Open Full Page

[CANCER RESEARCH 61, 3033–3038, April 1, 2001] 1,2-Bis(methylsulfonyl)-1-(2-chloroethyl)-2-(methylamino)carbonylhydrazine (101M): A Novel Sulfonylhydrazine Prodrug with Broad-Spectrum Antineoplastic Activity1

Rick A. Finch, Krishnamurthy Shyam, Philip G. Penketh, and Alan C. Sartorelli2 Department of Pharmacology and Developmental Therapeutics Program, Cancer Center, Yale University School of Medicine, New Haven, Connecticut 06520

ABSTRACT likely results from the chloroethylation and subsequent cross-linking of DNA (12), the other biological reactivities may contribute little to the Our laboratory has synthesized and evaluated the anticancer activity of a therapeutic value of these compounds and actually may contribute sig- number of sulfonylhydrazine DNA modifying agents. As a class, these com- nificantly to increasing the toxicity of these drugs. In fact, animal studies pounds possess broad spectrum antitumor activity, demonstrating significant activity against a variety of experimental murine tumors, including the P388 have shown that hydroxyethylation of DNA is largely a carcinogenic and L1210 leukemias, B16 melanoma, M109 lung carcinoma, and M5076 and/or mutagenic event (13, 14) and vinylation produces no known reticulum cell sarcoma, as well as against the human LX-1 lung carcinoma therapeutic benefit. The potential therapeutic contribution of the car- xenograft. The current report describes the activity of a more recently bamoylating species (i.e., the isocyanate) generated by the CENUs is synthesized member of this class, 1,2-bis(methylsulfonyl)-1-(2-chloroethyl)-2- unclear. Isocyanates, which can react with thiol and amine functionalities (methylamino)carbonylhydrazine (101M). 101M was active in mice against in proteins, have been shown to inhibit DNA polymerase II at the enzyme the i.p. implanted L1210 leukemia over a wide range of doses and produced level rather than acting directly on the DNA (15), thereby preventing the long-term survivors when administered as a single i.p. bolus of 10, 20, 40, 60, repair of DNA strand breaks (16). BCNU has also been shown to inhibit or 80 mg/kg, demonstrating a wider margin of safety than the nitrosourea, the thiol-containing enzymes, glutathione reductase, ribonucleotide re- 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). Curative therapy was achieved ductase, and thioredoxin reductase (17). Some of these properties un- with doses of 101M that did not produce depression of the bone marrow. 101M was also highly effective against the L1210 leukemia when adminis- doubtedly contribute to the toxic side effects of the CENUs (18–21); tered by the oral route. The ability of 101M to penetrate the blood-brain however, it is possible that isocyanate generation in tumor cells may act barrier and eradicate leukemia cells in the brain was remarkable (>6 log to prevent the repair of DNA alkylations and cross-links, thereby poten- kill). This agent was also curative against L1210 variants resistant to cyclo- tiating the cytocidal activity of the chloroethylation induced by the phosphamide, BCNU, or melphalan. Mice implanted with the murine C26 CENUs. Thus, one can envision the potential improved usefulness of colon carcinoma were also cured by two injections of 10 or 20 mg/kg of 101M. agents that possess both chloroethylating and carbamoylating properties Administration of 101M by two different well-tolerated regimens caused but lack hydroxylating and vinylating activities. complete regression of established human glioblastoma U251 xenografts in Our laboratory has been involved in the design, synthesis, and biolog- 100% of treated mice, and significant responses were also obtained with ical evaluation of a new class of antitumor compounds, the 1,2-bis(sul- 101M against advanced murine M109 lung carcinomas in mice. The broad fonyl)hydrazines (also called SHPs), which generate, through activation, spectrum of anticancer activity of the sulfonylhydrazine prodrug 101M coupled with the wide range of therapeutic safety exhibited by this agent, makes the reactive electrophilic structures that are capable of alkylating and 101M particularly attractive for further development and clinical evaluation. cross-linking DNA. The chloroethylating species generated by these agents are similar to, but not identical to, the ones produced by the INTRODUCTION CENUs. Penketh et al. (22) have recently compared the DNA lesions produced by SHPs with those produced by the CENUs. One of the SHPs, Tumor cell DNA is the primary target of many effective anticancer 101M, which was designed to retain chloroethylating and carbamoylating