Probabilistic approach to the establishment of maximal content limits of impurities in drug formulations: the case of parenteral diphenylhydantoic acid.

Diphenylhydantoic acid (DPHA) is a degradation product in parenteral formulations of the anticonvulsant phenytoin and the prodrug fosphenytoin. DPHA has also been reported to be a minor metabolite of phenytoin. Levels found in the urine of various species, including humans, after oral or intravenous (iv) phenytoin ranged from undetected to a few percent of administered dose. In the present analysis, the toxicologic profile of DPHA was integrated with exposure data in order to characterize its safety under recommended clinical regimens of fosphenytoin administration. In preclinical safety studies, DPHA was without effect in the Ames assay and at concentrations up to 3000 microg/plate in the presence or absence of metabolic activation, and in the in vitro micronucleus test with acute and 2-week repeated dose studies in Wistar rats at iv doses up to 15 mg/kg. In 4-week studies conducted in rats and dogs receiving fosphenytoin containing DPHA levels up to 1.1%, and in an in vitro structural chromosome aberration test with DPHA levels up to 2.0%, all findings were consistent with known effects of phenytoin (such as CNS signs and increased liver weight), and none were attributed to DPHA. Reports in the literature indicate that in murine in vivo and in vitro models, DPHA has much lower potential for reproductive toxicity than phenytoin. A no-observed-effect level (NOEL) of 15 mg/kg established from the 2-week study in rats was used with probabilistic techniques to estimate tolerable daily doses (TDDs) of DPHA. In this approach, interspecies correction was performed by allometrically scaling the NOEL based on a distributional power of body weight while intraindividual variability was accounted for by selecting the lower percentiles of the population-based distribution of TDDs. The results indicate that a DPHA content limit of 3.0% in an administered dose of fosphenytoin is unlikely to cause adverse effects in patients.

In vitro induction of micronuclei and chromosome aberrations by quinolones: possible mechanisms.

The bacterial gyrase inhibitors, ciprofloxacin and PD 124816, were tested for clastogenic and aneugenic activity in V79 Chinese hamster lung cells in vitro. Cells were exposed for 3 h, washed free of drug, and subcultured for assessment of various endpoints. For structural chromosomal aberration (SCA) analysis, cells were incubated for 18 h, and treated with Colcemid for 2 h before harvest. For micronucleus (MN) analysis, treated cells were incubated with cytochalasin B (CYB) for 16 h. Aneugenicity was assessed by utilizing antikinetochore antibody to detect kinetochore-containing (K +) MN. Both quinolones induced significant increases in SCAs and MN, indicating clastogenic activity. With both compounds, however, the MN response was apparent at lower doses, and remained much higher throughout the dose range than the SCA response. The induced MN were predominantly K --, indicating that aneugenicity was not playing a major role in their induction. A possible explanation for the chromosome effects is that cross-reactivity of the gyrase inhibitors with mammalian topoisomerase II interferes with the separation of chromatids at anaphase leading to chromosome breaks and MN. Quinolones are known to inhibit resolution of the normally transient topoisomerase II-DNA cleavable complex, which may result in chromosome stickness. Thus, SCAs detected in metaphase cells may be attributed to quinolone-induced inhibition of topoisomerase II prior to mitosis while MN arise in binucleated cells as a result of this effect which interferes with chromatid separation during anaphase.

The genotoxicity profile of atorvastatin, a new drug in the treatment of hypercholesterolemia.

