Clinical and Clinicopathological Changes in 6 Healthy Ponies following Intramuscular Administration of Multiple Doses of Imidocarb Dipropionate

INTRODUCTION Equine piroplasmosis is a tick-transmitted , disease caused by the intra-erythrocytic protozoa Babesia equi and Babesia caballi. The geographical distribution of this disease is limited to regions in which the tick vectors occur 14,37,42. To prevent the spread of infection to unex-posed, susceptible equids, endemic countries are subject to rigid exportation restrictions with substantial economic implications for equine industries. The spread of infection to non-endemic regions could be contained if a suitable and safe sterilisation protocol became instituted to eliminate infection in equids destined for exportation. Imidocarb dipropionate [3, 3' bis-(2-imidazolin-2-yl) carbanilide dipropionate] and the less readily available dihydrochloride derivative are aromatic diamidines used as chemotherapeutic agents for the treatment of Babesia equi and the less patho


INTRODUCTION
Equine piroplasmosis is a tick-transmitted, disease caused by the intraerythrocytic protozoa Babesia equi and Babesia caballi.The geographical distribution of this disease is limited to regions in which the tick vectors occur 14,37,42 .To prevent the spread of infection to unexposed, susceptible equids, endemic countries are subject to rigid exportation restrictions with substantial economic implications for equine industries.The spread of infection to non-endemic regions could be contained if a suitable and safe sterilisation protocol became instituted to eliminate infection in equids destined for exportation.Imidocarb dipropionate [3, 3' bis-(2-imidazolin-2-yl) carbanilide dipropionate] and the less readily available dihydrochloride derivative are aromatic diamidines used as chemotherapeutic agents for the treatment of Babesia equi and the less patho-genic Babesia caballi infections in equids 12,19,29,30 .A study performed by Frerichs et al. imidocarb dihydrochloride was used experimentally to sterilise Babesia equi infections in horses and donkeys 19,20 .They showed that only those animals treated with 4 doses of 4 mg/kg imidocarb dihydrochloride at 72-hour dosage intervals were successfully sterilised.Further, animals treated with the same dose and number of treatments, but at different dosage intervals, failed to eliminate the infection, which demonstrated that timing of administration of imidocarb is crucial when attempting to sterilise piroplasmosis.
Dose-dependent nephrotoxic and hepatotoxic effects of imidocarb have been documented in sheep, goats, cattle, dogs and horses 2 .In a study by Adams, increasing levels of imidocarb were correlated with increasing rates of morbidity and mortality attributed to acute renal cortical tubular necrosis and acute periportal hepatic necrosis 2 .Although in the study by Frerichs et al. the anticholinesterase effects of imidocarb, including salivation, restlessness, moderate colic and hypermotility of the gastrointestinal tract as well as local reactions at the injection sites were described, clinicopathological tests addressing the potential nephro-and hepatotoxic effects of the drug were not performed 20 .
The primary aim of this study was to examine the clinical and clinicopathological effects of a multiple dose treatment regime of imidocarb dipropionate in healthy ponies with particular emphasis on evaluation of renal and hepatic function.

Experimental animals
Three healthy Nooitgedacht ponies and 3 part-bred Welsh Mountain ponies aged 2-4 years and weighing 244-294 kg (mean 272.5 kg) were used for the trial.The 3 Nooitgedacht ponies and 2 of the partbred Welsh Mountain ponies were mares.The remaining part-bred Welsh Mountain pony was a gelding.The animals were maintained in a tick-free isolation unit at the Faculty of Veterinary Science, University of Pretoria, Onderstepoort, for the duration of the trial.
The 6 ponies used in the trial were introduced into their new environment approximately 1 week prior to the commencement of treatment.Two of the ponies were stabled individually and the remaining ponies were housed in pairs in large 4 × 5 m 2 box stalls.The stalls were in an insect-proof, artificially ventilated building with continuous air filtration.The ponies were maintained in these stalls for most of the day and at night.They were turned out into a sand enclosure for a short period during the day to allow them to exercise.Diets were standardised and a routine was established to eliminate the potential effects of nutrition and stress on clinicopathological testing.The animals received mixed grass hay and water ad libitum, as well as 500 g of a 12 % protein pelleted ration (Epol Rider Cubes, Epol (Pty) Ltd, Pretoria, South Africa) twice daily.Clean pine shavings were used for bedding.
A complete physical examination and results of complete blood counts and select serum and urine indices established that all of the ponies were healthy at the onset of the trial.