agents. Thus, DNA alkylating agents occupy a key position in the activities but not hydroxylating and vinylating activities, demonstrated currently available chemotherapeutic arsenal. The cross-linking of DNA pronounced antitumor activity in mice in preliminary studies (23). We is believed to be the primary event responsible for the anticancer activity now report the results of preclinical studies with this promising new of most of the clinically useful alkylating agents (1, 2). One series of antineoplastic agent, which include several direct comparisons of the agents, shown to be capable of alkylating and cross-linking DNA (3–5) therapeutic efficacy of 101M with that of the clinically useful CENU, and which are highly effective in model tumor systems, are the CENUs.3 BCNU. Clinically useful members of this class are BCNU and 1-(2-chloroethyl)- 3-cyclohexyl-1-nitrosourea (CCNU). These agents have been used effec- MATERIALS AND METHODS tively in the treatment of lymphomas (6) and a variety of human solid tumors including brain tumors (7); however, their utility is limited by Drugs and Mice. 101M and 90CE were synthesized according to published serious toxicities. Studies on the decomposition of the CENUs have methodology (23, 24). The chemical structure of 101M is shown in Fig. 1. All demonstrated the generation of several reactive species, with chloroethy- of the other agents were purchased from Sigma Chemical Co. (St. Louis, MO). ϫ lating (8, 9), carbamoylating (8–10), hydroxyethylating (11), and viny- BALB/c DBA/2 (CD2F1) mice, 8–10 weeks of age, were used with the lating (10) activities. Because the antitumor activity of the CENUs most L1210 leukemia, the P388 leukemia, and the M109 lung carcinoma. Athymic nu/nu mice were used with the human U251 glioma xenograft. BALB/c mice were used with the colon 26 carcinoma. Mice were treated i.p. with drugs dissolved in Received 8/30/00; accepted 1/26/01. 100% dry DMSO and delivered at a rate of 0.001 ml/g of mouse body weight or The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with in some studies with drugs prepared in solutions of 10% DMSO-90% water and 18 U.S.C. Section 1734 solely to indicate this fact. delivered at a rate of 0.01 ml/g of mouse body weight, or in pure water and 1 This research was supported in part by USPHS Grant CA-74970 from the National delivered at rates of 0.01–0.03 ml/g of mouse body weight. For oral administra- Cancer Institute. R. A. F. is a Special Fellow of the Leukemia and Lymphoma Society. tion, 101M was prepared as a fine suspension in corn oil and delivered by catheter 2 To whom requests for reprints should be addressed, at Department of Pharmacology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06520. Phone: gavage. (203) 785-4533; Fax: (203) 737-2045; E-mail: [email protected]. Leukemia Studies. In studies with the parental L1210 leukemia, female 3 The abbreviations used are: 101M, 1,2-bis(methylsulfonyl)-1-(2-chloroethyl)-2- ϫ 5 CD2F1 mice were inoculated i.p. with 1 10 cells. For studies with drug-resistant (methylamino)carbonylhydrazine; CENU, (chloroethyl)nitrosourea; BCNU, 1,3-bis(2- variants of the L1210 or with the P388 leukemia, the inoculum level was increased chloroethyl)-1-nitrosourea; SHP, sulfonylhydrazine prodrug; i.p., intraperitoneal; BBB, ϫ 6 blood-brain barrier; CNS, central nervous system; AGT, O6-alkylguanine-DNA alkyl to 1 10 cells. Twenty-four h after tumor inoculation, treatment was initiated on transferase; 90CE, 1,2-bis(methylsulfonyl)-1-(2-chloroethyl)hydrazine. the schedule indicated. Postinoculation life span was monitored for 60 days. 3033

As shown in Table 2, 101M was also highly effective against the i.p. implanted L1210 leukemia when administered by the oral route, curing mice with single treatments of 60, 80, 100, 200, and 300 mg/kg. Although we do not have pharmacological bioavailability information on 101M, we can estimate by comparison of the mini- Fig. 1. Chemical structure of 101M. mum curative single doses when given i.p. with that when given p.o., that a single oral dose of 60 mg/kg has roughly the equivalent curative To determine the effects of 101M on leukemia cells in the brain, a bioassay potential as an i.p. dose of 10 to 20 mg/kg. As gauged by body weight change, curative oral doses did not produce acute toxicity. procedure developed by Skipper et al. (25, 26) was used. Briefly, CD2F1 mice were given i.p. injections of drug or of 0.9% NaCl 24 h after intracerebral It should be noted from the data reported in Table 1, that the high inoculation of 1 ϫ 105 L1210 cells. Approximately 24 h later, the mice were bolus doses of 80 and 100 mg/kg of 