«: AGROCHEMICALS: FATE IN FOOD AND THE ENVIRONMENT PROCEEDINGS OF A SYMPOSIUM, ROME, 7 - 1 1 JUNE 1982 JOINTLY ORGANIZED BY IAEA AND FAO l^J I N T E R ...»
It was recognized at that meeting that the lack o f sufficiently sensitive methods o f drug assay has been a major factor delaying pharmacokinetic studies on existing anti-trypanosomal compounds ; radiotracer techniques are clearly applicable and it was recommended that synthesis o f radioactively labelled drugs should be given high priority. Today, in the pharmaceutical industry, synthesis o f radioactive 'lead compounds' is an accepted step in drug development, but many o f the drugs used for the treatment o f African trypanosomiasis in man and animals were marketed before such procedures were commonplace, consequently most o f the published data on their pharmacokinetics is based on chemical, physical or biological assays. While all these methods have yielded valuable information each has its limitations: small amounts o f drug distributed in tissues are difficult to determine by chemical methods; cell constituents not easily separated f r o m the drug may interfere with fluorometric and spectroscopic measurements; a biological assay may have the advantage o f measuring an activated or modified form o f the drug but may yield spurious results if other biologically active substances (e.g. antibodies) are present. Use o f radioactive drugs does not overcome all these problems but the obvious advantage o f high sensitivity makes this the method o f choice, particularly f o r investigation of transport across membranes.
Many early studies on the pharmacokinetics o f antitrypanosomal drugs were concerned with the ability o f compounds to penetrate the 'blood-brain barrier' in order to assess their value for the treatment o f advanced cases o f sleeping sickness in man, but it is n o w recognized that the African trypanosomes do not just confine themselves to the bloodstream and cerebrospinal fluid. Losos  has emphasized that T. congolense and T. rhodesiense differ considerably in their distribution in the body o f their hosts; T. congolense and T. vivax are generally parasites o f the bloodstream, whereas T. rhodesiense and other members of the brucei group become more widely distributed in the intercellular fluids o f the connective tissue stroma o f various organs. These variations in distribution may account for some o f the observed differences in drug sensitivity between these species. Clearly a better understanding o f the distribution, and the factors affecting the distribution, o f existing antitrypanosomal drugs is highly desirable;
recent progress in this field is reviewed here, with particular reference to the use o f radiotracer techniques in the study o f two important representatives (Fig. 1 ) of the phenanthridine and aromatic diamidine trypanocides used in the control of African trypanosomiasis in animals.
F1C.1. Structure of: (A) ethidium (homidium) bromide (3, 8-diamino-5-ethyl 1.6 phenyl phenanthridinium bromide); (B) berenil (diminazene) aceturate (4, 4 -(diazoamino) benzamidine).
T h e trypanocidal activity o f phenanthridines was first reported by Browning and co-workers  in 1938 and ethidium ( h o m i d i u m ) bromide (3, 8-diamino-6phenyl-5-ethyl phenanthridinium b r o m i d e ) has been used t o treat T. congolense and T. vivax infections in ruminants in Africa since 1948. Field trials [ 4 ] showed that ethidium has limited prophylactic activity and it has been r e c o m m e n d e d t o provide short-term protection to trade cattle en route t o market through tsetseinfested areas [ 1 ]. Mass treatment with ethidium resulted in the 1960s in the appearance o f resistant strains o f T. congolense in both East and West Africa, but Whiteside  reported that ethidium-resistant strains were not resistant to the aromatic diamidine berenil (dirriinazeneaceturate), which had been introduced by Jensch  in 1955, and he p r o p o s e d that either these t w o drugs, or samorin (isometamidium) and berenil, should be used as 'sanative' pairs, i.e. used alternately in the field when resistance to either drug appears.
During the last 25 years much has been learnt about the mechanisms o f action o f phenanthridine and diamidine drugs; ethidium is k n o w n to be a selective inhibitor o f D N A synthesis [ 7 ] combining with the D N A molecule by intercalation into the helix [ 8 ] so causing it t o unwind . Ethidium has been widely used as a biochemical p r o b e [ 1 0 ] and there is n o w a very detailed knowledge o f the molecular basis o f its interaction with D N A . Similar investigations o f the m o d e o f action o f aromatic diamidines have revealed that these drugs also c o m b i n e with 258 NEWTON and GILBERT DNA but in a different manner to phenanthridines [12—14]; diamidines do not intercalate into the DNA helix but are probably situated in the small groove, astride the t w o complementary strands, slightly distorting the helical structure .