Experimental design
The 6 ponies were given 4 intramuscular injections of 4 mg/kg imidocarb dipropionate [3, 3'-bis-(2-imidazol-2 yl) carbanilide dipropionate] (Forray 65, Schering-Plough, Isando, South Africa) at 72-hour dosage intervals.The drug was administered alternately into the left and right cervical muscles with 2 treatments on each side at different sites.Immediately prior to each treatment the ponies were administered 0.02 mg/kg atropine sulphate (Atropine 10, Centaur Laboratories, Bryanston, South Africa) intravenously, to minimise adverse reactions caused by the anticholinesterase effects of imidocarb.Injection sites were examined daily by the same investigator and local reactions were subjectively classified as mild, moderate, severe or not observed.A reaction was classified as mild if the affected area was less than 5 cm in diameter, required firm palpation to elicit pain or resulted in a barely discernible elevation in skin temperature over the affected area.A reaction was termed moderate if the affected area was between 5 and 10 cm in diameter, if pain was evident with gentle palpation or if heat was easily detected over the affected area.A reaction was classified as severe if the local reaction was larger than 10 cm in diameter, the animal resisted or avoided palpation or heat was evident with the hand raised above, but almost touching, the affected area.Oedema was recorded if these areas were seen to 'pit on pressure' while more persistent swellings were firm and were noted as such.The moderate to severe local reactions observed following the 1st intramuscular injection, prompted the investigators to divide the remaining doses in half and administer each half at a different site.
At the onset of the trial and 24 hours after each treatment on Days 2, 5, 7 and 11 as well as on Days 18 and 38, blood samples were collected from the jugular vein of each animal into a serum and an EDTA tube.At the same time a urine sample was collected via sterile urinary catheterisation.To facilitate urinary catheterisation of the gelding, 15 mg acepromazine maleate (Aceprom, Centaur Laboratories, Bryanston, South Africa) was administered intravenously half an hour prior to collection.All parameters were measured on samples collected on Days 0, 2, 5, 7 and 11 with the exception of bile acids, which owing to financial constraints were only measured on Days 0, 5 and 11.In addition, due to its long half-life serum AST was also determined on Day 38.All samples were stored at 2-8 °C and were analysed within 24 hours of collection.

Analytical techniques
Blood cell counts were measured using an automated cell counter (Cell-Dyn 3500, Abbott Diagnostics, Abbott Park, USA).Blood smear examinations for babesia parasites were done manually using thin blood smears stained with Cam's Quick-Stain (Milsch, Krugersdorp, South Africa).Urinalyses consisted of macroscopic examination of the urine, urine dipstick analysis (Combur 9 Test, Boehringer, Mannheim, Germany), measurement of urine specific gravity and microscopic evaluation of urine sediment.
Serum and urine creatinine analysis was based on the Jaffe alkaline picrate reaction as described by Rossignol using a Technicon RA-1000 analyser 36 (Technicon Instrument Corporation, Tarrytown, USA).Sodium and potassium were measured using a NOVA Biomedical (Nova Biomedical, Waltham, USA) electrolyte analyser 1 with sodium and potassium ion selective electrodes 16 .Serum and urine inorganic phosphate was measured using the phospho-molybdate reaction as described by Amador and Urban using a Technicon RA-1000 analyser 3 .Serum and urine GGT were analysed by the optimised, kinetic method of Szasz using a Technicon RA-1000 analyser 40 .
Total serum protein analysis was based on the biuret reaction of Weichselbaum as described by Skeggs and Hochstrasser using a Technicon RA-1000 analyser 38 .AST analyses was performed using an optimised, kinetic assay based on the Karmen method described by Bergmeyer et al. using a Technicon RA-1000 analyser 5,7 .A Sigma Diagnostics bile acid reagent kit was used to measure serum bile acid concentrations enzymatically using the method described by Mashige et al. on a Technicon RA-1000 analyser 32 .Glucose concentrations were measured using the hexokinase procedure for glucose determination as described by Léon et al. using an automated analyser (Technicon RA-1000) 31 .
Fractional clearances of creatinine, sodium, potassium and phosphate (%) as well as urine GGT : urine creatinine ratios (IU/g) were calculated.The fractional clearance of electrolytes were calculated using the following equation:

