Safety assessment, in vitro and in vivo, and pharmacokinetics of emivirine, a potent and selective nonnucleoside reverse transcriptase inhibitor of human immunodeficiency virus type 1.

Emivirine (EMV), formerly known as MKC-442, is 6-benzyl-1-(ethoxymethyl)-5-isopropyl-uracil, a novel nonnucleoside reverse transcriptase inhibitor that displays potent and selective anti-human immunodeficiency virus type 1 (HIV-1) activity in vivo. EMV showed little or no toxicity towards human mitochondria or human bone marrow progenitor cells. Pharmacokinetics were linear for both rats and monkeys, and oral absorption was 68% in rats. Whole-body autoradiography showed widespread distribution in tissue 30 min after rats were given an oral dose of [(14)C]EMV at 10 mg/kg of body weight. In rats given an oral dose of 250 mg/kg, there were equal levels of EMV in the plasma and the brain. In vitro experiments using liver microsomes demonstrated that the metabolism of EMV by human microsomes is approximately a third of that encountered with rat and monkey microsomes. In 1-month, 3-month, and chronic toxicology experiments (6 months with rats and 1 year with cynomolgus monkeys), toxicity was limited to readily reversible effects on the kidney consisting of vacuolation of kidney tubular epithelial cells and mild increases in blood urea nitrogen. Liver weights increased at the higher doses in rats and monkeys and were attributed to the induction of drug-metabolizing enzymes. EMV tested negative for genotoxic activity, and except for decreased feed consumption at the high dose (160 mg/kg/day), with resultant decreases in maternal and fetal body weights, EMV produced no adverse effects in a complete range of reproductive toxicology experiments performed on rats and rabbits. These results support the clinical development of EMV as a treatment for HIV-1 infection in adult and pediatric patient populations.

Screening for potential toxicity of antiviral therapies.

Preclinical screening for the toxicity of antiviral drugs has thus far proven quite successful. Twenty-three antivirals spanning a variety of chemical classes and including a combination product, have been safely developed and are listed in the 1999 Physician's Desk Reference. Several of these antivirals have been administered for many years and in various combinations and we are currently unaware of any being withdrawn for safety-related reasons. Progress in protecting this record includes advances in rational drug design, the development of in vitro and in vivo models available to study both efficacy and safety, and improved bioanalytical capabilities. Coupled with approximately 25 years of experience since vidarabine was approved for the treatment of herpes encephalitis, these advances set the stage for even more precise and focused development of pharmaceuticals for the treatment of life-threatening viral infections, including those caused by HIV, HBV and emerging viruses. Experience to date suggests that each antiviral, particularly the nucleoside analogs, may have unique attributes of efficacy and safety. This brief review is an attempt to highlight a combination of established and recent methods that may be helpful in screening antivirals for potential toxicity. The earliest phases of drug development are mentioned in the introduction, followed by contributions provided by in vitro and in vivo testing in models of viral disease. Finally, current and new methods applicable to screening for toxicity are discussed.

Preclinical development of antiviral drugs.

The early preclinical development of an antiviral agent is accomplished essentially in two stages. The first stage consists of gathering data to estimate the potential therapeutic index of the agent. This process includes testing for antiviral activity and for cytotoxicity in vitro and performing preliminary pharmacokinetic and toxicology studies in vivo. The second stage consists of carrying out more-extensive safety (toxicology) studies to determine any significant potential toxicities before the agent is used in humans. Chronic-toxicity studies, reproductive toxicity studies, and carcinogenesis bioassays are performed to support clinical trials of longer duration and, ultimately, approval of efficacious antiviral agents. It is essential to identify toxicities early in the developmental process for both safety and economic reasons. Progress in determining the mechanisms of specific toxicities will greatly aid in risk assessment and in our ability to predict and avoid these toxicities.

Inactivators of herpes simplex virus ribonucleotide reductase: hematological profiles and in vivo potentiation of the antiviral activity of acyclovir.