In contrast to this detailed knowledge at a molecular level of drug/DNA interactions little is known about the pharmacokinetics of either ethidium or berenil. Results o f the earliest pharmacokinetic studies on ethidium, carried out by the manufacturers (The Boots Co. Ltd), were not published; 14 C-ethidium of very l o w specific activity was used (ca 0.12 juCi/mg) and the results indicated that, in cats, the drug is rapidly excreted (Watkins, personal communication). 1 Kandaswamy and Henderson [ 16] came to a similar conclusion from experiments with mice in which intraperitoneal injection of 14 C-ethidium (15 mg/kg) resulted in 51% being excreted unchanged in the urine in 24 h. Similarly MacGregor and Clarkson , using a chemical assay, found 5 0 - 5 5 % of an intravenous dose ( 15 mg/kg) to rats was recovered in the bile in 16—18 h, about a quarter of the recovered drug being in an unchanged form and the remainder as a monoacetylamino conjugate.
As mentioned above, field studies  have indicated that ethidium may have some prophylactic activity and protection for periods o f up to 20 weeks has been reported . Such protection is surprising if the drug is excreted as rapidly in cattle as it has been reported to be by laboratory animals. There are similarly anomalous results in the literature for berenil; it has frequently been stated that this drug is noteworthy among trypanocides in being rapidly excreted, all o f a parenterally administered dose being cleared through the kidneys within 24 h .
However, Van Hoeve and Cunningham , using a biological assay, detected trypanocidal activity in bovine blood 3 weeks after intramuscular injection of berenil and Cunningham et al. [ 2 0 ] have reported protection of cattle exposed to high infection by fortnightly treatment with this drug. Clearly there is a need for more detailed investigation o f the pharmacokinetics of both ethidium and berenil, preferably in bovines, if we are to explain these anomalies. We have recently used 14 C-ethidium (3,8-diamino-5-ethyl 1.6 14 C-phenyl phenanthridinium bromide, specific actvitiy 45 mCi/g, obtained from Modichem Ltd., UK, courtesy o f The Boots Co., UK) and berenil (bis-phenyl-U- 14 C, specific activity 18.4 mCi/g, the generous gift o f F. Bauer, Hoechst Farbwerke, A G, Frankfurt-am-Main) for such studies. With ethidium initial experiments were carried out in rabbits to establish suitable techniques and the work was then extended to uninfected and calves; preliminary reports o f this work have been T. congolense-infected published [21—23] and a full account is in press . The experiments with berenil have only been performed in rabbits and, due to the limited amount o f labelled drug available, are only at a preliminary stage.
2. MEASUREMENT OF THE DISTRIBUTION OF TRYPANOCIDES INBODY FLUIDS A technique developed by Calnan et al.  involving subcutaneous implantation o f small plastic cages permits serial samples o f blood-free tissue fluid to be collected. Calnan and his collaborators used this method to study the concentrations attained by antibiotics in tissue fluid  and Goodwin and Tierney  used it in conjunction with a microbiological assay to compare the trypanocidal activity o f tissue fluids and blood from drug-treated uninfected and T. brucei-infected rabbits. T w o o f the drugs studies were the cattle trypanocides berenil and samorin (isometamidium) and it was found that trypanocidal activity persisted longer in both plasma and tissue fluids o f drug-treated trypanosomeinfected animals than in uninfected animals. This difference was attributed to the immune response of the infected animals and the experiment serves to illustrate the sort o f problem in interpretation that can rise when a biological assay is used as a basis for pharmacokinetic studies. Nevertheless, these experiments provide the first detailed comparison o f the b l o o d and tissue fluid levels o f anti-trypanosomal drugs and clearly illustrate the value o f this elegant technique devised by Calnan and co-workers . It should be more widely used in pharmacokinetic studies.
