However, the lack of correlation of the plasma sodium on day 3 with any of the potassium indices tends to invalidate

the sick-cell concept. Although we found no increased osmolal gaps in plasma (Figure 1E), Groups 2 and 3 on day 1 showed increased osmolal gaps in urine (Figure 1G), which could be taken as evidence for sick-cell concept. However, again, the concept was not supported by the lack of correlation of the value of the plasma sodium on day 3 with the osmolal gap in urine. Also, there was no correlation of the decrease in plasma sodium with the osmolal gap in urine or plasma, although the plasma sodium on day 3 did correlate with the osmolal gap in plasma on day 2 (Table 1). In theory (19), the osmolal gap and potassium data should produce the strongest evidence for the sick-cell mechanism. On this basis, the data in this study have not produced convincing evidence for this mechanism operating in the development of hyponatremia after AM!. if the sick-cell mechanism is operating at all, this study suggests the most likely time is during the first day of the illness. The other mechanistic possibility is that there is a sustained release of AVP over the three days after AM!; some recent studies (4-6,20) support this view. We thank Dr. J. A. Tulloch for permission to study his patients, and Christine Bell, Lindsay Crowe, Joe Gibb, Graham Matthew, and Avril Nessfor carrying out the majority of the assays.We thank the medical and nursing staff ofthe Coronary Care Unit, Stracathro Hospital, for collecting blood and urine samples for this study.
References

dinfarctus myocardique aign: implications cliniques. Schweiz Med Wochenschr 1986;116:1727-9. 5. Schaller MD, Nussberger J, Fiehl F, et al. Clinical and hemodynamic correlates of elevated plasma arginine vasopressin after acute myocardial infarction. Am J Cardiol 1987;60:1178-80. 6. McAlpine HM, Morton JJ, Leckie B, Rumley A, Gillen G, Dargie HJ. Neuroendocrine activation after acute myocardial infarction. Br Heart J 1988;60:117-24. 7. Flear CTG, Hilton P. Hyponatraemia and severity and outcome of myocardial infarction. Br Med J 1979;i:1242-6. 8. Evans JR. Osmolal gaps in urine [Tech Brief]. Clin Chem
1986;32:1415.

9. Neary RH. More on osmolal gaps in urine [Letter]. Clin Chem 1986;32:2225-6. 10. Evans JR. Yet more on osmolal gaps in urine [Letter]. Clin

1. Oliver MF. Metabolic responsesduring myocardial infarction. II. Clinical implications. Circulation 1972;45:491-500. 2. Vetter NJ, Strange RC, Adams W, Oliver MF. Initial metabolic and hormonal response to myocardial infarction. Lancet 1974;i:284-8. 3. Ceremuzynski L. Hormonal and metabolic reactions evoked by acute myocardial infarction. Circ Res 1981;48:767-76. 4. Schaller MD, Nicod P, Nussberger, J, et al. Vasopressine lors

Chem 1987;33:746. 11. Technicon Instrument Corp., Tarrytown, NY. Total CO2 method, N-8b(IfH), 1970. 12. Marsh WI!, Fingerhut B, Miller H. Automated and manual direct methods for the determination of blood urea. Clin Chem 1965;11:624-7. 13. Varley H, Gowanlock All, Bell M, eds. The estimation of plasma and urine creatinine by a continuous-flow method. In: Practical clinical chemistry, 5th ed., Vol. 1. London: Heinemann, 1980:481-2. 14. Gommorri SJ. The measurement of plasma phosphate. J Lab Clin Med 1942;29:955. 15. Varley H, Gowanlock Al!, Bell M, eds. The estimation of plasma and urine ammonia using the Berthelot reaction. Op.cit. (ref. 13):835-6. 16. Chan 5, Radcliffe A, Johnson A. The serum sodium concentration after surgical operation: precision permits predictions. Br J Surg 1980;67:711-4. 17. Snedecor GW, Cochran WS. Comparison of means of independent samples with different population differences. In: Statistical methods, 7th ed., Ames, IA: The Iowa State University Press: 1980:96-8. 18. Ibid. 158-62. 19. Gill GV, Flear CTG. Hyponatraemia. Recent Adv Clin Biochem 1984;3:149-76. 20. Col J, Petain M, Van Eyll C, Cheron P, Charlier AA, Panleur H. Earlychanges in sodium and water balances in patients with acute myocardial infarction: relationship to haemodynamics and creatine kinase. Eur J Clin Invest 1984;14:247-54.

