National Toxicology Program

EVIDENCE FOR POSSIBLE CARCINOGENIC ACTIVITY

Human Data: No epidemiological studies or case reports investigating the association of exposures to pyridostigmine bromide (PB) and cancer risk in humans were identified in the published literature. Adverse effects of PB are chiefly those of exaggerated response to parasympathetic stimulation and include adverse muscarinic effects such as nausea, vomiting, diarrhea, increased peristalsis, miosis, excessive salivation and sweating, increased bronchial secretions, abdominal cramps, bradycardia, and bronchospasm. Nicotinic side effects include weakness, muscle cramps, and fasciculation. Extremely high doses may produce central nervous system symptoms of agitation, restlessness, confusion, visual hallucination, and paranoid delusions. Overdosage can cause cholinergic crisis and death. As with other drugs containing bromide, skin rash may occasionally occur during therapy (McEvoy, 1992; PDR, 1995).

Several studies have examined the adverse effects of PB use during the Persian Gulf War. In one such study, Keeler and coworkers (1991) reported that about half the study population of 41,650 soldiers instructed to take the drug at the onset of hostilities noted gastrointestinal changes that included increased flatus, abdominal cramps, soft stools, and nausea. While under the threat of nerve-agent attack, the drug was self-administered by the troops (30 mg orally every 8 hours for 1 to 7 days). Other reported effects were urinary urgency, headaches, rhinorrhea, diaphoresis, and tingling of the extremities. Fewer than 0.1% of the soldiers had effects sufficient to discontinue the drug. Seventy-five percent of 213 Israeli soldiers surveyed by Sharabi and coworkers (1991) reported at least one symptom following PB use. The most frequent complaints were nonspecific and included dry mouth, general malaise, fatigue, and weakness. Typical effects, such as nausea, abdominal pain, frequent urination and rhinorrhea, were infrequent. The severity of symptoms was generally mild and no correlation was found between levels of cholinesterase and type or severity of complaints. Both of these studies found that the pyridostigmine regimen followed by soldiers under wartime conditions caused a higher incidence of adverse physiologic events than had been reported in earlier peacetime evaluations. It was felt that the combined stresses of anticipated combat, sleep deprivation, and life in the field may have affected or modified many of these responses.

Gouge and coworkers (1994) observed exacerbation of asthma symptoms in 7 of 10 asthmatic soldiers given a single 30 mg PB dose. The authors postulated that the increased irritant effect of desert dust might have predisposed these asthmatics to worsen after PB treatment, an effect not seen in the laboratory.

The unexplained illnesses experienced by Gulf War veterans, as well as research findings of a 10-fold PB-enhancement of the toxicity of DEET in cockroaches, has led to concern regarding the synergistic effects of PB and insecticides used by the soldiers in the field (Anon., 1994; Ember, 1995). Recent news reports cite a Duke University study in which these synergistic effects resulted in neuropathies in chickens (Washington Post, 1995).

Few data are available regarding the effects of cholinesterase inhibitors, including PB, on the fetus because of the rarity of maternal conditions requiring the use of these drugs during pregnancy. Transient muscular weakness has occurred in 10-20% of neonates whose mothers received anticholinesterase drugs for the treatment of myasthenia gravis, although similar symptoms have also been reported in infants whose mothers were not treated with these drugs (McEvoy, 1992). Anticholinesterase drugs may cause uterine irritability and induce premature labor when given iv to pregnant women near term. Although PB is not known to cause fetal injury or malformation, there are no adequate studies to support its safety during pregnancy (Flagg, 1991; McEvoy, 1992; PDR, 1995).

Animal Data: No 2-year carcinogenicity studies of PB in animals were identified in the published literature. Toxicity information identified was limited to acute and subchronic studies.

