Bedrich Friedecky

Comparison of different immunoassays for CA 19-9.

Abstract We compared six routinely employed immunoassay kits: Architect i2000 and AxSYM, Abbott Laboratories; Elecsys 2010, Roche Diagnostics; ELSA, CIS-BioInternational; Immulite 1, Diagnostic Products Corporation; and IRMA-mat, Byk-Sangtec Diagnostica. Using all analytical systems, we measured identical groups of clinical samples completed with selected control samples. The repeatability of measurements (coefficient of variation) ranged from 2.1% (Elecsys 2010) to 6.7% (ELSA). The parameters of Passing-Bablok regression show significant systematic differences among analytical systems. Data from a Bland-Altman diagram suggest that these differences project onto other, still more significant individual differences among individual samples. Though the cut-off values differ between various systems, no similar clinical efficacy appears to be attained. The behavior of individual systems is quite different for identical control materials and does not necessarily duplicate the calibration for biological samples. The results of determining CA 19–9 cannot be extrapolated from one analytical technique to another, even in cases where the same monoclonal antibody is used. Standardization of CA 19–9 measurement systems is necessary to allow use of the results for the purposes of evidence-based medicine.

Why do different EQA schemes have apparently different limits of acceptability

It would be logical to assume that for various EQA systems in this situation, there are comparable evaluation criteria necessary to reach equivalent conditions for accreditation organizations and for patient safety in various geographical areas. Above all, this applies to the value of the limits of acceptability (limit of acceptability means maximal allowed deviation of the participants individual result from the target value), where achieving cross compatibility should not present any fundamental problem. However, in reality, the value of the acceptance limits used in various EQA programmes differ widely. The analytes which were evaluated, as well as the acceptance limits of several selected EQA programmes, are presented in Table 1. From the data in Table 1, it is evident that the acceptance limit values differ significantly. The results from 380 to 402 participants in one control cycle SEKK (EQA system Pardubice, Czech Republic, survey AKS4/09) were evaluated. The assigned reference method values (RMV) obtained using reference method procedures were the criteria for success of the measurement results. Cholesterol determination was an exception, where for all laboratories, trimmed mean (ALTM) results were used. What are the reasons for the dramatic differences among the acceptance limits of various EQA programmes that are

Interference of IgM-λ paraprotein with biuret-type assay for total serum protein quantification

Milos Tichy*, Bedrich Friedecky, Marek Budina, Vladimir Maisnar, Tomas Buchler, Magdalena Holeckova, Dagmar Gotzmannova and Vladimir Palicka 1 Institute of Clinical Biochemistry and Diagnostics, Charles University Hospital, Hradec Králové, Czech Republic 2 SEKK, Pardubice, Czech Republic 3 2nd Department of Internal Medicine, Division of Clinical Hematology, Charles University Hospital, Hradec Králové, Czech Republic 4 Department of Oncology, Thomayer University Hospital and 1st Faculty of Medicine, Charles University, Prague, Czech Republic 5 Laborex, Ostrava, Czech Republic

Quality of serum creatinine measurement in light of EQA programs.

Abstract Background: The aim of our study was to identify the role of External Quality Assessment (EQA) programs in improving the quality of serum creatinine measurement and glomerular filtration rate (GFR) estimation. Comparison of results achieved during EQA with National Kidney Disease Education Program and College of American Pathologists guidelines identified an urgent need for an improvement in measurement quality. We compared actual results for serum creatinine measurement within the Czech Republic EQA with the requirements of EC Directive 98/79. Methods: We used the results for 2005–2006 EQA programs. There were seven surveys involved with two samples each, and a 2006 questionnaire on the post-analytical phase survey. Results: Bias depended strongly on the creatinine concentration. However, this dependence varied for different in vitro diagnostic manufacturers, although they should all follow the same directive. We chose biological variation as the significance rate for bias and a resulting overall error of 6.9%. The proportion of results with total error <6.9% ranged from 11% to 80%. The total error for a reference sample of 94.8 μmol/L also showed significant dependence on the working calibrator used and ranged from 1% to 17%. Conclusions: The main role of EQA programs in improving the quality of creatinine measurement results and GFR calculation should be in monitoring the quality of IVD products, enabling users to adapt their use of these products accordingly. EQA programs can also educate on performing GFR estimation in a unified way. Highly commutable control materials with certified creatinine values or, alternatively, lyophilized materials with sufficient commutability proved by comparison with native frozen human sera, should become an important EQA tool. Clin Chem Lab Med 2007;45:685–8.

The problems of proteinuria measurement in urine with presence of Bence Jones protein

OBJECTIVES Protein concentration measurement in the urine can be problematic in the presence of Bence Jones protein. We have carried out an external quality control assessment with the participation of 79 clinical biochemistry laboratories from the Czech Republic and Slovakia. DESIGN AND METHODS The laboratories received a reference urine sample obtained from a patient with multiple myeloma and lambda free light chain proteinuria and were asked to type the paraprotein using immunofixation and to measure total urinary protein using their established method, most commonly turbidimetry, pyrogallol red assay, and biuret assay. RESULTS There was a very wide inter-laboratory variability in the protein concentration readouts with up to three-fold difference in some cases. High-resolution two-dimensional electrophoresis and linear mass spectrometry showed that a high proportion of the urinary paraprotein was composed of lambda light chain fragments with molecular weight of 12kDa. CONCLUSIONS Our results highlight the challenges of reliable and reproducible measurement of urinary protein concentration in the presence of Bence Jones protein.