CLIN. CHEM. 36/2, 325-329 (1990)

Falsely Low Estimation of Triglycerides in Lipemic Plasma by the Enzymatic Triglyceride Method with Modified Trinders Chromogen
Mark D. S. Shephardand MalcolmJ. Whiting
The enzymatic assay of triglyceride, based on the use of L-glycerol-3-phosphate oxidase (EC 1.1.3.21) and a modified Tnnders chromogen involving 4-chlorophenol, is subject to strong negative interference at concentrations of triglyceride >20 mmol/L, such as occur in grossly lipemic plasma. This interference is caused by the rapid utilization of oxygen, resulting in the reaction becoming transiently anaerobic. The dye product already formed may then be reduced (bleached) by acting as an alternative electron acceptor for glycerol-3-phosphate oxidase. Reduction of the dye leads to a marked decrease in final absorbance at 505 nm. Grossly underestimated values for triglyceride concentrations, apparently within the linear range of the assay, may therefore be inadvertently obtained withequilibrium methods.We suggest that samples giving unexpectedly low results for lipemic plasma should be re-assayed after dilution or with use of a smaller volume of sample.
AdditIonal Keyphrases: glycerol-3-phosphate indicator dye oxidase

electron

acceptor Department of Biochemistry and Chemical Pathology, Flinders Medical Centre, Bedford Park, South Australia, 5042, Australia. Received August 28th, 1989; acceptedOctober 19, 1989.

Colorimetry

linked

of hydrogen peroxide by use of peroxidasechromogenic systems is an indicator reaction used CLINICAL CHEMISTRY, Vol. 36, No. 2, 1990 325

for many analytes. These include glucose (1), uric acid (2), and cholesterol (3), which are oxidized by specific oxidases to form hydrogen peroxide. In addition, the enzymatic assay oftriglycerides has been recently linked to this type of indicator reaction, in preference to monitoring NADH production (4). This development became possible with the purification of L-ce-glycerol phosphate oxidase (5), and the use of lipase and glycerol kinase to convert triglyceride to glycerol 3-phosphate, as shown below. Triglyceride
+ 3 H20
lipase

61236), final concentrations of components of the reagents were as follows: (Reagent 2) Tris buffer, 100 mmol/L, pH 7.6; magnesium salt, 4 mmolJL; 4-chlorophenol, 5.4 mmoll L; and (Reagent 3) ATP, 0.8 mmol/L; 4-aminophenazone, 0.4 mmoLIL; lipase, 100 kUIL; glycerol phosphate oxidase, 2.0 kUIL; glycerol kinase, 200 U/L; and peroxidase, 200 UJL. To study the separate steps in the triglyceride assay, we used individual assay components at the same concentrations as in the Boehringer-Mannheim reagent. Instrumentation
Automated

(EC 3.1.3.1)

glycerol + 3 RCOOH

Glycerol + ATP
Glycerol

glycerol kinase

(EC 2.7.1.30)
+ 02

glycerol 3-phosphate + ADP

glycerol phosphate oxidase (EC 1.1.3.21)

trifugal analyzer with Boehringer

assays were performed in a Cobas-Bio cen(Roche Diagnostics Systems, Nutley, NJ)

Mannheim

reagents and in a Model 717

3-phosphate

analyzer (Hitachi Ltd., Tokyo, Japan) with reagents from bioMerieux. Some manual assays were monitored in a Model UV-160 recording spectrophotometer (Shimadzu

H202 + dihydroxyacetone H2O2 + 4-aminophenazone

+ 4-chlorophenol
+

phosphate
peroxidase

Corp., Kyoto, Plasma

Japan).

(EC 1.11.1.7)

4(benzoquinone-monoamino)-phenazone

2 H20

HC1

In a recent batch of patients specimens presented for routine estimation of triglyceride, one plasma sample was observed to be grossly lipemic. Upon enzymatic analysis, a surprisingly low value (for triglyceride) of 5.4 m.mol/L was obtained for the undiluted sample. This result, although well within the linear range of the assay (upper limit 12 mmoIfL), was clearly inappropriate for this sample. We describe investigations into the reason for grossly underestimated triglyceride results obtained with the enzymatic assay of lipemic plasma when a modified Trinders chromogen (6) involving 4-chlorophenol is used in an indicator reaction.

The plasma sample under study was a grossly lipemic specimen collected from a 52-year-old woman. Its actual triglyceride concentration (see below) was 40 mmol/L. To study individual steps in the reaction sequence, we prepared 40 nunol/L solutions of glycerol, glycerol-3-phosphate, and hydrogen peroxide in distilled water as substitutes for plasma.