Statistical methods
The data in this study were analysed using Sigma Stat 2.0 statistical software (Jandel Scientific Software, San Rafael, USA).Significant changes with time within the treatment group were analysed using 1-way analysis of variance (ANOVA) for repeated measures followed by Dunnett's test to examine deviation from pre-treatment values.Where the data were either not normally distributed or the equal variance test failed, the data were analysed using Friedman repeated measures analysis of variance (ANOVA) on ranks followed by Dunnett's method to examine deviations from pre-treatment values.P < 0.05 was considered significant.Data are reported as mean ± standard error; α = 0.050.

Clinical manifestations and observations
Mild colic, depression and anorexia were seen in some animals after each treatment.The type and duration of these systemic signs are summarised in Table 1.One of the ponies became severely depressed and anorectic following the administration of the 3rd intramuscular injection.This behaviour coincided with a serum glucose and elevated serum creatinine level of 5.4 mmol/ and 195 µmol/ , respectively (reference ranges 3.5-6.3µmol/ ; 76.8-174.5 µmol/ ) 18 .These variables had returned to normal by the following day.The incidence of the systemic reactions following the 1st injection did not differ from the incidence of reactions following subsequent injections.Local reactions were noted at the injection site in all of the animals following the 1st treatment.Distribution of the drug across various sites, as well as the administration of a smaller amount per site, resulted in a marked reduction in the local tissue reactions seen.The characteristics of the local reactions are summarised in Table 2.

Clinicopathological results
The red cell count and haematocrit were significantly lower than their respective pre-treatment values on Days 2, 5 and 18 of the study.No significant changes were observed in the haemoglobin concentration.No significant changes were observed in the total white cell count or in the mature and immature neutrophil counts at any time during the trial.The lymphocyte count was significantly lower than the pre-treatment value on Day 5 while the eosinophil count was significantly reduced on Day 18.No significant changes were observed in monocyte counts.The total serum protein concentration was significantly lower than the pre-treatment values on Day 11 and 18. Serum aspartate aminotransferase activity was significantly elevated above pre-treatment values on Days 2, 5, 8, 11 and 18, but had returned to pre-treatment levels by Day 38 of the trial (Fig. 1).Serum and urine creatinine concentrations did not differ significantly from the pretreatment value at any time during the trial although there appeared to be a trend towards an elevated creatinine concentration and 2 of the 6 treated animals developed mild transient azotaemia (Fig. 2).The serum sodium concentration was significantly lower than the pre-treatment value on Day 5 of the trial.There were no significant differences between the fractional clearance of sodium (reference range 0.032-0.52) 10, the serum potassium concentration, the fractional clearance of potassium (reference ranges 23.3-48.1) 10 , the serum inorganic phosphate or the fractional clearance of phosphate (reference ranges 0-20) 10 at any time during the trial and their respective pre-treatment values.The serum and urine GGT activities and urine GGT : urine creatinine ratios were all significantly elevated above pre-treatment values on Day 5.The serum GGT activity was also elevated on Day 18 of the trial (Fig. 3).The total bile acid concentration was significantly lower than the pre-treatment values on Day 5 of the trial (Fig. 4).No significant changes were observed in urine specific gravity during this trial.In contrast, urine pH was significantly lower than the pre-treatment value on Day 18 of the trial (Table 3).No abnormalities or changes were detected on macro-or microscopic examination of the urine and urine sediment.Glucosuria was present in 2 of the animals following the 3rd intramuscular injection.The concurrent serum glucose and serum creatinine levels in these animals were 5.2 µmol/ and 5.4 µmol/ , and 168 µmol/ and 195 µmol/ , respectively (reference ranges 3.5-6.3µmol/ ; 76.8-174.5 µmol/ ) 18 .

