of fructosamine values between these two populations are not significantly different. These results imply two different distributions for measurements of Hb A1 but the same gaussian (anamorphosis) distribution for measurements of fructosamine. H,qerglycemk (Mann-Whitney tosamine impaired patients. Statistical analysis of the data test) shows a significant difference of frucaverages between normal subjects and those with glucose tolerance (IGT) with hemoglobinopathies (P <0.01 the difference is weaker for Hb nonsignificant and P <0.05). As expected, fructosamine assay undiagnosed diabetes in patients (2) or diabetic patients each); the significance of A1 data (respectively, more efficiently detects with hemoglobinopathies and seems to have good potential to be used as a screening test in populations in which hemoglobinopathies are common. The assay does not require subjects to be fasting, is easily automated, and is low cost. References 1. Moutet JP. Etude epidemiologiquede l’hypertension arterielle, du diabete et de l’alcoolismeen Guadeloupe.Enquete INSERM 1987; in press. 2. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categoriesofglucoseintolerance.Diabetee 197928:1039-57. 3. Kilzer P. Glycated hemoglobin and diabetes: a case and an overview of the subject.Clin Chem 1985;31:1060-67. 4. Lloyd D, Marples J. Serum fructosamine and thyroid function [Tech. Brief]. Clin Chem 1986;33:1985. 5. Van Dieijen-Viaser MP, Seynaeve C, BrombacherPJ. Influence of variations in albumin or total protein concentrations on serum fructosamine concentrations [Letter]. Clin Chem 1986;32:1610. 6. Johnson RN, Metcalf PA, Baker JR. Fructosamine: a new approach to the estimation of serumglycosylproteins.An index of diabetic control. Clin Chim Acts 1982;127:87-95. 7. BernStein RE. Glycosylated hemoglobin:hematologicconsiderations determine which assay for glycohemoglobin is advisable. Clin Chem 1980;26:174-5. 8. Aleysseine H. Low proportions of glycosylated hemoglobin ease- dated with Hb S and Hb C. Clin Chem 1979;25:1484-6. immunoturbidimetric C-Reactive Protein KIt Adapted to the Technicon RA-1 000, Stephen P. Harrison and Ian M. Barlow (Dept. of Biochem., Bradford Royal Infirmary, Duckworth Lane, Bradford, BD9 6RJ, U.K.) The “Sera-Pak” kit (Miles Laboratories Limited, Stoke Poges, Slough, SL2 4LY, U.K.) for estimating C-reactive protein in serum is based on the formation of immunocomplexes (1), in the presence of polyethylene glycol, resulting from the reaction of C-reactive protein in the sample with a specific antibody. The resulting turbidity is proportional to the concentration of C-reactive protein and is measured as an increase in absorbance at 340 nm and is compared with a series of calibrators to give the C-reactive protein concentration. The reagents were used in the RA-1000 according to the manufacturer’s instructions. The calibration curve, which should be prepared for each new batch of reagents, is stable for at least six weeks. The range of the assay is up to 120 mg/L, with no evidence of antigen excess up to 650 mg/L. The CVs for 10 determinations at a concentration of 11.5 mg/L were 3.7% within-batch and 6.0% between-batch; at 25.1 mgIL, 2.1% and 5.7%; and at 58.5 mg/L, 2.1% and 4.3%. 