Forensic Science International. 47 (19901 l- 15 Elsevier Scientific Publishers Ireland Ltd.

1

EVALUATION OF HIGH PERFORMANCE LIQUID CHROMATOGRAPHY (HPLC), ENZYME LINKED IMMUNOSORBENT ASSAY (ELISA) AND PARTICLE CONCENTRATION FLUORESCENCE IMMUNOASSAY (PCFIA) METHODS FOR THE SCREENING, QUANTITATION AND PHARMACOKINETIC STUDY OF FUROSEMIDE IN HORSES

A.K. SINGH, C. McARDLE*, M. ASHRAF. K. GRANLEY, U. MISHRA and B. GORDON**

Minnesota Racing Laboratory, Department of Veterinary Diagnostic Investigation, College of Veterinary Medicine, University of Minnesota, St. Paul MN 55108 IV2XA.I

(Received October 12th. 19891 (Revision received November 27th. 19891 (Accepted February 12th, 19901

Summary

Equine plasma and urine samples were analyzed by using a high-performance liquid chroma- tography (HPLCl, enzyme-linked immunosorbent assay (ELISAl and particle concentration fluorescence assay (PCFIA). Although ELISA and PCFIA were rapid, simple and sensitive for the screening of furosemide, they did not give reproducible quantitative results. The HPLC method, which required relatively longer analysis time, provided simple and reproducible quanti- tative analysis of furosemide in plasma and urine. The performance of the three methods was compared for the quantitation of furosemide in plasma obtained from thoroughbred mares dosed intravenously with furosemide (500 pg/kg (n = 71 and 1.0 mg/kg (n = 51) . Although the plasma furosemide profiles determined by ELISA, PCFIA and HPLC were similar, ELISA and PCFIA methods exhibited considerable variation in values. At high furosemide concentrations, the PCFIA method gave better quantitative values than ELISA. However, at trace furosemide concentrations the PCFIA method gave false positive values which were not confirmed by HPLC or ELISA. The pharmacokinetic values obtained from the HPLC data and the pharmacok- inetic values obtained previously from the gas chromatographic data [4.7] were comparable. The data obtained by ELISA and PCFIA were not suitable for the pharmacokinetic calculations.

Key words: Furosemide; Horses; Comparative analysis; Enzyme-linked immunosorbent assay; High-performance liquid chromatography; Particle concentration fluorescence assay

Introduction

Furosemide is a potent diuretic agent used for the treatment of edema and ascites in humans, and epistaxis in race horses. Although the use of furo- semide for the purpose of controlling bleeding in race horses is allowed in

*Commission Veterinarian Minnesota Racing Commission. **Large Animal Clinical Sciences College of Veterinary Medicine.

0379-0738/90/$03.50 0 1990 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland

2

many racing jurisdictions, there are concerns that its use may mask the detection of other drugs present in horse urine [2]. Several studies have pro- posed that quantitation of furosemide in plasma may be necessary to assure a controlled use of this drug in race horses [3]. Presently, the use of furosem- ide in race horses is controlled by a detention barn system which has signifi- cant disadvantages to horses, race tracks and the regulatory authorities. An alternate method in which the level of furosemide in plasma will be quantitated and compared with an accepted cut-off level has been proposed for the regulation of furosemide use [ll]. The assays presently used for the quantitation of furosemide are gas chromatographic (GCl, high performance Iiquid chromatographic (HPLC) and immunological assays [4-61. As reported previously [‘7], the GC method is laborious and requires derivatization of the samples. Immunological methods such as enzyme linked immunosorbent assay (ELISAl and particle concentration fluorescence immunoassay (PCFIA), are simple to use. However, the quantitation of drugs by using these methods may be difficult [5]. Several investigators have used the fluorimetric-HPLC method for the quantitation of this drug in plasma and urine obtained from humans [8 - lo]. Although the HPLC methods are simple and sensitive, previous studies have suggested that the method may not be suitable for the quantitation of furosemide in horses [7]. Therefore, the objective of this study was to evaluate the fluorimetric HPLC, ELISA and PCFIA methods IS] for the screening and quantitation of furosemide in plasma and urine, and for

Materials and Methods

Materials

pharmacokinetic studies in horses.