The 180-day subchronic oral toxicity of PB was evaluated in 69 male Sprague-Dawley rats. PB was administered in the diet at doses of 0, 1, and 10 mg/kg/day every day, and 10 mg/kg/day 5 days a week for 180 days. Following the 180-day dosing period, subgroups of animals from the control and both 10 mg/kg groups were subjected to a 30-day recovery period during which the test compound was not administered. No morphologic evidence of PB-induced toxicity was observed. All gross lesions were considered to be incidental findings commonly observed in Sprague-Dawley rats. Microscopic lesions with significantly increased incidence in pyridostigmine-treated groups compared to controls included chronic, multifocal hepatic inflammation found in the 10 mg/kg/daily group necropsied at 180 days (P < 0.05) and brown pigment, probably hemosiderin, within splenic macrophages found in the 10 mg/kg 5 days a week group necropsied at 210 days (P < 0.05). These microscopic lesions were also considered to be incidental findings unrelated to treatment. After 180 days, doses of PB that produced up to 63% cholinesterase inhibition in plasma and 49% acetylcholinesterase inhibition in erythrocytes did not have toxic effects other than increased startle reflex associated with the decrease in cholinesterase activity. Increases in aspartate aminotransferase, lactate dehydrogenase, and creatine phosphokinase were observed at 210 days but the changes could not be attributed to compound administration/ withdrawal (Morgan et al., 1990a).

The same researchers studied the 90-day subchronic oral toxicity of PB in 104 male and 104 female Sprague-Dawley rats. Administration in the diet at 0, 1, 10, 30, 60, and 90 mg/kg a day for 90 days resulted in dose-related decreases in plasma cholinesterase and erythrocyte acetylcholinesterase activity ranging from 5% to 76% and from 18% to 95%, respectively. Toxic signs associated with the decrease in cholinesterase activity included muscarinic (perianal, perioral, and periocular stains or material, diarrhea, and increased salivation) and nicotinic (hypertonia and tremors) effects. No compound-related gross or microscopic lesions were observed. Blood samples taken at necropsy for hematological and serum chemistry analyses exhibited no significant abnormalities (Morgan et al., 1990b).

Three short-term oral dosing studies were conducted with male and female beagle dogs in order to evaluate the preclinical safety of repeated PB administration. The drug was administered by capsule gavage once a day at 5, 10, or 20 mg/kg for 14 days to 10 dogs of each sex; every 8 hours at 2 or 5 mg/kg for 28 days to 6 dogs of each sex; or every 8 hours at 0.05, 0.5, or 2 mg/kg for 3 months to 37 dogs of each sex. A small portion of the dogs receiving PB for 3 months were allowed an untreated recovery period of an additional 3 months. In the 14-day study, signs of acute anticholinesterase intoxication occurred in all three dose groups. These included lacrimation, hypersalivation, diarrhea or soft stools, occasional emesis, muscle fasciculation, tremors, and occasional convulsions. No lesions were observed at necropsy except for the dogs that died or were euthanized during the study. These four animals exhibited a reddened mucosa in the large and small intestines, occasional ulcerations in the small intestine or colon and ileal intussusception. No morphological abnormalities were observed upon microscopic examination of the diaphragm muscle, a potential target organ. Signs of toxicity in the 28-day and 3-month studies were generally limited to the gastrointestinal tract and included diarrhea or soft stools and reddened or mucoid-containing stools. A single dog given 2 mg/kg every 8 hours developed an apparent intussusception. There were no pathological changes in clinical chemistry, hematology, or urinalysis parameters associated with PB administration for up to 3 months, nor were any drug-related lesions observed upon gross necropsy and microscopic evaluation of the major tissues and organs. These studies suggest that prolonged oral administration of PB at doses sufficient to cause as high as 70% inhibition of red blood cell acetylcholinesterase activity cause mainly local, gastrointestinal distress related to altered intestinal motility (Kluwe et al., 1990).

Gebbers and coworkers (1986) assessed the morphological changes in 26 male and female Tif:RAI f rats following single, sub-lethal gavage doses of PB. Within 24 hours of 20 or 40 mg/kg doses, acute focal necroses, leukocytic infiltrates, and marked changes in the motor endplates appeared in the skeletal muscle. Changes were more evident in the diaphragm than in the quadriceps muscle. Bowman and coworkers (1989) also reported myopathic changes in the diaphragm of 18 male Sprague-Dawley rats following administration of 90 mg/kg pyridostigmine in the diet for 15 days. Within the first day of dosing, 1% of the myofibers in the diaphragm were damaged. By 7 days, although myofibers were damaged as evidenced by centralized nuclei, dilated sarcoplasmic reticulum and disruption of Z-bands, they appeared less severely damaged than those examined earlier, indicative of some mechanism of accommodation that minimizes continued muscle injury.

