Journal of Analytical Toxicology,Vol. 20, March/April 1996

Effectsof Sodium Bicarbonate and Sodium Chloride on the Elimination of Etorphine in Equine Urine D.R. Lloyd1, R.J. Rose1,*, A.M. Duffield 2, and C,l. Suann 2 7Departmentof Veterinary Clinical Sciences, University of Sydney, NSW 2006, Australia and 2AustralianJockey Club Laboratory, Randwick, NSW 2031, Australia

Abstract t The combination of large doses of sodium bicarbonate and the potent narcotic, etorphine, has reportedly been given to racehorses in attempts to improve their performance and also to "mask" the presence of etorphine in urine samples. The increased urinary output and pH associated with sodium bicarbonate (approximately 500 g) administration may reduce the urinary concentration of etorphine, making it more difficult to detect. Our experiment was designed to examine the effects of this combination. Six Thoroughbred horses were used in a latin-square design with three horse pairs and three treatments consisting of the following: etorphine (20 pg), etorphine (20 pg) plus sodium bicarbonate (1.0 g/kg), and etorphine (20 pg) plus sodium chloride (0.7 g/kg). Sodium chloride was used to distinguish between the urinary alkalinizing effects of sodium bicarbonate and the diuretic effects associated with the large electrolyte load. Venous blood and urine samples were collected prior to and for 24 h posttreatment. Sodium bicarbonate produced a significant metabolic alkalosis and an increase in urine pH. Both sodium bicarbonate and sodium chloride produced a profound diuresis. After sodium bicarbonate and sodium chloride treatments, the urinary concentration of etorphine, measured by radioimmunoassay (RIA), was reduced and in some cases could not be detected. Extraction of the urine samples, prior to RIA analysis, increased the sensitivity of the assay and in most cases gave a positive result. We conclude that the coadministration of etorphine and sodium bicarbonate or sodium chloride can make the detection of etorphine more difficult because of the dilutional effects associated with the administration of a large electrolyte load.

Introduction In recent years, some athletes have resorted to techniques that make it more difficult to detect drugs in urine samples. An example of this is the use of diuretics to increase urine volume and thus reduce the concentration of drugs in urine. This practice has been recognized by the International Olympic *Author to whom correspondenceshould be addressed.

Committee's Medical Commission, which has banned the use of substances and methods that alter the integrity and validity of urine samples used in doping controls (1). The technique of diluting urine samples to hinder drug detection by the administration of diuretics has also been recognized as a potential problem in the horse racing industry, particularly in those racing jurisdictions that permit the use of the diuretic frusemide (2). Several studies have demonstrated that frusemide decreases the concentrations of various drugs in urine (3-7). A popular routine for racehorse trainers has been the administration of large doses of sodium bicarbonate (approximately 500-600 g), via nasogastric tube, to their horses prior to racing. Often a number of other ingredients are administered in conjunction with sodium bicarbonate, and together this concoction has been termed a "milkshake" (8). This practice has now been prohibited in most racing jurisdictions. Sodium bicarbonate has been administered with the belief that it may improve the horses' performance by delaying the onset of fatigue because of an enhanced blood buffering capacity. We observed that the administration of sodium bicarbonate (0.5 g/kg) can produce diuresis (9). Consequently, it is possible that sodium bicarbonate may be used in a similar way as frusemide to interfere with the detection of other drugs in urine samples (8,10). There are also anecdotal reports that sodium bicarbonate has been used in racehorses to mask the presence of potent opioid drugs, in particular etorphine. When etorphine (4,5-epoxy-3hydroxy-6-methoxy-oc,17-dimethyl-r phinan-7-methanol), also known as "elephant juice", and a sodium bicarbonate based milkshake are administered together, it is known as a "supershake"(11). In the horse, opiate narcotic analgesics depress pain and, in low doses (approximately 50-100 pg per horse), etorphine has stimulatory or excitatory locomotor effects (12,13). In addition, etorphine increases the performance time to exhaustion during treadmill exercise (14), making it a particularly effective drug for stimulating racehorses. The rationale behind the supershake is that etorphine acts on the central nervous system to stimulate the horse, and sodium bicarbonate increases the horse's

Reproduction(photocopying)of editorialcontentof thisjournalis prohibitedwithoutpublisher'spermission.

