Clinical and analytical performance of automated immunochemiluminometric assays for determination of cardiac markers in serum

Three new immunochemiluminometric assays for quantitation of cardiac markers, i.e., creatine kinase isoenzyme MB (CK-MB), myoglobin and cardiac troponin I (cTnI), were evaluated with the Sanofi Access analyser. The complete profile requires 20 min to perform, the method being suitable in true stat situations. In patients with early myocardial infarction (median time of sample collection: 210 min from onset, range 30-450; n = 44), the diagnostic sensitivity of Access cTnI was 66%, compared with 80% for myoglobin, and 43% for CK-MB. For comparison, cTnI, with an automated immunofluorimetric assay was also measured (sensitivity, 45%; p < 0.05 vs. Access cTnI). Our data confirmed myoglobin as the first biochemical marker to appear elevated after infarction. However, cTnI may be a more sensitive marker for early detection of cardiac damage than initially thought, when determined by an ultrasensitive method such as an immunochemiluminometric assay.

Semi-automated analysis of NMDA-mediated toxicity in digitised colour images from rat hippocampus

The evaluation of neuronal cell survival after, for example, mechanical, hypoxic or drug-mediated injury requires the analysis of a high number of histological specimens. Since this is a time-consuming occupation, we have developed a semi-automated analysis routine for the determination of the distribution of live and dead cells. After digitalization of the histological preparations, 8-bit colour bitmaps were assessed using a compiled image-analysis programme of the software package Khoros. In the current study a detailed example of the application of this image-processing approach is described for the investigation of the cell survival after intraventricular application of N-methyl-D-aspartate (NMDA). The samples were prepared as fuchsin acid/toluidine blue stained hippocampal thin slices. The calculated areas of the live and dead cells were highly correlated with manual counts of live and dead cells in the 100 samples examined in this study. Twenty-four hours following NMDA-treatment animals (n = 5) were found to have significantly fewer live and more dead hippocampal cells than the saline-treated animals (n = 5), using either automated or manual examination techniques. The automated technique also revealed that NMDA treatment resulted in a reduction in the density of live cell distribution.

Chromogenic assays for equine coagulation factors VII, VIII:C, IX, and X, and C1-esterase inhibitor

OBJECTIVES

To adapt manual human chromogenic assays for coagulation factors VII (F.VII), VIII:coagulant (F.VIII:C), IX (F.IX), and X (F.X), and C1-esterase inhibitor (C1-INH) for use with an automated analyzer, and to measure the activity of these proteins in horses.

ANIMALS

10 healthy horses were used to determine ranges for the assays. Pooled plasma for standards was collected from an additional 20 healthy horses.

PROCEDURE

A computer-assisted analyzer was programmed from the manual method for commercially available human F.VII, F.VIII:C, F.IX, F.X, and C1-INH chromogenic assay kits. Standards were prepared from pooled citrated equine plasma for the F.VII, F.VIII:C, and F.X assays, and from commercial pooled citrated human plasma for F.IX and C1-INH assays.

RESULTS

Mean +/- SD activities in citrated equine plasma from 10 horses were 226 +/- 19% for F.VII; 209 +/- 31% for F.VIII:C; 149 +/- 38% for F.IX; 88 +/- 12% for F.X; and 18.4 +/- 8.4% for C1-INH. Intra-assay coefficients of variation (CV) were 5.3% for F.VII; 2.1% for F.VIII:C; and 3.0% for C1-INH. Interassay CV were 5.7% for F.VII; 7.4% for F.VIII:C; 3.8% for F.IX; 14.4% for F.X; and 22.0% for C1-INH.

CONCLUSIONS

Human chromogenic assay kits can be automated and used to measure F.VII, F.VIII:C, F.IX, F.X, and C1-INH activities in citrated equine plasma.

CLINICAL RELEVANCE

Human chromogenic assays can be routinely used to measure F.VII, F.VIII:C, F.IX, F.X, and C1-INH in horses, and may be useful in evaluation of horses with disorders of hemostasis.