Results and Discussion

Reaction Kinetics with Lipemic Plasma Absorbance data for the enzymatic triglyceride assay of the lipemic plasma were obtained with both the Cobas Bio and the Hitachi 717. The reaction kinetics are shown in Figure 1. With the instrument settings listed in Table 1, the Cobas Bio monitored the reaction for 5 mm and calculated the triglyceride result on a final absorbance reading taken at 300 s after the plasma was added. The Hitachi 717 monitored the reaction for 10 mm, but also derived the triglyceride concentration from a single absorbance value taken at 600 s. An obligatory stirring of the reaction mixture took place in the Hitachi 717 at 300 s, even though a start reagent was not used. The time course of absorbance changes showed that an initial rapid increase in the formation of the Trinder-type

Materials and Methods

Reagents The reagent kits for enzymatic determination of triglyceride were from Boehringer Mannheim GmbH Diagnostica, Mannheim, F.R.G., and bioMerieux, France. The enzymatic cholesterol kit also was from Boehringer Mannheim. Individual components of the triglyceride assay-glycerol 3-phosphate, glycerol-3-phosphate oxidase, 4-chlorophenol, 4-aminophenazone, and peroxidase-were from Sigma Chemical Co., St. Louis, MO 63178, as were the dyes p-iodonitrotetrazolium violet and 2,6-dichlorophenolindophenol. Glycerol was obtained from May & Baker Australia Pty. Ltd., Melbourne, Victoria, Australia, and hydrogen peroxide was supplied by BDH Chemicals, Melbourne. Reagent kits were prepared according to the manufacturers instructions. For the Boehringer Mannheim Periodochrom Triglyceride GPO-PAP kit (cat. no. 701904), concentrations of components of the reagents used in the test were as follows: (Reagent 1) Tris buffer, 150 mmolJL, pH 7.6; magnesium sulfate, 17.5 mmol/L; EDTA disodium salt, 10 mmoLIL; 4-chlorophenol, 3.5 mmollL; sodium cholate, 1.5 g/L; potassium hexacyanoferrate, 6 mol/L; and (Reagent 2) ATP, 0.5 mmollL; 4-aminophenazone, 0.35 mmol/L; lipase, 3 kUIL; glycerophosphate oxidase, 2.5 kU/L; glycerol kinase, 2 kU/L; and peroxidase, 150 UIL. For the bioMerieux Triglyceride PAP 150 kit (cat. no.

a
0

Time

Fig. 1. Absorbance readings obtained with Cobas-Bio (A)and Hitachi 717 ( analyzers during assayoftriglyceride inlipemic plasma Reactionswere monitored at 30-s intervals for 300 s on the Cobas Bio or at
12-s intervals for 600 s on the Hitachi 717. In the Hitachi 717 the reaction mixture was stirred at 300 s (arrow)

326 CLINICAL CHEMISTRY, Vol. 36, No. 2, 1990

Table 1. Instrument Settings for the Cobas Bio and HItachi 717 for Enzymatic Assay of TriglycerIdes In Plasma
Cob..

Blo

HitachI 717

2. 3.
4.

5. 6. 7. 8. 9. 10.
11.
12.

Units Calculation factor Standard 1 concn Standard 2 concn Standard 3 concn

Limit Temperature (#{176}C) Type ofanalysis

mmol/L 0 1.86 1.86 1.86 12 37.0 505 05 40

300

Assay code

13.

14.
15.

16. 17. 18. 19.

Wavelength (nm) Sample vol (&L) Diluent vol (ML) Reagent vol (pL) Incubation time (s) Start reagent vol (LL) Time of first reading (s) Time interval (s) Number of readings Blanking mode Printout mode

60 0 0.5 30 11

Sample vol Ri volume R2 volume Wavelength Calib. method STD (1) concn-pos STD (2) concn-pos STD (3) concn-pos STD (4) concn-pos STD (5) concn-pos STD (6) concn-pos SD limit Duplicate limit
Sensitivity limit