3. DISTRIBUTION A N D EXCRETION OF 14 C-ETHIDIUM IN RABBITS
After intramuscular injection o f 1 mg 14 C-ethidium(specific activity
22.5 mCi/g)/kg, radioactivity was measured in blood and tissue fluid. Peak levels occurred less than 1 h after injection and, assuming that all the radioactivity was present as ethidium, corresponded to 180 ng ethidium/ml in blood and 50 ng/ml in tissue fluids. These concentrations fell rapidly to less than 10 ng/ml after 96 h and over this period one-third o f the drug administered was excreted in urine and two-thirds in faeces. Cannulation o f the bile duct showed that biliary excretion o f radioactivity corresponded closely to the amount excreted in the faeces o f uncannulated animals over a similar time period. These results confirm earlier reports o f biliary excretion of phenanthridines [17, 27, 28]. A ten-fold increase in the amount o f drug injected resulted in only a two- to three-fold increase in blood and tissue fluid levels and did not affect the proportion o f drug excreted in urine and faeces. Rabbits dosed with 1 mg/kg were sacrificed after 9 d and the distribution o f radioactivity in tissues was studied; only 2—3% o f the radioactivity injected was present at this time, the highest residues per unit weight o f tissue being present in the liver and kidney.
260 NEWTON and GILBERT
4. DISTRIBUTION A N D EXCRETION OF 14 C-ETHIDIUM IN CALVES
Similar results were obtained in uninfected and T. congolense-infected calves.
Radioactivity in b l o o d and tissue fluids (Fig.2) reached a maximum (equivalent to ca. 1 2 0 - 1 7 0 ng/ml) within 1 h o f injection o f 1 mg 14 C-ethidium/kg. The levels o f radioactivity fell rapidly during the next 24 h and then more slowly during the next 8 d. Within 2 d o f injection 20% o f the administered drug was excreted in urine and 50% in faeces (Fig.3) and after 8 d radioactivity in blood and tissue fluids was equivalent to 15 ng ethidium/ml. Tissue residues (Table I) were determined in animals sacrificed after 10 d; it is estimated that about 4% o f the initial dose was retained at this time and, as in the rabbit experiments, the highest levels were observed in the liver and kidney. These findings are in broad agreement with other studies on the distribution o f phenanthridines [16, 28, 29] in other experimental animals.
Chromatography o f serum collected from animals after intramuscular injection o f 14 C-ethidium has demonstrated that all the radioactivity is in a single substance, which chromatographs in the same position as a 14 C-ethidium marker in the gel and thin-layer systems used. Chromatography o f urine and bile from ethidium-treated animals has shown that up to 46% o f the radioactivity separates from an ethidium marker. T w o excretion products have been detected, which are acid labile and are not m o n o - or diacetyl derivatives o f the drug; they have not yet been identified.
These findings seem to be incompatible with field studies, which indicate that ethidium can provide protection for up to 20 weeks . The very low levels o f ethidium (ca. 10—20 ng/ml) detected in blood and tissue fluids are about two orders o f magnitude less than the trypanocidal level measured in vitro, but this may not be relevant; we have no idea what the in vivo trypanocidal level is. It has been suggested by a number o f workers that host defence mechanisms play an important role in potentiating the action o f anti-trypanosomal drugs and, in the case o f ethidium, Leach et al.  and G o o d w i n and Tierney  hint at this possibility. Our own experiments [ 2 3 ] with T. brucei and T. congolense-infected rabbits indicate that there is a short 'apparent' prophylactic period when infected animals are treated with ethidium and subsequently re-infected; this has been correlated with the host immune response. In view o f these results it is interesting to speculate whether the success o f ethidium as a prophylactic for slaughter cattle is related to the time they b e c o m e infected after administration of the drug, or to whether they have a sub-patent infection at the time the drug is given. Our findings also bring into question the suggestion that failure o f ethidium to protect cattle in the field is always due to drug resistance. Clearly the efficacy of this drug in field cattle should be re-examined and the possibility of slow release formulations explored.
IAEA-SM-263/44 261 FIG.2. Levels of radioactivity in blood f *J and tissue fluid (о o) of calf after intramuscular injection of 1 mg 14C-ethidium bromide (specific activity 22.5 mCi/gj/kg.
Results expressed as ng ethidium/ml, assuming all radioactivity is present as ethidium; each point represents the mean of three determinations. Standard deviations were in the range 3-9% of the values given. (1 Ci = 3.70 X 10iOBq.J
drug/g wet weight tissue, assuming all the radioactivity t o be present as ethidium.
Figures in parentheses represent an estimate o f the drug residue in each tissue (as a percentage o f the total drug injected). Standard deviations are in the range 1 — 10% o f the value shown.)
Tissue fluid levels did not rise as rapidly or t o such high levels (ca. 2 0 0 ng max.) but remained at approximately twice the level detected in b l o o d for up to 7 d.