Fully automated assay for total homocysteine, cysteine, cysteinylglycine, glutathione, cysteamine, and 2-mercaptopropionylglycine in plasma and urine

We describe a 6-min HPLC method to measure the total concentrations of the most important thiols in plasma and urine--cysteine, homocysteine, cysteinylglycine, and glutathione--as well as the concentrations in plasma and urine, respectively, of cysteamine and 2-mercaptopropionylglycine, two compounds used to treat disorders of cysteine metabolism. Precolumn derivatization with bromobimane and reversed-phase HPLC were performed automatically by a sample processor. Throughput was up to 100 samples in 24 h. The within-run CV ranged from 0.9% to 3.4% and the between-run CV ranged from 1.5% to 6.1%. Analytical recovery was 97-107%, with little difference between plasma and urine samples. The detection limit was approximately 50 nmol/L for all the analytes studied. Thiol concentrations were determined in the plasma of 206 healthy donors and in the urine of 318 healthy donors distributed for age and sex. Mean values of plasma cysteine and homocysteine were significantly lower in infants (ages, <1 y) compared with other age groups (P <0.005). In adults, mean plasma homocysteine values were higher in males than in females (9.2 vs 6.7 micromol/L, P <0.0001) and in the 6- to 10-year-old group (P <0.05). Mean values for glutathione and cysteinylglycine were not sex- and age-dependent. In urine, both cysteine and homocysteine showed a wide range of variation.

Evaluation of the analytical performance of the Boehringer Mannheim Elecsys 2010 immunoanalyzer

OBJECTIVES

We evaluated the analytical performance of the Elecsys 2010 immunoanalyzer (Boehringer Mannheim Canada), which is based on a new detection technology, electrochemical luminescence.

DESIGN AND METHODS

We used six representative assays from the initial launch menu of the instrument: thyroid stimulating hormone (TSH), free thyroxine (FT4), troponin T, human chorionic gonadotropin (hCG), carcinoembryonic antigen (CEA), and prostate specific antigen (PSA). Within-run and between-day imprecision were evaluated using pools of human specimens at low, mid and high concentrations. Linearity was evaluated by diluting specimens with high analyte concentrations with specimens that had a low level of this analyte. Carry over and hook effect were investigated using specimens with high concentrations of hCG. Functional sensitivity was studied by running low TSH specimens over 21 daily runs, and by comparing the scatterplot of FT4 as a function of TSH. Over 100 specimens distributed across the analytical range were analyzed with two comparison methods: ES 300 (Boehringer Mannheim Canada) and AxSYM (Abbott Laboratories).

RESULTS

Within-run and between-day imprecisions were less than 4% and 10%, respectively, for most assays. All assays were linear over the whole analytical range. Carry over was minimal (< 0.0002%). A hook effect was present for hCG levels greater than 560,000 U/L. The functional sensitivity of the TSH assay was lower than 0.02 mlU/L. Correlation coefficients were all > 0.94. Small proportional errors were observed in comparison studies for the CEA and PSA assays.

CONCLUSIONS

The Elecsys 2010 system was shown to have an acceptable analytical performance for the rapid analysis of a wide variety of analytes. The hook effect observed with hCG assay would imply that the laboratory informs the clinicians of the possibility of falsely low values in trophoblastic diseases or that all specimens with values greater than 3000 U/L are reassayed after dilution.

An automated chromogenic peptide substrate assay for coagulation factor XII

We have developed an automated chromogenic peptide substrate assay for factor XII (FXIIcs) on a Cobas Mira S Plus clinical chemistry analyser using a new commercially available kit. This was used to determine factor XII (FXII) levels in plasma samples from 320 blood donors, 206 patients with a history of venous thrombosis and 74 lupus anticoagulant positive (LA+) patients. Results were compared with those obtained in a clotting assay for FXII (FXIIct) and an immunochemical assay (FXIIag). A satisfactory correlation coefficient of 0.92 and a regression line equation of y = 7.898 + 0.871x was obtained between FXIIcs and FXIIct in the 320 blood donors. Levels of FXII below the calculated normal range were found in nine blood donors (2.8%) and 16 venous thrombosis patients (7.8%). The blood donors and patients with venous thrombosis with low FXIIcs values had FXII levels below our lower limits of normal for both FXIIct and FXIIag; all were lupus anticoagulant negative. When FXII levels were determined in the 74 LA+ patients, 27 (36.5%) gave markedly lower FXII values in the FXIIct when compared with the FXIIcs. FXIIag levels corresponded with FXIIcs. The automated FXIIcs assay is therefore lupus anticoagulant insensitive and allows us to measure FXII levels accurately and routinely in large numbers of patient samples.

