WHO FOOD ADDITIVES SERIES: 49
First draft prepared by Dr Pamela L. Chamberlain
Center for Veterinary Medicine
Food and Drug Administration, Rockville, Maryland, USA
Special studies on the genetic basis for sensitivity to the toxicity of avermectins
Relative sensitivities of mice, rats, rabbits, dogs and non-human primates to the toxicity of avermectins
Polymorphisms in the human MDR-1 gene coding for P-glycoprotein
Doramectin is a member of the avermectin class of compounds, which includes abamectin and ivermectin. It is a semisynthetic avermectin that has close structural similarity to abamectin and ivermectin. It is used as an endoparasitic agent in non-lactating cattle.
Doramectin was previously evaluated by the Committee at its forty-fifth meeting (Annex 1, reference 119), when it established an ADI of 0–0.5 µg/kg bw on the basis of a NOEL of 0.1 mg/kg bw per day for mydriasis in a 3-month study in dogs treated by gavage and using a safety factor of 200. An additional safety factor of 2 was applied because doramectin was not tested in CF-1 mice, which is the test animal most sensitive to the neurotoxic effects of this family of drugs. In 1997, the Joint FAO/WHO Meeting on Pesticide Residues (JMPR) concluded that the sensitivity to avermectins of CF-1 mice was due to a genetic variation that causes reduced expression of P-glycoprotein in the blood–brain barrier (FAO/WHO, 1998). The JMPR further concluded that the results of studies with CF-1 mice were not appropriate for establishing ADIs for avermectins.
P-glycoprotein was expressed in the brain and jejunum of all species studied. P-glycoprotein is a cell membrane protein that acts to remove a wide variety of lipophilic compounds from cells, including avermectins. In the capillary endothelium of the central nervous system, it serves as a functional component of the blood–brain barrier. In intestinal epithelium, P-glycoprotein can limit intestinal absorption of a range of compounds.
The Committee at its fiftieth meeting (Annex 1, reference 134) accepted the conclusions of the JMPR and considered that it was no longer necessary to apply an additional safety factor of 2 for avermectins and milbemycins that had not been tested in CF-1 mice. Doramectin was re-evaluated by the Committee at its present meeting in order to determine whether removal of the additional safety factor of 2 was appropriate. On the basis of the Committee’s decision taken at its fiftieth meeting, the present Committee concluded that use of an additional safety factor of 2 in establishing the ADI for doramectin was no longer necessary.
No new data were provided to the Committee. The literature was reviewed for published information on the toxicity of avermectins considered relevant to this evaluation. The Committee reviewed information on the mechanism of the toxicity of ivermectin in a subpopulation of collie dogs and observations of its toxicity in a subpopulation of Murray Grey cattle. The Committee also considered a published review of the relative sensitivities of mice, rats, rabbits, dogs and non-human primates to avermectins. The relative potencies of doramectin, ivermectin and abamectin were also considered. The Committee examined information about variants of the human gene that codes for P-glycoprotein and reviewed observations in humans.
(a) Collie dogs
The genetic basis for the sensitivity of collies to avermectins was studied in 13 clinically normal collies, previously identified as being sensitive or insensitive to ivermectin. Seven animals were identified as sensitive after displaying typical clinical signs of neurotoxicity, including depression, ataxia, mydriasis, salivation or drooling, after receiving a single oral dose of 120 µg/kg bw. The objective of the study was to determine whether altered gene expression of P-glycoprotein or a polymorphism of the canine Mdr1 gene that codes for P-glycoprotein exists in avermectin-sensitive collies. The sensitivity of the CF-1 mouse strain to the neurotoxic effects of avermectin has been traced to a polymorphism of the murine Mdr 1 gene resulting in decreased expression of P-glycoprotein (Umbenhauer et al., 1997).
The level of Mdr1 expression was similar in sensitive and insensitive collies, as determined by semi-quantitative reverse transcriptase polymerase chain reaction (RT-PCR) analysis. Sequence analysis of canine Mdr1 by RT-PCR was conducted on RNA isolated from blood leukocytes obtained from the sensitive and insensitive collies and also from other breeds (one beagle, two golden retrievers and one Staffordshire terrier cross-bred dog). Sequence analysis of clones from three ivermectin-sensitive collies revealed an identical four-base pair deletion in the first 10% of the transcript. This deletion causes a frame-shift mutation resulting in the production of a truncated, non-functional protein. The same four-nucleotide deletion was detected in all samples from ivermectin-sensitive collies, which were also homozygous for the deletion. Insensitive collies had a heterozygous genotype, with one mutant allele and one wild-type allele. Blood samples from all the other breeds showed homozygosity for the wild-type. The investigators concluded that their study provided evidence that the sensitivity of collies to ivermectin results from a frame-shift deletion of four base pairs in the canine Mdr1 gene (Mealey et al., 2001).
