Hexokinase Method with a Reflectance ... - Clinical Chemistry

1 downloads 0 Views 423KB Size Report
Frank H. Wians, Jr., Glenn L. Jenkins, Nona Staples, and James I. Heald (Dept. .... References. 1. Henry RJ, Sobel C, Segalove M. Turbidimetric determinationof.
Technical Briefs (-‘300 words text) summarize findings that are of’ interest to a relatively limited audience. Readers desiring fuller details may obtain them by writing directly to the author(s) at the address given or, if this is impossible for any reason, to the Editorial Office of this

journal.

Measuring Serum and Plasma Glucose by the

Hexokinase Method with a Reflectance Photometer in a High-Risk Environment, A. Stott, B. Clark, and I. Barwise (Dept. of Biochem., Fazakerley District General Hosp., Liverpool L9 7AL, U.K.) The Ames Seralyzer reflectance photometer has proved useful for performing biochemical analyses both in the laboratory and in clinical units (1-3). We have found it suitable for analyzing samples within a contained environment, i.e., for “high-risk” patients positive for HIV antibody, hepatitis B antigen, etc. Ames has recently introduced a method for measuring serum glucose by using the hexokinase reaction, the colored product (a reduced formazan) being measured kinetically at 630 rim. We have evaluated and compared this method with the glucose oxidase method for glucose, performed with the Technicon RA-1000. For precision studies we used control sera (Gilford QCS, Normal and Abnormal) and patients’ samples. The results indicated acceptable performance. Concn, mmolfL No.

Range

QCS within 10 10

batch

Mean

CV. % Seralyzer

RA -

-

tCS between batch 14 14 -

4.0

3.1

16.1

1.9

5.1

3.6

16.8

3.8

Patients’ samples within batch 41 0-5.0 3.9 51 5.1-10 6.6 62 >10 15.8

2.3 2.2

3.6

1.5 3.8 3,9 -

-

There was good agreement between the two instruments for serum samples (n = 92, r = 0.99, slope = 0.99, intercept = 0.25) and plasma samples (n = 48, r = 0.99, slope = 0.99, intercept = 0.08), indicating that either serum or plasma may be used on the Seralyzer. We analyzed several patients’ samples that had various degrees of frank hemolysis. Positive interference was observed with even the minimum degree of hemolysis, although the effect was less pronounced at higher glucose concentrations. Lipemic samples with triglyceride concentrations ranging from 2.2 to 40 mmol/L and nonlipemic samples, analyzed before and after filtration through a 0.22jm (pore size) filter, showed no significant difference between the results, thus indicating no interference from lipemia. To some samples we added ascorbate to give final concentrations of 20 and 30 mg/L; these additions had no 424

CLINICAL CHEMISTRY, Vol. 34, No. 2, 1988

effect on the assay. The effect ofjaundice was determined by comparing the Seralyzer glucose results for 27 icteric samplea with those obtained on the RA-1000, the latter having

been found to be relatively free from such interference. The Seralyzer results showed a positive bias of approximately +6% for such samples, within a glucose range of 2.6-28.0 mmol/L, the degree of interference being independent of the degree of jaundice. This should be of little clinical significance. In conclusion: the Ames hexokinase method performs well, although the use of franidy hemolyzed samples should be avoided. References

1. Stevens JF, Tsang W, Newall RG. Measurement of the enzymes lactate dehydrogenase and creatine kinase using reflectance spectroscopy and reagent strips. J Clin Pathol 1983;36:1371-6. 2. Gibb I, Barton JR, Adams PC, Pratt D, Dean CR, Tarbit IF. Rapid measurement of creatine kinase activity in a coronary care unit using a portable benchtopreflectance photometer.Br Med J

1985;290:1381-3. 3. Gibb I. Evaluation and assessmentof new disposablestrip for determination of plasma potassium concentration. J Clin Pathol 1987;40:298-301.