81

Journal of Analytical Toxicology, Vol. 20, March/April 1996

stamina. In addition, because of the diuretic effect of sodium bicarbonate, etorphine in the urine will be diluted, thereby making detection more difficult. Sodium bicarbonate also increases urinary pH (15), which may influence the rate of excretion of etorphine in the urine. Basic drugs, such as etorphine, are less likely to transfer into an alkaline urine and will preferentially remain in a more acidic solution (2). This effect, combined with the diuresis, may result in etorphine administration going undetected. The current study was designed to examine whether the diuretic and urinary alkalinizing effects resulting from sodium bicarbonate administration (1.0 g/kg) would affect the concentration and detection of etorphine in urine samples over a 24-h period. The effects of a dose of sodium chloride (0.7 g/kg) were also examined to determine if the concentration of etorphine was affected by a substance that increased urine volume but did not increase urine pH. In this way it was possible to distinguish between the urinary alkalinizing effects of sodium bicarbonate and the dilutional effects associated with an increased urine output. We decided to standardize the times of administration of etorphine and sodium bicarbonate or sodium chloride by giving them simultaneously. An alternative would have been to administer etorphine so that the time of its peak urinary concentration coincided with the peak urine pH or diuretic effect. However, because each horse is likely to experience peak urine pH or diuresis at different times, this would introduce another variable into the experiment that would be difficult to control.

Methods Pilot study A pilot study was performed to determine the effects of different dosesof etorphine on the behavior of the horses, as well as the length of time etorphine could be detected in the urine. Anecdotal reports indicated that the doses of etorphine likely to be given are approximately 100 lJg per horse. However, for practical purposes it was necessary to administer a dose of etorphine that would not excite the horses and that would allow urine and blood samples to be collected. Two horses were used, eachwearing urine collection harnesses,and placed in specially constructed stocks. One horse was given an intravenous dose of etorphine (Large Animal Immobilon; C-Vet, Suffolk, England) of 10 IJg, and the second horse was given a doseof 20 IJg. Urine was collected for 72 h after administration, and the concentration of etorphine determined on hydrolyzed urine by radioimmunoassay (RIA) is described. No excitatory effects were observedwith either dose. Becausethe 20-1Jgdose did not excite the horses and was closer to the dose of etorphine reportedly used in racehorses, it was considered the more appropriate dose for the proposed study.

Experimental design Six horses (three pairs) were used in a latin-square design involving three treatments: etorphine (20 I~g),etorphine (20 IJg) plus sodium bicarbonate (1.0 g/kg), and etorphine (20 IJg) plus

82

sodium chloride (0.7 g/kg). Water was freely available at all times. Each horse wore a urine collection harness and was placed in a specially constructed stocks 2 h prior to treatment. Immediately prior to treatment a catheter was inserted into the jugular vein, and venous blood and urine samples were collected to give control samples. The dose of etorphine (20 IJg) was injected via the catheter, which was then flushed with 20 mL of heparinized saline (4 IU/mL). Venous blood and urine samples (when freely voided) were collected every hour up to 15 h postadministration and then at 20, 21, 22, 23, and 24 h postadministration. Blood samples were collected into 2-mL heparinized syringes and immediately placed in a crushed ice slurry for blood gas and acid-base analysis (ABL300;Radiometer, Copenhagen, Denmark) within 1 h of collection. Urine volume and pH (PHM83 Autocal pH meter; Radiometer) were measured immediately after collection. Portions were stored for specific gravity (Reichert refractometer; Cambridge Instruments Inc., Buffalo, NY) and osmolality (vapor pressure osmometer 5500; Wescor Inc., Logan, UT) measurements. An extra 10 mL was stored at-20~ for etorphine analysis (via RIA), and another 150-200 mL of urine from each collection time was stored at 4~ for gas chromatographic-mass spectrometric (GC-MS) analysis. It was not possible to perform statistical analysis on the urine data between treatment groups as the horses urinated infrequently when etorphine only was given, providing inadequate data for statistical comparisons to be made. The correlations between urinary specific gravity and etorphine concentration and also urinary specific gravity and urinary osmolality for the three treatment groups were determined using least-squares regression. The venous blood pH results were analyzed by analysis of variance with time as the repeated measures factor. Where the F values were significant a post-hoc Tukey test was used. All results are reported as the mean plus or minus the standard error of the mean.