(b) Observations in Murray Grey cattle
Murray Grey cattle on one farm in the central tablelands of New South Wales, Australia, were reported to be sensitive to the toxicity of avermectin B1. The first cases were noted in October 1985, in 50 Murray Grey heifers aged 18–26 months treated with avermectin B1 at an estimated dose of 175–200 µg/kg bw by injection. Three of the heifers died within 2 days of treatment.
Two weeks later, 144 Murray Grey cattle aged 4–18 months and weighing 200–450 kg were treated with avermectin B1 at an estimated dose of 120–200 µg/kg bw by injection. The numbers of males and females treated were not stated. Within 48 h of treatment, three steers weighing 400–450 kg developed severe neurological signs, and all three were slaughtered for necropsy. A fourth steer in this group showed mild neurological signs and was slaughtered 19 days after treatment.
A field trial was conducted on this farm with 208 Murray Grey cattle, comprising 90 steers that had been treated with avermectin B1 1–2 months earlier and a second group of 118 cattle that had not been treated previously. The animals were weighed and treated with the recommended therapeutic dose (200 mg/kg bw) or injected with the vehicle only. One steer in the group that had not been treated previously developed neurological signs 42 h after treatment and was slaughtered for necropsy.
Brain, spinal cord, liver, kidney, lung, heart, spleen, intestines, skeletal muscles, adrenals, lymph nodes and peripheral nerves from the four initial cases, the case found in the field trial and one normal treated animal were examined macroscopically and histologically. No pathological changes were found that would explain the severe neurological syndromes observed. The concentrations of avermectin B1 in plasma and/or serum, liver, brain and spinal cord from the five clinically affected animals and the normal animal were assayed by high-performance liquid chromatography with fluorescence detection. The average concentration of avermectin B1a was 56 µg/ml in brain tissue from affected animals and 4 µg/ml in brain tissue from the normal animal.
Two additional field trials were undertaken with Murray Grey cattle in other areas of New South Wales and in Victoria. A total of 83 cattle were treated with at least twice the normal therapeutic dose of avermectin B1. No adverse reactions occurred. The authors stated that no other incidents of toxic effects of avermectins have been reported in Murray Grey cattle. They also noted that the farm on which the adverse reactions were seen had maintained a virtually closed herd for approximately 15 years (Seaman et al., 1987).
The relative sensitivities of the central nervous system in mice, rats, rabbits, dogs and non-human primates to avermectins have been reviewed (Lankas & Gordon, 1989). The studies conducted with ivermectin addressed acute toxicity in mice, rats, dogs and rhesus monkeys treated orally; short-term studies of toxicity in rats, dogs and rhesus monkeys; and studies of developmental and reproductive toxicity in mice, rats and rabbits. The studies with abamectin given by oral administration comprised long-term studies of toxicity and carcinogenicity in mice and rats, a 1-year study of toxicity in dogs and a study in rhesus monkeys given single doses.
The authors concluded that clear species differences exist in the sensitivity of the central nervous system to the toxicity of avermectin, rodents being the most sensitive. A dose of 0.2 mg/kg bw in mice and slightly higher doses in rats resulted in clinical signs of central nervous system toxicity, comprising tremors and ataxia, while these doses caused no adverse effects in rabbits, dogs or rhesus monkeys.
The authors described a study of acute toxicity in which groups of two male and two female rhesus monkeys were given abamectin or ivermectin at single oral doses of 0.2, 0.5, 1, 2, 8, 12 or 24 mg/kg bw. The time between administration of the next higher dose to the same group of monkeys was 2–3 weeks. The authors noted that the minimum single oral dose of ivermectin or abamectin that was toxic (2 mg/kg bw) was approximately 10-fold greater than the human clinical dose of ivermectin. Emesis was the only toxic effect observed in rhesus monkeys after an oral dose of ivermectin of 2 or 8 mg/kg bw; the clinical signs of toxicity observed after a dose of 24 mg/kg bw were emesis, mydriasis and sedation. The authors compared the effect at 8 mg/kg bw with effects seen in a child (age and sex not stated) after the apparently accidental ingestion of approximately 8 mg/kg bw. The toxic effects observed in the child were emesis, mydriasis and sedation. In view of the similarity of the toxic effects observed in rhesus monkeys and the child, the authors proposed that rhesus monkeys are an appropriate model for predicting the acute toxic effects of ivermectin in humans. The NOEL for acute effects after administration of abamectin or ivermectin to rhesus monkeys was 1 mg/kg bw.