Quantification of Urinary Albumin and Globulin by the Suifosalicylic AcidlTrlchloroacetlc Acid and DuPont “aca III” Analyzer TurbIdlmetric Total Protein Methods, Frank H. Wians, Jr., Glenn L. Jenkins, Nona Staples, and James I. Heald (Dept. of Pathol., Wilford Hall USAF Medical Center/SGHQLCC, 78236-5300) The labor intensity

associated

Lackland AFB, TX with high-volume testing

for total protein in urine by our present manual turbidimetnc method, involving a combination of sulfosalicylic/trichloroacetic acids (SSAiTCA) (1), caused us to examine the automated alkaline benzethonium chloride (BZC) procedure (2) performed on the DuPont “aca ifi” analyzer (ACA/BZC). Because the principal advantage of the SSAITCA procedure is the equal turbidity provided by albumin and globulin (1), we compared the ability of the SSAJTCA and ACAJBZC methods to quantifr albumin or globulin, or both, in supplemented urine samples. In addition, we compared total

protein values obtained by both methods for 24-h urine specimens from 56 patients selected without consciousbias. To normal human urine containing a negligible amount of endogenous protein we added a solution of purified human albumin (Sigma Chemical Co., St. Louis, MO 63178) or globulin (“Sandoglobulin”; Sandoz, Inc., East Hanover, NJ 07936) to give supplemented samples containing various concentrations (100-2000 mgfL) of albumin (Alb), globulin (Glob), or both. The purity of the albumin and globulin preparations was confirmed by immunofixation electrophoresis (WE), which demonstrated a single albumin band and globulin of the polyclonal IgG kappa class. We assayed all supplemented urine samples for total protein by the

SSA/TCA sults:

and ACA/BZC methods, with the following re-

Added

[ProteinJ’,found/expected

IAlb/GlobP mg/L 100/0 0/100

SSA/TCA

ACA/BZC

0.96 0.96

100/100 200/0

0.92 0.90

0/200

1.10

1.35 1.51 1.16 1.12 1.18

200/200 400/0

0.86

1.04

0.86 1.07 0.90

1.01 1.11 1.00

0.87

0.94

1.05 0.86 0.84

1.02 0.91 0.92

0/400 400/400 800/0 0/800

800/800 1000/0

0/1000 1000/1000

1.05

1.00

0.85

0.90

8[AlbJ+ [Globi = expected [Proteini. bmparison of totalproteinresults (mg/L)obtained by the present (c) and theautomatedmethods(y)yieldedthe followinglinearregressionequation:y = 1.15x + 36.24 (n = 15, , = 0.994). We similarly compared results for total protein in 24-h urine samples containing 0 to 8 g of protein per 24-h specimen and obtained the following linear regression equation: ACAJBZC method = 1.04 (SSAITCA method) + 0.09 (n = 56, r = 0.976). We examined test packs after assay of each patient’s urine sample in the “aca 1ff’ analyzer, but saw none with protein aggregates that might indicate a falsely low value for total protein (3). We evaluated the accuracy and intra-assay precision of both methods by assaying within a single run 20 samples of a protein standard (Quantimetrix, Hawthorne, CA 90250) containing 1000 mg of protein (70% albuminl30% gamma globulin) per liter, with the following results: SSAJTCA method mean protein concentration, mg/L 1004 (SD = 1.8; CV = 0.2%); ACAIBZC method mean protein concentration, mgfL = 1055 (SD = 2.1; CV = 0.2%). We used as single-point calibrator for the SSA/TCA method a solution of “Urine Protein Reference Material” (College of American Pathologists, Skokie, IL 60077) containing 2170 mg of human albumin per liter, and for the ACA/BZC method a five-point calibration curve obtained with a set of “DuPont aca urinary protein calibrators” containing a maximum of 2271 mg of human albumin per liter. We conclude that the SSAJTCA and ACAIBZC methods have comparable ability to quantiIr urinary albumin and globulin; the values for total protein obtained with the ACA/BZC method are approximately 5% to 10% higher than those obtained by the SSA/TCA method; the SSAITCA method is slightly more accurate than the ACA/BZC method; and the intra-assay precision (CV, %) of both methods is excellent. The small bias between these methods and the slight inaccuracy of the ACAIBZC method as compared with the SSAJTCA method are unlikely to provide clinically significant differences between values for urinary total protein obtained by these methods. Moreover, the ACAIBZC method offers advantages of automation and convenience.