Pyridostigmine bromide was a nonirritant in a modified Draize dermal irritation assay in New Zealand white rabbits (Magnuson et al., 1990). In guinea pig skin sensitization studies, PB was found to be a potential contact sensitizer that showed a potentiated response in the presence of surfactants. The formulations tested included 50% pyridostigmine bromide, 30% pyridostigmine bromide with 0.198% sodium lauryl sulfate, and 30% pyridostigmine bromide with 0.21% of a proprietary surfactant (Harris & Maibach, 1989).

The reproductive and developmental toxicity of PB was evaluated through the following gavage studies in Sprague-Dawley rats: fertility study, a) male rats received doses of 0, 5, 15, or 45 mg/kg a day for at least 70 days prior to mating with untreated females or b) female rats received doses of 0, 5, 15, or 45 mg/kg a day for at least 14 days prior to co-housing with untreated males; perinatal/postnatal study, sperm-positive female rats received doses of 0, 3, 10, or 30 mg/kg a day from gestation day 15 until lactation day 21; teratology study, sperm-positive female rats received 0, 3, 10 or 30 mg/kg a day on gestation days 6-15 and were killed on gestation day 20. Dose levels in each study were sufficient to result in overt cholinergic tremors at the high dose. PB administration did not affect fertility or reproductive performance in male or female rats. In the perinatal/postnatal studies, treatment did not alter reproduction indices and did not result in abnormal treatment-related effects in the offspring. Pups born to treated-dams did show slight, transient decreases in body weight gain, apparently secondary to the nursing behavior of dams demonstrating overt tremors. In the teratology study, a significantly increased rate of early resorptions, (approximately 2-fold over the control group, P < 0.05), was seen at the high-dose level. PB did not result in significant increases in either visceral or skeletal malformations. Skeletal variations indicative of delayed ossification such as hypoplastic supraoccipital, poor ossification of the cervical vertebrae, and missing vertebrae were slightly but significantly increased at the high-dose level (P < 0.05). These effects, however, were considered secondary to maternal toxicity (Levine & Parker, 1991).

Inhibition of chicken embryo kynurenine formamidase (KFase) results in a decreased concentration of NAD in the embryo and abnormal feathering and micromelia. Many potent avian teratogens produce prolonged inhibition of this enzyme in mice (Moscioni et al., 1977). PB was tested for in vitro and in vivo mouse liver KFase inhibition at doses of 10 mM and 1 mg/kg intraperitoneal (ip), respectively. In addition, teratogenic potency was assessed following injection of 1 mg into white Leghorn eggs at day 4 of incubation. PB was found to be without effect on KFase and it was not teratogenic. Further details of the results with this compound were not provided (Eto et al., 1980). PB at 10 mg or more per egg injected into the yolk sac of chicken eggs at 96 hours of incubation resulted in short and crooked necks as well as muscular hypoplasia of the legs. The experiment was not reported in detail (Landauer, 1975). At 15 mg/egg injected at 96 hours of incubation, the incidences of short/crooked neck and muscular hypoplasia were 95% and 61%, respectively (Landauer, 1976).

Short-Term Tests: The in vivo clastogenic potential of PB was evaluated with the rat micronucleus assay. Male and female Sprague-Dawley rats were administered 1, 10, or 30 mg/kg in the diet for 180 days. No differences were found between the treated and vehicle control groups in the numbers of micronuclei or in the percentages of polychromatic erythrocytes. The selected doses produced a dose-dependent inhibition of cholinesterase activity and toxic signs associated with the decreased activity were noted, indicating that pyridostigmine is not a clastogen at doses that produce significant pharmacological activity and/or toxicity in the rat (Orner & Korte, 1990). No further information on the genotoxicity or mutagenicity of PB was found in the available literature.

Pyridostigmine bromide (PB) has been reported to be negative in the Ames/Salmonella mutagenicity assay conducted for the Short-Term Test Program (STTP) of the National Cancer Institute's Division of Cancer Etiology (NCI/DCE). PB was negative at doses up to 10,000 mg/plate in strains TA98, TA100, TA1535, TA1537, and TA1538, both with and without S9 activation. PB has been selected for the mouse lymphoma assay conducted for the STTP of NCI/DCE (NCI/DCE, 1995).

Metabolism: Metabolism of PB has been studied in both humans and animals.

Human Data: PB is poorly absorbed from the GI tract. After oral administration, onset of action is 30-45 minutes and the duration of action is 3-6 hours (McEvoy, 1992).

