X-ray Structural Characterization of Anhydrous Metronidazole Benzoate and Metronidazole Benzoate Monohydrate

RESEARCH ARTICLES

Degradation and Inactivation of Antitumor Drugs JOHN A. BENVENUTO*', THOMAS H. CONNOR*, DAVIDK. MONTEITH*', JANA L. LAID LA^, STEPHEN c. ADAMSYA, THOMAS s. MATNEY'I, AND JEFFREY c. THElSSS5 Received August 17, 1992, from the *Department of Medical Oncology, Universiiy of Texas M. D.Anderson Cancer Center, Box 052, 1515

Holmmbe Blvd., Houston, T X 77030,$Environmental Sciences, University of Texas Health Science Center at Houston School of Public Health, Houston, TX, the 'Department of Pharmacy, Universiiy of Texas M. D.Anderson Cancer Center, Houston, TX, and the IlUnivemity of Texas Accepted for publication February 1, Health Science Center at Houston Graduate School of Biomedical Sciences, Houston, TX. Present addresses: §Warner Lambert Company, Ann Arbor, MI, and ASt.Luke's Hospital, Houston, TX. 1993. Abstract 0 Chemical methods for the degradation of 11 antineoplastic drugs [etoposide, teniposide. bleomycin, mitomycin C, cisplatin, cisdichloro-trans-dihydroxy-bis(isopropy1amine) platinum IV (CHIP),cyclo-

phosphamide, ifosfamide, carmustine, lomustine, and methotrexate] were investigated. The success of the degradation procedures was assessed by HPLC and degree of biological inactivation by mutagenicity assays. The most widely applicable procedure was oxidation with potassium permanganate or 5.25% sodium hypochlorite solution (bleach). Oxidation completely degraded and inactivated etoposide, teniposide, bleomycin, mitomycin C, and methotrexate. In addition, oxidation followed by nucleophilic substitution resulted in the complete degradation and inactivation of cyclophosphamide and ifosfamide. Although carmustine and lomustine were chemically degraded by treatment with acidic potassium permanganate, the resulting reaction mixtures remained mutagenic. Therefore, this procedure cannot be recommended. The platinum-containingcompounds, cisplatin and CHIP, were rendered nonmutagenicby reaction with sodium diethyldithiocarbamate. These easily performed, relatively safe procedures can be used to prevent exposure to mutagenic wastes and spills in the hospital setting.

Identifylng and avoiding environmental mutagens is a major concern. An insidious source of these environmental hazards is cancer chemotherapeutic drugs. Although these drugs are used to treat life-threatening diseases, most are carcinogenic and/or mutagenic and pose a threat of secondary malignancies in patients. Of the drugs included in this study, cyclophosphamideis considered to be a human carcinogen by the International Agency for Research on Cancer, whereas carmustine and cisplatin are probably human carcinogens.1 Bleomycin and mitomycin C are regarded to be possible human carcinogens.1 These drugs are also health hazards to others who are exposed to thern.2.3 Significant quantities of waste materials containing mutagenic antitumor agents are generated through manufacturing, research, therapy, and patient excretion. Data are available to indicate the danger to workers in the health care fields in which potential exposure may occur or to the families and friends of patients with cancer who also may be exposed. The possibility of exposure of pharmacy, nursing, and other hospital personnel to mutagenic anticancer drugs has been discussed.2.3 Reducing or eliminating exposure to mutagens is one way to decrease toxic risk, but it is not always feasible. On the other hand, destruction of mutagenic compounds is a feasible and practical method of protection. In most instances, these methods have been inadequate because they only measure the disappearance of the chemical entity and must, as has been 980 1 Journal of Pharmaceutical Sciences Vol. 82, No. 10, October 1993

recommended, determine the biological activity (usually mutagenicity) of the reaction products.4-6 Other methods, whereas they may eliminate mutagenicity, may result in hazardous by-products.7 We have devised procedures for the chemical deactivation of mutagenic antitumor agents and have monitored the success of these procedures by HPLC and the Ames mutagenicity assay. These procedures, executed primarily with potassium permanganate and bleach, are readily amenable for inactivating the drugs as solids, in aqueous solutions, and in spills. Some of these procedures (oxidation of methotrexate and complexation of cisplatin) have been previously described in an International Agency for Research on Cancer (IARC)monograph.8 This monograph is one of many to arise out of a joint effort of the National Institutes of Health and IARC to provide methods for the safe inactivation and disposal of genotoxic compounds.

