in Plasma and Urine - Clinical Chemistry

20 downloads 0 Views 744KB Size Report
trimethoprim, sulfamethoxazole, metronidazole, nystatin, amphotericin. B, flucytosine, ketoconazole, acyclovir, ganci- clovir, and vidarabine. Only metronidazole.
CLIN. CHEM. 34/8, 1565-1568

(1988)

A Sensitive Liquid-Chromatographic (AZI) in Plasma and Urine Moheen

A. Hedaya

Method for Determination

of 3’-Azido-3’-deoxythymidine

and Ronald J. Sawchuk

We

describe a liquid-chromatographic assay for AZT in human plasma and urine. This assay involves the use of two internal standards, allowing reference of AZT peaks to the appropriate internal standard, the choice depending on the range of concentrations encountered. This method is isocratic, specific, sensitive enough to allow quantification of AZT in concentrations observed clinically, and requires only 13 mm of chromatographic time. We saw no interference from various over-the-counter and prescription drugs often used in treating the infectious complications of AIDS. The antiviral agent 3’-azido-3’-deoxythymidine (AZT; BW A509U; Zidovudine; Retrovir#{174}) is a synthetic thymidine analog, first synthesized by Lin et. al (1).’ Because of its inhibitory effect on the human immunodeficiency virus (HIV), AZT was approved by the U.S. Food & Drug Administration for use in treatment of acquired immune deficiency syndrome (AIDs) (2). After administration, ATZ is converted to the triphosphate derivative by cellular enzymes. When this triphosphate form is incorporated into DNA by the H1V DNA polymerase (reverse transcriptase), the azido group at the 3’ position prevents further 5’- to 3’- phosphodiester linkage and terminates the viral DNA synthesis (3,4). The H1V reverse transcriptase is about 100 times more susceptible to inhibition by AF than is the cellular DNA polymerase (3, 4). AZT has been shown to significantly decrease the HIV core antigen concentration in the serum of AIDS patients after 16 to 20 weeks of treatment (5). Also, AZT-treated AIDS patients have shown clinical, immunological (6), and neurological improvement (7). In a double-blind placebo-conrolled trial, AZT administration was found to significantly Lecrease mortality and the frequency of opportunistic infecions in selected groups of patients with &ms, or AIDS-related omplex (8). Adverse reactions associated with AZT adminstration include bone-marrow suppression (9), nausea, myilgia, headache, anemia, and neutropenia. Patients with nore advanced disease were more likely to have hematologc toxic effects (10). Because the benefits in AIDS patients utweigh the toxic effects, AZT is currently used to treat this atient population. Studying the disposition of AZT in AIDS patients and nonitoring AZT in body fluids during therapy may provide nformation that can be helpful in assessing the efficacy and oxicity of AZT. A sensitive, specific, and reliable method of malysis for AZT in plasma and urine is needed for these tudies. An HPLC assay for AZT was briefly described in a

Clinical Pharmacokinetics Laboratory, Department of Pharmautics, College of Pharmacy, University of Minnesota, Minneapo;, MN 55455. ‘Nonstandard abbreviations: AZT, 3’-azido-3’-deoxythymidine; ne, acquired immune deficiency syndrome; HIV, human immunoflciency virus; BHET, f3-hydroxyethyl theophylline; BHPT, (3ydroxypropyl theophylline. Received March 4, 1988; accepted April 12, 1988.

recent publication (11). We describe here in detail a more sensitive procedure currently used in our laboratory to quantify AZT in plasma and urine in concentrations as low as 7.5 and 75 tgfL, respectively.

MaterIals

and Methods

We used a high-pressure liquid chro1084B; Hewlett-Packard, Palo Alto, CA 94303) equipped with a variable-wavelength (190-660 nm) detector, and an automatic sampling system. We used a 15 cm x 4.6 mm (i.d.) Supelcosil reversed-phase column with Instrumentation: matograph (Model

an average particle size of 5 .an (LC-18; Supelco Inc., Bellefonte, PA 16823). The flow rate of the mobile phase, 9 mL of acetonitrile per 91 mL of 10 mmol/L monobasic ammonium phosphate, is 1.5 mLlmin. The temperature is maintained at 30#{176}C. The column effluent is monitored at 266 nm (the X,,. of AZT), and the peak heights are measured with an electronic integrator (Model 3390A; Hewlett-Packard).