Detection of etorphine in the urine RIA detection. The concentration of etorphine was determined using the method described in the Etorphine [l~Sl] Double Antibody Radioimmunoassay kit (Diagnostic Products Corp., Los Angeles, CA). All samples, calibrators, and controls were analyzed in duplicate, and background values for etorphine equivalentswere obtained from the control urine samples. Extraction of etorphine for RIA analysis. The screening of urine samples for etorphine by the method outlined above revealed that there was a remarkable decrease in the urinary etorphine concentrations when either sodium bicarbonate or sodium chloride was coadministered with etorphine (Figure 1). A basic back extraction procedure, rather than direct analysis of urine, was used to measure the etorphine concentrations in 17 of these urine samples and four preadministration urine samples. Urine samples were prepared for RIA screening by the following method: Urine samples (5 mL) were adjusted to pH 5 and hydrolyzed with I drop per milliliter of concentrated glucuronidase-aryl sulfatase (Boehringer Mannheim, Mannheim, Germany) overnight at 37~ Each urine sample was adjusted

Journal of Analytical Toxicology, Vol. 20, March/April 1996

to pH 9.5 + 0.5, dichloromethane (5 mL) was added, and the capped tubes were rotoracked for 10 rain and centrifuged at 2000 rpm for 10 min. The aqueous layer was aspirated to waste, and the dichloromethane was vortex mixed with 1 mL 0.2M hydrochloric acid. After centrifugation at 2000 rpm for 10 min, the acidic layer was transferred to another tube, and the pH was adjusted to 9.5 • 0.5 with 10% ammonium hydroxide. Hexane (3 mL) was added, and the tube was vortex mixed (30 s). The organic phase was dried (anhydrous sodium sulfate), transferred to a Kimble tube, and evaporated at 50~ under N2 gas. The residue was reconstituted in 50 pL of buffered saline (pH 7) and transferred to polypropylene tubes for RIA screening as previously outlined.

Preparation of urine for GC-MS analysis Extraction Enzymatically hydrolyzed urine (10 mL; pH 9.5 • 0.5) was extracted with hexane (three times with 2-mL portions) by rotoracking for 15 rain, and then it was centrifuged at 2000 rpm for 10 rain. The combined hexane extracts were vortex mixed (30 s) twice with 2 mL 0.25M sulfuric acid and centrifuged at 2000 rpm for 10 min. The aqueous phase was vortex mixed (30 s) twice with hexane (2 mL) and centrifuged at 2000 rpm for 10 rain, and the hexane was discarded. The pH of the aqueous fraction was adjusted to 9.5 • 0.5 and extracted with hexane (four times with 2-mL portions) and centrifuged at 2000 rpm for 10 min. The combined hexane extracts were dried (anhydrous sodium sulfate), and the organic phase was dried under N2 gas at 50~

and a 2-pL aliquot was used for positive ion chemical ionization (PCI) GC-MS analysis. Pentafluorobenzoylation. Hydrolyzed urine was extracted and base back extracted as previously described, except that diisopropyl ether was used as the organic solvent. The dry residue was reacted with 2,3,4,5,6-pentafluorobenzoyl chloride (Sigma Chemical Co., St Louis, MO) (10 drops of a 10% solution in ethyl acetate) at 50~ for 30 rain. The reaction mixture was evaporated to dryness under N2 gas at 50~ the residue was reconstituted in dry ethyl acetate (50 tJL), and a 2-pL aliquot was used for negative ion chemical ionization (NCI) GC-MS analysis.

GC-MS analysis A Varian Instruments 3400 gas chromatograph interfaced to a triple stage quadruple mass spectrometer (model 700; Finnigan MAT,San Jose, CA) was used for GC-MS analysis. The GC injector temperature and MS transfer line were both maintained at 275~ and the MS ion source was set at 180~ Samples were injected at an initial GC oven temperature of 100~ (3 min), and the temperature was programmed to 300~ (3 min) at 20~ For NCI, the final temperature was main-

A 4 O

E -5 > t"E

Derivatization of urine extract Acetylation. The dry residue was acetylated with acetic an-

3 2 1 0

hydride (3 drops) in pyridine (3 drops) at 75~ for 45 rain, and the reaction mixture was evaporated to dryness under N2 gas at 50~ The residue was reconstituted in ethyl acetate (50 pL),

I

I

I

I

I

I

I

I

I

1.05

I

I

I

I

)

I

I

i

i

i

B

1.04 >

1.03

5000 O

1.02

r O ffl

4000

E

3500

1.01 1.00 I

I

I

I

I

I

I

I

I

I

3000

e 1200

2500

E

0

9

C

O

2000

g 1000

1500

8OO

1000 (D

500

.m

600

~

4

6

8 10 12 14 16 18 20 22 24

Time postadministration (h) Figure 1. Changes (mean plus or minus standard error of the mean) in urinary etorphine concentrations after the administration of etorphine (20 pg) ( I - t ) , etorphine (20 pg) plus sodium bicarbonate (1.0 g/kg) (O-O), and etorphine (20 IJg) plus sodium chloride (0.7 g/kg) (A-A) (n = 6). Etorphine concentrations were measured by RIA analysis.