DISCUSSION
The use of enzymuria in screening for renal disease has been actively pursued in human medicine and is considered a non-invasive, sensitive indicator of early renal dysfunction 5,21,24 .Elevations in urine GGT activity are considered to be renal in origin, as the high molecular weight of this enzyme prevents its filtration from the blood by the normal glomerulus 24 .Therefore no correlation exists between increased serum GGT activity and increased urine GGT activity originating from the brush border of the proximal renal tubular epithelium 24 .
In horses, urine GGT : urine creatinine ratios less than 25 IU/g are considered normal, but the small sample sizes on which these estimates are based may not be representative of a larger population and there is some indication that ratios between 25 and 100 IU/g should be interpreted with care 24,35 .In this study, the serial determination of this variable allowed comparison with individual baseline activity with the significant elevation observed on Day 5, confirming renal involvement.However, the rapid return to previous baseline values supported reported observations that changes between 25 and 100 IU/g may be a function of drug excretion and are not necessarily indicative of significant nephrotoxicity 5,21,24,35 .The lack of evidence of significant renal damage may, however, in part be due to the extensive regenerative capacity of the kidney 11 .Consequently, in response to these changes, reduction of dose or frequency of drug administration may prevent manifestation of clinical nephrotoxicity.This conclusion can be supported by Adams' study which showed that increasing levels of imidocarb were correlated with increasing local and systemic reactions, increasing levels of blood urea nitrogen, serum aspartate amino transferase, serum sorbitol dehydrogenase, serum creatine phosphokinase, neutrophilia and increasing severity of renal, hepatic and pulmonary lesions.Mortalities were attributed to acute renal cortical tubular necrosis and acute periportal hepatic necrosis induced by 2 injections of 16 or 32 mg/kg of imidocarb dipropionate 2 .
In concurrence with previous studies, the urine GGT : urine creatinine ratio therefore appears to be one of the earliest practical indicators of renal involvement and may thus have clinical value in the therapeutic monitoring of potentially nephrotoxic substances, including imidocarb derivatives 5 .Serial monitoring of renal function may be particularly significant in diseased animals as it has previously been shown that significant alterations in the disposition kinetics of imidocarb were evident in diseased goats. 1 In these goats, changes in the volume of distribution and drug clearance, related to the pathophysiology and febrile reactions of various experimentally induced disease states, were reported 1 .Altered drug disposition increases the susceptibility of treated animals to the potential toxic effects of various substances 1,39 .
The effect of this treatment regime on liver function was examined by sequential analysis of serum bile acid and serum GGT concentrations.Bile consists of several components of which bile acids make up 90 % of the organic portion 4 .These acids act as detergents, which facilitate the uptake of lipid soluble compounds from the gastrointestinal tract as well as excretion of cholesterol and phospholipids from the liver 4,34 .Bile acids are produced solely by the liver and elevation of total plasma bile acid concentration is considered a highly sensitive and very specific indicator of hepatic disease and provides a measure of hepatobiliary transport function 33,39 .
The majority of bile acids are restricted to the enterohepatic circulation resulting in normally low plasma bile acid concentrations 43 .A study by Hoffman et al. determined reference values for serum bile acid concentrations from 15 fasted horses to be 6.06 ± 2.56 µmol/ , from 5 ponies to be 5.64 ± 3.09 µmol/ , and for all equids, 5.94 ± 2.27 µmol/ 25 .There was no significant difference between the mean bile acids of ponies compared with values for horses 25 .It has been estimated that bile acids are circulated approximately 38 times per day between the gastrointestinal tract and the liver in healthy ponies 25,34 .The age and sex of animals does not influence bile acid concentrations, nor is there a significant diurnal variation in these concentra-tions 43 .There is also no significant difference in results of bile acid concentrations obtained from serum or plasma samples and examination of either blood component is acceptable 43 .An additional analytical advantage is that plasma bile acids remain stable during storage at -20 °C for at least 1 month, enhancing the practicality of this test 43 .
Serum gamma glutamyl transpeptidase occurs as a membrane-bound enzyme on the canalicular surface of hepatocytes and on the biliary epithelial cell membrane.The only other potential source of serum GGT is the pancreas, but pancreatic disease in equids is rare and an elevated serum GGT is more likely to be of hepatic origin 8,14,20,34 .There is no correlation between serum bile acid concentrations and serum GGT activity 34 .In contrast to serum bile acids, which may increase significantly within 24-48 hours of acute hepatic damage, increases in serum GGT are more consistent in chronic disease 8,20,34 .Similarly, fasting for 2 to 4 days results in elevations of serum bile acids to approximately 3 times base-line concentrations, with little or no effect on serum GGT activity 14 .Thus, although not as sensitive as serum bile acids, serum GGT activity may be used in combination with measurement of bile acids and a particular diagnosis is made only if all the tests performed are positive 22 .This method of testing increases the specificity or number of animals without a disease that test negative.The positive predictive value, or probability of a disease being present in an animal that tests positive, is also increased 22 .Normal serum GGT activity may thus rule out suspected hepatic damage in animals with elevated serum bile acid concentrations which have fasted for longer than 3 days 20,25 .The significant increase in serum GGT in conjunction with decrease in serum bile acid concentrations observed on Day 5 may have been due to drug-induced hepatic enzyme induction.Clearly, further studies would be required to elucidate such speculation.Total plasma bile acids have been shown to be the most sensitive indicator of hepatic disease compared with other liver function tests 17,22,33,43 .Owing to the absence of any increase in this variable, it was concluded that this treatment regime had no detectable deleterious effect on hepatic function in healthy ponies.