Evaluation of a haematological analyser (Sysmex F-800) with equine blood

A semiautomatic electronic blood cell counter (Sysmex F-800) was evaluated with equine blood, according to the protocol of the International Committee for Standardization in Haematology (ICSH, 1984). The precision and overall reproducibility were acceptable for all the parameters studied except for the platelet count, in which a coefficient of variation of 18.8% and 21.7% was obtained for within and between batch precision and 26.76% for overall reproducibility. Carry-over for the haematocrit value and platelet count was unsatisfactory, thus the use of a blank diluent sample between different analyses was necessary. Linearity of the analyser was acceptable in the range studied. A total of 100 equine blood samples were studied with the Sysmex F-800 and the manual method. The correlations between both Methods were acceptable for all the parameters. In the storage study, the red blood cell count and the haemoglobin concentration were the parameters with longest stability (72 h) at 4 degrees C and 25 degrees C. Platelet count decreased significantly at 3 h post-extraction in those samples stored at 4 degrees C. During the evaluation period, Sysmex F-800 was easy to handle and it is well suited for routine equine blood cell analysis.

Evaluation of some fluorogenic substrates for continuous assay of aminopeptidase P

Three potential fluorogenic substrates for assay of aminopeptidase P (AP-P) have been prepared and evaluated, using enzyme purified from porcine kidney. They are based on internal quenching of the synthetic, fluorescent amino acid (R, S)-2-amino-3-(7-methoxy4-coumaryl)propanoic acid ((R,S)-Amp) by a 2, 4dinitrophenyl (DNP) group. The compounds are X-Pro-Pro-(R, S)-Amp-NH2, where X is H-Lys(epsilon-DNP), H-Orn(delta-DNP), or L-2-amino-3-(DNP)aminopropionic acid. The first two were found to be excellent substrates for AP-P, with respective Km values of 4.8 and 5.2 microM. An advantageous feature is that under the conditions of assay, using 4-mm2 cells, the substrates are without noticeable quenching effect on the fluorescence of Pro-Pro-(R,S)-Amp-NH2 (the product liberated by the action of AP-P). At concentrations greater than about 30-50 microM, both substrates appear to inhibit the enzyme, but this has little practical consequence since assays can be carried out at substrate concentrations, giving up to approximately 80% of Vmax without this inhibitory effect being noticeable. The Lys derivative was found to be a very useful substrate for a continuous assay for AP-P and equally good in a discontinuous assay of multiple samples using microtiter plates. The racemic center at the Amp residue did not prevent total hydrolysis of the Lys derivative, suggesting that subsite specificity in AP-P does not extend as far as the P3' position.

Determination of trehalose by flow injection analysis using immobilized trehalase

A new method for the determination of trehalose by flow injection analysis (FIA) is described. The basic principle is the hydrolysis of the disaccharide trehalose into its monomer d-glucose by trehalase, a periplasmic enzyme of Escherichia coli. d-glucose is quantified spectrophotometrically after reaction with hexokinase and glucose-6-phosphate dehydrogenase. Trehalase is prepared by osmotic shock from a recombinant E. coli strain and precipitated with ammonium sulfate. The enzyme is immobilized on VA-Epoxy Biosynth from Riedel-de-Haën. The immobilization rate is about 60%. The FIA signals show a nonlinear dependence on the trehalose concentration. The resulting curve corresponds to a second-order polynomial that serves as a calibration function for test samples. Immobilized trehalase was used during a period of 4 months without any loss of suitability. Several samples of fermentation broth were tested. The results are verified by HPLC. Within an interval of 2 to 10 g/L trehalose the recovery is about 100-120% with a precision of 7% (coefficient of variation).

Fully automated determination of pesticides in wine

A fully automated solid-phase extraction gas chromatographic/mass spectrometric (SPE/GC/MS) method was developed for determination of pesticides in wine. All steps from aspiration of infiltrated wine to printout of the integrated chromatogram were performed without human interaction. A dedicated robot performed addition of internal standard, application of wine onto the SPE cartridge, elution of analytes, drying and concentrating of eluate, and passing of concentrate to the GC sampler. All steps were performed in standard liquid chromatography/GC vials, using a minimum of organic solvent. The method permits determination of 21 different pesticides. Individual detection limits were 0.005-0.01 mg/L. The regression coefficients relating to linearity were > 0.99; only 4,4-dichloro-benzphenone and dicofol showed lower coefficients. The recoveries for 17 pesticides ranged from 80 to 115%.