Experimental Section Warning-The drugs included in this study may be acutely toxic and most are considered to be human carcinogens. Therefore, appropriate protective clothing and equipment should be employed whenever handling these drugs. Antineoplastic Drugs-The antineoplasticdrugs etoposide [injectbleoable; 4'-demethyl-9-(4,6-0-~D-ethylidene-glucopyranoside)l, mycin sulfate (powder), mitomycin C (powder), cyclophosphamide [powder; NJV-bis(2-chloroethyl~tetrahydro-W-1,3,2-oxaphosphorin2-amine 2-oxide], cisplatin (powder, cis-diamminedichloro-platinum 111, carmustine [powder;N,N'-bis[2-chloroethyll-N-nitrosoureal, and lomustine [powder; N-[2-chloroethyll-N'-cyclohexyl-N-nitrosoureal were obtained from Bristol Laboratories (Syracuse, NY). Methotrexate [powder; N-[p[[(2,4-diamino-6-pteridinyl) -methyl]-methylaminol-benzoyll-glutamicor 4-amino-10 methyl folic acid1 was obtained from Lederle Laboratories (Pearl River, NY). Teniposide [injectable;epipodophyllo-toxin-4'-dimethyl-,9-(4,6-0-2-thenylidine~-D)glucopyranoside],ifosfamide [powder;N,3-bis(2-chloroethyl)tetrahydro-2H-l,3,2-oxaphosphorin-2-amine 2-oxide], and cis-dichlorotruns-dihydroxy-bis(isopropy1amine)platinum IV (CHIP, powder) were investigational drugs that were obtained from the National Cancer Institute. CHIP was prepared at a concentrationof 5 mg/mL in normal saline. All drug solutions were prepared according to the manufacturers' package inserts. Injectable solutions (etoposide, teniposide) were treated without dilution. Any other variations are described in the text. Chromatographic Procedures-HPLC assays were performed with a Waters Associates (Milford, MA) chromatograph equipped with an M6000A pump, a U6K injector, a model 450 variable wavelength W detector, and a model 720 Data Module. All separations were achieved on pBondapak C18 columns (Waters Associates). Chromatographic solvents (HPLC grade) were distilled in glass (Burdick and Jackson Laboratories, Muskegon, MI). The separation This article is not subject to U S . copyright. Published 1993, American Pharmaceutical Association

Table I-HPLC

Condltlons for Analysls of Drug Solutions

Drug

Solvent (vol:vol)

Detector Wavelength, nm

Flow Rate, mumin

Retention Time, min 6.93 4.50 1.80 peak 1 2.50 peak 2 4.58 peak 3 5.41 peak4 4.13 2.86 5.50 1.77 3.13 3.33 1.81 2.59

Etoposide Teniposide Bleomycin sulfate

CH,OH/H,O (50:50) CH,OH/H,O (60:40) 5 mM HSA,a pH 3.5/CH30H (5050)

254 254 265

1.5 2.0 1.8

Mitomycin C Methotrexate Cyclophosphamide lfosfamide Cisplatin CHIP Carmustine Lomustine

CH,OH/H,O (50:50) 5 mM NH40H, pH 3.5/CH30H (60:40) CH,OH/H,O (60:40) CH,OH/H,O (60:40) CH,OH/H,O (50:50) CH,OH/H,O (50:50) 5 mM NH,CO,H, pH 3.5/CH30H (60:40) 5 mM NH,C02H, pH 3.5/CH30H (60:40)

254 254 210 210 250 250 254 254

1.o 1.o 1.o 1.o 1.o 1.o 1.o 1.o

a

HSA, Heptanesulfonic acid.