Fractional clearance of phosphate (FCP) is reportedly a sensitive indicator of early renal dysfunction in horses and elevations in this variable precede increases in fractional clearance of sodium (FCNa) in this species 6,9,15,27,28,41    centrations for either variable suggests that the nephrotoxic effects of imidocarb dipropionate may be limited at this dose and dosage interval.
In horses, elevations above normal, or an increase in serum creatinine within the normal reference range of greater than 26.5 µmol/ in the absence of an alternative underlying cause of renal damage, is considered an indicator of nephrotoxicity 24 .Nephrotoxicity in man has similarly been associated with elevations in serum creatinine concentration greater than 44.0 µmol/ 2 4 .Elevated serum creatinine levels outside the normal reference range were observed in 2 of the ponies on Day 11 of the trial.Changes in concentrations of this variable from pre-treatment values for the group were, however, not significant and supported the observation that the renal effects of this treatment regime were limited, although individual variation may occur.Glucosuria has also been documented as a non-specific indicator of renal disease 5 .
There should be no glucose present in normal urine, although in the presence of serum hyperglycaemia the renal threshold for glucose is exceeded resulting in glucosuria in the absence of underlying renal disease 13 .In this study the glucosuria, in the presence of normal serum glucose levels and elevated serum creatinine levels, was probably a consequence of the drug-induced renal changes.As previously reported, the poor sensitivity and specificity of these 2 variables render them inappropriate as individual spot tests for detection of early renal dysfunction 13,23,24 .
Although in the trial performed by Frerichs et al., imidocarb dihydrochloride was used, the locally available aromatic diamidine, imidocarb dipropionate was substituted for the less readily available dihydrochloride derivative.This may, however, be preferable, because while these 2 compounds have the same active ingredient, the dipropionate salt [3, 3'-bis-(2-imidazol-2 yl) carbanilide dipropionate] is reported to be less irritating 20,29 .All ponies in this study showed uniform moderate to severe local responses at the site of injection following the administration of multiple doses of imidocarb dipropionate.Focal myonecrosis induced by repeated intramuscular injections of imidocarb dipropionate may explain the transient elevation in aspartate aminotransferase (AST) activity seen in this trial.AST is also present in liver and heart tissue and examination of sera for AST enzyme activity in the absence of additional serum biochemistry, is not a specific indicator of tissue source 6 .The plasma half-life of AST in the horse is 2-4 days with persistent eleva-tions indicative of ongoing damage and a continuous release of isoenzymes 6 .The absence of an ongoing myonecrosis in this study could be confirmed by the return of AST levels to within normal limits (reference range: 10-240 U/ ) 18 approximately 2 weeks after administration of the final intramuscular injection.
Although this drug is routinely associated with adverse effects such as salivation, restlessness, moderate colic and hypermotility of the gastro-intestinal tract the results of this trial, in concurrence with previous reports, showed that administration of parasympatholytics such as atropine sulphate significantly reduced these effects 12,29,30 .In addition, varying injection sites and administering smaller amounts of drug per site substantially reduced the mild to moderate local reactions previously reported with this drug 12 .
The haematological results obtained in this study indicated that the dosage regime of imidocarb dipropionate used had minimal effect on cellular blood components.The changes in haematological variables, including the mild leukocytosis and concurrent lymphocytosis observed were considered a normal response for healthy, young and excited animals and often occur when a struggle develops during restraint for sample collection 26 .This physiological leukocytosis occurs in response to adrenaline release during which the marginal pools of neutrophils and/or lymphocytes are mobilised into the general circulation raising the absolute neutrophil and/or lymphocyte counts 26 .However, the increase in white blood cell parameters may also have been due in part to a response to the transient myonecrosis occurring at the injection sites 26 .The decrease in total eosinophils over time was interesting and may have been due to the limited exposure of the animals to insects or allergens while housed in an insect-proof, artificially ventilated building with continuous air filtration.This observation may be supported by a concurrent significant decrease in total serum proteins, which in turn may have resulted from a decrease in antigenically stimulated immunoglobulin production 26 .
In conclusion, mild changes in indicators of renal function observed in this study were indicative of renal involvement.The nephrotoxic properties of imidocarb have been reported in previous studies and although the renal changes were mild and transient, the potentially harmful effects of imidocarb dipropionate should not be disregarded as individual variation may occur 24 .These effects may also be of particular signifi-cance when prior renal or generalised cardiovascular compromise is evident.The severe hepatotoxic effects of imidocarb previously reported were not evident at the dose and dosage interval used in this study.