I-L

4OO i

I

pre 2

-

i

i

i

i

pre 2

4

6

8 10 12 14 16 18 20 22 24

i

i

i

i

i

Time postadministration (h)

Figure2. Changes(meanplusor minusstandarderrorofthe mean)in (A) urine volume, (g) specificgravity,and (C) urine osmolalityafter the administrationof etorphine(20 pg)(O-e),etorphine(20 pg) plus sodium bicarbonate(1.0g/kg)(O--O),and etorphine(20 pg) plussodiumchloride (O.7~g)(A-A)(n = 6). 83

Journal of Analytical Toxicology, Vol. 20, March/April 1996

tained at 300~ for 10 rain. The GC column was an Ultraphase 1 (Hewlett-Packard,Palo Alto, CA),and helium was the carrier gas (flow rate, 1 mL/min). For PCI, the spectrometer was scanned from m/z 60 to 550 in 0.5 s, and for NCI, the spectrometer was scanned from rn/z 100 to 620 in 0.5 s. Methane was used as the chemical ionization reagent gas (0.12 kPa).

Results Effects of sodium bicarbonate and sodium chloride on fluid and acid-base balance The administration of sodium bicarbonate and sodium chloride produced an increase in the urine volume (Figure 2), which was maintained for approximately 20 h postadministration. The diuretic effectwas pronounced after sodium chloride administration, and there was a peak urine volume of 3.3 • 0.9 L/h at 5 h postadministration. Sodium bicarbonate administration resulted in a lower average urine output than sodium chloride, but was higher than the horses that received only etorphine; at 5 h postadministration, the average value

A

7.4-8

was 1.3 • 0.2 L/h compared with 0.5 • 0.2 L/h for the etorphine-treated horses. There was a marked drop in urine specific gravity and osmolality (Figure 2) after both sodium bicarbonate and sodium chloride administration, but this was more pronounced with sodium chloride, indicating the dilute nature of the urine associated with the excretion of large electrolyte loads. The values returned to normal levelsby 22 h. In the control horses and horses given sodium chloride, there was a drop in urine pH of approximately 0.7-1 pH units over the first 16 h postadministration (Figure 3). Urinary pH then increased and returned to the resting level after 20 h. The administration of sodium bicarbonate caused a rise of approximately 0.6 pH units, which was sustained for approximately 12 h and returned to resting levels by 24 h. Both sodium chloride and sodium bicarbonate administration caused considerable alterations to the horses' acid-base status, as shown by the changes in blood pH (Figure 3). Sodium bicarbonate administration caused an alkalosis, resulting in a significant increase in blood pH from a resting value of 7.382 • 0.006 to a peak value of 7A54 • 0.006 at 7 and 10 h postadministration, whereas sodium chloride caused a mild acidosiswith blood pH decreasing from a resting value of 7.361 • 0.01 to 7.312 • 0.006 at 2 h after treatment. Blood pH had returned to resting levelsby 7 and 20 h in sodium chloride and sodium bicarbonate treated horses, respectively.

7.46 7.4-4-

Table I. Urinary Etorphine (Et) Concentrations (and Etorphine Equivalents) Derived from RIA Analysis Directly on Urine and on the Same Urine Samples after Solvent Extraction and Purification from Treated and Untreated Horses

7.#2 -r 7.4,0 r.~ 7.38 7.36

Urinaryetorphine concentration(pg/mL)

7.3# 7.32

Sample (treatment)

7.30 7.28

i

i

[Dre 2

i

i,

i

i

i

i

i

i

1

i

8.8

,,

8.2 8.0

o

6.6 6.4 pre

l

l

J

~

I

I

I

I

~

~

I

2

4 6 8 10 12 14 16 18 20 22 24 (h) Time postadministration

I

Figure 3. Changes (mean plus or minus standard error of the mean) in (A) venous blood and (8) urine pH after the administration of etorphine (20 IJg) ( I - I ) , etorphine (20 iJg) plus sodium bicarbonate (1.0 g/kg) (O-O), and etorphine (20 pg) plus sodium chloride (0.7 g/kg) (A-A) (n = 6).