Furosemide and bumetanide were purchased from the Veterinary Pharmacy, University of Minnesota. The HPLC consisted of two Beckman model-Gold pumps, Spectrophysics autosampler and Spectrovision FD-300 fluorescence detector. The column used was reverse phase, 3 pm, C8, 4.5 x 150 mm column purchased from the Al Tech Scientific. The ELISA kits for furosemide were obtained from the Tri Tee Corporation, PA and the Interna- tional Diagnostic Systems Corporation, MI. The PCFIA kit was obtained from the International Diagnostic Systems Corp., MI.

Methods

Analysis of plasma by HPLC One milliliter plasma was mixed with 50 ng bumetanide as internal

standard. The sample was acidified with 1.0 ml HCl (0.1 Nl and extracted with 10 ml dichloromethane (DCMl as described previously [6]. The sample was centrifuged at 1500 g for 15 min, the dichloromethane layer was sepa- rated and dried at 50°C under nitrogen. The dried residue was reconstituted in 1.0 ml of acetonitrilel phosphate buffer (pH 3.01 (35:651, and 20 ~1 of the

3

extract were injected into the HPLC. The chromatographic conditions were as follows: composition of the mobile phase, acetonitrilel phosphate buffer (pH 3.0) (35:651; flow rate, 1.0 ml/min, the detector excitation wavelength, 235 nm; and the emission wavelength, 409 nm. Quantitation was done by adding known amounts of furosemide and bumetanide (as the inter- nal standard) in plasma and measuring the area under the peak for each con- centration. The standard/internal standard ratio was used for the calculation of drug concentration in plasma as described previously [6].

Analysis of urine by HPLC Fifty microliters of urine sample was mixed with 1.0 ml of the mobile

phase. Twenty microliters of the mixture was directly injected into the HPLC. The chromatographic conditions for urine were similar to the one described for plasma.

Analysis of plasma by ELBA Fifty microliters of blank plasma, calibrator standard, and plasma samples

were pipetted into individual ELISA plate wells (provided with the kit). To each well, 50 pl of lasix conjugate were added. The contents in the plate were mixed and incubated at room temperature for 60 min. Thereafter, the wells were emptied and washed 3-4 times. After washing, 200 pl of sub- strate (O-phenyldiaminel were added to each tube and the color was meas- ured at 492 mm. For quantitation, 10, 20, 50, 100, 500 or 1000 ng furosemide were added to individual wells and the color was developed as described ear- lier. A standard curve was prepared by plotting log [standard con.] in the ‘X’ axis and absorbance in the ‘Y’ axis by using a program built into the ELISA reader. The amount of furosemide present in each sample was quantitated by using the standard curve.

Analysis of plasma by PCFIA One milliliter of plasma was mixed with one drop of con. HCl and 500 pl of

acid trisolvent (ether/dichloromethane/hexane, l:l:l, 10% isopropanol). The mixture was mixed and centrifuged at 40,000 rev./min for 1 min. The clear supernant was poured into a clean tube and dried at 45% under nitrogen. The dried residue was redissolved in 25 ~1 of methanol. Twenty microliters of serum extract, negative control and positive control samples were added into the PCFIA wells. To each well, 20 pl of drug-phycoerythrin (BPE) conju- gate and 20 pl of drug antisera were added. The mixture was incubated for 15 min at room temperature and placed in the PCFIA equipment. The wells were washed and fluorescence was quantitated at 5451575 nm setting. Quantitation was done by using a standard curve prepared from the positive control values. Each sample was analyzed 3 - 5 times.

Treatment of horses Twelve female, thoroughbred horses, weighing around 500 kg, were

divided into two groups:

4

: .

n

1

TIME bin)

Fig. 1. Chromatogram obtained from 1.0 ml of horse plasma spiked with 50 ng of furosemide and 50 ng bumetanide as the internal standard. The peak at RT 5.69 min is furosemide and that at RT 14.08 is bumetanide.

5

Group 1 (n = 7). Horses of this group received 500 pg furosemideikg body weight by intravenous (i.v.1 injection. Blood samples were collected at 1, 2, 5, 10, 15, 20, 30 and 60 min, and 2, 3, 4, 6, 8 and 10 h after injection. Plasma was separated and stored at - 70 OC for the quantitation of furosemide. At the time of the analysis, plasma was thawed and was analyzed by HPLC, ELISA and PCFIA as described earlier.

Group 2 In = 51. Horses of this group received 1.0 mg furosemide/kg by i.v. injection. Blood samples were collected at various time intervals and the plasma sample described in Group 1.