172 CLINICAL CHEMISTRY, Vol. 34, No. 1, 1988 Comparison of results obtained with the RA-l000 (y) and a radial immunodiffusion (x) method (2) gave the regression equation: y = 1.07x 0.12 mgfL (r = 0.99). Bilirubin and hemoglobin do not affect the assay, but grossly lipemic specimens (sample blank absorbance >1.0) can give erratic results, and sosuch specimens should not be analyzed. The assay is quick (10-mm throughput) and the cost compares favorably with other kits on the market. Settings for the RA-1000 are as follows: - Inununoassay IA Table 1 20 Type %Sample vol Filter pos Delay Default Blank %Reagentvol Units Umtfactor 2 60 1 WL 340 700 Decimal point RBL1ow RBL high Range low Rangehigh Normal low Normalhigh Slope Intercept EP Lim 1 60 7mgfL 1.0 1 0 0.12 0 120 0 10 1.00 0 STD 4* STD 5* STD 6* STD7* LIM 1 LIM2 LIM 3 LIM4 LIM5 LIM6 LIM 7 SLM1 SLM2 SLM 3 SLM4 SLM 5 SLM6 SLM7 10 RSP % 10 RSP lAType 0.01 0 #STD 6 %5ORSP 14.8 29.5 59.0 118.0 99 99 25 5 5 5 5 2 2 2 2 1 1 5 0 #ASP 2 9ORSP 99 0 99 0.0 STD1* STD 2* 0.00 3.7 %9ORSP RSS LIM 99 0.1 S 7.4 * 3* 5ORSP These values may vary between kits. We thank Miles Laboratorieslimitedfor providingkitsfor this study. References 1. Dona V, Maierna M, Tarenghi G, Berti G. Poster session, Xffl hit. Congress ofClin. Chem./Vll Eur. CongressofClin. Chem.,The Hague, The Netherlands, 1987. 2. Harrison SF. Pre-albumin and C-reactive protein after acute myocardial infarction. Med Lab Sc 1987;44:15-9. Effect of Captoprli and Other Inhibitors of AngiotensinEnzyme on Plasma Prolidase Activity, I. Myara,’ C. Cosson,’ P. F. Plouin,2 and N. Moatti’ (1 Lab. de Biochimie and 2 Service de Med. interne et Hypertension, Hopital Broussais, F-75674 Paris Cedex ConvertIng 14, France) Captopril, an inhibitor ofangiotensin-converting enzyme, is used in treating hypertensive diseases. Recently, Ganapathy et al. (1) showed that it inhibits the activity of proline dipeptidase (prolidase, EC 3.4.13.9) in various tissues, but no studies on its effect on this enzyme activity in plasma have been reported. Our laboratory has proposed that such determination in plasma be used as a marker in both chronic liver disease (2) and prolidase deficiency (3). There- fore, we studied the effect of captopril on prolidase activity in plasma. Using a previously described method (2), we showed that plasma prolidase activity in vitro was slightly decreased (residual activity: 85%) in the presence of 3 mmol of captopril per liter, a concentration about 500-fold larger than the C determined in pharmacokinetic studies (4). In patients receiving a single high oral dose of captopril (1 mg/kg body wt), prolidase activity was not significantly different before and 2 h after drug administration (n = 31, paired t-test). Similarly, prolonged treatment with captopril (50-100 mg/day, in two equal daily doses) for six months or more did not change the activity of this enzyme in plasma (controls: n = 46, mean 770, SD 195 U/L; patients: n = 13, mean 720, SD 155 UIL). We also studied the in vitro effect of other angiotensinconverting enzyme inhibitors. The following tabulation shows that these drugs were not or less inhibitory than captopril: Residualprolidas. acty inhibItor, 3 mmol/L Manufacturer (%) In plasma Captopnl Squibb 85 Enalapril Merck, Sharp & Dohme- 99 Perindopnl Perindoprilatea Ramipnl Ramiprilatea Chibret Servier Servier Hoechst Hoechst 98 101 93 99 Active metaboilte. Eachvaluerepreeentsthe median ofthreedeterminations. We conclude that plasma prolidase activity is not altered in patients treated with captopril and other angiotensinconverting enzyme inhibitors. References 1. Ganapathy V, PashleySJ,RoeselA, Pashley DH, Leibach FH. Inhibition of rat and human prolidases by captopril. Biochem Pharmacol 1985;34:1287-91. 2. Myara I, Myara A, Mangeot M, Fabre M, Charpentier C, Lemonnier A. Plasma prolidase activity: a possible index of collagen catabolism in chronic liver disease.Cliii Chem 1984;30:211-5. 3. Myara I, Stalder ,JF. Plasma prolidase and prolinase activity in prolidase deficiency [Tech Brief]. Olin Chem 1986;32:562. 