Penetration of pyridostigmine into the central nervous system is poor. The drug crosses the placenta and small amounts are excreted in breast milk (Reynolds, 1993). Maternal doses of 180-300 mg/kg PB a day resulted in maternal plasma and breast milk concentration ranges of 6-100 ng/ml and 2-25 ng/ml, respectively. The drug was not identified in infant plasma (Hardell et al., 1982).

3-Hydroxy-N-methylpyridinium (HNM) has been identified as the main metabolite of pyridostigmine in man. Using bidirectional radiochromatography (BDRC), Kornfeld and coworkers (1970) detected as many as eight metabolites in the urine of eight myasthenic patients and five control subjects following iv injection of 2 mg radiolabeled PB. The investigators suggested that the following pyridostigmine (P) biotransformations could account for six of these (unidentified except for HNM) metabolites.

Neither pyridostigmine, its chief metabolite (HNM) nor the other metabolites found in plasma were protein bound. Somani and coworkers (1972) confirmed that pyridostigmine and HNM are the two main compounds in the urine of patients taking oral pyridostigmine iodide. Two additional urinary metabolites were found following intramuscular (im) administration of the radiolabeled drug to a myasthenic patient. The authors proposed that the first metabolite was formed from HNM and could be the 3,4- or 3,6-dihydroxy-N-methylpyridinium compound. Either could be present as the tertiary amine in its resonance form. Methoxy-N-methylpyridinium or acetoxy-N-methylpyridinium were suggested as the other metabolite.

Pyridostigmine undergoes hydrolysis by cholinesterases. It is also metabolized by microsomal enzymes in the liver (McEvoy, 1992).

PB is fairly quickly metabolized and excreted. When the radiolabeled drug was injected iv into patients and volunteers, levels of radioactivity in the urine varied widely among myasthenics and controls. Forty-seven to seventy-seven percent of the injected radioactivity appeared in the first-hour urines and an average of 88% of the radioactivity was excreted in the urine within 24 hours (Kornfeld et al., 1970). The mode of excretion is apparently primarily via renal tubules (Eiermann et al., 1993).

Animal Data: In vitro studies of pyridostigmine iodide (PI) with rat liver homogenates demonstrated that hydrolysis predominantly occurs in the soluble fraction of the liver cell, and is independent of the cofactor NADPH. In addition to the major hydrolysis compound, HNM, an additional metabolite was detected but its structure was not identified. However, it was suggested that it probably contained a carbamate group, since it was not formed when 3-hydroxy-N-¹⁴C-methylpyridinium was used as a substrate (Burdfield et al., 1973; Burdfield & Calvey, 1974).

Metabolism and urinary excretion proceeded more slowly than noted for human subjects after administration of 500 mg of ¹⁴C-labeled PI to rats (strain not reported) by gavage. After 24 hours 42% of the dose was absorbed and excreted in the urine. About 75% of the radioactivity in the urine was present as unchanged pyridostigmine, the remainder was a metabolite (Husain et al., 1968).

In rats (strain not reported), after im administration of ¹⁴C-labeled PI, radioactivity was rapidly excreted in the urine, mostly as pyridostigmine. About 45% of the dose was excreted in the first hour. The excretion of metabolite, HNM, steadily increased and after 3 hours was greater than that of pyridostigmine. The concentration of radioactivity in the liver reached its peak of 70% 20 minutes after injection and rapidly decreased during the next 40 minutes. The peak concentration of HNM occurred after 30 minutes and from 45 minutes onward its concentration exceeded that of pyridostigmine. The authors postulated that the liver is probably the main site of pyridostigmine metabolism and the source of HNM in urine. A second metabolite was detected but was not identified. Radioactivity was detected in most tissues except the brain, intestinal wall, fatty tissue, and the thymus gland (Birtley et al., 1966).

Other Biological Effects: Twenty-two drugs, whose active agents contain dimethylamino groups, were tested for their ability to form N-nitrosodimethylamine under simulated human gastric conditions. No measurable amounts of this carcinogen were formed in vitro by incubation of 60 mg PB and nitrite with human gastric juice (Ziebarth & Schramm, 1984).