Pharmacokine tic calculations Pharmacokinetic analysis of the data was performed by applying a one

compartment model, KINA computer program [12]. The parameters calcu- lated were the area under the curve (AUC), elimination rate constant (BETA), total body clearance (TBC), mean residence time (MRT), volume of distribution at steady state (VDsS1, and plasma half life (T1/21.

Results

Analysis of furosemide in plasma by using the HPLC, ELISA or PCFIA methods

This study indicated that the chromatographic conditions used provided a clear separation of furosemide from the endogenous compounds present in horse plasma (Fig. 1). The standard curve obtained by HPLC was Iinear in the concentration ranging from 10 nglml to 1000 nglml and the repeated analysis exhibited good precision (Fig. 21. The standard curves obtained by both PCFIA and ELISA methods were non-linear at the lower (10 nglmll and higher ends of the curve (Figs. 3,41. The recovery curves obtained by using the HPLC, PCFIA and ELISA methods are shown in Fig. 5. For each con- centration, the individual recovery data obtained by HPLC were of rela- tively tighter fit than the individual recovery data obtained by ELISA or PCFIA.

Analysis of furosemide in urine by using the HPLC method This study indicated that the fluoremetric-HPLC method was suitable for

the quantitation of furosemide in urine (Fig. 61. The standard curve for urine was linear in the concentration range of 100 - 1000 ng/ml

Furosemide level in plasma samples obtained from horses dosed with the drug

Analysis by HPLC. In horses receiving 500 pg/kg furosemide, the plasma furosemide level declined from an initial value of 2310 -C 320 nglml (l-2 min after dosing1 to 95 f 20 nglml within 1 h (Fig. 7). The drug was not detectable in plasma after 3 h from dosing. In horses receiving 1.0 mg/kg of furosemide. the drug was detectable for up to 6 h from dosing (Fig. 71.

6

3.a

HPLC

10 100

FUROSEMIDE CONCENTRATION (ng/ml)

Fig. 2. Standard curve for furosemide obtained by using the HPLC procedure. Values are mean of 5 determinations f S.D.

Analysis by ELBA. In horses receiving 500 rg/kg furosemide, the plasma furosemide level declined from an initial value of 6100 f 1810 nglml to approximately 60 k 50 nglml in an hour (Fig. 81. Several high values were obtained by using ELISA which were not confirmed by HPLC or by further ELBA analysis (Fig. 7 (*)I. Furosemide was not detectable in plasma after 3 h from dosing. In horses receiving 1.0 mglkg of furosemide, the drug was detectable for up to 4 h from dosing.

Analysis by PCFIA. The plasma furosemide profile determined by PCFIA (Fig. 91 was similar to the plasma profile determined by HPLC or ELISA. As with the ELISA method, the PCFIA method also exhibited consider- able variation in the furosemide levels. For the 500 pg/kg dose, the furosem- ide level at 1 h after dosing was approximately 80 nglml. Furosemide was detectable for up to 5 h in horses receiving the 500 rg/kg dose (Fig. 91.

ELISA

0.4LI I I 10 100 1000

CONCENTRATION ha/ml)

Fig. 3. Standard curve for furosemide obtained by using ELISA method. Values are mean of 8 determinations +

.6- PCFIA

.l

/ I / I / I I I I I / I, I I, , , 1

0 20 40 60 60 100 120 140 160 160 200 CONCENTRATION btg/mO

Fig. 4. Standard curve for furosemide obtained by using PCFIA method. Values are mean of 10 determinations k S.D.

I (1230*460)

AMOUNT ADDED (no)

Fig. 5. Recovery of furosemide from plasma determined by the HPLC, PCFIA and ELBA meth- ods (n = 5). Values are mean 2 S.D. (0, HPLC, 0, ELISA, A, PCFIA).

Pharrnacokine tics of furosemide This study indicated that the AUC, VDsB, MRT, BETA and T1/2 values for

furosemide in horses receiving 500 pg/kg furosemide were different from the values in horses receiving 1.0 mglkg furosemide (Fig. 101. However, the TBC values calculated from the data obtained from the two groups of horses did not differ significantly (Fig. 101.