While HMG-CoA reductase inhibitors such as fluvastatin, lovastatin, pravastatin and simvastatin demonstrate lack of in vitro and in vivo mutagenicity and clastogenicity in bacterial and mammalian cells, long term rodent carcinogenicity studies resulted in an increased incidence in neoplasms at high doses. These effects may be attributable to an exaggeration of the desired biochemical effect of the drug and/or a tumor promoting effect. The genotoxicity of atorvastatin, a newly developed HMG-CoA reductase inhibitor, was evaluated in a variety of test systems. In bacterial mutagenicity tests, the E. coli tester strain WP2(uvrA) and S. typhimurium strains TA98, TA100, TA1535, TA1537, and TA1538 were exposed to concentrations of atorvastatin as high as 5000 micrograms/plate both in the absence (S9-) and presence (S9+) of metabolic activation. Atorvastatin was not mutagenic in either E. coli or S. typhimurium. Chinese hamster lung V79 cell cultures were exposed to atorvastatin at concentrations of 50-300 micrograms/ml (S9-) and 100-300 micrograms/ml (S9+) and structural chromosome aberrations were assessed. Mutation at the hgprt locus was assessed at concentrations of 100-300 micrograms/ml (S9-) and 150-275 micrograms/ml (S9+). Atorvastatin was neither mutagenic nor clastogenic in the absence or presence of S9. The lack of in vitro genotoxicity was corroborated in vivo in a mouse micronucleus study in which single oral doses of atorvastatin were administered to male and female CD-1 mice at 1, 2500, or 5000 mg/kg. No biologically significant increases in the frequency of micronucleated polychromatic erythrocytes in bone marrow at 24, 48, or 72 h postdosing were observed. Thus, atorvastatin, as with the other tested HMG-CoA reductase inhibitors, is not genotoxic.

Chemical purity and mutagenicity: case study of a drug in development.

During a routine Ames assay of a potential antipsychotic drug candidate, the compound appeared to be a frameshift mutagen in Salmonella typhimurium strains TA98 and TA1538. Additional testing indicated the mutagenic activity was due to one or more contaminants incurred during synthesis. While the compound was initially shown to be greater than 98% pure by high-performance liquid chromatography, the presence of small amounts (0.01-0.1%) of a highly mutagenic impurity produced positive mutagenicity results. The need to assess for chemical purity before discontinuing development of drug candidates found positive in the Ames assay is discussed.

Genotoxicity assessment of pirmenol, a new antiarrhythmic drug.

The genotoxicity of pirmenol was tested in the E. coli and S. typhimurium mutagenesis assay, an in vitro mammalian cell chromosome-aberration assay and an in vivo mouse micronucleus assay. The E. coli tester strain WP2s was exposed to concentrations of pirmenol as high as 10,000 micrograms/plate both in the absence (S9-) and presence (S9+) of metabolic activation. Five strains of S. typhimurium (TA98, TA100, TA1535, TA1537, TA1538) were exposed to concentrations of pirmenol as high as 5000 micrograms/plate in the absence and presence of S9. Pirmenol was not mutagenic toward either E. coli or S. typhimurium. Chinese hamster lung V79 cell cultures were exposed to pirmenol at concentrations of 500-2500 micrograms/ml (S9-) and 500-3000 micrograms/ml (S9+). Pirmenol increased the frequency of structural chromosome aberrations (SCAs). The minimum clastogenic concentration was 1500 micrograms/ml (both S9- and S9+) with a peak clastogenic response of 6% (S9-) and 34% (S9+) cells with aberrations. Although there were statistically significant results in the S9- experiment, the percent cells with aberration values for treated groups were within the historical control range (0-6%) of this laboratory. The observed effects in both the absence and presence of S9 appear at high concentrations compared to human circulating plasma levels of 1-3 micrograms/ml and the clastogenicity was confined to chromosome gaps and breaks. Consequently, this in vitro effect would not be expected to be reflected by either in vivo clastogenic or carcinogenic activity. This was supported by findings in the mouse micronucleus study of pirmenol in which single oral doses administered to male CD-1 mice at 5, 55, or 115 mg/kg (80% LD50) produced no statistically significant increases in the frequency of micronucleated polychromatic erythrocytes in bone marrow at 24, 48 or 72 h postdosing. Additionally, no evidence of carcinogenicity was seen in a mouse or rat bioassay.

Simultaneous micronucleus and chromosome aberration assessment in the rat.