Reagents: We used AZT (Burroughs Wellcome Co., ReThangle Park, NC 27709), f3-hydroxyethyl theophylline, (3-hydroxypropyl theophylline (Sigma Chemical Co., St. Louis, MO 53178), acetonitrile, chloroform (Burdick and Jackson Labs., Inc., Muskegon, MI 49442), ammomum phosphate monobasic (All grade), isopropyl alcohol, and methyl alcohol (Mallinckrodt, Inc., St. Louis, MO 63160). All solvents were of “HPLC” grade. Standard solutions of AZT: AZT is dissolved in methanol to prepare the first of two AZT standard solutions at a concentration of 7.5 mgtL. Ten milliliters of this solution is diluted to 100 mL with methanol to prepare the second AZT standard solution, which therefore has a concentration of 0.75 mg/L. Internal-standards solution: f3-Hydroxyethyl theophylline (BHET) and /3-hydroxypropyl theophylline (BHPT) are dissolved in methanol to prepare the internal standards solution containing both, in concentrations of 90 and 3 mg/L, respectively. Extraction solvent: Chloroform is used to dilute 50 mL of isopropyl alcohol to 1000 mL. Sample extraction and chromatography: Use the second AZT standard solution (0.75 mg/L) to add AZT to a series of 12 13-mL ground-glass-stoppered tubes (cat. no. 4100550; Kontes, Evanston, IL 60204) in amounts of 0 (blank), 0.0075, 0.015, 0.03, 0.075, 0.15, and 0.3 g to tubes 1 to 7, and use the first AZT standard solution (7.5 mgfL) to add AZT in amounts of 0.375, 0.75, 1.5,3.0, and 6.0 g to tubes 8 to 12. These solutions are used in preparing the standard curve. Add 20 jL of the internal standards solution (90 mg/ L and 3 mgfL of BHET and BHPT, respectively) to each of these tubes, except the blank. Add the same amounts of the internal standards to another series of 13-mL tubes for use with the samples. Evaporate the methanol from the standard and sample tubes under reduced pressure, with an evaporator (Evapo-Mix; Buchier Instruments, Fort Lee, NJ 07024). Add 1 mL of blank human plasma (or 0.1 mL of blank search

CLINICAL CHEMISTRY,

Vol. 34, No. 8, 1988

1565

urine and dilute to 1 mL with distilled water) to each of the standard-curve tubes, and 1 mL of plasma sample (or 0.1 mL of urine sample and dilute to 1 mL with distilled water) to the appropriate sample tubes. Into all the tubes, pipette 8 mL of the extracting solvent, stopper, and shake horizontally at 180 cycles/mm on a mechanical shaker (Eberbach Corp., Ann Arbor, MI 48106) for 10 mm. Centrifuge for 10 mm at 750 x g. Aspirate and discard the aqueous phase. Transfer as much as possible of the organic phase to a clean 13-mL centrifuge tube, and evaporate the solvent at 60#{176}C with an evaporator. Reconstitute the residue with 75 L of mobile phase, vortex-mix, and transfer the solution into microvials (cat. no. 3-3208; Supelco Inc.) for automatic injection. Inject 40 iL of the reconstituted samples and chromatograph, using the conditions described above. Calculations: The peak-height ratios of AZT/BHPT, for the range of concentrations between 7.5 and 300 zgfL, and of AZT/BHET for the range of concentrations between 0.375 and 6.0 mg/L are calculated. Two standard curves are constructed by simple linear regression of the peak-height ratios for AZT/BHPT vs concentration, and peak-height ratios for AZT/BHET vs concentration. Unknown AZT concentrations are determined by referring AZT peaks to one of the two internal standards, depending on the range of concentrations in the samples, and the AZT concentrations are calculated from the appropriate regression equation.