84

extractionof etorphine

I

4. 6 8 10 12 1# 16 18 20 22 24. Time postadministration (h)

8.#

directanalysis on urine

Preadministration Preadministration Preadministration Preadministration Et + NaHCO3 (4 h) Et + NaHCO3 (4 h) Et + NaHCO3 (9 h) Et + NaHCO3 (9 h) Et + NaHCO3 (10 h) Et + NaHCO3 (11 h) Et + NaHCO3 (11 h) Et + NaHCO 3 (14 h) Et + NaHCO 3 (20 h) Et + NaHCO3 (22 h) Et + NaC] (3 h) Et + NaCI (4 h) Et + NaCI (10 h) Et + NaCI (11 h) Et + NaCI (20 h) Et + NaCI (20 h) Et + NaCI (20 h)

429.78 215.21 162.53 133.82 594.28 405.25 77.71 105.10 141.47 108.89 77.23 119.26 30.35 75.63 576.64 622.01 63.06 51.45 104.71 96.01 38.03

373.92 196.67 308.20 115.87 81,003 64,643 1603.1 377.50 2284.0 19.49 2448.1 2779.1 1335.7 2886.0 34,431 30,057 482.16 2072.7 2482.9 1950.2 904.44

Journal of Analytical Toxicology, Vol. 20, March/April 1996

RIA detection of etorphine The variations in urinary concentration of etorphine (20 pg) with time are shown in Figure 1. It should be noted that not all the horses urinated at each of the collection times, and therefore, the points on the graphs only indicate the values that were obtained at each time. This was especially the case when the horses were given etorphine only, which may account for the larger standard error bars. In horses that received etorphine only as well as etorphine plus sodium bicarbonate, the peak urinary concentration of etorphine occurred at 2 h postadministration. The peak concentrations that were found at 2 h were 3896 • 779 and 3530 • 513 pg/mL for horses that received etorphine only (n = 3) and etorphine plus sodium bicarbonate (n = 4), respectively. This was followed by a profound decrease in urinary etorphine concentrations in horses that received sodium bicarbonate; the mean concentration was 405 ___86 pg/mL at 5 h. In the etorphine-only treated group, the concentration had decreased to 1312 • 228 pg/mL by 6 h postadministration and slowly declined to a mean concentration of 733 • 141 pg/mL by 24 h. The administration of sodium chloride with etorphine produced an increase in urinary etorphine concentration at 2 h postadministration that was less than that for etorphine plus sodium bicarbonate and the etorphine-only treated groups; the mean concentration was 1730 • 432 pg/mL at 2 h (n = 4). The urine etorphine concentration remained at this level for the next 2 h and then fell to a concentration of 399 • 145 pg/mL at 4 h. By 6 h postadministration the urinary etorphine concentration in the horses that received sodium bicarbonate and sodium chloride was between 20 and 260 pg/mL, and it remained at this level until

IE+086

436 [MH-H20]

~176 1

1133 13:14

m/z454

21 h when it rose by approximately 300 pg/mL. The background concentration of etorphine equivalents in the preadministration urine samples varied between 133 and 429 pg/mL. Therefore, when sodium bicarbonate or sodium chloride was administered with etorphine, a negative result could be returned between 6 and 20 h postadministration, whereas positive results would be obtained for the entire 24-h period in horses that received only etorphine. In some instances during the 6-20-h period, the values were lower than those obtained before etorphine was administered, suggesting that the background was diluted to a lower level in addition to decreasing the etorphine concentration. The increase in the etorphine concentration at 21-24 h in the horses given sodium bicarbonate and sodium chloride coincided with an increase in the urine specific gravity (Figure 2), indicating that there was a return by this time to the normal urine concentrating capacity. Although this seems a likely explanation for these changes, it is not known how the background levels change over time in an untreated horse. However, the urinary specific gravity and etorphine concentration were poorly correlated, and there were correlation coefficient (r) values of 0.351, 0.542, and 0.536 for control, sodium bicarbonate, and sodium chloride treated horses, respectively. There was a good correlation between the urinary specific gravity and the urinary osmolality for all three treatment groups, and the r value was 0.951. Solvent extraction of etorphine from urine samples was used to increase the etorphine concentration prior to RIA analysis,

3J

§

E+05 ' 8.286

100

80

50 60

+ E RIC

475 7:24

~" 100

E+08 1.299

40

454

IMHI+ 20 50

366 258

I 482

164

....

87 121 a.,78 .~ 286:+ + + ._s 2.a~. LL..~. . . . ,

. . . . .

1O0 ....

,+-+-,

200

,

400

,

9

i+---,.

,

.......

, . . . . . .

600 800 1000 Scan number

,

......

1200

,.,

1400

Figure 4. PCI (methane) GC-MS analysisof an extract from equine urine 1 h postadministrationof etorphine (20 pg). Etorphineacetate eluted at a retention time of 13 min and 14 s (scan 1133).