Using a cellulose column fractionation procedure to eliminate nucleated cells for micronucleus assessment, micronucleus and chromosome aberration endpoints in the same animal were compared in male and female rats following i.p. injection with cyclophosphamide (CP). Groups of 5 Wistar rats per sex were given single doses of CP at 0, 20, or 40 mg/kg. Two hours prior to sacrifice, animals were given colchicine (4 mg/kg) to arrest cells in metaphase. One femur from each animal was used for micronucleus assessment and the other for chromosome aberration assessment. In the micronucleus assessment, 2000 polychromatic erythrocytes (PCEs) per animal and in the chromosome aberration assessment, 50 metaphase cells per animal were scored. This experiment was repeated once. In both experiments, significant increases in micronucleated PCEs and chromosome aberrations were noted at both doses of CP in both sexes. In general, the clastogenic effects of CP were more pronounced in males than females. Both doses of CP caused a decrease in the proportion of PCEs and in mitotic index in both experiments, indicating toxicity of CP to the bone marrow. These results show the usefulness of this rat model for simultaneous evaluation of two cytogenetic endpoints in the same animal and indicate that assessment of MNPCE frequency in the bone marrow of male rats may be an appropriate model for screening test substances for in vivo clastogenic activity in this species.

Dimethylnitrosamine-induced micronucleus formation in mouse bone marrow and spleen.

The present study was designed to obtain information on the kinetics of micronucleus (MN) formation following dimethylnitrosamine (DMN) treatment in mice. Male mice were injected once intraperitoneally with 50 or 100 mg/kg DMN. Bone marrow and spleen were obtained at various sacrifice time-points and processed for micronucleus analysis. The vehicle control group had 0.6 and 0.9 MN polychromatic erythrocytes (PCEs)/1000 PCEs in bone marrow and spleen, respectively. DMN, at 50 mg/kg, caused 3.8, 7.8, 8.5 and 10.2 MN PCEs/1000 PCEs in bone marrow and 8.0, 9.2, 19.3 and 32.8 MN PCEs/1000 PCEs in spleen at 12, 24, 36 and 48 h sacrifice times, respectively. A similar time-related elevation of micronucleus frequency was noted for 100 mg/kg DMN. At each sacrifice time-point, spleen PCEs had a higher micronucleus frequency than bone-marrow PCEs. In general, DMN decreased the proportion of PCEs to total erythrocytes, suggesting toxicity. Thus, this study demonstrates the clastogenic activity of DMN in both bone-marrow and spleen PCEs of mice and shows a time-related pattern in elevating DMN-induced MN PCE frequency.

Use of the cytokinesis-block method for the analysis of micronuclei in V79 Chinese hamster lung cells: results with mitomycin C and cyclophosphamide.

The cytochalasin B (CYB)-blocked binucleated cell assay has been explored to analyze micronuclei and cell cycle kinetics using 2 known mutagenic carcinogens in V79 Chinese hamster lung cells. To determine the optimum time to obtain the maximum number of binucleated cells for micronucleus analysis, duplicate cultures of exponentially growing cells were treated with 3 micrograms/ml CYB for varying durations (8-48 h). A peak appearance of binucleated cells at 16 h in the presence of CYB suggested this as an optimum time for micronucleus analysis in binucleated V79 cells. To evaluate the capacity for induction of micronuclei in V79 cells, 2 mutagenic carcinogens, mitomycin C (0.125-1.0 micrograms/ml) and cyclophosphamide (2-12 micrograms/ml) were tested in duplicate cultures. Mitomycin C, a direct-acting alkylating agent, caused approximately an 18-fold increase in micronucleus frequency over controls at the highest concentration tested (1.0 micrograms/ml), and this increase occurred in a dose-related manner (r = 0.92). The concentrations of mitomycin C tested also caused a significant dose-related cell cycle delay, thus suggesting cytotoxicity to V79 cells. Cyclophosphamide, an indirect-acting alkylating agent, requiring the presence of S9 mix, caused approximately a 17-fold increase in micronucleus frequency over controls at the highest tested concentration (12 micrograms/ml), with a clear dose response (r = 0.99). The various concentrations of cyclophosphamide also caused cytotoxicity in a dose-related fashion. Thus, this study demonstrates the usefulness of the cytokinesis-block method in V79 cells as a possible screen to analyze micronucleus induction and cytotoxicity. Because this approach is much less labor intensive than conducting a structural chromosomal analysis, this assay has great potential both as an initial screen for clastogenic activity and as a tool for investigating the underlying mechanisms for clastogenicity.

Lack of in vivo and in vitro genotoxicity with the nonsteroid, antiinflammatory agent sodium meclofenamate.