Multicenter study of Abbott AxSYM Digoxin II assay and comparison with 6 methods for susceptibility to digoxin-like immunoreactive factors

Performance characteristics of the Abbott nonpretreatment AxSYM Digoxin II assay were evaluated for quantification of digoxin at four independent sites. Correlation of digoxin measurements with the Abbott pretreatment AxSYM, Baxter Stratus II, Abbott TDx/ TDxFLx II, Abbott IMx, Emit 2000, and Beckman Synchron CX digoxin assays showed acceptable agreement, as indicated by: slope values > 0.84, r > 0.90, y-intercepts for all comparisons at or below the assay detection limit, and Sy/x ranging between 7.5% and 15.4% of the average digoxin value. Susceptibility to interference from digoxin-like immunoreactive factors (DLIFs) was examined in 233 samples from renal patients, liver disease patients, cord blood, and third-trimester pregnancies; the AxSYM Digoxin II assay demonstrated the least DLIFs interference. DLIF susceptibility for four of the methods was significantly greater (P < 0.05) than in the AxSYM Digoxin II assay; susceptibilities of the Stratus II and Emit 2000 methods were similar to the AxSYM Digoxin II assay.

Interferences of o-raffinose cross-linked hemoglobin in three methods for serum creatinine

The interferences of o-raffinose cross-linked hemoglobin (Hemolink) were examined and compared in two serum creatinine methods on the Hitachi 717 [Boehringer Mannheim (BMC) and Synermed] and in an enzymatic creatinine method on the Vitros 750 (Johnson & Johnson). Interference was considered significant when the change in creatinine concentration from the control exceeded the 95% confidence limits of each method. Significant interference was observed for the BMC/Hitachi 717 method with Hemolink > or = 5 g/L. No interference was observed for the Synermed/Hitachi 717 assay with Hemolink up to 30 g/L in normal samples and up to 50 g/L in samples with increased creatinine. No interference was observed for the Vitros 750 assay with Hemolink up to 50 g/L at both normal and increased concentrations of creatinine. Although the BMC/Hitachi 717 method was considered unacceptable, the Synermed/Hitachi 717 and the Vitros 750 methods allow accurate quantification of serum creatinine in the presence of Hemolink.

Multicenter evaluation of a homogeneous assay for HDL-cholesterol without sample pretreatment

We evaluated a new homogeneous assay for the measurement of HDL-cholesterol (HDL-C) in six European laboratories. The assay includes two reagents and is applicable to most autoanalyzers, which allows full automation. The total CVs of the new method ranged between 1.3% and 6.7%. Thereby determined HDL-C values were in good agreement with those obtained by precipitation with phosphotungstic acid/MgCl2 or by a combination of ultracentrifugation and precipitation (0.956 < r < 0.994). The assay was linear up to at least 1500 mg/L HDL-C. Hemoglobin did not interfere, whereas icteric samples with bilirubin > 100 mg/L showed discrepancies between the homogeneous and the precipitation assay. Lipemia up to total triglyceride concentrations of 8000 mg/L did not interfere with the homogeneous HDL-C assay. The homogeneous HDL-C assay was easy to handle and produced similar results in all laboratories participating in this study. This method will significantly facilitate the screening of individuals at increased risk for cardiovascular disease.

Microchip device for performing enzyme assays

An automated enzyme assay was performed within a microfabricated channel network. Precise concentrations of substrate, enzyme, and inhibitor were mixed in nanoliter volumes using electrokinetic flow. Reagent dilution and mixing were controlled by regulating the applied potential at the terminus of each channel, using voltages derived from an equivalent circuit model of the microchip. The enzyme beta-galactosidase (beta-Gal) was assayed using resorufin beta-D-galactopyranoside (RBG), a substrate that is hydrolyzed to resorufin, a fluorescent product. Reaction kinetics were obtained by varying the concentration of substrate on-chip and monitoring the production of resorufin using laser-induced fluorescence. Derived Michaelis--Menten constants compared well between an on-chip and a conventional enzyme assay. Bias in the derived K(m) and kcat was primarily due to the limited solubility of RBG and the associated lack of measurements at substrate concentrations exceeding the K(m). A Ki of 8 microM for the inhibitor phenylethyl beta-D-thiogalactoside (PETG) was determined from plots of initial rate versus substrate concentration obtained at three concentrations of PETG. The relative inhibition of beta-Gal by lactose, p-hydroxymercuribenzoic acid, and PETG was determined by varying the inhibitor concentration with constant enzyme and substrate concentration. An enzyme assay performed on the microchip within a 20-min period required only 120 pg of enzyme and 7.5 ng of substrate, reducing the amount of reagent consumed by 4 orders of magnitude over a conventional assay.