N

= 0

(A)

(Dl

Resutts Assay

Characteristics

Specificity and reproducibility: Figure 1 shows typical chromatograms obtained for AZT in blank plasma, for blank plasma supplemented with AZT and the internal standards, and for a serum sample obtained after oral AZT dosing. Figure 1 also shows chromatograms for blank serum and blank urine, and for blank urine supplemented with AZT and the internal standards. No interfering peaks were observed in these chromatograms. We examined the chromatographic behavior of some overthe-counter drugs and drugs that are commonly prescribed to treat the infectious complications of sins: acetaminophen, salicylic acid, caffeine, isoniazid, rifampin, streptomycin, trimethoprim, sulfamethoxazole, metronidazole, nystatin, amphotericin B, flucytosine, ketoconazole, acyclovir, ganciclovir, and vidarabine. Only metronidazole interfered, and only with the BHET peak. However, if samples from patients who are receiving metronidazole are to be analyzed, BHFF can be used as the internal standard for the whole range of concentrations. Reproducibility of the retention time for AZT and the two internal standards was calculated for 40 consecutive injections during the analysis of a series of AZT samples. The coefficients of variation were found to be 0.32%, 0.45%, and 0.45% for AZT, BHET, and BHPT, respectively. Sensitivity: We calculated the sensitivity criteria for seven different standard curves, using the method described by Oppenheimer et al. (12). The critical level (the assay response above which an observed response is reliably recognized as detectable) was 1.7 (SD 0.4) gfL. The detection limit (the actual net response which may a priori be expected to lead to detection) was 3.3 (SD 0.9) zg/L. The determination limit (the concentration that can be measured with a coefficient of variation of 10%) was 7.8 (SD 2.1) gfL.

1566

CLINICAL

CHEMISTRY,

Vol. 34, No. 8, 1988

FIg. 1. Representative

chromatograms

for the analysis for AZT in

plasma, serum, and urine (A) blank plasma; ( blank plasma with added AZ and the internal standards (0.1 mg Air per liter); (C) serum sample obtained from a normal volunteer after AZT oral dosIng (61 Air per liter); (C)blank serum; (E) blank urine; (F) blank urine with added AZT and the Internal standards (1 mg Air per liter). Sample volumes: 1 mL of plasma, 1 mL of serum, and 0.1 mL of urine

and precision: Standard curves obtained during for AZT in plasma and urine were linear in the range of concentrations they covered. For example, the equations for standard curves expressing peak-height ratio as a function of concentration of AZT in plasma over the range of 7.5 to 300 ug/L and 0.375 to 6.0 mgfL from the within-run precision analysis were respectively Linearity

the analysis

peak-height peak-height

ratio = ratio

0.0035 =

0.0016

+ 13.83(concn) +

and

0.307(concn)

where “concn” is the concentration (mg/L) of AZT in plasma. Within-run precision was determined by analysis of four different standard curves on the same day. Run-to-run precision was determined from standard curves prepared on each of seven different days during 60 days. The precision of the assay was determined from the variability in the peakheight ratios at each concentration (Table 1). Accuracy: Known amounts of AZT were added to blank plasma and urine to prepare samples of concentrations ranging from 0.015 to 3.0 mg/L and 0.3 to 30.0 mg/L for plasma and urine, respectively. The samples were stored at -20#{176}C until analysis. These samples were analyzed during a period of 60 days and the accuracy of the assay was determined by comparing the nominal concentrations with the measured concentrations (Table 2). These experiments also confirmed the stability of AZT in plasma and urine

Table

1. AnalytIcal

Precision

of AZT Assay

in Plasma

and Urine Urine

Plasma WIthin-run

Concn, mg/I

0.000

0.0075

0.115

0.015

0.205 0.427 1.037

0.150 0.300

=

0.238 0.456 0.904 1.847

3.000

6.000 Slope lAnalyzed

Within-run (n

7)b

±

SD

0.000

CV, % -

Concn, mg/L

0.000

0.075 0.150

0.124

10.3

0.116

±

0.011

9.9

±

0.017

±

0.010

4.6

0.011 0.033

±

0.025 0.050

5.8 4.8

0.300

±

8.3 2.5 3.1

0.217

±

±

0.081

3.9 3.8

1.500

3.1

±

0.004

3.2 3.0 3.4 1.7

±

0.012

2.7

0.024 0.019 0.003

2.7

±

4.174 ± 0.160 13.89 ± 0.469 0.118 ± 0.005 0.235 ± 0.008

0.461 0.950 1.899

±

0.020 0.028 0.040 0.007

3.4 3.9 3.4 4.3

mean

0.000

0.012

0.430 1.047 2.091

Peek-height

0.750

2.123 4.193

3.000 Slope 3.75 7.50

±

2. Accuracy samples

Measured

=

2.9

of AZT Quality-Control

0.013

±

0.003

±

0.013

±

0.025

±

0.046 0.032 0.012 0.002 0.003 0.022 0.021 0.031

±

0.004

± ± ± ± ±

1.835 0.306

4)