Discussion

This study indicated that the fluorimetric HPLC method, described in this study, was simple and sensitive for the screening and absolute quantitation of furosemide in horse plasma and urine. Contrary to the observations of Roberts et al. [4], fluorescent compounds present in horse urine did not interfere with the HPLC assay. Previous studies have shown that the immu- nological methods, such as PCFIA and ELISA, were rapid and sensitive for

9

TIME (mid Fig. 6. Chromatogram obtained from 50 pl of horse urine containing furosemide (100 ng/ml) and bumetanide (50 nglml). The peak at RT 5.69 is furosemide and that at RT 13.44 is bumetanide.

10

?? Mean Fvrosemlde Values-250 mg dose

0 lnaiwdual Furosemlde Values-250 mg dose

A Mean Furosemtde Values-500 mg dose

A lndlvldual Furooemlde Values-500 mg dose

1 2 3 4 5 6 TIME (hrs)

Fig. 7. Plasma levels of furosemide in horses determined by the HPLC method. The solid circles (0) are mean furosemide values and the open circles (0) are individual furosemide values in horses receiving 500 pg furosemide per kg (total dose approx. 250 mg) (n = 7). The solid trian- gles (A) are mean values and the open triangles (A) are individual values in horses receiving 1.0 mg furosemideAg (total dose approx. 500 mg) tn = 4). The vertical line is S.D.

11

ELISA

??Mean-furosemlde Values - 250mQ dose o Individual Furosemido Values - 250mg dose A Mean Furosemide Values - 500mg dose D Individual Furosemlde Values - 500mg dose

Time (Hrs)

Fig. 8. Plasma levels of furosemide in horses determined by the ELBA method. The solid circles fe) are mean furosemide values and the open circles (0) are individual furosemide values in horses receiving 500 pg furosemideRg (n = ‘7). The solid triangle (A) are the values in horses receiving 1.0 mg furosemidentg (n = 4). The vertical line is S.D.

PCF IA

0 Mean Furosemide Value 250mg dose

0 Individual Horse Value8 (Mean Of 5 8nafySe8) 250mg dose

A Mean Furosemide Value 500mg dose

A Individual Horse Value8 (Mean of 5 analyses) 5OOmg dose

TIME (hr)

Fig. 9. Plasma levels of furosemide in horses determined by the PCFIA method. The open circles (0) are mean furosemide values and the solid circles (0 1 are individual horse values (each solid circle is mean of 5 determinations) for horses receiving 500 pg/kg furosemide (n = 5). The open triangle (A) are the mean furosemide values and the solid triangle are individual horse values (each solid triangle is mean of 5 determinations) for horses receiving 1.0 mg/kg furosemide (n = 4). The vertical line is S.D.

13

0.6 1.0 DOSE (ma/k DOSE (ma/kg)

DOSE (mg/kg) DOSE (mgikg)

0.6 1.0 DOSE (mg/ kg)

3 3o

w 2o

g 10

0 0.6 1.0

DOSE (ma/kg)

Fig. 10. Pharmacokinetics of furosemide is horses. AUC: Area under the curve, BETA: Elimina- tion rate constant, TBC: total body clearance, MRT: Mean residence time, VD_: volume of distribution and Tl/Z: half life.

14

the screening of furosemide in plasma [4]. This study indicated that, although the ELISA and PCFIA procedures were simple and rapid, they lacked quan- titative reproducibility. Similar to the results of this study, Leavitt et al. [5] have also suggested that ELISA or PCFIA was not suitable for the absolute quantitation of furosemide in equine plasma.

In this study the performances of HPLC, ELISA and PCFIA methods were evaluated for the quantitation of furosemide levels in plasma samples obtained from horses receiving different doses of furosemide. At 1 h after dosing of the horses with 500 pglkg furosemide, plasma levels of the drug ranged from 20 nglml to 300 nglml when ELISA procedure was used, whereas, the level ranged from 60 to 120 nglml when HPLC method was used. The 1 h values obtained by using HPLC or PCFIA were similar to those obtained by using gas chromatography as described in the study by Chay et al. [7]. However, the 4 h values obtained in this study were different from the 3 h values obtained by Chay et al. [7]. Both the HPLC and ELISA methods failed to detect the presence of furosemide in plasma at 3 h after dosing the horses with 500 pg/kg furosemide. However, furosemide was detectable by both methods for up to 5 h in horses receiving 1.0 mg/kg furosemide. Unlike the HPLC and ELISA, PCFIA gave lo- 50 nglml values for samples collected between 3 and 5 h after dosing the horses with 500 ,ug/ kg furosemide. Therefore, it is proposed that while the ELISA and PCFIA methods may be suitable for screening the positive and negative samples, the HPLC method is suitable for both screening and the absolute quantita- tion of furosemide. Although 6-8 h may be required for the analysis of 30 samples by HPLC, the manpower time required may be only 1.5-2 h for the extraction and loading of the samples to HPLC autoanalyzer. The cost of purchasing an HPLC system may be similar to that of the ELISA equipment or less than for the PCFIA system, and the cost of the HPLC mobile phase may be less than the cost of the ELISA or PCFIA kits. Also, the ELISA and PCFIA analysis will depend upon the supply of the kits by commercial companies, whereas, the HPLC solvents are universally available. However, because of the longer analysis time, HPLC method may not be suitable for the screening of pre-race samples.