Thiocyanate in smokers interferes with the Nova magnesium ion-selective electrode

Thiocyanate found in serum ordinarily is the metabolite of cyanide that is inhaled with tobacco smoke and ingested with cyanogenic foods. We investigated the effect of the thiocyanate ion (SCN-) on the ionized magnesium (iMg) and ionized calcium (iCa) results determined with the AVL and Nova magnesium and calcium ion-selective electrodes (ISEs). We analyzed saline and pooled serum with added SCN-, and serum from apparently healthy nonsmokers (n = 20) and smokers (n = 20). The mean (and range) of the measured serum SCN- concentration was 0.019 (0.008-0.046) mmol/L for nonsmokers and 0.077 (0.020-0.138) mmol/L for smokers. Only the Nova iMg results decreased with increasing SCN- concentration, and the change was dependent on the baseline iMg concentration. In the absence of Mg, SCN- decreased the voltage response of the Nova Mg ISE to calcium ions. At apparently normal serum iMg and iCa concentrations, the interference by SCN- appeared to be equimolar (iMg = -1.04 x SCN- + 0.52). Thus, the serum SCN- commonly found in smokers causes a significant (P < 0.0001) decrease in the Nova iMg results.

Evaluation of an automated photometric fibrinogen assay

A recently-introduced automated method for the determination of plasma fibrinogen is based on the principle of von Clauss, combined with photometric detection: after addition of thrombin, the coagulation time is determined by measuring the change in absorption at 405 nm. This method was evaluated and compared with the original coagulometric Clauss assay and with the prothrombin time (PT)-derived automated method. The inter-assay coefficient of variation of the Clauss-derived assay was lower (14.1, 3.8 and 4.6%) than the PT-derived assay (16.1, 7.5 and 10.5%, respectively) at all three fibrinogen levels tested (1.2, 4.0 and 7.5 g/l). The correlation between the assays was investigated according to the method of Passing and Bablok and could be described as follows: Clauss-derived = 0.79 (PT-derived) + 0.66; Clauss-derived = 1.12 (Clauss) + 0.143. The interference of heparin (< 1.5 U/ml), haemoglobin (< 30 micromol/l), bilirubin (< 200 micromol/l) and triglycerides (< 5.5 mmol/l) in the Clauss-derived assay was negligible. The effects of fibrinogen degradation products on the Clauss-derived assay were comparable with the effects on the Clauss assay, in contrast to the effects on the PT-derived assay. In conclusion, the Clauss-derived assay is a specific and precise automated method to determine fibrinogen concentrations in plasma, which is not liable to interference from different pathophysiological substances.

Automated procedure for estimation of blood cholinesterase activities in rabbits

An automated procedure for estimation of blood cholinesterase activities was evaluated and used to determine baseline cholinesterase activities in rabbits. In addition, the kinetics of cholinesterase inhibition and recovery from diazinon challenge were determined by use of this procedure. The procedure developed is a modification of the colorimetric Ellman technique, adapted for use with the Instrumentation Laboratory Multistat III Plus Centrifugal Analyzer. The substrate used in this technique was acetylthiocholine. Baseline cholinesterase activities in domestic rabbits were plasma cholinesterase, 360 mU/ml; red blood cell cholinesterase, 2,658 mU/ml; and whole blood cholinesterase, 1,332 mU/ml. Rabbits were challenged with 0.3 LD50 (40 mg/kg) of diazinon, and blood samples were obtained at hourly, then daily intervals. Cholinesterase activity in these samples was determined to monitor the effect of the cholinesterase inhibitor diazinon in vivo, as estimated by the Multistat III Plus Centrifugal Analyzer procedure. Plasma cholinesterase activity decreased to 4% within 6 h, whereas red blood cell cholinesterase activity decreased to 51%. By 10 days, plasma and red blood cell cholinesterase activities had returned to 100 and 91%, respectively, of pre-challenge activities.