250

- , - - ,

. . . . .

200

- - - - i

. . . .

300 m/z

tl!

'

9

- . - i - - -

400

9

- - -

494 r

5 2 1

r

. . . . .

500

Figure 5. The PCI (methane) mass spectrum of etorphine acetate from a sample identified in Figure 4 at 1 h postadministration of etorphine (20 lag). The protonated molecular ion occurred at m/z454 with the elimination of water providing the base peak at m/z 436.

85

Journal of Analytical Toxicology, Vol. 20, March/April 1996

reducing the background interference. Table I shows the urinary etorphine concentrations derived from RIA analysis directly on urine and on the same urine samples after solvent extraction and purification. The samples chosen for this study were preadministration urine specimens, as well as urine samples that produced negative results or were considered to be borderline between a positive result and background on RIA analysis of urine. A positive screen for etorphine would be indicated if the concentration was greater than 500 pg/mL after extraction. It is evident that in most cases the extraction procedure resulted in the RIA screening of etorphine, although in several instances a negative result was still obtained (Table I). All RIA screening of etorphine in equine urine must be confirmed by GC-MS before etorphine is declared to be present in that urine specimen.

GC-MS detection of etorphine Two urine samples from horses given etorphine were analyzed using GC-MS to confirm that the positive results derived from the RIA analysis were due to the presence of etorphine. One sample was acetylated using acetic anhydride in pyridine and analyzedusing PCI (methane) GC-MS analysis.Etorphine acetate eluted at a retention time of 13 min and 14 s (scan 1133) (Figure 4), and its PCI mass spectrum is shown in Figure 5. The protonated molecular ion [MH § occurred at mlz 454, and the elimination of water provided the basepeakat mlz 436. The second urine sample was derivatized using 2,3,4,5,6pentafluorobenzoylchloride and analyzedusing NCI (methane) GC-MS analysis. Etorphine 2,3,4,5,6-pentafluorobenzoate eluted at 15 rain and 59 s (scan 1431) (Figure 6), and its mass

100

1159 13:33

m/z 605

1431 15:59

spectrum is reproduced in Figure 7; the molecular anion [M-] occurred at m/z 605.

Discussion

The administration of sodium bicarbonate to racehorses has been a controversialpractice in the horse racing industry in recent years. It is apparent that sodium bicarbonate has been widely administered in the Standardbred racing industry (8,15), and authorities in several countries have now prohibited its use, alone or in the form of a milkshake. Most countries have now instigated testing procedures focusing on the measurement of bicarbonate or total carbon dioxide concentrations in venous blood. The administration of sodium bicarbonate (1.0 g/kg) or an equimolar dose of sodium chloride produced changes in the fluid and acid-base balance that affected the concentration of etorphine in equine urine. Sodium chloride administration resulted in a slight metabolic acidosis, whereas sodium bicarbonate treatment resulted in a profound metabolic alkalosis. The decrease in urinary etorphine concentration can be attributed to the diuresis caused by the administration of the large electrolyte loads. Although it has been suggested that the administration of sodium bicarbonate may be able to reduce the rate of etorphine excretion into the urine by increasing the urine pH (8), it seems that the decrease in urinary etorphine concentration is more likelydue to a dilutional effect.Sodium bicarbonate produced an increase in the urinary pH, whereas sodium chloride produced a decrease 100 "

605 E+06 [M-]

E+06 1.150

1,03

80" 50

60

E

100

RIC

E+08 "1.203

167

9

148

n.- 40 9

541

196 523

50 20 .

I

498 ' 203

130

I

I

I-

' .

5OO

. .

, . . . .

1000 Scan number

, . . . .

283

07

4r~

T9

I

3~ 1. s

I

I,

, ....

15oo

Figure 6. NCI (methane) GC-MS analysis of an extract from equine urine I h postadministration of etorphine (20 pg). Etorphine 2,3,4,5,6-pentafluorobenzoate e]uted at 15 min and 59 s (scan 1431).

86

393

238

100

200

300

400

500

600

m/z

Figure 7. NCI (methane) mass spectrum of etorphine 2,3,4,5,6-pentafluorobenzoate from a sample identified in Figure 6 at I h postadminis-

trationofetorphine(20pg).Themolecularanion[M-]occurredat m/z605.