conditions are listed in Table I. Chromatographic conditions for mitomycin C, cyclophosphamide, ifosfamide, cisplatin, and CHIP were taken from previously referenced techniques.6.B Chromatographic conditions for the other drugs tested in this study were developed in our laboratory, and degradation was determined by the decrease of the parent drug peak. In all test procedures, the HPLC techniques detected decreases of the parent drug peaks. When complete degradation occurred, there was a total absence of integrated area at the retention time of the parent drug. There were no interferences from degradation products. Mutation Assay-The mutagenicity assay used was that of Maron and Ameslo with slight rnodifications.4.6.10 The system used for metabolic activation of the mutagens was the supernatant from a 9000 x g Centrifugation of rat liver homogenate (S9 fraction).The S9 fraction (40mg protein/mL) was also prepared according to Maron and Ames and was required only for the metabolic activation of cyclophosphamide and ifosfamide. The choice of the Salmonella typhimurium test strain for each drug (Table 11) was based on previously established evidence of strain sensitivity to the mutagenic activity of the drug.4.6JlAll drug solutions and treated solutions were assayed at 100 pL per plate. Although methotrexate is nonmutagenic in the Ames assay, the reaction products were assayed without and with S9 to rule out the formation of mutagenic products. Mutagenicity assays were started at the time indicated for chemical analysis and performed in duplicate. Revertant colonies were scored after 48 h of incubation at 37°C. A control set of plates contained untreated cells to determine the spontaneous mutant frequency. Two sets of plates were prepared with each drug solution. One set was treated with various concentrations of the drug and a second set contained drug at the same concentrationsas the first set but was assayed after chemical treatment. No apparent toxicity was noted for the data from assays reported in this paper. Controls for all drug inactivation procedures were diluted with distilled water or as specified with the procedure to the same final

volume as the drug being treated chemically (i.e., same final theoretical concentration). These samples were also kept at room temperature concurrent with the chemically treated samples. Final data for both the amount of parent drug present and mutagenicity are presented as percent of these concurrent controls. Mutagenicity presented as zero percent would mean that mutations occur only at the background reversion (i.e., mutation) frequency and not a fraction of the concurrent controls. All concurrent controls exhibited greater than a twofold increase in background reversion. Under the circumstances of these experiments, room temperature was 20-23 "C. Inactivation Procedures-Potassium Permangunate (KMnOS Oxidation-KMnO, was added to aqueous solutions of the anticancer drugs in the proportions indicated in Table 111. The reaction mixtures were allowed to stand at room temperature for up to 24 h. Immediately upon completion of the inactivation procedure and at 24 h, aliquot8 were removed and treated with 1% sodium bisulfite (NaHSO,)until color of the solution was clear to opaque to reduce the remaining KMnO,. The samples were centrifuged (1500 x g for 10 min) to remove the manganese dioxide (MnO,) and analyzed for amount of parent compound and mutagenic activity. Acidic Potassium Permangunate Oxidation-Carmustine (100 mg) was dissolved in 3.0 mL of ethanol diluent. Then, 1.5 mL of this solution was treated by diluting with 11 mL of sterile water and slowly adding 20.8 mL of 3 M H,S04 and 2 g of KMnO,. After 24 h, the reaction mixture was neutralized with 10 M KOH, and excess KMnO, was destroyed by adding 1% NaHSO,. The remaining 1.5 mL (control) was diluted appropriately with distilled water and allowed to stand for 24 h. Lomustine (100mg) was dissolved in 3.6 mL of dimethylformamide. To 1.8 mL of this solution were slowly added 21 mL of 3 M H,SO, and 2 g of KMnO,. The reaction mixture was neutralized by adding of 14.5 mL of 10 M KOH, and excess KMnO, was reduced by adding an excess of 1% NaHSO,. The MnOz formed was removed by centrifugation (1500X g for 10 min). The remaining 1.8 mL (control) was appropriately diluted with 50% ethanol.

Table Il--selmone//a fyph/mur/umStralns Used for Monltorlng Mutagenlclty Drug

Strain

Etoposide Teniposide Bleomycin sulfate Mitomycin C Methotrexate Cyclophosphamide lfosfamide Cisplatin CHIP Carmustine Lomustine

UTH8413 UTH8413 TA102 TA102 TA100, TA1535 (&S9)a TA100 (+S9) TA100 (+S9) UTH8414 TA100 TA100 TA100 ~~

Methotrexate was assayed without and with S9; cyclophosphamide and ifosfamide were assayed with S9; all other drugs were assayed without S9.

Table Ill-Drug Degradatlon and lnactlvatlonwlth Potassium Permaganate (KMnOJ

Etoposide Teniposide Bleomycin sulfate Mitomycin C Methotrexate Cyclophosphamide lfosfamide

2.42 2.65

0.34* 0.91 1.04 1.70 1.82

Immediate Immediate Immediate Immediate Immediate 24 24

0 0 0 0 0 24.1 0

0 0 0 0 Ob

16 33

a

a 1 mg is equivalent to 1 unit. Methotrexate is nonmutagenic and the reaction products are nonmutagenic under the assay conditions.