4. Either C, Giroux B, Plouin PF, Maarek B, Giudicelli Jr. Captopril: pharmacokinetics, antihypertenaive and biological effects in hypertensive patients. Br J Cliii Pharmacol 1984;17:24350. Errors In Blood Glucose Determinations in Hemolysates by the Glucose Dehydrogenase Method on the Cobas Mira Analyzer, E. Raabo (Dept. of Clin. Chem., Central Hospital, Holbaek, 4300 Denmark) Measuring blood glucose in hemolysates by using glucose dehydrogenase is quite popular, owing to its rapidity and ease of automation (1). This method is easily adapted to the Cobas Mira analyzer (Hoffmann-La Roche, Switzerland). However, we encountered erroneous results for five specimens from patients with high leukocyte counts. We used a kit from E. Merck, Darmstadt, F.R.G. (cat. no. 13886; Gluc-DH5), based on the method of Banauch et al. (2). Procedure: To sample capillary blood, we used 20-pL heparinized micro pipets (Hounisens Laboratorieudstyr, Riiskov, Denmark). The samples were put into polypropylene cups (Sarstedt, no. 72.690 PP), containing 1000 zL hemolyzing solution and analyzed according to the Cobas Mira program. For the five specimens with leukocyte counts ranging from 67 to 270 x i0/L the double determinations differed considerably and deviated highly from the values for plasma. Re-sampling gave still other results, as did re-assaying the same hemolysate. When we examined the hemolysates, we found them to be unusually viscous, apparently due to gel formation, and in the cuvettes there was insufficient mixing ofthe hemolysate with the Gluc-DH reagent. Various hemolyzing solutions, included one from Roche, gave variable results, displaying atypical reaction curves. With regard to the hemolyzing solution from Merck (1, 3), this phenomenon seems to be caused by the high concentration of lysed leukocytes in the hemolysates, in the presence ofa high concentration ofsodium chloride (1.0 molJL). Thus, by decreasing the concentration of sodium chloride in the hemolyzing solution to 0.5 moIJL, but keeping the concentrations ofphosphate at 0.05 molIL, EDTA at 2.8 mmol/L, and, as detergent, Triton X-100 at 2.0 g/L, expected values are obtained for glucose, with at least one-week stability, stored at 4#{176}C. Although errors in glucose determinations are seldom encountered until cell counts exceed about 50 x i0fL, the chance for misleading results is present with the commercial reagents mentioned above. References 1. Dick W. Glucosebestimniung in Venen-mid Kapillarblut mit der Glucosedehydrogenase-Methodein einer neuen Hamolysierlosung am AutoAnalyzer IL J Cliii Chem Cliii Biochem1980;18:255-6. 2. Banauch D, Bruinmer W, Ebeling H, et al. Eine GlucoseDehydrogenase fiir die Glucose-Bestimmung in KOrperfiussigkeiten. Z Klin Chem Klin Biochem 1975;13:101-7. 3. Vormbrock R, Helger R. Eine neue Hamolysierlosung zur Gewinnung von Hamolysat fur die Bestimniungder Glucoeekonzentration im Blut. J Cliii Chem Clin Biochem1979;17:196. Stimulating Effects of Calcium and Magnesium on Serum Pseudocholinesterase Activity, Kunio Kobayashi, Takafumi Sakoguchi, and Akira Matsuoka (Dept. of Clin. Pathol. and Clin. Labs., Hyogo College of Medicine, 1-1, Mukogawa-cho, Nishinomiya 663, Japan) Serum pseudocholinesterase (pchE, EC 3.1.1.8) activity is in the diagnosis of hepatic diseases, although the biological roleof the enzyme is not understood. The activity can be reversibly or irreversibly inhibited by eserine, histidine, quarternary ammonium bases, organic phosphorus compounds, and fatty acids. Recently, we reported on the effects of heavy metals and alkaline earths on the pchE activity (Jpn J Clin Chem 1986;15:304-9). It was strongly inhibited by Cu2, (Jj2, and Cr3, which are physiologically distributed in low concentrations (10_6_10_9 mmol/L) in the body, but was stimulated by Ca and Mg, which are present in a relatively high concentration (10 mmol/L) in body fluids and tissues. The pchE activity in clinical serum samples containing important CLINICAL CHEMISTRY, Vol. 34, No. 1. 1988 173