1 point : 50-0 5:2

300:100: NO o : 20: NO

700 : 505 Linear : 0 : 0 0.00 :1 1.86:2 0:0

0:0

0:0 0:0
0.1

200
0

Abs limit (inc/dec) Prozonelimit Expected value

Instrument factor

0 : Increase o : Lower 0.3:1.70 mmol/L

1.0

quinonemonoimine dye during the first 100 s was followed by a large decline in absorbance until 300 s of reaction time (Figure 1). In the Hitachi 717, a sudden increase in absorbance was recorded after the 300-s point, coincident with stirring of the reaction mixture. In both instruments the triglyceride concentrationwas calculated from the final absorbance reading, so it became evident why a low result of 5.4 mmol/L was obtained for this sample. However, the reason for the steep decline in absorbance was not immediately apparent. The atypical reaction kinetics seen with the lipemic plasma were shown to be concentration dependent. Serial dilutions of the plasma specimen, corresponding to 20%, 40%, 50%, 60%, 80% and 100% of the original plasma, were prepared in isotonic saline and re-analyzed in the Cobas Bio and the Hitachi 717. Absorbance data obtained for each dilution are plotted for the range 50% to 100% in Figure 2. The results show that the true triglyceride concentration
k COSAS 810

in this specimen was -40 mnmol/L. More importantly, the abnormal reaction kinetics were observed only at triglyceride concentrations >20 mmol/L. In addition, we found that the linear range of the bioMerieux reagent (upper limit 15 mmol/L) exceeded that of the Boehringer-Mannheim reagent (upper limit 12 mmol/L). Possible Reasons for the Observed Interference The following investigations were carried out to explain the negative interference observed in the enzymatic triglyceride assay. First, spectral analysis of the reaction products at 1-mm intervals in a recording speetrophotometer revealed that the decrease in absorbance at 505 rim was not ascribable to a wavelength shift in the absorbance maximum of the quinonemonoimine dye product, but instead to a loss of dye

(bleaching).
Second, each step in the reaction sequence was examined separately. A 40 mmol/L glycerol solution was substituted for the lipemic plasma and analyzed as a patients sample; that is, the first step in the reaction sequence was bypassed. Absorbance data displayed in Figure 3A illustrate that interference identical to that observed with the lipemic plasma was still observed. Glycerol solutions-2, 5, 10, 20, and 30 minol/L-were also analyzed and interference was shown to be concentration dependent, commencing at glycerol concentrations >20 mmoIIL. Glycerol phosphate solutions of the same concentrations were then prepared and analyzed to bypass the second step in the reaction sequence. Interference at concentrations >20 mmol/L was again observed. At 40 mmolJL (Figure 3B), the effect was seen within the first 30s of the reaction. Finally, a series of solutions containing 2, 5, 10, 30, and 40 mmol of hydrogen peroxide (H2O2) per liter were prepared and analyzed; that is, the reaction was commenced at step 4. At all H2O2 concentrations tested (including 40 mmol/L, see Figure 3C), a constant absorbance plateau was attained, thereby isolating the interference to the third step of the reaction
sequence.

a
S
8. HITACHI
717

Tim.(sec)

200

400

600

200

400

600

200

400

600

200

400

600

Time s

Fig. 2. Absorbance readings obtained with Cobas-Blo (A)and Hitachi 717 ( analyzers dunng assay of triglyceride in serial dilutions of
lipemic plasma The dilutions were madeIn saline and corresponded to 50%, 60%, 80%, and 100% of the original plasma (a, b, c, and d, respectively)

It has been suggested that the bleaching of Trinders CLINICAL CHEMISTRY, Vol. 36, No. 2, 1990

dye
327

A.Glycerol

B.GlycerolPhosphate

C.Hydrogen Peroxide

5.0
4.0

I
a

100

200

300

too

200

300

100

200

300 10 TIme (mm) 15 20 25

Fig. 3. Absorbance during incubation of the triglyceride assay reagent with a 40 mmoVL sample of glycerol (A), glycerol 3-phosphate (, or hydrogen peroxide (C) Reactions were monitored at 30-s intervals for 300 S is produced by reducing agents such as ascorbate (7,8). We predicted that excess glycerol 3-phosphate might be responsible for reducing the quinonemonoimine dye in our reactions if oxygen became limiting in the reaction sequence; that is, the dye could act as an alternative electron acceptor to oxygen from the glycerol phosphate oxidase-glycerol phosphate complex if reactions became anaerobic, as shown in the following equation:
Glycerol 3-phosphate + quinonemonoimine dye
-*