Journal of Analytical Toxicology,Vol. 20, March/April 1996

in urinary pH; yet there was a greater reduction in the urinary etorphine concentration associated with sodium chloride. In addition, there was a greater urine output when the horses received sodium chloride compared with sodium bicarbonate. This finding was surprising, as equimolar doses of sodium chloride and sodium bicarbonate were administered. It seems likely that the increase in urinary volume was responsible for the lower etorphine concentration rather than the effects of a change in urine and blood pH. To overcome the dilution effect of large electrolyte loads upon the excretion of etorphine into urine, an extraction step was used prior to RIAanalysis. This preconcentrated etorphine and effectively increased the sensitivity of the RIA assay, also reducing the levels of background substances. The presence of background levels of substances that register as etorphine equivalents has also been recognized as a problem in immunoassay tests on equine urine (16,17). In all urine samples that had borderline amounts of etorphine (concentrations in the range of 400-600 pg/mL), the extraction procedure gave positive results. In those samples that had low etorphine concentrations and would have been regarded as negative, extraction resulted in positive results using RIA in nearly all instances. There were several cases where the administration of the electrolyte load reduced the urinary concentration of etorphine and etorphine equivalents to levels less than those found in urine samples prior to etorphine administration. Other researchers (17), using RIA analysis, confirmed the results found in the current experiment: hydrolysis and extraction increase the sensitivity of the assay and therefore increase the length of time a single dose of etorphine can be detected in equine urine samples. A previous study that examined the relationship between specific gravity and background etorphine equivalents concluded that there was no correlation between these measurements (18). This study only examined the relationship between specific gravity and the level of background etorphine equivalents in postrace horse urine samples where etorphine had not been administered. It was thought that a relationship between specific gravity and etorphine concentration may be evident after a dose of etorphine and with the production of a very dilute urine. However, in the current study, urinary specificgravity and etorphine concentration were poorly correlated in control, sodium bicarbonate, and sodium chloride treated horses. Despite this poor correlation, it can be observed in Figure 2 that the decrease in specific gravity when sodium bicarbonate and sodium chloride are administered relates to the time in which a decrease in urinary etorphine concentration occurs (Figure 1). The increase in etorphine concentration, which occurs after 21 h, in sodium bicarbonate and sodium chloride treated horses is likely to be due to a return of the renal concentrating capacity after the diuretic effectsof the electrolyte load have passed. Similar trends have also been noted after the diuretic effects of a dose of frusemide have passed with the urinary concentrations of fentanyl, phenylbutazone, and morphine (5,6). Although the coadministration of etorphine and sodium bicarbonate or sodium chloride resulted in a decreased urinary etorphine concentration, other dosing schedules may produce similar or greater decreases in urinary drug concentrations. It

is possible that sodium bicarbonate and sodium chloride could produce a more profound decrease in the urinary etorphine concentrations if the administration of etorphine and the electrolyte solution were timed so that the peak diuresis coincided with the time of peak etorphine excretion, between 1 and 3 h postadministration. A similar approach was used by other researchers (7) when investigating the excretion of several drugs in conjunction with frusemide administration. Another factor to consider is that in the current study the horses were allowed free access to water at all times. If water was withheld for a period of time after the administration of etorphine and sodium bicarbonate or sodium chloride, the reduction in urinary etorphine concentration may be less marked because of less dilution, resulting from a decreased urine output. However, it should be noted that after the administration of sodium bicarbonate (0.5 g/kg) there is still an increase in the volume of urine produced if water is withheld for a period of 6 h posttreatment (9). The current studies were performed in resting horses. Urine samples for detection of etorphine were collected following racing, which could affect urine volume and pH. This may have compounding effects on urinary drug excretion patterns. Further studies investigating the effects of strenuous exercise may be of value. In conclusion, we have demonstrated that the administration of the so-called supershake, consisting of sodium bicarbonate and etorphine, was capable of reducing the urinary concentrations of etorphine. Although in most cases it was still possible to detect the presence of etorphine by solvent extraction of urine prior to RIAanalysis, it seems likely that there is an increased risk of some cases of etorphine administration going undetected when urine is used for screening. However, the dose of etorphine administered to the horses in the present study was lower than that reportedly illegally given prior to racing, and it is possible that the supershake would not be as effective masking the presence of larger doses.

Acknowledgments We would like to thank Claire Ward, Sally Martyr, Pamela Manning, Michael Eaton, and Shirley Ray for their valuable technical assistance. This study was supported by the New South Wales Racing Research Fund, and laboratory facilities were provided by the Racecourse Development Fund.