The review of the developmental and reproductive toxicity of ivermectin included the results of studies conducted in mice, rats and rabbits. The reported NOELs for maternal toxicity in mice, rats and rabbits were 0.1, 5 and 3 mg/kg bw per day, respectively. The reported NOELs for developmental toxicity in mice and rabbits were 0.2 and 1.5 mg/kg bw per day, respectively. A NOEL of 0.2 mg/kg bw per day was reported for neonatal and developmental toxicity in a multigeneration study in rats (Lankas & Gordon, 1989).
The Committee evaluated the relative potencies of doramectin, ivermectin and abamectin by comparing the NOEL values reported in evaluations by the Committee at its firty-fifth meeting, the JMPR in 1997 and Lankas and Gordon (1989). The studies of reproductive and developmental toxicity in rats and rabbits and the 90-day studies of toxicity in dogs, all treated orally, were the only ones in which all three compounds could be compared. On the basis of these data, the Committee concluded that the potency of these compounds was similar.
In a study of 461 white volunteers in Germany (294 men and 167 women aged 18–65), DNA samples were analysed for polymorphisms in the MDR1 gene by polymerase chain reaction–restriction fragment length polymorphism assays. Eleven polymorphisms of MDR1 were identified in this population. The author stated that 16 polymorphisms were identified. The polymorphism identified in exon 26 of the MDR1 gene is associated with a specific alteration in drug transport. Volunteers with this polymorphism showed enhanced uptake of an oral dose of digoxin and a steady-state concentration that was 38% higher than that in volunteers without this polymorphism. The difference was statistically significant. Quantitative immunohistochemistry and western blot analysis were used to analyse the level of expression of P-glycoprotein in the duodenum of the volunteers. A correlation was found between decreased expression of P-glycoprotein and the polymorphism in exon 26. Thus, both the expression level and the functionality of P-glycoprotein are affected by this polymorphism. Controlled clinical trials and in-vitro studies are needed to determine the clinical relevance of other polymorphisms of the MDR1 gene. Data on the distribution of MDR1 polymorphisms in populations of other ethnic origins are not available (Hoffmeyer et al., 2000; Cascorbi et al., 2001).
Ivermectin has been administered to several million patients in Africa and Latin America for the treatment of onchocerciasis. A WHO report (WHO, 2001) stated that in 2000 alone, over 23 million people were treated. The reported adverse reactions in treated patients are allergic or inflammatory responses resulting from the killing of microfilarariae, referred to as the ‘Mazotti reaction’ (Pacqué et al., 1989). No cases of acute central nervous system toxicity have been reported after treatment of patients for onchocericasis.
When ivermectin was first distributed in 1987 to selected populations for the treatment of onchocerciasis, the drug sponsor contraindicated its use in pregnant women, mothers who were breastfeeding infants under 3 months of age, children under 12 years of age and people with active disease of the central nervous system, such as meningitis or epilepsy. These contraindications were listed as it was not known with certainty whether the blood–brain barrier of neonates and those with central nervous system disease would prevent entry of ivermectin. It was also unknown whether the placenta could prevent transfer of ivermectin to the fetus. The sponsor’s evaluation of information that had become available since 1987 caused them to change the original list of contraindications. The new information includes elucidation of the genetic basis for the sensitivity of the CF-1 mouse, identification of P-glycoprotein in the human placenta and in the human fetus from week 28 of gestation (Cordon-Cardo et al., 1989), the finding of no significant risk to the fetuses of pregnant women who had been inadvertently treated with ivermectin (Pacqué et al., 1990) and the demonstration that ivermectin is safe for patients with epilepsy (Kipp et al., 1992). As a result, use of ivermectin in pregnant or epileptic patients is now permitted, and treated mothers can breastfeed infants as young as 1 week of age (Brown, 1998). The usual dose administered to humans for the treatment of onchocerciasis is 150 µg/kg bw once every 12 months (Anon, 1999).