Journal of Pharmaceutical Sciences / 989 Vol. 82, No. 10, October 1993

Table IV-Degradatlon and Drug Inactivation wlth 5.25% Sodlum Hypochlorite (Bleach) Proportion Parent Drug Mutagenicity Drug (mL bleach1 Remaining, after mg drug) % Reaction, %

:iF:

Etoposide Teniposide Bleomycin Mitomycin C Methotrexate

20.1 28.0 7.0e

3.0 3.9

Immediate immediate Immediate immediate Immediate

0 0 0 0 0

0 0 0 0

~~~

1 mg is equivalent to 1 unit. Methotrexate is nonmutagenic and the reaction products are nonmutagenic under the assay conditions.

Etoposide Teniposide Bleomycin Carmustine Lomustine a

25.5 22.8 7.1a 30.4 25.7

24 24 24 24 24

36 0 70 62 57

Mutagenicity after Reaction, % 38 3 113 73 62

1 mg is equivalent to 1 unit.

Alkaline Potassium Permanganate Oxidation and Nucleophilic Substitution-The pH of the drug solution was increased to alkalinity (>7.0)with 5 M KOH, and the proportion of KMnO, was added. After 2.4 h, an excess of 1% NaHSO, was added and MnO, was removed by centrifugation. If the solution was not basic, more 5 M KOH was added. Sodium thiosulfate (1.5g/250 mg of drug) was added, and the reaction allowed to continue for 30 min, neutralized with HC1, and analyzed. Bleach Oxidation-Aqueous drug solutions were treated with the proportions of bleach (sodium hypochlorite) as shown in Table IV. After mixing, the reactions were quenched by adding an excess of 1% NaHSOs. Acid Hydrolysis-1 N HCl was added to the drug solution in the proportions shown in Table V. Carmustine (100mg) was dissolved in 3.0 mL of ethanol diluent; 1.5 mL of this solution was treated with 1 N HCI and the remaining 1.5 mL was diluted with distilled water. Lomustine (100mg) was dissolved in 3.6 mL of dimethylforamide; 1.8 mL of this solution was treated with 1 N HC1 and the remaining 1.8 mL was diluted with 50% ethanol. The reaction mixtures and controls were allowed to stand for 24 h. The reaction mixtures were neutralized with 1N NaOH, and the controls were diluted with an equivalent amount of distilled water. Nucleophilic Substitution (10% Na$320J-Etoposide, teniposide, and mitomycin C were treated with the proportions of 10% Na$,O,, as indicated in Table VI. The reactions were allowed to proceed at room temperature for 24 h. Alkaline Hydrolysis (0.5M KOH/CH30H)-To 5 mL of a 5-mg/mL solution of cyclophosphamidein sterile water was added 1.5mL of 0.5 M KOH in methanol. The solution was allowed to stand at room temperature for 24 h, neutralized, and then analyzed. To 10 mL of a 5-mg/mL solution of ifosfamide in sterile water was added 1.5 mL of 0.5 M KOH in methanol. The reaction was allowed to stand at room temperature for 24 h, neutralized, and then analyzed.

Table VCDegradation and Drug lnactlvation with 10% Sodlum Thlorulfate (Na,O,S,) Proportion Parent Drug Mutagenicity Drug (mLNa3O3S2t Remaining, after g drug) % Reaction, %

F:z,iy

Etoposide Teniposide Mitomycin C

13.4 48.2 45.3

24 24 24

990 I Journal of Pharmaceutical Sciences Vol. 82, No. 10, October 1993

82 50 37

added DDTC solution (100 mg/mL in 0.1 N NaOH) and saturated sodium nitrate solution. The final solution contains equal volumes of platinum solution, DDTC solution, and saturated nitrate solution. Controls were prepared by dissolving the platinum compounds in normal saline. An additional control was prepared by adding normal saline to equal volumes of the DDTC solution and NaNOS solution. The reaction mixtures were allowed to stand at room temperature for 24 h.

Ob

~ _ _ _ _ _ _ _ ~

Table V-Degradatlon and Drug lnactlvation with 1 N Hydrochlorlc Acld (HCI) Proportion Parent Drug Dfllg (mL 1 N Remaining, % HCltg drug)

Sodium Diethyldithiocarbamate (DDTC) Complexatwng-To normal saline solutions of cisplatin (1mg/mL) and CHIP (5mg/mL) were

109 155 26