Determination of 18 beta-glycyrrhetinic acid in human serum using the fully automated ALCA-system

We report a method for the determination of 18 beta-glycyrrhetinic acid (glycyrrhetinic acid) in human serum using the ALCA-system. The technology of the ALCA-system is based on the principles of adsorptive and desorptive processes between liquid and solid phases. The assay is run fully automated and selective. Procedural losses throughout the analysis are negligible, thereby allowing for external calibration. The calibration curve is linear up to 10 mg/l and concentrations as low as 10 micrograms/l are detectable. CV is 2.5% for within- and 7.5% for between-assay precision at a level of 50 micrograms/l and 1.2% for within- and 8.5% for between-assay precision at a level of 500 micrograms/l. Specific and expensive reagents are not necessary and time-consuming manual operations are not involved. This assay can be selected from a wide spectrum of methods at any time. Thus, the present method is well-suited for drug monitoring purposes in the routine laboratory. In a pharmacokinetic study we measured serum levels of glycyrrhetinic acid in ten healthy young volunteers after ingestion of 500 mg glycyrrhetinic acid. Maximum levels of glycyrrhetinic acid were 6.3 mg/l 2 to 4 hours after ingestion. Twenty-four (24) hours after ingestion seven probands still had glycyrrhetinic acid levels above the detection limit with a mean level of 0.33 mg/l.

Evaluation of the ABX Cobas Vega automated hematology analyzer and comparison with the Coulter STKS

An evaluation of the new automated hematology analyzer was performed in comparison with the Coulter STKS on 1,694 blood samples coming from the different departments of Nice University Hospital. The Cobas Vega showed very satisfactory results in terms of repeatability, reproducibility and linearity. Correlation with the STKS was excellent with the exception of the following parameters: red blood cell distribution index and the absolute values for eosinophils and basophils. Two qualities were particularly appreciable: absence of leukocyte carryover, and stability of the complete blood count and leukocyte differential count over a long period. Analysis of qualitative flags showed that the overall blood smear review rate was 47% for the Cobas Vega, not forgetting that optical microscopy detects 37% of all abnormalities. The STKS's review rate was 49.5%. Flags commonly concerned the granulocytic lineage, 61% for the STKS and 48% for the Vega, with a false positive rate of 43.4% for the STKS compared with 22% for the Vega. The opposite phenomenon was observed with the flag for atypical lymphocytes which represented 11% of flags for the STKS and 25.6% for the Vega, with a false positive rate of 25.5% for the STKS and 34% for the Cobas Vega. This may be explained by the fact that lymphocyte abnormalities sometimes generated "granulocytic" flags on the STKS. Studies of the false negative rate carried out using light microscopy on 505 blood samples without flags on either system, detected the presence of a slight myelemia, and a few hyperbasophilic lymphocytes or plasmocytes in 18.6% of all cases. Finally, the Cobas Vega's practicality was greatly appreciated and there was no trouble with breakdowns throughout the whole period of its use.

A modification of the acid diazo coupling method (Malloy-Evelyn) for the determination of serum total bilirubin

A simple and reliable method for the determination of total bilirubin from human serum is described. In this method, indirect bilirubin is liberated by the tenside in 0.12 mol l-1 HCl (R1), and the total bilirubin is coupled with a 2,5-dichlorobenzene diazonium (DBD) salt to obtain the corresponding azobilirubin having a lambda max of about 520-522 nm. The method can easily be applied to the KONE Delta, a fully automated, discrete random access clinical analyser, and also to less modern instruments. A sample volume of 5 microliters, R1 volume of 180 microliters, and R2 volume of 36 microliters was used on the KONE Delta. After a 5-min incubation at 37 degrees C, measurement at 575 nm was done (main wavelength). The within-run imprecision (CV%) varied from 2.9 to 0.3% within the serum total bilirubin range of 14-290 mumol l (n = 10). The between-run imprecision was from 2.2 to 1.3% within the range 13-97 mumol l-1 (n = 8). The method is linear up to at least 340 mumol l-1 (19.8 mg dl-1), and dilution extends the test limit to 3400 mumol l-1 (198.8 mg dl-1). The linearity of dilution was good over the practical measuring range. The present method had a strong linear correlation with the Boehringer 2,5-dichlorophenyl diazonium (DPD) method on the Hitachi 717 analyser: y(DBD) = 1.018x(DPD)+0.758, r = 0.9955 (n = 61). The stability of R2 (diazo reagent) in the analyser reagent compartment lasts at least 2 weeks.