Estimation of trueness of measurement results obtained in external quality assessment.

Abstract Background: We demonstrated that lyophilised EQA/PT control materials, under certain circumstances, provide equal information about bias values as pools of native patient sera, and that in some cases, long-term reliable work with such materials is possible. Methods: Bias values, estimated from results of routine surveys of EQA SEKK (Czech Republic) programmes for eight basic blood serum analytes using lyophilised control materials, were compared with bias values reached in the CAP (USA) programme where pools of native and lyophilised patient sera were used. Results: Results for the components Na, K, Mg, Cl, P, urea, glucose, and uric acid were assessed. No significant differences were found between the bias values estimated by CAP using native sera, and those estimated by SEKK using lyophilised sera. Conclusions: It can be concluded that well-prepared lyophilised control materials may show the required commutability level. If so, they constitute a practical and cost-efficient alternative to native or fresh frozen sera when assessing bias. Clin Chem Lab Med 2009;47:489–90.

Comparability of Eight Immunoassay Procedures for the Determination of CA 15-3 and Related Markers

Abstract MUC1 mucins are tumour markers that are frequently indicated and examined, particularly as part of the treatment of breast cancer. Relatively large differences were observed in external quality assessment (EQA) between the results that were obtained by different immunoassay technologies. Thus, we compared eight routinely employed immunoassay sets for the determination of MUC1 mucins in the serum: six closed automated systems (AxSYM, Centaur, ECi Vitros, Elecsys 2010, Immulite 2000 and Kryptor), and two IRMA kits (ELSA CIS and IRMA-mat Byk-Sangtec). Using all analytical systems, we measured identical groups of clinical samples complete with selected calibrator and control samples. The repeatability of measurements (presented as coefficients of variation) ranged from 0.7% (Kryptor) to 6.9% (Immulite 2000). Even though the cut-off values differ among various systems, no similar clinical efficacy appears to be attained. In the region of cut-off values, the highest specificity that was set as a standard was found for the AxSYM analyser, while the sensitivity was highest for the Elecsys 2010. Data from Bland-Altman differential plots suggest the presence of significant individual differences among individual samples, mainly in the region of high concentrations of MUC1 mucins. The parameters of Passing-Bablok regression show significant systematic differences between some of the analytical systems as well as an increase of the differences with increasing MUC1 mucin concentrations. The effect of the combination of antibodies used on the extent of differences among results obtained with individual systems is more obvious than the effect of the matrix of analysed materials.

Selected markers of bone biochemistry

Publisher Summary Monitoring bone turnover and bone remodeling using biochemical methods is possible with the help of two types of parameters. This chapter reviews the selected markers used in bone biochemistry. Mature cortical and trabecular bone consistently undergo bone remodeling. Bone remodeling is in progress with variable intensity throughout life, and the connection of resorption to subsequent bone formation is essential. Some of the hormones and local agents used as markers are cytokines, dehydroepiandrosterone (DHEA), hormones of the thyroid gland, anabolic steroids, prolactin, and parathormone–related peptide (PTHrP). Systemic markers of bone remodeling are divided into: markers of bone formation produced by osteoblasts, markers of bone resorption produced by osteoclasts, and products of collagen type I degradation. The markers of bone remodeling change their concentration and activity in various diseases of the bone tissue, which are tabulated. The chapter also discusses the general potential, variability of the results, serum markers produced by osteoblasts and osteoclasts, and clinical usefulness of the markers of bone remodeling. Follow–up of the response to treatment is the principal current application of the measurement of the markers of bone remodeling.

Serum cholesterol: required and actual bias values.

There is insufficient experimental data about bias for serum cholesterol measurements. In 1986, ‘‘standard’’ (lyophilized) external quality assessment/proficiency testing (EQA/PT) control materials and frozen mixtures of native patient sera were sent simultaneously to routine clinical US laboratories. The differences between results depended on the reagents and equipment used, and ranged from –8.8% to q4.4%. In addition, some analytical systems did not find any difference at all between the ‘‘standard’’ and native control materials. The authors considered the differences that were observed to be caused by the influence of the matrix of the ‘‘standard’’ EQA/PT control materials (1). The maximum required bias (b) for serum cholesterol measurements, set according to basic medical recommendations, is b-3.0%. The value shown at the Westgard web site, derived from biological variability data, is a bit higher: b-4.0% (2). Other experiments to monitor bias were performed using native patient serum mixtures to exclude any potential matrix influences. To determine bias, the experiment organized by the Institute for Reference Materials and Methods (IRMM) Geel in 2002–2003 International Measurement Evaluation Programme (IMEP-17) used control material based on native human sera with a reference value of 5.111 mmol • L, obtained by means of liquid chromatography mass spectrometry (LC-MS); gas chromatography mass spectrometry (GCMS). Of the 991 participating laboratories, the required bias of b-3% was achieved by only 50% of the IMEP-17 participants. The results are available at the web site (3). Based on a 2003 study of native samples obtained from blood donors, an average bias value bs5.1% was found for the routine cholesterol methods, which is also markedly higher than what is required (4). The Scandinavian Nordic Reference Interval Project (NORIP) 2000, performed in 104 routine clinical laborato-