A1110U (BW 1110U81) is an inactivator of herpesvirus ribonucleotide reductases and a potentiator of the antiviral activity of acyclovir (ACV) (T. Spector, J. A. Harrington, R. W. Morrison, Jr., C. U. Lambe, D. J. Nelson, D. R. Averett, K. Biron, and P. A. Furman, Proc. Natl. Acad. Sci. USA 86:1051-1055, 1989) that was subsequently found to cause hematological toxicity at high oral doses in rats. Eleven structurally related inactivators of herpes simplex virus (HSV) ribonucleotide reductase were therefore tested in vivo for hematological toxicity and for potentiation of ACV. None of the novel ribonucleotide reductase inactivators was hematologically toxic to rats following oral dosing at 60 mg/kg/day for 30 days. Four of these inactivators statistically improved the antiviral topical potency of ACV on HSV type 1-infected nude mice. A promising compound, 2-acetylpyridine 5-[(2-chloroanilino)thiocarbonyl]thiocarbonohydrazone (BW 348U87), was studied more extensively in two in vivo models: dorsum-infected athymic nude mice and snout-infected hairless mice. BW 348U87 significantly potentiated the antiviral activity of ACV against all virus strains tested, i.e., wild-type (ACV-sensitive) HSV type 1 and HSV type 2 strains and three mutant (ACV-resistant) HSV type 1 strains. The latter included a virus expressing a DNA polymerase resistant to inhibition by ACV triphosphate, a virus deficient in thymidine kinase (the enzyme responsible for phosphorylating ACV), and a virus expressing an altered thymidine kinase, which catalyzes the normal phosphorylation of thymidine but not of ACV. BW 348U87 and ACV are currently being developed as a combination topical therapy for cutaneous herpes infections.

Neurotoxic effects of the anti-varicella zoster virus agent 2'-valeryl-6-methoxypurine arabinoside in monkeys and rats dosed orally for 90 days.

The 2'-valerate ester of 6-methoxypurine arabinoside (170U88), a nucleoside analog with anti-varicella zoster virus (VZV) activity, was given to monkeys and rats. In subchronic preclinical toxicity studies, dosing was by gavage to monkeys (distilled water vehicle) and rats (0.5% methylcellulose vehicle) for 90 days. Groups of 5 male and 5 female monkeys (Macaca fascicularis) were given 170U88 at 0, 25, 50, or 100 mg/kg/day. The daily dose was given in two equal portions with 6 hr between doses. Monkeys in the high-dose group lost weight. Food consumption was decreased for mid- and high-dose monkeys and for low-dose female monkeys. Slightly decreased values for erythrocyte and leukocyte counts at the mid- and high dose were fully reversed during an 8-week recovery period. Two high-dose male monkeys and a middose female monkey developed signs of central nervous system toxicity and were necropsied before dosing was complete. These signs were first observed in the fifth week of dosing and included body tremors, incoordination, reduced activity, sleepiness, stupor, and lack of eye tracking. Axonal lesions were observed in histologic sections of sciatic nerve in monkeys at all dose levels. Neither the signs of central nervous system toxicity nor the axonal lesions reversed during the 8-week recovery period. Groups of 14 male and 14 female CD rats (Sprague-Dawley derived) were given single daily doses of 170U88 at 0, 150, 300, or 600 mg/kg. Body weights were decreased at all dose levels and food consumption was decreased for mid- and high-dose rats.(ABSTRACT TRUNCATED AT 250 WORDS)

Nail loss and footpad erosions in beagle dogs given BW 134U, a nucleoside analog.

The oral administration of BW 134U to Beagle dogs in a 1-month study was associated with lameness, footpad erosions, and nail loss. Groups of dogs received 60, 120, and 240 mg/kg/day. Compound-related effects were observed in the high dose group and only during the postdose period. Light microscopy revealed prominent cell maturation or radiomimetic defects in the stratum germinativum of the distal phalanges and footpads. The mechanism by which BW 134U produced these cell maturation defects is unknown. There were no signs of adverse effects in other keratin-producing or keratin-containing tissues.

Preclinical toxicology studies with acyclovir: teratologic, reproductive and neonatal tests.

Five studies were done to define the potential of Acyclovir (ACV), a new nucleoside analog for antiviral chemotherapy, to produce adverse effects on reproduction and development in laboratory animals. ACV produced no adverse effects when given by gavage to F0 generation mice at 50, 150 and 450 mg/kg/day in a two generation reproduction/fertility study. Some mice were evaluated for teratologic effects and others for postnatal development, including behavior, with negative results. ACV was not embryotoxic and did not increase the incidence of fetal malformations when given by subcutaneous injection to pregnant rats and rabbits at dose levels of 12, 25 and 50 mg/kg/day during the periods of major organogenesis. A comparative LD50 study revealed that 3-day-old rats were not more sensitive to acute toxic effects of ACV than more mature rats. Finally, in a comprehensive multidose toxicity study ACV was given subcutaneously to neonatal rats at 5, 20 and 80 mg/kg/day for 19 consecutive days. There was minimal effect on body weight gain in neonates treated at 20 mg/kg/day and a significant decrease in body weight gain at 80 mg/kg/day. Minimal renal lesions occurred at 80 mg/kg/day but no other signs of adverse effects on developing organ systems were observed. Except for decreased body weight gain in neonatal rats treated at 80 mg/kg/day, ACV did not produce adverse effects on mammalian development when tested in a variety of preclinical toxicology studies.

