Drug therapy is usually optimized by concentration measurement in patient serum. High-performance liquid chromatography (HPLC) is one of the most important analytical techniques used for therapeutic drug monitoring (TDM) of drugs for which no immunoassay kits are available. HPLC has been frequently used for screening purposes in toxicology, too. The Merck Tox Screening System (MTSS) has been developed for the identification of substances by a combination of gradient HPLC with diode-array detection and identification with a database system. For routine TDM an isocratic HPLC system is more suitable because of shorter analysis time, better reproducibility of retention index and better precision of results. Therefore we defined a set of methods in steps of 10% of the two MTSS eluents. Three examples are shown: Amiodarone, Indometacine and Thiopental. New applications to test for other substances can be transferred to an isocratic system after a complete MTSS gradient run.

T Bertsch, R EBanks!, M A Forbes', J Aufenanger, M Storr, J M Illingworth-, T J Perren, P J Selby! and R Kattermann From the /nstitut fur Klinische Chemie, Klinikum der Stadt Mannheim der Universitdt Heidelberg, Theodor Kutzer Ufer 1-3, 68167 Mannheim, Germany, I/CRF Cancer Medicine Research Unit, St James's University Hospital, Beckett Street, Leeds LS9 7TF, UK and 2Ares Serono, Geneva, Switzerland

Determination of activities of phospholipase A2 (PLA2) in human sera was based on the hydrolysis of phospholipids from [1-14C]oleic acid-labeled Escherichia coli biomembranes. The E. coli membranes served as substrate specifically for the PLA2 of human serum and were essentially resistant to other lipases in human sera, i.e., lipoprotein lipases, hepatic triacylglycerolipase, or pancreatic lipase in acute pancreatitis. Exchange of phospholipids between the serum and the biomembrane compartment aggravates the determination of PLA2 activity in human serum, which is naturally rich in phospholipids. In our modified E. coli assay, which overcomes these difficulties, the main substrate components phosphatidylethanolamine (70%) and cardiolipin (25%) were > 90% labeled in the sn-2 position. Fatty acids released by PLA2 activity were eluted from an aminopropyl solid-phase column directly into scintillation vials, where the radioactivity was counted. The ratio of [1-14C]oleic acid to released total fatty acids was used to calculate true enzymatic activity. The linear assay range extended from 0 to 3.6 U/L (0-60 nkat/L), with a detection limit of < 0.03 U/L (< 0.5 nkat/L). Within-assay imprecision (CV) was < 6% and between-assay is < 10% over the whole activity range. The normal range for men was 0-0.44 U/L (0-7.33 nkat/L) and for women 0.044-1.11 U/L (0.73-18.4 nkat/L). Patients with septicemia, pancreatitis, acute respiratory distress syndrome, or other severe diseases had PLA2 values up to 540 U/L (9000 nkat/L).

Sera from the routine of therapeutic drug monitoring were assayed for phenobarbital, phenytoin, and theophylline with three different methods: fluorescence polarization immunoassay as the standard procedure, the new CEDIA assays within a multicenter evaluation and HPLC which is known to yield results with a high specificity. CVs for between-day imprecision ranged from 2.6-8.6%, depending on the concentration of the drugs. There was a tendency to lower CVs for the HPLC procedure. Accuracy was verified with commercial control materials and spiked sera and proved to be satisfactory for all three methods and parameters. The linear range was approx. twice as wide for the HPLC compared with the other methods. The method comparisons were quite favorable. Deviations occurred mainly in the subtherapeutic concentration range.

The cholesterol of lipoproteins, separated electrophoretically on thin layer agarose films, is visualised and quantitated by incubating the gels in an enzymic reagent containing cholesterol esterase and cholesterol dehydrogenase. The individual fractions are quantitated by scanning densitometry. No sample pretreatment is necessary. All major fractions are detected readily. The accuracy of the determination is similar to that of ultracentrifugation. On average, imprecision is 3.1% for beta-, 7.0% for pre beta-, and 4.8% for alpha-lipoprotein cholesterol. Concentration and colour development are linear up to 8 mmol/l cholesterol in a given lipoprotein fraction. The results from the direct enzymic procedure for beta-, pre beta- and alpha-lipoprotein cholesterol are compared with those from quantitative lipoprotein electrophoresis after precipitation with phosphotungstic acid and bivalent cations and with those from different precipitation methods using dextran sulphate and polyethylene glycol. The new method has several advantages: high specificity; lack of dependence on the actual composition of the lipoproteins; lack of interference from coprecipitated proteins in the gel, e.g. fibrinogen or paraproteins; and insensitivity to lipolysis and high free fatty acid concentrations caused by heparin application or ageing of the specimen (at least for alpha-lipoprotein cholesterol quantitation). In its convenience and simplicity of operation, and the simple calculation of results, the method is similar to standard protein electrophoresis. The proposed method is therefore suggested as a standard method for elucidating lipoprotein disorders.

We describe a simple and rapid, but nevertheless precise and accurate method for the enzymatic determination of the main lipid constituents in human bile. Interfering bile pigments, especially bilirubin are eliminated by the use of aminopropyl bonded phase columns ("Bond-Elut") prior to the enzymatic measurement of cholesterol and lecithin. Intra-assay imprecision was between 3.1 and 4.9% CV, while the inter-assay figures were rather higher at 4.6 to 7.5% CV. Recoveries of bile salts, lecithin and cholesterol were between 94 and 103%. In contrast, the direct enzymatic determination in native bile produces falsely low results: lecithin from 5 to 20%, cholesterol from 25 to 40% of the true value. The results of both enzymatic methods correlated well with commonly accepted procedures for phospholipid and cholesterol determination. When compared with methods of bile lipid analysis involving solvent extraction, the column separation followed by enzymatic determination has the advantage of being simpler and less time consuming, without need of high-cost equipment, e.g. gas chromatography.

