Journal of clinical chemistry and clinical biochemistry. Zeitschrift fur klinische Chemie und klinische Biochemie最新文献

In contrast to therapeutic methods no general guidelines have as yet been formulated for the investigation of diagnostic measures. Previous studies have often delivered results that have not been interpretable. This situation was obviously unsatisfactory since staff, apparatus and financial resources were not being meaningfully employed. For these reasons the present memorandum was drawn up by a working group of the GMDS. It is intended to provide a framework for the evaluation of diagnostic measures. The process for the evaluation of diagnostic tests has by analogy to the procedure for the registration of drugs been divided into four clearly defined phases: Phase 1: preliminary technical and methodological investigations; Phase 2: estimation of test parameters in selected patients; Phase 3: Controlled diagnostic study; Phase 4: Investigation of the effectiveness; continuing risk-benefit analysis. Studies on new diagnostic tests should be designed according to the guidelines formulated in this memorandum. Many already established tests will require examination according to the criteria described here.

An enzyme immunoassay (EIA) for the measurement of serum testosterone has been developed using nitrocellulose paper discs as the solid support. The paper discs (6 mm diameter) coated with testosterone-specific antibody were incubated with testosterone and testosterone-peroxidase conjugate in glass tubes. The amount of testosterone present in samples could be estimated from the bound peroxidase activity. The assay was validated by comparison with a microtitre plate-based ELISA and a commercial radioimmunoassay (RIA) kit. The correlation coefficient between RIA and EIA was 0.84. No significant cross reactivity was observed with other steroids, except for dihydrotestosterone. The inter- and intra-assay coefficients of variation were 4.4 and 9.6% respectively. The preservation and transport of the coated paper discs are convenient and the overall cost of the method is less than that of other methods.

We evaluated a new immunoluminometric technique (ILMA) for the measurement of the cancer antigens, CA 19-9 and CA 125 in serum. A satisfactory reproducibility was found, coefficients of variation ranging from 4.2 to 7.3% in within-run and between-run assays. The linearity of tests was maintained over a wide concentration range. Mean analytical recovery was 94% for CA 19-9 and 98% for CA 125. A significant agreement between results obtained by immunoradiometric assays and evaluated methods was found both for CA 125 (ILMA = 0.917 IRMA + 1.048; r = 0.966; n = 98) and CA 19-9 (ILMA = 1.156 IRMA + 0.996; r = 0.995; n = 100).

MINILAB is a small photometer designed for use in small labs and physicians' offices. The instrument is preprogrammed for the determination of several analytes either in whole blood, plasma or serum. Each method is provided as a kit containing disposable cuvettes and capillaries for sample or reagent addition. We examined the precision of the MINILAB using the cholesterol and bilirubin methods for serum, following the recommendations provided for each method, and found that the precisions were not acceptable (CV up to 16% for cholesterol and 43% for bilirubin). Suspecting that these variations were due to sample and reagent pipetting, we tried a simple modification using a Microcap for pipetting and were able to improve the precisions, resulting in CVs of 1.4% for cholesterol and 3.5% for bilirubin.

The performance characteristics of an extraction/enzymatic procedure for serum cholesterol measurement were evaluated. The procedure is substantially derived from the accepted reference method as standardized by the Centers for Disease Control, substituting the enzymatic reaction for the Liebermann-Burchard reaction. Imprecision (CV) was consistently less than 1.5%, and accuracy was comparable to that of the definitive isotope dilution mass spectrometry method and the accepted reference method. Direct comparison of the enzymatic with the Liebermann-Burchard reaction, using a set of 50 human sera, revealed about -0.05 mmol/l constant bias of the former versus the latter, this being possibly due to higher specificity of the enzymatic reaction. As compared with the accepted reference method, the method described is characterized by higher practicability, the reagent being easier to prepare and to handle, and generating a more stable, chemically defined end-product.

The determination of methotrexate on the aca clinical analyser (aca MTHO method) was evaluated and the method compared with a fluorescence polarisation immunoassay (FPIA) and a specific high-performance liquid chromatography (HPLC) procedure. The within series and the between days coefficients of variation of measurements in human pool serum spiked with methotrexate (0.07 mumol/l to 15 mumol/l) ranges from 11.7% to 2.5% and 14.0% to 5.8%, respectively. The calibration was found to be stable for 9 weeks. There was a good correlation between the results of the MTHO method when compared with FPIA and HPLC. At low methotrexate concentration, the results of HPLC were on average 11% and 14% lower than those obtained by the MTHO assay and FPIA respectively.