Structure/Activity Relationships: Nine compounds structurally related to PB, as well as the hydrolysis product HNM, were screened for data relevant to the possible association, either positively or negatively, of mutagenicity or carcinogenicity with compounds of this structural type. Pertinent information was identified for only two of these compounds, neostigmine bromide and neostigmine methylsulfate. These synthetic quaternary ammonium compounds, which behave pharmacologically similarly to pyridostigmine bromide, have demonstrated significant inhibition of chemically-induced liver, stomach, or colon tumors in rats, presumably through a parasympathomimetic mediated mechanism. A summary of the carcinogenicity and mutagenicity information on PB, neostigmine bromide and neostigmine methylsulfate is shown in Table 2. No information on carcinogenicity or mutagenicity for the following structurally related compounds was found in the available literature: pyridostigmine chloride [7681-22-3]; pyridostigmine iodide [4685-03-4]; 3-[[(diethylamino)carbonyl]oxy]-1-methylpyridinium bromide [67465-54-7]; 2-bromo-3-[[(dimethylamino)carbonyl]oxy]-1-methylpyridinium bromide [51581-39-6]; 3-[[(dimethylamino)carbonyl]oxy]-1-(1-methylethyl)pyridinium bromide [69440-43-3]; 3-hydroxy-N-methylpyridinium bromide [31034-86-3]; edrophonium bromide [302-83-0]; and edrophonium chloride [116-38-1]. Structures of these compounds are shown in Table 3.

Chemical Name [CAS RN]	Carcinogenicity Data	Mutagenicity Data
Pyridostigmine bromide [101-26-8]	NDF	negative in a rat micronucleus assay (Orner & Korte, 1990)
Neostigmine bromide [114-80-7]	significant inhibition of chemically-induced liver tumors in rats (Gurkalo & Zabezhinski, 1982)	negative in a DNA-cell binding assay, E. coli Q 13 cells (Kubinski et al., 1981)
Neostigmine methylsulfate [51-60-5]	significant inhibition of chemically-induced stomach and colon tumors in rats (Tatsuta et al., 1988, 1989, 1992)	NDF

NDF:No data found.

Carcinogenic Effects

The following studies examined the role of the autonomic nervous system in the mechanisms of chemical carcinogenesis and the ability of pharmacological neurotropic compounds to modify the carcinogenic process. Gurkalo and Zabezhinski (1982) suggested that compounds that enhance the activity of the sympathetic nerves stimulate carcinogenesis while those that enhance cholinergic functions inhibit carcinogenesis. Neostigmine, an acetylcholinesterase inhibitor, demonstrated significant inhibition of carcinogenesis in these studies.

Neostigmine bromide administered subcutaneously (sc) at 50 mg/kg 3 times a week significantly decreased (P < 0.05) both the incidence and size of liver tumors in noninbred male rats treated with 0.7 mmol/l N-nitrosodiethylamine (NDEA) in drinking water for 4 months. At 6 months, 11 of 15 NDEA-treated rats had liver tumors while the incidence was 2 of 11 rats in the neostigmine group (Gurkalo & Zabezhinski, 1982).

The incidence of gastric cancers induced by N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) in Wistar rats was significantly (P < 0.02) decreased by administration of neostigmine (salt not specified). MNNG in drinking water (50 mg/ml) for 25 weeks followed by neostigmine (0.1 mg/kg/day) for 27 weeks after MNNG treatment resulted in a gastric cancer incidence of 7/19 (37%) versus 15/18 (83%) for the control group (olive oil only from week 25 on). The route of neostigmine administration was not stated (Tatsuta et al., 1989). In a second study of gastric cancers in MNNG-treated Wistar rats (75 mg/ml drinking water for 25 weeks), the sc administration of neostigmine methylsulfate (0.075 mg/kg every other day for 27 weeks after MNNG) significantly (P < 0.05) inhibited both the incidence (39% vs 80% for the control group) and multiplicity (0.4 vs 1.1 for the control group) of gastric cancers (Tatsuta et al., 1992).

Azoxymethane (AOM) induced colon tumors in 18 of 20 Wistar rats following sc administration of 7.4 mg/kg per week for 10 weeks. The tumor incidence was significantly decreased (P < 0.02) to 9 of 19 rats by sc administration of 0.1 mg/kg neostigmine methylsulfate every other day beginning 2 weeks before AOM. Colon tumor multiplicities were also reduced to 0.7 in the neostigmine group versus 1.9 in the control group (P < 0.001) (Tatsuta et al., 1988).

Mutagenic Effects

Neostigmine bromide, tested at 100 or 1,000 mM, was negative in the DNA-cell binding assay with metabolically activated E. coli Q13 cells (Kubinski et al., 1981).

Return to Table of Contents page