3. Lacher DA, DeBeukelaer M. Falsely low value for total protein in urine as measured in the DuPont aca discrete clinical analyzer

[Tech.Brief]. Clin Chem 1986;32:203.

of High-Density Lipoprotein Subfractions in Stored Plasma, Anne Matthew and P. Finbarr Duggan (Biochem. Lab., Regional Hospital, Cork, Ireland) Stability

There is an inverse relationship between the risk of coronary heart disease (CHD) and the concentration of highdensity lipoprotein (HDL) cholesterol in plasma (1-3). Sepa-

ration of HDL into its two major subfractiorts, HDL and HDL3, has indicated that the smaller of the two, HDL, is particularly low in patients with CHD (4). Thus there is increased interest in the routine determination of HDL and its subfractions in many laboratories. Consequently, methods based on precipitation with polyanions have become popular, because they are fast, simple, and inexpensive. Lipoprotein concentrations are most accurately measured in fresh samples, because storage reportedly results in changes that are reflected in alteration in precipitability with polyanions (5). However, in practice, storage of samples may be unavoidable.

We wanted to determine the possible effects of storage on plasma H1)L-cholesterol subfractions, using a dextran sulfatefMgCl2 precipitation procedure (6). We used 54 heparmnized plasma samples, from healthy volunteers and patients with CHD. Cholesterol was determined enzymatically (cholesterol esterase, cholesterol oxidase, peroxidase) in fresh plasma and in plasma that had been stored for seven days at either 4 #{176}C or -20 #{176}C. Total HDLcholesterol

-

-

References 1. Henry RJ, Sobel C, Segalove M. Turbidimetric determinationof proteins with sulfosalicylicand trichloroaceticacids.ProcSocExp Biol Med 1956;92:748-51. 2. Iwata J, Nishikaze 0. New micro-turbidimetric methodfor the determination of protein in cerebrospinal fluid and urine. Clin Chem 1979;25:1317-9.

HDL3cholesterol mmol/L

1:

n

=

26

Fresh plasma Plasmaat 4 #{176}C, 7 days P value’ 2: n

=

1.38

(0.05)

1.31

(0.05)

0.002

a

0.89

(0.03)

0.97 0.006

(0.04)

28

Fresh plasma

1.48 (0.07) 0.91 (0.05) Plasma at -20 ‘C, 7 days 1.46 (0.07) 0.97 (0.04) P value >0.05 0.001 Figures are ±SEM. b Probabilityvaluesby paired Students f-test. at 4 #{176}C the total HDLwas 5% lower (P = 0.002). However, HDL3 cholesterol actually increased by 9%. Storage at -20 #{176}C for seven days resulted in no significant change in total HDL cholesterol (P >0.05), but there was a significant 7% (P = 0.001) increase in HDL3 cholesterol. Thus storage at either temperature resulted in some apparent H1)L2IHDL3 interconversion. Although further study may prove shorter-term storage to be feasible, these results suggest that determination of HDL subfractions by the dextran sulfate/MgCl2 procedure (6) (and possibly other precipitation methods as well) should be performed as soon as possible after the sample is taken.

After

cholesterol

seven days of concentration

storage

References 1. Miller GJ, Miller NE. Plasma high density lipoprotein concentration and development of ischaemic heart disease. Lancet

1975;i:16-19. CLINICAL CHEMISTRY, Vol. 34, No. 2, 1988

425