Zusammenfassung:Das selektive Analysensystem Boehringer Mannheim/Hitachi 717 wurde an vier Prüfstellen bei der Arbeitstemperatur 25 °C evaluiert.Acht für die verschiedenen Analysenabläufe charakteristische Methoden wurden unter weitgehender Beachtung des ECCLS-Protokolls untersucht.Während der viermonatigen Prüfphase wurden über 40000 Analysendaten erhoben, deren statistische Auswertung folgende Ergebnisse brachte: 1. Die Präzision in der Serie, gemessen in Kontrollseren, wurde für die untersuchten Methoden durch je 36 Variationskoeffizienten charakterisiert, deren Mediane zwischen V K = 0,6% und V K = 1,5% liegen.Für die Präzision von Tag zu Tag finden sich die entsprechenden Mediane zwischen VK = 1,0% und VK = 2,3%. Durch die großen linearen Meßbereiche allerMethoden fallen Wiederholungsanalysen nur sehr selten an. 3. Drifteffekte traten während des achtstündigen Prüfteils bei keiner der untersuchten Methoden auf. 4. Bei den Verschleppungsuntersuchungen waren keine Beeinflussungen von Probe zu Probe erkennbar.Modellversuche zur Reagenzverschlepung wurden an sensiblen Methodenpaaren wie GPT/LDH, Amylase/ Phosphat und Cholesterin/Harnsäure durchgeführt.Bei keinem dieser Paare wurde eine relevante Übertragung von Reagenz beobachtet. 5. Die Auswertung zur Wiederfindung in Kontrollseren erbrachte für alle untersuchten Methoden Werte zwischen 96% und 107% der deklarierten Sollwerte.6. Die Methodenvergleiche in Humanseren zeigen eine sehr gute Übereinstimmung zwischen Hitachi 717 und der jeweiligen Routineanalytik.Die Regressionsanalysen• ergaben in 29 von 32 Methodenvergleichen Steigungen zwischen 0,95 und 1,05.Die berechneten Achsenabschnitte waren generell vernachlässigbar.Während der gesamten Pfüfzeit kam es zu keiner Störung oder Fehlfunktion der untersuchten Geräte.Die Teilnehmer beurteilten die analytische Zuverlässigkeit und die Handhabung des Analysensystems Boehringer Mannheim/Hitachi 717 als sehr gut.

Lipoproteins are separated electrophoretically and cholesterol is visualised with an enzymic reagent specific for cholesterol in which the gels are incubated. Quantitation of the individual fractions is accomplished by scanning densitometry. No sample pretreatment is necessary. All major fractions are detected readily. Accuracy agrees favorably with results from the ultracentrifugation. On the average, imprecision is 3.1% for beta-, 6.9% for prebeta-, and 5.2% for alpha-lipoprotein cholesterol. Concentration and color development are linear up to 8 mmol/l cholesterol in a given lipoprotein fraction. The results of the direct enzymic procedure for beta-, prebeta- and alpha-lipoprotein cholesterol are compared to the quantitative lipoprotein electrophoresis after precipitation with phosphotungstic acid and bivalent cations.

A new enzymatic method for the determination of cholesterol in serum and plasma was evaluated in 8 separate laboratories in comparison with routine and reference methods. Investigation of the analytical reliability in the 2-26 mmol/l measurement range showed the following results: At the set reading points (10 min at 25 degrees C and 5 min at 37 degrees C) the reaction shows complete substrate conversion. The colour complex is stable over a period of 60 min. The response to cholesterol is linear up to 26 mmol/l. Precision within the series was 0.6-2.8% in 20 determinations (coefficient of variation). Day to day precision was 0.5-3.3% in triple determinations of 10 days (coefficient of variation). Accuracy was studied with 2 samples (assigned value: 3.52 and 6.70 mmol/l respectively). In the case of sample 1 the mean for the 8 laboratories was 3.44, with a median of 3.44; for sample 2 the values were 6.68 and 6.72. The results demonstrate an excellent transferability. In comparison with other enzymatic procedures, the values found with the new test were 5-10% higher; these results agree at all concentration ranges with the reference methods of Abell & Kendall and with those from mass spectrometry.

An enzymatic method is described for the determination of total cholesterol in serum using a single aqueous reagent which can easily be prepared from commercial substrates and enzymes. The determination is carried out manually, the resulting stable chromogen is measured at a wavelength of 500-550 nm. The cholesterol concentration may be calculated either using a primary cholesterol standard or a constant factor for a given wavelength. The reliability of the method is reported: Data for the imprecision are given on the basis of a survey in 3 laboratories, the accuracy is established by comparison with a definitive and a reference method. Analytical and biological interferances are briefly discussed and results with this enzymatic method are reported concerning the reference values for serum cholesterol.

Diabur-test 5000, a new test strip for estimation of urinary glucose, was compared with the hexokinase glucose-6-phosphate dehydrogenase method in more than 2500 urine samples. By combination of two test ranges glucose concentrations of up to 5% can be detected by the strip test. After a reading time of 2 minutes, very precise estimation of urinary glucose is possible in eight steps from negative to 5%. False estimations of more than one color step virtually do not occur. Ketone bodies, salicylic acid and several antibiotics do not influence the test strip. Ascorbic acid shows a slight influence only in concentrations above 40 mg/dl. This influence disappears with glucose concentrations of more than 0.5%. Good correlation with the reference method, wide range of readings and simple handling make the test strip suited for the laboratory and particularly for self control of diabetic patients.