We investigated the effect of the pretreatment (sonification or centrifugation) of saliva samples on the concentration of several steroid hormones as measured with highly specific RIA after extraction and chromatography. It appeared that sonification of saliva resulted in significantly higher values for progesterone, cortisone, 17-hydroxyprogesterone, testosterone and oestradiol (10-49% increase), compared with the levels recorded after centrifugation. No differences were demonstrated for the concentrations of cortisol and androstenedione, except that a sex-dependent difference effect was observed in the values for androstenedione: concentrations measured in sonificated male saliva were lower than those measured in supernatant saliva.

The present paper describes a multicentre evaluation of a one step enzyme-immunoassay for the determination of cortisol in serum or plasma. Data from the investigation were analysed in terms of imprecision, detection limit, and correlation with other test methods. Within-run and between-run imprecisions (coefficient of variation) of Enzymun-Test Cortisol were less than 8% and 12%, respectively. The detection limit was 30 nmol/l (11 micrograms/l). With the exception of prednisolone, only low interference was found with other endogenous steroids. A good correlation between Enzymun-Test Cortisol and HPLC, LIA, FPIA and RIA was registered, although the latter two methods showed a scattering of regression lines from the different evaluators. The results show that Enzymun-Test Cortisol can be recommended as an alternative for the measurement of cortisol. As the method is calibrated against isotope dilution-mass spectrometry, results obtained with Enzymun-Test Cortisol are in agreement with the reference method.

The concentration of amniotic fluid acetylcholinesterase activity is elevated in cases of foetal open malformations, the levels being higher in cases of open neural tube defects than in cases of abdominal wall defects. Determination of amniotic fluid acetylcholinesterase activity is an established procedure for the antenatal diagnosis of foetal neural tube defects. Performance data, technical advantages and limitations for three procedures for the determination of acetylcholinesterase activity are reviewed in this paper: an immunoassay, a gel electrophoretic procedure and a spectrophotometric procedure. An immunoassay using the monoclonal antibody 4F19 and the gel electrophoretic procedure show nearly identical diagnostic performances, with detection rates for open spina bifida close to 100% and overall false positive rates of approximately 0.2%. The spectrophotometric procedure is not suitable for the antenatal diagnosis of foetal open neural tube defects and abdominal wall defects. It is possible to distinguish open neural tube defects from abdominal wall defects by determination of the ratio of acetylcholinesterase activity to butyrylcholinesterase activity, either by combining the 4F19 immunoassay with a butyrylcholinesterase immunoassay or by gel electrophoresis followed by densitometry, on samples that display elevated levels of acetylcholinesterase activity.

The interference of immunoglobulins in the radioimmunoassay (RIA) of human beta-endorphin was investigated. Human IgM showed no cross-reactivity. Human IgA showed a weak cross-reaction, but the dilution curve of IgA did not show parallelism with the standard curve of beta-endorphin, thus indicating its antigenic difference. The dilution curves of human IgG showed 0.18% displacement with respect to the human beta-endorphin standard curve, with good parallelism. Moreover, five patients with multiple myeloma of the IgG type showed falsely elevated beta-endorphin levels. We investigated the possibility that certain IgGs may be responsible for the displacement of [125I]beta-endorphin in the beta-endorphin kit. The apparent beta-endorphin level of plasma from multiple myeloma patients was markedly decreased after affinity chromatography of the serum on protein A-Sepharose. In another 3 patients with multiple myeloma, we examined IgG interference by measuring the beta-endorphin levels in their lyophilized IgG diluted with saline. The results demonstrated high values of 20.2, 25.5 and 21.2 pmol/l respectively, also showing good parallelism. These immunological parallels to human beta-endorphin verify that a part of the amino acid sequence of human IgG is similar to that of human beta-endorphin. Consequently, in the measurement of beta-endorphin with polyclonal antibody, the results may sometimes be spuriously high due to cross-reaction with IgG, e.g., in patients with IgG myeloma. To avoid IgG interference, a specific monoclonal antibody to synthetic beta-endorphin should be used rather than polyclonal antibodies.