Comparative evaluation of three assay systems for automated determination of hemoglobin A1c

We evaluated three newly introduced systems for automated determinations of hemoglobin (Hb) A1c, which allow the processing of large amounts of samples in a routine clinical laboratory. We compared these methods--the Variant HPLC, the Hi-Auto A1c analyzer system, and the Roche immunoassay--with the Diamat HPLC system. All showed good precision and good concordance with the Diamat HPLC. The reference range for Hb A1c has to be determined by the laboratory for each assay system. Interference study showed no statistically significant influence of anemia, polycythemia, rheumatoid factor, or chronic hemodialysis, although individual Hb A1c values can be influenced by polycythemia (when measured with the Hi-Auto A1c analyzer) and by chronic hemodialysis (when measured with the Variant HPLC). HPLC was not suitable for measuring Hb A1c in the examined cases of hemoglobin variants; assaying fructosamine seems to be better for monitoring these patients.

An automated procedure for cluster analysis of multivariate satellite data

This paper studies the applicability of available cluster analysis methods on multivariate satellite data. Four different methods are studied, these are the Principal Component Analysis (PCA), K-means, Self Organizing Maps (SOM) and the Adaptive Resonance Theory (ART). Special focus is placed on the usefulness of these cluster analysis methods, that is, if the results generated by these methods are relevant and are of help to the physical analysis of the data. A combined SOM adaptive dynamic K-means procedure suitable for automated cluster analysis is presented. This new method is capable of achieving useful partition of multivariate satellite data which has not previously been studied.

Fluid elements--a concept for automation of diagnostic tests

Constructs consisting of a channel, a membrane, and an absorber are designed for autonomously carrying out various liquid-handling functions of analytical tests. These so-called fluid elements can be used to set up various circuits for conducting several kinds of analytical tests. To demonstrate the feasibility of this concept, we constructed such a circuit and used it to perform, with two handling steps, an ELISA of hepatitis B surface antigen. The detection limit of the assay was comparable with those of state-of-the-art ELISAs for screening blood, and results could be obtained within a total test time of 20 min. We anticipate that this concept of automation may also serve as a basis for new, highly simplified immunoanalyzers.

Performance evaluation of automated immunoassays on the Technicon Immuno 1 system

We performed an immunoassay evaluation for various analytes on a fully automated random-access analyzer, the Technicon Immuno 1 system from Bayer Corp. This system involves latex agglutination, magnetic separation sandwich, and magnetic separation competitive immunoassay configurations. The evaluated analytes were thyrotropin (TSH), triiodothyronine, thyroxine, free thyroxine, follitropin, lutropin, prolactin, beta subunit of human chorionic gonadotropin, cortisol, ferritin, alpha-fetoprotein, carcinoembryonic antigen, and prostate-specific antigen. We tested the assay precision, linearity, and correlation with comparison methods for these analytes. The functional sensitivity of the TSH assay and the sample-to-sample carryover were also studied. Excellent results were obtained for within-run and between-day precision studies, with most assays showing within-run CVs <4% and between-day CVs <6%. The linearity for all assays was acceptable and the correlation between Immuno 1 assays and comparison methods showed satisfactory results. The functional sensitivity of the TSH assay was estimated at 0.04 mU/L. No sample-to-sample carryover was detected.

Design and assessment of average of normals (AON) patient data algorithms to maximize run lengths for automatic process control

Achieving high quality and high productivity with automated testing processes will require process control systems that are optimized for the necessary error detection, minimum false rejection, and maximum run length. This study investigates whether run length could be monitored by average of normals (AON) algorithms that truncate the patient test distribution and estimate the average of a suitable number of patient results. The design of AON algorithms for individual analytes is facilitated by computer-simulated power curves that consider the ratio of the population biological variation (Spop) to the test method variation (Smeas), represent a range of Spop/Smeas ratios from 2 to 15, and include numbers of patient test results from 10 to 600. The potential applications of AON algorithms are assessed for 38 tests whose quality requirements represent the total error criteria from the Ontario Medical Association Laboratory Proficiency Testing Program, Spop/Smeas ratios from 0 to 32, critical systematic shifts from 0.02 to 10.85 Smeas, and test workloads representative of a regional reference laboratory. Approximately half of these tests provide high potential for applying AON algorithms to monitor run length.