References 1. Drugsand Sport, 2nd ed. L. Badewitz-Dodd, Ed., MIMS Australia, Sydney, Australia, 1992. 2. T. Tobin. Drugs and the Performance Horse. Charles C. Thomas, Springfield, IL, 1981. 3. B.L. Roberts, J.W. Blake, and T. Tobin. Drug interactions in the horse: effect of furosemide on plasma and urinary levels of phenylbutazone. Res. Commun. Chem. Pathol. Pharmacol. 15: 257-65 (1976).

87

Journal of Analytical Toxicology,Vol. 20, March/April 1996 4. J.R. Miller, B.L. Roberts, J.W. Blake, R.W. Valentine, and 1. Tobin. Drug interactions in the horse. III. Effects of furosemide on urinary concentrations of glucuronide metabolites of pentazocine. Res. Commun. Chem. Pathol. Pharmacol. 17:447-56 (1977). 5. J. Combie, T. Nugent, and T. Tobin. The pharmacology of furosemide in the horse. V. The duration of reduction of urinary concentration of drugs. J. Equine Vet. Sci. 1:203-207 (1981). 6. UR. Soma, K. Korber, T. Anderson, and J. Hopkins. Effects of furosemide on the plasma and urinary concentrations and the excretion of fentanyh model for the study of drug interaction in the horse. Am. J. Vet. Res. 45:1743-49 (1984). 7. A.J. Stevenson, M.R Weber, E Todi, M. Mendonca, J.D. Fenwick, E. Kwong, L. Young, R. Leavitt, R. Nespolo, P. Beaumier, S. Timmings, and S. Kacew. The influence of furosemide on plasma elimination and urinary excretion of drugs in Standardbred horses. J. Vet. Pharmacol. Ther. 13:93-104 (1990). 8. R.J. Rose and D.R. Lloyd. Sodium bicarbonate: more than just a "milkshake"? Equine Vet. J. 24:75-76 (1992). 9. D.R. Lloyd and R.J. Rose. Effects of sodium bicarbonate on fluid, electrolyte and acid-base balance in racehorses. Br. Vet. J. 151: 523-45 (1995). 10. N. Ulecia. Sodium bicarbonate in doping. Irish Vet.J. 33:99-100 (1979). 11. D.R. Lloyd, R.J. Rose, and R Reilly. "Milkshakes". Aust. Equine Vet. 10: 119, 122 (1992). 12. Y. Bonnaire, P. Plou, N. Pages, C. Boudene, and J.M. Jouany. GC/MS confirmatory method for etorphine in horse urine. J. Anal. Toxicol. 13" 193-96 (1989). 13. J. Combie, T. Shults, and T. Tobin. The pharmacokinetics and behavioural effects of fentanyl and other narcotic analgesics in the horse. Proceedings of the Third International Symposium on

88

14.

15.

16.

17.

18.

Equine Medication Control. T. Tobin, J.W. Blake, and W.E. Woods, Eds., International Equine Medication Control Group and the Department of Veterinary Science and College of Agriculture, University of Kentucky, Lexington, KY, 1979, pp 311-20. C.J. Suann, R.J. Rose, C. Plummer, and P.K. Knight. The effects of propranolol, buprenorphine and etorphine on performance during graded and maximum exercise in the Thoroughbred horse. Proceedings of the Eighth International Conference of Racing Analysts and Veterinarians. J.P. Rodgers and T. Toms, Eds., Jockey Club of South Africa, Johannesburg, Republic of South Africa, 1990, pp 215-28. D.R. Lloyd, P. Reilly, and R.J. Rose. The detection and performance effects of sodium bicarbonate in the racehorse. Proceedings of the Ninth International Conference of Racing Analysts and Veterinarians, Vol. 2. C.R. Short, Ed., International Conference of Racing Analysts and Veterinarians, Baton Rouge, LA, 1992, pp 131-38. S. Stanley, A. Jeganathan, T. Wood, R Henry, S. Turner, W.E. Woods, M. Green, H.-H. Tai, D. Watt, J. Blake, and T. Tobin. Morphine and etorphine: XlV. Detection by ELISA in equine urine. J. Anal. Toxicol. 15:305-10 (1991). C.L. Tai, C. Wang, T.J. Weckman, M.A. Popot, W.E. Woods, J.M. Yang, J. Blake, H.-H. Tai, and T. Tobin. Radioimmunoassay for etorphine in horses with a 12Slanalog of etorphine. Am. J. Vet. Res. 49:622-28 (1988). W.E. Woods, T. Weckman, T. Wood, S.L. Chang, J.W. Blake, and T. Tobin. Radioimmunoassay screening for etorphine in racing horses. Res. Commun. Chem. Pathol. Pharmacol. 52:237-49 (1986). Manuscript received December 16, 1994; revision received June 26, 1995.