Time s readings obtained with a Cobas-Bio analyzer

Fig. 4. Formation and bleaching of quinonemonoimine dye by various relative concentrations of glycerol 3-phosphateand oxygen The reaction was initiated by adding glycerol3-phosphate at time zero (A). After 5 mm, the reaction mixture was oxygenated by shaking the cuvette(. At 16 mm,moreglycerol3-phosphate was added (C) (Figure 5). However, no subsequent bleaching of dye product occurred with reagent bubbled with 100% oxygen, whereas a 60% and 90% reduction in absorbance was subsequently observed with reagent bubbled with air and untreated reagent, respectively (Figure 5). Oxygenation upon mixing would explain the increase in absorbance seen at 300 s for reagent monitored on the Hitachi 717 (Figure 1B). 4. To demonstrate a requirement for glycerol-3-phosphate oxidase to mediate the bleaching effect of glycerol 3-phosphate, we generated the same quinonemonoimine dye with use of an enzymatic cholesterol-assay kit (Boehringer Mannheim Cholesterol CHOD-PAP kit, cat. no. 23669) by incubating a normal patients sample with the reagent for cholesterol. The specific oxidase in this reagent system was cholesterol oxidase. A 15-FL aliquot of 40 1.8
C. 95% 02

dihydroxyacetone

phosphate + reduced dye

The following experiments gave results consistent with this explanation. 1. All oxygen was removed from an aliquot of assay reagent in a sealed vial by degassing the solution under reduced pressure and filling the vial with nitrogen. When a sample of 40 mmol/L glycerol 3-phosphate was injected into the solution while anaerobic conditions were maintained, no color developed. However, injection of a sample of a 40 minol/L solution of hydrogen peroxide resulted in rapid formation of a red color, followed by slower, complete bleaching of the color, a process that was dependent on the prior addition of glycerol 3-phosphate. Control assays showed that degassing solutions did not per se inactivate the enzymes in the assay reagent. 2. A 15-oL sample of a 40 mmol/L solution of glycerol 3-phosphate was added to 900 pL of triglyceride reagent in a recording-spectrophotometer cuvette. The absorbance of 505 nm was read at regular intervals. Chromogen was formed, but was then bleached (Figure 4A). After 5 mm, the cuvette was removed from the spectrophotometer and vigorously shaken to reoxygenate the reagent. The cuvette was replaced and the absorbance was observed to increase rapidly and reach a constant value, which was stable for 10 mm (Figure 48). Further bleaching did not occur because the glycerol 3-phosphate was depleted. However, addition of a further 15-L aliquot of glycerol 3-phosphate again resulted in a near complete loss of dye absorbance at 505 nm (Figure 4C). 3. The oxygen content of the assay reagent was increased by bubbling air (20% oxygen) or 100% oxygen through it for 30 s. On the addition of a sample of 40 mmol/L glycerol 3-phosphate, the absorbance increased after 30s to a maximum value of 1.7 for all assay mixtures 328 CLINICAL CHEMISTRY, Vol. 36, No. 2, 1990

1.5 E
C U, 0 IL, (0

0, .0
0
U C (0

1.0

.0

I,

8.20% 02 0.5
(Air)

A. Normal
Reaction

Conditions Time (mm) Fig. 5. Formation and bleaching of quinonemonoimine dye underthe usual reaction conditions (A) or after oxygenation of the assay reagentwith air ( or oxygen (C)

mmolJL glycerol 3-phosphate was added to 900 L of dye produced from the cholesterol kit. In the absence of glycerol-3-phosphateoxidase, the glycerol 3-phosphate did not bleach this dye; evidently, the enzyme is indeed required as a catalyst. Further studies revealed that not only the presence but also the activity of glycerol-3-phosphate oxidase was important, because the quinonemonoimine dye was bleached only when the enzyme activity in the assay was 2.5 kUIL. Glycerol-3-phosphate oxidase from three different bacterial sources (Streptococcus thermophilus, Pediococcus spp. and Aerococcus uiridans) behaved similarly. 5. The effect of substituting other redox dyes for the Trinder-type chromogen was investigated. The colorless, oxidized form of iodonitrotetrazolium dye at a concentration of 160 moI/L was reduced by <10% to produce the red reduced dye after the 40 mmol/L solution of glycerol 3phosphate was added. In contrast, oxidized dichlorophenolindophenol at a concentration of 70 jimol/L was completely bleached after 1 mm by the added 40 mmol/L glycerol 3-phosphate. The addition of a 5-xL sample of a 20 mmol/L solution of glycerol 3-phosphate had no effect on either dye. These two dyes act as weak electron acceptors for purified glycerol-3-phosphate oxidase, with the indophenol being a fivefold better acceptor than the tetra.zolium (5). The behavior of other dyes that have been used as chromogens in triglyceride assays, such as the quinonemonoimine formed with 3,5-dichloro-2-hydroxybenzenesulfonic acid (9) or the triarylimidazole leuco dye (10), has not been investigated. In summary, we have shown that the enzymatic assay for triglyceride involving a modified Trinders chromogen is subject to significant negative interference when the concentration of triglyceride in the sample exceeds 20 mmol/L. Interference arises when oxygen becomes limiting in the reaction and dye is bleached by acting as an alternative electron acceptor for glycerol-3-phosphate oxidase. As a