Preclinical toxicology studies with acyclovir: ophthalmic and cutaneous tests.

Topical formulations of acyclovir (ACV) were tested in animals to define potential for tissue irritation and systemic toxicity. Acyclovir ointments (5 and 10% concentrations in polyethylene glycol vehicle) produced no sign of dermal irritation or systemic toxicity when applied to shaved abraded and intact skin of guinea pigs for 24 consecutive days. Solutions (0.9% normal saline vehicle) of ACV did not sensitize guinea pigs when 10 sensitizing doses and a challenge dose were injected intradermally. Petrolatum base ophthalmic ointments containing 1 and 3% ACV did not produce significant ocular irritation when applied to the corneas of New Zealand White rabbits 5 times each day for 21 consecutive days. A 6% petrolatum base ointment produced mild conjunctival irritation but no sign of corneal or iridic toxicity. Mean concentrations of 2.53 microM ACV were found in aqueous humor 2 hours after a 1 cm ribbon (21 mg) of 3% ophthalmic ointment was placed in the eyes of rabbits. A single treatment with a topical ointment containing 5% ACV in polyethylene glycol vehicle produced minimal irritation when placed in the eyes of New Zealand White rabbits.

Preclinical toxicology studies with acyclovir: acute and subchronic tests.

Acyclovir (ACV), a new antiherpes drug, was evaluated for toxicity in a series of acute and subchronic toxicity tests. Oral LD50 values were greater than 10 000 mg/kg in male ICR mice and greater than 20 000 mg/kg in male Long Evans rats. When ACV was given iv, the LD50 was 405 mg/kg for male mice and greater than 600 mg/kg for male rats. Additionally, LD50 values for male rats treated sc were 1070, 790, 678, and 650 mg/kg in rats that were respectively, 3, 10, 28 and 71 days old indicating that very young rats were not more sensitive to acute toxic effects of ACV. There were no signs of toxicosis in CD-1 mice given ACV by gavage at dose levels of 50, 150 and 450 mg/kg/day for 1 month. Obstructive nephropathy occurred in rats given 20, 40 and 80 mg/kg/day once each day by rapid iv injection for 3 weeks. Both 5 and 10 mg/kg/day were no effect dose levels. Renal damage caused by precipitation of drug crystals in renal tubules and collecting ducts in rats given ACV by rapid iv injection was readily reversible within 2 weeks. Beagle dogs were given doses of 10, 20, 25, 50 and 100 mg/kg b.i.d. by rapid iv injection for 1 month. All 8 dogs given 100 mg/kg b.i.d. died by the 8th day of treatment; 5 of 8 dogs given 50 mg/kg b.i.d. died after 21 to 31 days of treatment.(ABSTRACT TRUNCATED AT 250 WORDS)

Preclinical toxicology studies with acyclovir: carcinogenicity bioassays and chronic toxicity tests.

Acyclovir (ACV), a nucleoside analog that is a new herpes-specific antiviral drug, was given by gavage at 50, 150 and 450 mg/kg/day to Sprague Dawley rats and Swiss mice for most of their lifetime to assess chronic toxicity and carcinogenicity. Treatment with ACV did not shorten the lifespan of either rats or mice. In fact, female mice given 150 and 450 mg/kg/day had significantly longer mean durations of survival than control female mice when analyzed by the life table technique. There were no signs of toxicosis produced by chronic exposure to ACV in either the rats or mice, and there was no drug-related increase in neoplasms in either species. Four groups of Beagle dogs were initially given daily oral doses of 15, 45 or 150 mg/kg ACV in a 1 year chronic toxicity study. Dogs treated at 150 mg/kg/day vomited, had diarrhea, consumed less feed and lost weight within 2 weeks. Dogs treated at 45 mg/kg/day also had minimal signs of gastrointestinal toxicosis. These dose levels were then decreased to 60 and 30 mg/kg/day for the rest of the one year test period. With the exception of occasional and inconsistent emesis and diarrhea, the 60 mg/kg/day dose level was well tolerated. Some mid and high dose dogs had sore paws due to erosion of footpads and cracking, splitting and loosening of the nails first becoming evident during the 13th week of the study.(ABSTRACT TRUNCATED AT 250 WORDS)

Postnatal behavioral effects of ochratoxin A in offspring of treated mice.