Results and Discussion In all test procedures, the HPLC techniques deteded decreases of the parent drug chromatographic peak. When complete degradation occurred, there was a total absence of the integrated area at the retention time corresponding to the parent drug. No interference from the degradation products was observed. The inactivation technique for cisplatin and CHIP was also the chromatographic technique that was accomplished by a n irreversible complexation reaction that involves the platinum with little effect by the attached ligands. This complexation reaction was confirmed by atomic absorption and has been previously published.9 The results of the inactivation of the antitumor drugs are shown in Tables 111-VI. The most effective means for inactivation was oxidation, either by KMnO, (Table 111) or NaOCl (Table rv).With these reagents, immediate, complete chemical degradation occurred for etoposide, teniposide, bleomycin, mitomycin C, and methotrexate. In all of these cases, the reaction products were not mutagenic; presumably, the destruction of mutagenicity was also immediate. In addition, KMnO, in 3 M H,SO, chemically degraded carmustine and lomustine. However, the reaction products were mutagenic (data not shown). Oxidation of cyclophosphamide with KMnO, did not completely destroy this drug (Table 111). Ifosfamidewas destroyed by treatment with a threefold excess of KMnO,. However, the reaction mixture retained significant mutagenicity. When cyclophosphamide and ifosfamide were treated with KMnO, in 5 M KOH followed by reaction with sodium thiosulfate (a nucleophile), there was complete chemical destruction and loss of mutagenicity (data not shown). This inactivation procedure has been previously reported as effective for both chemical degradation and mutagenic inactivation.4 Attempts t o inactivate etoposide, bleomycin, carmustine, and lomustine with 1 N HCl were unsuccessful (Table V).It appears that bleomycin may have been activated by acid treatment, as indicated by a n increase in mutagenicity. The only drug that was inactivated by acid treatment was teniposide. This unique behavior, which is not seen with the close analogue etoposide, may be attributed to the increased lability of the 0-4,6 acetyl bond to acid hydrolysis because of the adjacent 2-thienyl group in teniposide. Nucleophilic substitution with sodium thiosulfate at pH 5.0 was also unsuccessful in inactivating etoposide, teniposide, and mitomycin C (Table VI).Again, a n increase of mutagenicity over controls was observed for etoposideand teniposide. Apparently, reaction with sodium thiosulfate at acidic pH produces “activated”compoundsof greater mutagenicity than the parent drugs. The two platinum compounds were inactivated by formation of complexes with sodium diethyldithiocarbamate. Because this is the same procedure used to quantitate platinum by HPLC, we could not determine the extent of drug destruction. Other assays for platinum, such as atomic absorption, only measure total platinum and the extent of complexation cannot be determined. However, we could confirm by HPLC that the complex was formed. The virtual absence of mutagenicity after treatment with DDTC indicated that complete

complexation occurred and that the complex was nonmutagenic. Although base hydrolysis for 1 h has been reported to completely detoxify cyclophosphamide,l2 our results of the KOH hydrolysis of cyclophosphamide and ifosfamide for 24 h demonstrated essentially no decrease of either drug (data not shown). Actually, there was two- to threefold increase in the mutagenicity of ifosfamide after base treatment (data not shown).This may reflect chemical conversion to a biologically active species. Procedures have been investigated for the degradation and inactivation of 11antitumor agents (Table VII), all of which are mutagenic except methotrexate, which is teratogenic. Generally, previous investigations of this type have been deficient in lacking an assessment of biological inactivation.8 Collaborative studies have emphasized the need for both chemical and mutagenicity assays.s.6 Thus, we have monitored the chemical degradation of the compounds by HPLC and the loss of mutagenicity of the parent drug and the lack of mutagenicity of degradation products by a modified Ames assay.4.5Jl The IARC of the World Health Organization has devised procedures for the inactivation of a number of mutagenic agents (reviewedin ref 8), and some of those procedures are assessed in this study for both chemical and mutagenic degradation of the drug. Oxidation with KMnO, or NaOCl inactivated the majority of drugs studied. Etoposide, teniposide, bleomycin, mitomycin C, and methotrexate were completely degraded and inactivated by both reagenta (Tablea III and IV).The degradationswere shown to be complete by HPLC and inactive by the S.typhimurium mutagenicity assay. Carmustine and lomustine were degraded by -0, in &So, solution.Becausebleach is readily available and relatively less hazardous than KMn04, it is recommended over KMnO, where applicable. Although it was not determined, it is possible that NaOCl could substitute for KMnO, in the Table VII-Recommended Proceduresfor Degradatlon and lnacthratlon of Antltumor Drugs