Sodium meclofenamate (Meclomen), CI-583, is an anthranilic acid salt developed as a nonsteroidal, antiinflammatory agent. A multitest battery of short-term tests was employed to characterize the genotoxic and mutagenic potential of the compound by measuring point mutations in bacteria, induction of sister-chromatid exchange, chromosome aberrations, and gene mutations in mammalian cells in vitro. In vitro assays included metabolic activation. The in vivo assay was for chromosome aberrations in bone marrow cells from rats. At toxicity-limited doses for each assay, no activity was detected in the bacterial or mammalian cell mutation assays, and sister-chromatid exchange frequencies were not increased over control rates. When sodium meclofenamate was evaluated in vitro, chromosome aberrations were induced under metabolic activation in CHO cells. Chromosome aberrations and clastogenic activity were not demonstrated after oral administration of Meclomen to male rats. It was concluded that sodium meclofenamate does not possess overt mutagenic potential under these conditions. The activity seen in vitro with CHO cells after metabolic activation did not correlate with the results from the other tests and was attributed to the formation of reactive metabolites not present or formed in in vivo systems.

Absence of in vitro genotoxicity of pyrvinium pamoate in sister-chromatid exchange, chromosome aberration, and HGPRT-locus mutation bioassays.

A single preparation of U.S.P. pyrvinium pamoate that exhibited promutagen activity in Salmonella/mammalian microsome reverse mutation assays was subjected to in vitro mammalian cell bioassays. Cytotoxicity limiting doses of dimethyl sulfoxide-solubilized drug were without effect in the absence and/or presence of liver S9 fraction from Aroclor-1254-induced rats in an in vitro Chinese hamster ovary (CHO) cell assay for chromosome aberration and sister-chromatid exchange induction at concentrations of 0.025-0.78 microgram/ml. Forward gene mutation at the hypoxanthine-guanine phosphoribosyl-transferase (HGPRT) locus in Chinese hamster (V-79) cells was studied with and without exogenous metabolic activation at concentrations ranging from 0.3 to 10 micrograms/ml. In concurrent promutagen controls, dimethylnitrosamine, benzo[a]pyrene, and cyclophosphamide caused significant (p less than 0.05) increases over controls in each respective assay endpoint under the metabolic activations used. These results suggest that intrinsic mutagenic activity manifested in lower microbial systems does not have a corresponding effect in mammalian cells. Coupled with the lack of solubility and poor absorption in vivo, we concluded that the drug, when used In the single oral dose mode of administration, possesses a negligible somatic and germinal genotoxicity risk.

Chemical carcinogen induction of DNA-repair synthesis in human peripheral blood monocytes.

Fresh ex vivo cultures of normal human peripheral blood monocytes, which are nonreplicative and known to possess cytochrome P-450 associated mixed-function oxidase activity, were used to assay DNA-excision repair manifested as augmented [3H]thymidine (dThd) incorporation following treatment in culture with diverse mutagenic carcinogens. Untreated monocyte cultures established from pools of 3-6 normal donors incorporated a low level of cytoplasmic [3H]dThd throughout a majority of the cells during an 18-h incubation. This background incorporation into whole cells was 80-90% inhibited by hydroxyurea (HU) at concentrations greater than 5 mM. Dose-related increases in the cumulative 18-h [3H]dThd incorporation in monocytes were observed following treatment with UV, N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), methyl methanesulfonate (MMS), mitomycin C (MMC), N-acetoxy-acetylaminofluorene (NA-AAF), aflatoxin B1 (AFB1), benzo[a]pyrene (BaP), and dimethylnitrosamine (DMN). The presence of HU during chemical treatment and throughout this 18 h of incubation with [3H]dThd did not influence the dose-response curves obtained with UV, MMS, NA-AAF and BaP but it increased the input dose of MNNG, MMC, DMN and AFB1 required to give peak repair incorporation. When HU was added to cultures following MNNG damage no interference with repair response was observed. HU apparently influences the extent of DNA damage by direct reactivity with these chemicals or their endogenously generated metabolites rather than inhibiting DNA-repair processes. These results provide evidence that monocytes are enzymatically proficient in base and nucleotide excision pathways and have endogenous capacity to metabolize BaP, AFB1 and DMN to DNA-damaging metabolites. As such, the monocyte is a potentially useful human cell type for detecting genotoxic chemicals and studying individuality in chemical-biological interactions.