Ochratoxin A is a toxic isocoumarin derivative produced by Aspergillus ochraceus and several Penicillium species, which are storage fungi. Many kinds of agricultural commodities can be contaminated with ochratoxin A, which has been reported in both animal and human foods. Pure crystalline ochratoxin A was dissolved in 0.1 N sodium bicarbonate solution and given intraperitoneally to pregnant ICR-derived mice at dose levels of 1.25 and 2.25 mg/kg on gestation days 15, 16, and 17 (day 0, day of insemination). Dams were allowed to deliver, and their litters were culled to eight pups. Dams postnatal development, selected pups were tested for surface righting (days 3-12), swimming (even days 6-20) and pivoting (days 7, 9, and 11). Statistically significant differences for all three tests indicated that a developmental delay had occurred. Brains from the tested offspring were examined by light microscopy; no treatment or dose-related pathoanatomic alterations were found.

Dose selection as it pertains to testing a prodrug in carcinogenicity bioassays.

Toxicologists need more information than is usually available in the early stages of development of a drug in order to choose proper dose levels for testing in the bioassays. The approach most likely to result in successful bioassays involves an early multidisciplinary effort in which there is pharmacokinetic characterization of the test material in both rats and mice. Preliminary 3 month studies are desirable. Periodic sampling of plasma is essential to detect possible non-linear kinetics (as in the example we report herein) reflected as accumulation of the test material or metabolites. This is true regardless of the test substance. However, if one tests prodrugs it may be particularly helpful to know if chemical or enzymatic conversion of the prodrug is linear and if there is reversion to prodrug or other abberant metabolism. Failure to rule out these possibilities could result in subsequent clinically irrelevant organ damage or could compromise longevity or the interpretation of results in lifetime studies. Pharmacokinetic considerations are as valid as the more traditional biologic or morphologic end points used to estimate maximum tolerated or no-effect dose levels.

Prenatal effects of 2,4,5-T, 2,4,5-trichlorophenol, and phenoxyacetic acid in mice.

The herbicide 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) and two structurally related compounds, phenoxyacetic acid and 2,4,5-trichlorophenol, were suspended in a 1:1 solution of honey:water and administered by gavage to pregnant mice on one of gestation days 8-15 (copulation plug day = day 1) or on three consecutive days (7-9, 10-12, or 13-15). Doses were 800-900 mg/kg for single and 250-300 mg/kg/day for multiple treatments. With the exception of 2,4,5-trichlorophenol treatment on day 14, only 2,4,5-T treatment significantly increased prenatal mortality, and only 2,4,5-T was associated with decreased fetal weight when comparisons were made with the solvent controls. Although low incidences were seen in all treatment groups, only 2,4,5-T significantly increased cleft palate or other gross malformations. Significant skeletal, visceral or histopathological defects were not observed. These results indicate that both the carboxyl group and chlorination of the aromatic ring are essential for an unambiguous teratogenic response.

Effects in mice of simultaneous prenatal exposure to ochratoxin A and T-2 toxin.

Aspergillus ochraceus and Fusarium tricinctum are food contaminating molds whose toxic metabolites, ochratoxin A and T-2 toxin, are known mammalian teratogens. In order to determine the possible effects of simultaneous exposure to such environmental agents, ochratoxin A (2 or 4 mg/kg) and T-2 toxin (0.5 mg/kg) were injected ip, either together or individually, in CD-1 mice on gestation days 8 or 10. Ochratoxin induced craniofacial malformations when given alone on day 8, but not on day 10. T-2 toxin induced tail and limb anomalies particularly when given on day 10. When the two toxins were given together on day 10, ochratoxin exacerbated the incidence of T-2 induced gross malformations. An increase in fetocidal effects was also noted in groups treated at the high dose combination on either day, and effects on fetal growth of the high dose combination given on day 10 were greater than those of the other treatments. Few skeletal or visceral malformations were noted. These results indicated that two teratogens with presumably differing mechanisms of teratogenesis may have additive effects when administered concurrently to the same animal. Such results could be due to generalized toxic effects on the fetus or to more specific mechanisms, but further information is needed to differentiate between the two possibilities.