Drug Etoposide Teniposide Bleomycin Sulfate Mitomycin C Methotrexate Cyclophosphamide lfosfamide Cisplatin CHIP Carmustine Lomustine

Procedure’

’

Sodium hypochlorite [Potassium permanganate] Sodium hypochlorite [Potassium permanganate] Sodium hypochlorite [Potassium permanganate] Sodium hypochlorite [Potassium permanganate] Sodium hypochloriie [Potassium permanganate] Alkaline potassium permanganate followed by sodium thiosulfate Alkaline potassium permanganate followed by sodium thiosulfate Sodium diethyldithiccarbamate (DDTC) Sodium diethyldithiocarbamate (DDTC) None None

* Recommendedprocedure is given along with an alternate procedure in brackets (see text for details of degradation).

inactivation of cyclophosphamide and ifosfamide by oxidation and nucleophilic substitution. Cisplatin and CHIP were inactivated by complexation with DDTC. These complexes rendered the drugs nonmutagenic. This complexation forms the basis of a HPLC assay for ionic platinum and, once formed, the complex is extremely stable and cannot be destroyed even under very harsh conditions.9 Other methods of inactivation, such as acid hydrolysis (Table V), base hydrolysis, and nucleophilic substitution (Table VI)were unsuccessful. The necessity for determining chemical and mutagenic degradation can be readily appreciated from the results of these latter experiments. Although chemical degradation may have occurred, the mutagenicity of the products formed were, in some cases, much higher than that of the parent drug. The procedures presented here are easily performed with readily available reagents and can reduce the exposure of laboratory and hospital personnel to the hazards of genotoxic or carcinogenic antitumor drugs. If a user is unfamiliar with the safety precautions for handling some of the inactivation chemicals, the authors strongly encourage individuals to check laboratory safety manuals for the safe use of these compounds.8 It would be appropriate when these inactivation procedures are used routinely that a system be established to periodically check the efficacyof degradation and inactivation by chemical and mutagenic assays.

References and Notes 1. International Agency for Research on Cancer, IARC mono aphs on the evaluationof the carcinogenicrisk to humans, Suppgment No. 7, Overall evaluations of carcinogenicity: An u ating of IARC monographs, Volumes 1 to 29,International g n c y for Research on Cancer: Lyon, France, 1987. 2. Stellman, J. M.; Zoloth, S. R. Cancer Investigation 1986, 4, 127-135. 3. Rogers, B. Seminars Occup. Med. 1987,2,83-89. 4. Monteith, D.K.;Connor, T. H.;Benvenuto, J. A.; Fairchild, E. J.; Theiss, J. C. Environ. MoE. Mutagen. 1987,10,341-356. 5. Monteith, D.K.;Connor, T. H.; Binvenuto,J. A.;Fairchild, E. J.; Theiss, J. C. Toxicol. Lett. 1988,40,257-268. 6. Lunn. G.:Sansone. E. B.: Andrews. A. W.: Keefer. L. K. Cancer Res. i988,48,5221526. ’ 7. Lunn, G.;Sansone, E.; Andrews, A.; Hellwig, L. J. Pharm. Sci. 1989,78,652-659. 8. Laboratory Decontamination and Destruction of Carcinogens in Laboratory Wastes: Some Antineoplastic Agents; Castegnaro, M.; Adams, J.;Annour, M. A.;Barek, J.;Benvenuto, J.; Confalonieri, C.; Goff, U.; Ludeman, S.; Reed, D.; Sansone, E. B.; Telling, G., Eds.; International ncy for Research on Cancer: Lyon, France, 1985;IARC &cation No.73. 9. Bannister, S. J.; Sternson, L. A.; Repta, A. J. J. Chromatogr. 1979,330342. 10. Maron, D.M.;Ames, B. N. Mutat. Res. 1983,113,173-215. 11. Laidlaw, J. L.; Connor, T. H.; Theiss, J. C.; Anderson, R. W.; Matney, T. S. Am. J. Hosp. Phann. 1984,41,2618-2623. 12. Ehrenberg, L.;Wachtmeister, C. A. In Handbook ofMutagenicity Test Procedures;Kiley, B. J.; Legator, M.; Nichols,W.; Ramel, C., Eds.; Elsevier Scientific: Amsterdam, 1977.

Acknowledgments This work was su ported in part by a grant from Bristol Laboratories, Syracuse,

NQ.

Journal of Pharmaceutical SciencesI 991 Vol. 82, No. 70, October 1993