Peak-height

SD

±

±

0.454 0.928

Plasma

of

Nominal

concn,

nominal concn

CV, %

mg/L

0.222

±

0.016

0.422

± 0.014

1.046

±

2.098

±

0.7 0.8

4.154

±

9.6

0.30

0.304

±

0.006

100

7.6

1.50

±

0.018

97

6.0

0.064 0.143

94 96

9.3

7.50 30.0

1.497 7.305

27.87

± ± ±

0.474

±

0.940

±

1.911

±

0.032

±

CV, % 13.4 7.3 3.3 2.7 3.6 3.4 3.1 2.8 2.2

0.029 0.075 0.139 0.0.43 0.003

0.005 0.016 0.026 0.047 0.0008

3.4 2.8 2.5 2.5

Samples 6)

=

concn,

mean ± SD, mg/I

105

±

1.5 3.1 2.4 2.1

and Urine

0.0015

SD

0.017

1.384 0.120 0.238

=

ratio, ±

±

2.1 1.1 4.7 2.3 1.7 1.4

Measured Percent

(

0.000 0.123

10.5

±

±

mean

CV, %

UrIne samples (n

concn,

mg/L

2.870

in the Analysis (n

± SD,

mean

0.0157 0.075 0.290 0.703

±

Run-to-run

ratio, ±

±

1.398 0.116 0.235

15.0

=‘r

-

0.210 0.424 1.046

30.5 60.0 1.0 ± 2.1 ± 2.2 Slope 0.307 ± 1.0 0.317 ± on the same day. bMalyzed on seven different days during 60 days.

concn, mg/I

0.300 0.750 3.000

=

±

Plasma

0.015 0.075

mean

-

Table

Nominal

(n

ratio,

Peak-height CV, %

2.073 ± 0.064 4.157 ± 0.137 13.83 ± 0.413 0.113 ± 0.004

Slope 0.375 0.750 1.500

l

Run-to-run

4)1

Peak-height ratio, mean ± SD

0.000

0.030 0.075

(n

Percent

nominal

±

0.016

101

±

0.060

100 97 93

± ±

0.125 1.216

of

concn

CV, %

5.1 3.8 1.5

4.5

5.0

Measured during 60 days.

stored at -20 #{176}C for 60 days, because the measured AZT concentrations in either plasma or urine did not show a decreasing trend over this period. We participated in a quality-control study undertaken by the developer of AZT (Burroughs Weilcome Co.). Five different unknown serum samples, prepared by an independent group, were analyzed by the present method. The measured concentrations were from 91% to 101% of the nominal concentrations of the quality-control samples, which ranged from 72 pg/L to 3.50 mg/L (13). Extraction efficiency: Here we compared the peak-height ratios measured for the extracted plasma samples containing three different concentrations with the peak-height ratios for unextracted samples. The internal standard was added to the samples just before injection into the chromatograph. AZT accounted for was independent of concentration in the range of concentrations covered by the standard curves. The mean proportion of the sample accounted for, without correction for the loss of organic phase during aspiration of the supernatant aqueous phase and subsequent transfer of the organic phase, was 69.4%. Analysis

for AZT in Serum

from Normal

Volunteers

The present method was used to measure AZT concentrations in serum sampled over a 4-h period from two normal volunteers who received a single 200-mg oral dose of AZT. The serum concentration-time profiles of AZT in the two subjects are shown in Figure 2. The lowest AZT concentration measured was 15 g/L (the first sample from subject 2), which is twice the lowest concentration on our standard curve. The highest measured concentration was 0.72 mg/L.

Discussion This suring

is a simple, sensitive, and specific method for meaAZT in plasma and urine over the range of concentrations observed clinically during therapy with AZT. Although there was no difference in quantification on using peak-area ratios at the high end of the standard curves, accuracy and precision were improved when we used peakheight rather than peak-area ratios at the low end of the standard curves. The use of two internal standards allows us to quantify AZT accurately in concentrations corresponding to the entire range of the standard curves. Although the method was initially developed to measure

I TIME

FIg.