The pharmacokinetics of orally administered ivermectin was studied in 12 healthy male (race not stated) volunteers aged 18–50 years. A single therapeutic dose of 12 mg (150–200 µg/kg bw) as a tablet resulted in an average maximal time to peak plasma concentration of 3.6 h, an average maximal plasma concentration of 46 ng/ml and an average area under the plasma concentration–time curve of 880 ng/h per ml. No clinical adverse effects were reported (Edwards et al., 1988).
The genetic basis for the sensitivity of collie dogs to the neurotoxic effects of ivermectin was studied in four males and three females previously identified as sensitive to ivermectin and in six which showed no increased sensitivity. Sensitive animals were identified as those that exhibited typical clinical signs of toxicity to the central nervous system after receiving ivermectin at an oral dose of 120 µg/kg bw. The levels of P-glycoprotein expression were similar in sensitive and insensitive test animals; however, a specific variant of the gene coding for P-glycoprotein was identified in the sensitive animals that caused production of a severely truncated, non-functional form of P-glycoprotein. The Committee noted that the sensitivity of CF-1 mice to the toxicity of avermectins has also been linked to a variant of the gene responsible for expression of P-glycoprotein. When the levels or functionality of P-glycoprotein are reduced, avermectin compounds may penetrate the blood–brain barrier and may be more extensively absorbed by the gastrointestinal tract.
Sensitivity to the toxicity of avermectin B1 was observed in a herd of Murray Grey cattle in Australia in 1985. Eight of 312 cattle treated with ivermectin at a therapeutic dose of 120–200 µg/kg bw by injection showed symptoms of hypersensitivity. The average concentration of avermectin B1a in brain tissue from the affected animals was 56 µg/kg, while that in brain tissue from a normal animal was 4 µg/kg. No adverse reactions occurred in 83 additional Murray Grey cattle from other areas of Australia that were tested for sensitivity to avermectins by treating them with at least twice the normal therapeutic dose of avermectin B1.
The Committee evaluated the relative potencies of doramectin, ivermectin and abamectin by comparing the NOEL values reported for reproductive and developmental toxicity in rats and rabbits and in 90-day studies of toxicity in dogs treated orally. These were the only studies with which such a comparison could be made. On the basis of these data, the Committee concluded that the potencies of these compounds are similar.
Eleven variants of the human gene coding for P-glycoprotein were identified in a sample population of 461 white volunteers in Germany. One of the variants was correlated with decreased levels of P-glycoprotein expression in the duodenum. Volunteers with this variant gene showed enhanced bioavailability of an oral dose of digoxin, with a steady-state concentration that was 38% higher than in volunteers without the variant gene. The difference was statistically significant. Whether this variant could result in enhanced bioavailability of orally administered avermectins is unknown. Studies of variations in the gene coding for P-glycoprotein in populations of other ethnic groups have not been reported. The Committee noted that, although the effects resulting from variation in the human gene coding for P-glycoprotein are modest, the evidence to date does not exclude the possibility that a subpopulation of humans sensitive to the toxic effects of avermectins may exist.
Ivermectin has been administered to several million human patients in Africa and Latin America since its introduction in 1987 as the main treatment for onchocerciasis at a recommended dose of 150 µg/kg bw administered once every 12 months. The adverse reactions that have been observed in treated patients have been described as allergic or inflammatory responses resulting from killing of microfilariae, referred to as the ‘Mazotti reaction’. No signs of acute central nervous system toxicity have been reported. Ivermectin is now considered safe for use in pregnant women, on the basis of the finding of P-glycoprotein in human placentae and in human fetuses by week 28 of gestation and the absence of adverse effects to the fetus when pregnant women were inadvertently treated with ivermectin.
The pharmacokinetics of orally administered ivermectin was studied in 12 healthy male volunteers of unspecified race. A single dose at a therapeutic level of 12 mg (150–200 µg/kg bw) resulted in an average maximal plasma concentration of 46 ng/ml and an average time to maximum concentration in plasma of 3.6 h. No adverse clinical signs were reported.
An ADI for doramectin of 0–1 µg/kg of bw was established on the basis of a NOEL of 0.1 mg/kg bw per day for mydriasis in a 3-month study in dogs treated by gavage, with a safety factor of 100. The Committee noted that removal of the twofold safety factor resulted in an ADI that still provided an adequate margin of safety for all other toxicological end-points of doramectin. The Committee also noted that the resulting ADI for dormectin is 150–200 times lower than the human therapeutic dose of the related compound ivermectin.
