught ML, Sarnak MJ. Update on cystatin C: new insights into the importance of mild kidney dysfunction. Curr Opin Nephrol Hy pertens 2006;15:270-5.
(6.) Westhuyzen J. Cystatin C: a promising marker and predictor of impaired renal function. Ann Clin Lab Sci 2006;36:387-94.
(7.) Protocols for determination of limit of detection and limit of quantitation; approved guideline-second edition. 2004. NCCLS document EP17-A (ISBN 1-56238-551-8).
(8.) User verification of performance for precision and trueness; approved guideline, EP15-A2. 2006. CLSI document EP15-A2 (ISBN 1-56238-574-7).
(9.) Evaluation of the linearity of quantitative measurement procedures: a statistical approach; approved guideline. 2003. NCCLS document EP6-A (ISBN 1-56238-498-8).
(10.) Method comparison and bias estimation using patient samples; approved guideline--second edition. 2002. NCCLS document EP9-A2 (ISBN 1-56238-472-4).
(11.) Al-Turkmani MR, Law T, Kellogg MD. Performance evaluation of a particle-enhanced turbidimetric cystatin C assay on the Hitachi 917 analyzer. Clin Chim Acta 2008;398:75-7.
(12.) Blirup-Jensen S, Grubb A, Lindstrom V, Schmidt C, Althaus H. Standardization of Cystatin C: development of primary and secondary reference preparations. Scand J Clin Lab Invest Suppl 2008;241:67-70.
Jinong Li,  Willard Dunn,  Autumn Breaud,  Debra Elliott,  Lori J. Sokoll,  and William Clarke  *
 Division of Clinical Chemistry, Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD 21287; * address correspondence to this author at: Division of Clinical Chemistry, Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD 21287. E-mail firstname.lastname@example.org.
 Nonstandard abbreviations: GFR, glomerular filtration rate; FDA, US Food and Drug Administration; CLSI, Clinical and Laboratory Standards Institute; LoB, limit of blank; LoD, limit of detection; LoQ, limit of quantification.
Previously published online at DOI: 10.1373/clinchem.2009.141531
Table 1. Cystatin C assays used in the comparison. Assay Siemens Roche (BNII) (Cobas c501) Test principle Nephelometric Particle-enhanced immunoturbidimetric Sample volume, [micro]L. 30 2 Reaction time, min 6 10 Quoted normal range, mg/L 0.53-0.95 0.47-1.09 Assay Genzyme Tosoh (a) (Cobas c501) (AIA-600II) Test principle Particle Two-site immunoturbidimetric immunometric Sample volume, [micro]L. 2 10 Reaction time, min 10 10 Quoted normal range, mg/L 0.61-1.17 0.52-0.97 (a) Investigational use only. Table 2. Comparison of performance characteristics. (a) This study Product insert/ manufacturer claim LoB, mg/L 0.0004 0.005 LoD, mg/L 0.004 0.05 LoQ, mg/L <0.05 NA Total imprecision Serum pool 1 Mean, mg/L 1 0.8 Within-run CV, % 2.5 2.5 Total CV, % 5 2.4 Serum pool 2 Mean, mg/L 3.9 2.3 Within-run CV, % 2.5 2.6 Total CV, % 3 4.3 Serum pool 3 Mean, mg/L 5.6 7.1 Within-run CV, % 2.8 1.7 Total CV, % 3.2 2.9 Linearity/recovery Measured concentration, 0.65-6.8 NA 10%-100% Slope 1.024 NA Intercept -0.052 NA SEs of estimates 0.07 NA Mean recovery, % (range) 100.05 (96.2-103.4) NA This study Product insert/ manufacturer claim LoB, mg/L 0.05 [less than or equal to] 0.3 LoD, mg/L 0.07 [less than or equal to] 0.4 LoQ, mg/L <0.39 (b) NA Total imprecision Serum pool 1 Mean, mg/L 1.2 0.75 Within-run CV, % 6.4 1.71 Total CV, % 6.6 2.83c Serum pool 2 Mean, mg/L 3.8 NA Within-run CV, % 1.1 NA Total CV, % 2.4 NA Serum pool 3 Mean, mg/L 4.9 5.14 Within-run CV, % 2.8 0.67 Total CV, % 3.3 2.05 (c) Linearity/recovery Measured concentration, 0.83-5.9 NA 10%-100% Slope 0.945 NA Intercept 0.55 NA SEs of estimates 0.163 NA Mean recovery, % (range) 118.65 (100.0-141.9) NA This study Product insert/ manufacturer claim LoB, mg/L 0.07 0.02 LoD, mg/L 0.08 0.05 LoQ, mg/L <0.2 0.2 (CV <10%) Total imprecision Serum pool 1 Mean, mg/L 1.3 0.72 Within-run CV, % 8.9 1.7 Total CV, % 8.9 2.3 Serum pool 2 Mean, mg/L 4.6 2.79 Within-run CV, % 3.5 0.7 Total CV, % 3.7 1.3 Serum pool 3 Mean, mg/L 6.2 5.06 Within-run CV, % 9 1.5 Total CV, % 9 2.4 Linearity/recovery Measured concentration, 0.76-7.8 NA 10%-100% Slope 0.978 NA Intercept 0.02 NA SEs of estimates 0.104 NA Mean recovery, % (range) 98.49 (97.4-101.7) NA This study Product insert/ manufacturer claim LoB, mg/L 0.0047 NA LoD, mg/L 0.005 NA LoQ, mg/L <0.02 NA Total imprecision Serum pool 1 Mean, mg/L 1.2 0.7 Within-run CV, % 2 3.3 Total CV, % 5.6 3.7 Serum pool 2 Mean, mg/L 4.1 2.15 Within-run CV, % 2.7 2 Total CV, % 5 2.6 Serum pool 3 Mean, mg/L 2.8 4.69 Within-run CV, % 2.8 3.7 Total CV, % 4.6 3.4 Linearity/recovery Measured concentration, 0.74-7.2 NA 10%-100% Slope 0.994 NA Intercept 0.028 NA SEs of estimates 0.06 NA Mean recovery, % (range) 100.67 (98.1-103.2) NA (a) Serum pools used to evaluate imprecision in this study are different from pools used by each manufacturer. NA, information not available. (b) Instrument minimal reading. (c) Between-run CV.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Brief Communications|
|Author:||Li, Jinong; Dunn, Willard; Breaud, Autumn; Elliott, Debra; Sokoll, Lori J.; Clarke, William|
|Date:||Aug 1, 2010|
|Previous Article:||Estimation of the minimal preanalytical uncertainty for 15 clinical chemistry serum analytes.|
|Next Article:||Rapid detection of reassortment of pandemic H1N1/2009 influenza virus.|