2.

volunteers

Serum after

concentration-time a single oral 200-mg

CLINICAL

CHEMISTRY,

(mu)

profile

for

Air

in

two

normal

dose of Air Vol. 34, No. 8, 1988

1567

AZT in plasma and urine, subsequent application of the technique for assay of AZT in human serum demonstrated that there are no interfering peaks in serum. This can be observed by comparing the chromatograms for blank serum, AZT-supplemented serum, and serum from subjects receiving AZT orally in Figure 1. Because of the special precautions required for handling body-fluid samples from ms patients, we did not measure AZT concentrations in urine or plasma samples obtained from AIDS patients. However, some drugs that are commonly prescribed for AIDS patients are shown not to interfere in the analysis, except for metronidazole, which precluded the use of one of the two internal standards. We thank Dr. Henry Balfour and Barbara Chinnock of Clinical Virology Research, University of Minnesota, for their technical assistance, and for providing serum samples from normal volunteers receiving AZT. We also acknowledge the assistance of Dr. Steven Good (Burroughs Weilcome Co.) for providing us with AZT and for coordinating our involvement in the analytical qualitycontrol

program.

References 1. Un I’S, Fischer agents: 1. synthesis

methylnitrosourea

PH, Shiau GT, Prusoff WH. Antineoplastic and antineoplastic activities of chioroethyland analogues of thymidine. J Med Chem

1978;21:130-3.

2. Brook I. Approval of Zidovudine (AZT) for acquired immunodeficiency syndrome [Commentary]. J Am Med Assoc 1987258:1517. 3. Mitauya H, Weinhold KJ, Furman PA, et al. 3’-Azido-3’-deoxythymidine (BW A509U): an antiviral agent that inhibits the infectivity and cytopathic effect of human T-lymphotropic virus type Illhlymphadenopathy-associated virus in vith,. Proc Natl Acad Sci USA 1985;82:7096-100.

1568

CLINICAL

CHEMISTRY,

Vol. 34, No. 8, 1988

4. Furman PA, Fyfe JA, St. Clair MH, et al. Phosphorylation of 3’azido-3’-deoxythymidine and selective interaction of the 5’-triphos-

phate with human immunodeficiency virus reverse transcriptase. Proc Natl Acad Sci USA 1986;83:8333-7. 5. Chaisson RE, Allain JP, Volberding PA. Significant changes in HIV antigen level in the serum of patients treated with azidothymidine. N Engi J Med 1986;315:1610-1. 6. Yarchoan R, Klecker RW, Weinhold KJ, et al. Administration of 3’-azido-3’-deoxythymidine, an inhibitor of HTLV-llh/LAV replication, to patients with AIDS or AIDS-related complex. Lancet 1986;i:575-80. 7. Yarchoan R, Berg G, Brouwers P, et al. Response of human immunodeficiency

virus

associated

neurological

disease

to 3’-azido-

3’-deoxythymidine.

Lancet 1987;i:132-5. 8. Fischl MA, Richman DD, Grieco MB, et al. The efficacy of azidothymidine (AZT) in the treatment of patients with AIDS and AIDS-related complex. N Engi J Med 1987;317:185-91. 9. Gill PS, Rarick M, Brynes RK, Causey D, Loureiro C, Levine A. Azidothymidine associated with bone marrow failure in the acquired immunodeficiency syndrome (AIDS). Ann Intern Med 1987;107:502-5. 10. Richman DD, Fischl MA, Grieco MB, et al. The toxicity of azidothymidine (AZT) in the treatment of patients with AIDS and Ama-related complex. N Engl J Med 1987;317:192-7. 11. Klecker RW, Collins JM, Yarchoan R, et al. Plasma and cerebrospinal fluid pharmacokinetics of 3’-azido-3’-deoxythymidine: a novel pyrimidine analog with potential application for the treatment of AIDS and related diseases. Clin Pharmacol Ther 1987;41:407-12. 12. Oppenheimer mining the lowest

L, Capizzi TP, Weppelman RM, Mehta H. Deterlimit of reliable assay measurement. Anal Chem

1983;55:638-43. 13. Personal communication with Dr. Steven S. Good, Department of Experimental Therapy, Burroughs Wellcome Co. Research Laboratories, Research Thangle Park, NC 27709.