The Committee took special note of the available information on reduced expression of P-glycoprotein in humans, which results in increased bioavailability of substrates for this transporter. However, the effects on the bioavailability of avermectins and their ability to penetrate the blood–brain barrier are unknown. The Committee recommended that human populations continue to be monitored for possible genetic predisposition to sensitivity to avermectins.
Anon. (1999) Product information for Stromectol® , ivermectin tablets. In: Physicians’ Desk Reference (electronic version), PDR® Electronic Library Medical Economics Company Inc., Montvale, NJ. URL: www.pdrel.com.
Brown, K.R. (1998) Changes in the use profile of Mectizan: 1987–1997. Ann. Trop. Med. Parasitol., 92 (Suppl. 1), S61–S64.
Cascorbi, I., Gerloff, T., Johne, A., Meisel, C., Hoffmeyer, S., Schwab, M., Schaeffeler, E., Eichelbaum, M., Brinkmann, U. & Roots, I. (2001) Frequency of single nucleotide polymorphisms in the P-glycoprotein drug transporter MDR1 gene in white subjects. Clin. Pharmacol. Ther., 69, 169–174.
Cordon-Cardo, C., O’Brien, J.P., Casals, D., Rittman-Grauer, L., Biedler, J.L., Melamed, M.R. & Bertino, J.R. (1989) Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brain barrier sites. Proc. Natl Acad. Sci. USA, 86, 695–698.
Edwards, G., Dingsdale, A., Helsby, N., Orme, M.L’E. & Breckenridge, A.M. (1988) The relative systemic availability of ivermectin after administration as capsule, tablet, and oral solution. Eur. J. Clin. Pharmacol., 35, 681–684.
FAO/WHO (1998) Pesticide residues in food—1997. Report of the Joint Meeting of the FAO Panel of Experts on Pesticide Residues in Food and the Environment and the WHO Core Assessment Group. FAO Plant Production and Protection Paper, No. 145; available on the Internet at www.fao.org/ag/agp/agpp/pesticid/jmpr/pm-jmpr.htm.
Hoffmeyer, S., Burk, O., von Richter, O., Arnold, H.P., Brockmüller, J., Johne, A., Cascorbi, I., Gerloff, T., Roots, I., Eichelbaum M. & Brinkmann, U. (2000) Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. Proc. Natl Acad. Sci. USA, 97, 3473–3478.
Kipp, W., Burnham, G. & Kamugisha, J. (1992) Improvement in seizures after ivermectin. Lancet, ii, 789–790.
Lankas, G.R. & Gordon, L.R. (1989) Toxicology. In: Campbell, W.C., ed., Ivermectin and Abamectin, New York: Springer-Verlag, pp. 89–112.
Mealey, K.L., Bentjen, S.A., Gay, J.M & Cantor, G.H. (2001) Ivermectin sensitivity in collies is associated with a deletion mutation of the mdr1 gene. Pharmacogenetics, 11, 727–733.
Pacqué, M., Muñoz, B., Poetschke, G., Foose, J., Greene, B.M. & Taylor, H.R. (1990) Pregnancy outcome after inadvertent ivermectin treatment during community-based distribution. Lancet, ii, 1486–1489.
Pacqué, M., Dukuly, Z., Greene, B.M., Munoz, B., Keyvan-Larijani, E., Williams, P.N. & Taylor, H.R. (1989) Community-based treatment of onchocerciasis with ivermectin: Acceptability and early adverse reactions. Bull. World Health Org., 67, 721–730.
Seaman, J.T., Eagleson, J.S., Carrigan, M.J. & Webb, R.F. (1987) Avermectin B1 toxicity in a herd of Murray Grey cattle. Aust. Vet. J., 64, 284–285.
Umbenhauer, D.R., Lankas, G.R., Pippert, T.R., Wise, L.D., Cartwright, M.E., Hall, S.J. & Beare, C.M (1997) Identification of P-glycoprotein-deficient subpopulation in the CF-1 mouse strain using a restriction fragment length polymorphism. Toxicol. Appl. Pharmacol., 146, 88–94.
WHO (2001) Programme for the Prevention of Blindness and Deafness. www.who.int/pbd/oncho/oncho-brochure.pdf.
See Also: Toxicological Abbreviations Doramectin (WHO Food Additives Series 36) DORAMECTIN (JECFA Evaluation)