In order to compare the performance of the HPLC method described in this study and the GC method described previously [4,7], the pharmacokinetic values calculated in this study by using HPLC data were compared with the pharmacokinetic values reported by Chay et al. [7] and Roberts et al. [4]. Previous studies have shown that furosemide is mainly secreted into the renal tubule and that the pharmacokinetics of this drug in the horse is dependent on its urinary secretion. The observations of this study that the VDBB, T1/2, MRT and BETA values obtained from the two groups of horses were significantly different, whereas, the TBC values obtained from the two groups of horses were similar, suggests the possibil- ity that the distribution of furosemide in tissues also played an important role in the overall clearance of furosemide from plasma. The values obtained

15

by using PCFIA or ELISA were not suitable for studying the pharmacoki- netics of furosemide.

Acknowledgements

This project was funded by a grant from the Minnesota Racing Commission and the Breeder’s Funds.

References

1

2 3

4

5

6

7

8

9

10

11

12

R.C. Sweeney, L.R. Soma, C.A. Bucan and S. Ray, Exercise induced pulmonary hemorrhage in exercising thoroughbreds, preliminary results with pre-exercise medication. Cornell Vet., 74 (1984) 263- 281. T. Tobin, Drugs and the Performance Horses, Charles C. Thomas, Springfield, IL., 1981. T. Tobin, Review of the Pharmacology of Furosemide in the Horse. Proc. 3rd Int. Doping Cont. Rome, Italy, 1977. L.B. Roberts, J.W. Blake and T. Tohin, The pharmacology of furosemide in the horse. II. Its detection, pharmacokinetics, and clearance from urine. J. Equ. Med. Surg., 2 (1978) 185- 194. R.K. Leavitt, V.L. Miehm and M. Beaumier, Immune-analytical techniques for determining drugs in the racehorse. I. ELISA and PCFIA for blood furosemide. Presented at the 41st Meeting of the Association of Official Racing Chemists, New Orleans, LA, March 20, 1989. A.K. Singh, C. McArdle, B. Gordon, M. Ashraf and K. Granley, Simultaneous analysis of furosemide and bumetanide in horse plasma by using a high pressure liquid chromatograph. Biomed Chvomatogr., 3 (1989) 262-265.

S. Chay., W.E. Woods, K. Rowse, T.E. Nugent, J.W. Blake and T. Tobin, The pharmacology of furosemide in the horse. V. Pharmacokinetics and blood levels of furosemide after intravenous administration. Drug Metab. Disp., 11 (1983) 226-231. A.C. Ray, T.D. Tanksley, D.C. LaRue and J.C. Reagor, Analytical evaluation of urinary excretion of furosemide in burros. Am. J. Vet. Res., 45 (1984) 1460-1463. S.E. Swezey, P.J. Meffin and T.F. Blaschke, Measurement of furosemide by high-perfor- mance liquid chromatography. J. Chromat., 174 (1979) 469-473. K. Uchino, S. Isozaki, Y. Suitoh and F. Nakagawa, Quantitative determination of furosemide in plasma, plasma water, urine and ascites fluid by high-performance liquid chromatogra- phy. J. Chromatogr., 308 (1984) 241- 249. A.A. Gabel, T. Tobin, R.S. Ray and G.A. Maylin, Furosemide (Lasix? in horses: A review. J. Equ. Mea!. Sura, 1 (1977) 215-218. R.J. Sawchuck and D.E. Zaske, Pharmacokinetics of dosing regimens which utilize multiple intravenous infusions of gentamicin in burn patients. J. Pharmacokinet. Biopharm., 4 (1976) 183 - 195.