consequence, falsely low triglyceride values may be generated in this assay system when results are calculated from the final absorbance of the reaction mixture. In practice, such errors should be avoided if lipemic plasma samples are assayed at two different dilutions or a smaller sample volume is used. We thank Michael Peake for helpful suggestions during this work. References
1. Bergmeyer HU, Bernt E. D-Glucose. Determination with glucose oxidase and peroxidase. In: Bergmeyer HU, ed. Methods of enzymatic analysis. Vol 3. New York and London: Verlag Chemie, 1974:1205-15. 2. Kiose S, Stoltz M, Mum E, Portenhauser R. Determination of uric acid on continuous flow (AutoAnalyzer II and SMA) systems with a uricase/phenol/4-aminophenazone color test. Clin Chem 1978;24:250-5. 3. Richmond W. Preparation and properties of a cholesterol oxidasefrom Nocardia sp. and its application to the enzymatic assay of total cholesterol in serum. Clin Chem 1973;19:1350-6. 4. Nagele U, Hagele EO, Sauer G, et al. Reagent for the enzymatic determination of serum total triglycerides with improved lipolytic efficiency. J Clin Chem Clin Biochem 1984;22:165-74. 5. Esders TW, Michrina CA. Purification and properties of La-glycerophosphate oxidase from Streptococcus faecium ATCC

12755. J Biol Chem 1979;254:2710-5. 6. Trinder P. Determination of blood glucose using 4-amino-

phenazone as oxygen acceptor. Ann Clin Biochem 1969;6:24-7. 7. Rodriguez-Castellon JA, Robinson CA, Smith MA, Frye JH. Evaluation ofan automated glucose-oxidase procedure. Clin Chem 1975;21:1513-4. 8. Farrance I. Comparative methodology of glucose [Review]. Clin Biochem Rev 1981;2:7-12. 9. Fossati P, Prencipe L. Serum triglycerides determined colonmetrically with an enzyme that produces hydrogen peroxide. Clin
Chem 1982;28:2077-80.

10. Spayd RW, Bruschi B, Burdik BA, et al. Multilayer film elements for clinical analysis: applications to representative chemical determinations. Clin Chem 1978;24:1343-50.

CLIN. CHEM. 36/2, 329-331 (1990)

Increased Sensitivity in a Radioimmunoassay for Measuring Cyclosporine in Lipoprotein Fractions

LaneJ. Brunner,1 JonathanC. Yau, JulIo Lautersztaln,3 and DavIdR. Luki34
We describe a modification of the commercially available cyclosporine (CSA) 3H-RIA kit, allowing detection of drug in lipoprotein fractions obtained by ultracentrifugation. Sodium bromide solutions containing the lipoprotein fractions were dried and reconstituted with drug-free plasma. Methanolic extractions were centrifuged, and 250 4 of the supernate
1 Department 2Department of

of Pharmaceutics, Hematology and3

University

of Houston,

and

rier Section, The University Center. 4Address correspondence to this author at Department

Immunobiology and Drug Carof Texas M. D. Anderson Cancer of Drug

Nut-

was dried and reconstituted with 50 4 of methanol. Buffer, radioactive tracer, and monoclonal antibody were then added and incubated for 2 h at 4#{176}C. Samples were subjected to charcoal de-activation and then centrifuged, and the radioactivity in the supernate was counted. The limit of detection of CSA was 2.5 /L; analytical recovery was between 97.5% and 101.3%. Within- and between-day CVs were <9%. We conclude that the present method may be beneficial for therapeutic drug monitoring of GSA in lipoprotein fractions.
Owing to the wide variability of cyclosporine (CSA) in pharmacokmnetics, its narrow therapeutic range, and serious toxic effects on the kidney, therapeutic monitoring of CSA concentrations in clinical settings is standard practice

Metabolism, Hoffmann-La Roche, Inc., 340 Kingsland Ave., ley, NJ 07110. Received September 11, 1989; accepted October 30, 1989.

CLINICAL CHEMISTRY, Vol. 36, No. 2, 1990 329