Fig. 1 :Fig. 2 :
Fig. 1: Temporal changes of mean and standard error of aspartate aminotransferase activity (IU/ ) in 6 ponies following intramuscular administration of 4 mg/kg imidocarb dipropionate on Days 1, 4, 7 and 10. indicates days on which imidocarb was administered.* = mean at time indicated was significantly different (P < 0.05) from the pre-treatment value.

Fig. 3 :
Fig. 3: Temporal variation of mean and standard error of serum gamma glutamyl transpeptidase (GGT) (IU/ ), urine GGT (IU/ ) and the urine GGT : urine creatinine ratios (IU/g) in 6 ponies following intramuscular administration of 4 mg/kg imidocarb dipropionate on Days 1, 4, 7 and 10. indicates days on which imidocarb was administered.* = mean at time indicated was significantly different (P < 0.05) from the pre-treatment value.

Fig. 4 :
Fig. 4: Temporal changes in mean and standard error of serum bile acids ( mol/ ) in 6 ponies following intramuscular administration of 4 mg/kg imidocarb dipropionate on Days 1, 4, 7 and 10. indicates days on which imidocarb was administered.* = mean at time indicated was significantly different (P < 0.05) from the pre-treatment value.

Table 3 : Mean and standard error of the urine specific gravity and urine pH and incidence of glucosuria measured on Days 0, 2, 5, 8, 11 and 18.
*Mean at time indicated was significantly different (P < 0.05) from the pre-treatment value.