Automated assay of plasma bromide after a single deproteinization step

This study aimed to simplify the spectrophotometric fluorescein method for measuring plasma bromide, improve its reproducibility, and automate it. After major modifications of the method, we obtained an essentially linear calibration curve for plasma concentrations of bromide between 0 and 5.0 mmol/L. The intraassay CV for measuring bromide in the supernatants of deproteinized plasma samples (initial plasma concentrations 2.5-5.0 mmol/L) was as low as 0.5-1.0% (n = 9). When all the procedures were incorporated, including deproteinization and dilution of plasma, the intraassay CV was 2% at 2.5 mmol/L (n = 5) and 1% at 5.0 mmol/L (n = 5). The interassay CV for measuring bromide in plasma supernatants (initial plasma bromide concentration > 2.5 mmol/L) was 1%. Analytical recovery of bromide added to plasma was 99.1% +/- 2.5%. The new method is simpler and more reproducible than other spectrophotometric methods. Conditions for its automation are described.

Water-soluble conductive polymer homogeneous immunoassay (SOPHIA). A novel immunoassay capable of automation

Conductive polymers are extensively conjugated macromolecules able to conduct electricity in their doped state and having a UV-visible spectrum which undergoes important chromatic modifications when subjected to pH changes or to oxido-reductive processes. This article describes a novel homogeneous immunoassay in which a water-soluble conductive polymer is used as the label. When antigen-antibody binding occurs, the local pH near the complex is modified. Such a pH change is in turn able to induce modifications in the absorbance at a characteristic wavelength of a conductive polymer covalently linked to either the antigen or the antibody. Consequently, the extent of tracer binding can be directly monitored by photometry during incubation. We present examples which validate the concept and exemplify its applicability in quantitative competitive immunoassays for human C-reactive protein and human serum albumin, as performed in a Cobas-Mira automated analyzer.

Semiautomated method for the quantitation of plasma or sera androstenedione, testosterone, and dihydrotestosterone in population studies

A method is presented that analyzes quantitatively and reproducibly the androgens testosterone, androstenedione, and dihydrotestosterone from human sera or plasma. The chromatographic separation step generates an unattended throughput of one preparative separation per hour. Controls are built into the method to account for changing chromatographic conditions that otherwise result in shifts in retention characteristics. Separation factors for the three androgens are as follows (mean +/- standard deviation): alpha = 1.23 +/- 0.011 between androstenedione and testosterone and alpha = 1.38 +/- 0.025 between testosterone and dihydrotestosterone. Sensitivities of the method are androstenedione 5 pg, testosterone 3 pg, and dihydrotestosterone 14 pg. A study of procedural losses associated with initial sample processing, a validation, and application to two sample sets which demonstrates the methods utility for the analysis of hypoandrogenic populations (postmenopausal women) and hyperandrogenic groups (prostate cancer patients) is also reported. The precision for replicate aliquots of control plasma is androstenedione and testosterone = 5-11% CV and dihydrotestosterone = 10-20% CV.

Enzymatic measurement of creatine in erythrocytes

We describe an automated enzymatic procedure for quantifying creatine in erythrocytes. In this assay, after sample clean-up, creatine kinase (CK; EC 2.7.3.2) and pyruvate kinase (EC 2.1.7.40) are used as auxiliary enzymes; lactate dehydrogenase (EC 1.1.1.27) is used in the indicator reaction. CK is also used as the starting reagent. Data obtained with the present method (y) correlated as follows with those from the comparison method (x) for creatine measurement in erythrocytes (colorimetric, diacetyl-alpha-naphthol): y = 1.017x - 27.5 (S(y/x) = 11.9, r = 0.997). Precision data were obtained by analyzing three different samples of erythrocytes nine times per day for 25 days with both methods. The total CVs determined by our method and the comparison method were respectively 3.0-4.8% (measured creatine range, 171-463 micromol/L) and 5.1-7.1% (creatine, 185-485 micromol/L). We established the reference interval (mean +/- 2SD) for creatine in erythrocytes from healthy individuals as 194-538 micromol/L for men (n = 71) and 241-677 micromol/L for women (n = 100). Using erythrocytes from patients with various hematological abnormalities and chronic renal failure, who were being treated with continuous ambulatory peritoneal dialysis or hemodialysis, we performed a clinical evaluation and obtained results in agreement with existing data on biochemical markers for the mean age of erythrocytes.