101M were not as effective as the killed and the brains of three mice/experimental group were collected. A brei lower curative doses (10–60 mg/kg) in producing 60-day survivors. was prepared from the three pooled brains, and the equivalent of one-half of a This was not attributable to acute toxicity, however, because the times brain was injected i.p. into each of five mice of the same strain. Mean of occurrence of death in these mice were in excess of 50 days after postinoculation life spans were then determined and viable L1210 cells/brain treatment. These late-dying mice displayed no evidence of tumor were estimated by using concomitantly derived inoculum response data that development yet showed signs of a wasting condition (loss of body defined the relationship between life span and inoculum size. Solid Tumor Studies. Three solid tumors were tested for responsiveness to weight, small spleen, and loss of inguinal fat). Evaluation of the extent treatment with 101M. For the murine M109 lung carcinoma, tumor brei were of bone marrow toxicity (gauged by depression of the peripheral WBC count) of normal mice treated with single i.p. injections of implanted s.c. into CD2F1 mice, and therapy was initiated 9 days later and continued on the schedule indicated. In studies with the human U251 glioma, 101M at doses that produced long-term survivors in the leukemia fragments of the tumor were implanted s.c. in athymic mice, and therapy was studies demonstrated that very little marrow depression occurred at initiated 21 days later and continued on the schedule indicated. For both of these the curative doses of 10 and 20 mg/kg (Fig. 2). At the higher doses, models, tumor growth delay/regression was the evaluation end point; thus, tumor the WBC counts were depressed with a nadir at about 4 days, but measurements were recorded on the initial day of treatment and at various times blood counts, even at the 80 mg/kg dose, uniformly recovered to thereafter. In studies with the colon 26 carcinoma, tumor brei were implanted i.p., normal levels by 3 weeks after treatment. Thus, curative therapy of and two treatments with the test agents were administered on postimplant days 1 L1210 leukemia-bearing mice can be achieved with 101M at doses and 5. Life span was the evaluation end point for this experimental tumor system. Determination of Blood Leukocyte Counts. Peripheral blood was ob- that produce little bone marrow toxicity (i.e., doses of 10 and 20 tained from the retroorbital sinus of mice at selected times after i.p. treatment mg/kg). with 101M. After lysis of RBCs with Zapoglobin II reagent (Coulter Electron- ics, Miami, FL), the total blood leukocyte counts were determined using a Table 1 Comparison of the effectiveness of a single i.p. treatment of 101M or BCNU Coulter Multisizer II particle counter. against i.p. implanted L1210 leukemia in mice ϫ 5 Cell Culture Studies with the Human HT-29 and BE Colon Carcino- Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. A single i.p. mas. BE and HT-29 human colon carcinoma cells were provided by Dr. Neil treatment was administered 24 h later. Drugs were dissolved in DMSO and delivered at a rate Gibson (Bayer Corporation, West Haven, CT). Stock cultures were grown at of 0.001 ml/g of mouse body weight. Data represent at least five mice per treatment group. 37°C as monolayers in DMEM supplemented with 15% FCS and containing 101M BCNU 100 units/ml penicillin and 0.1 mg/ml streptomycin. For the growth inhibition Dosage ⌬ wt.a T/Cb Long-termc ⌬ wt.a T/Cb Long-termc studies, 250 cells were seeded into each well of a 96-well tissue culture plate. (mg/kg) (%) (%) survivors (%) (%) (%) survivors (%) Approximately 24 h later, test agents were added to the wells, and incubation 1.25 ϩ7.4 117 0 ϩ10.5 103 0 was continued for 5 days. The numbers of viable cells were then estimated 2.5 ϩ4.6 131 0 ϩ8.9 114 0 using the CellTiter 96 AQ cell proliferation assay (Promega, Madison, WI). 5 Ϫ1.0 160 0 ϩ3.5 137 0 10 ϩ1.0 229 60 Ϫ1.8 209 0 20 Ϫ0.8 100 ϩ0.0 170 0 RESULTS 40 Ϫ0.9 100 Ϫ3.2 200 80 60 Ϫ2.8 100 Ϫ7.4 429 80 In preliminary studies in mice, of the SHPs tested, 101M was found to 80 Ϫ6.7 800 80 Toxic 77 0 Ϫ d possess superior activity against both the L1210 leukemia and the 100 17.4 814 33 nd nd nd a ⌬ B16F10 melanoma (23). In the present more detailed studies, 101M wt, change in weight from onset of therapy to day 6. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. demonstrated activity in mice against the i.p. implanted L1210 leukemia T/C calculations do not include long-term survivors. c Long-term survivors, mice alive and tumor-free 60 days after inoculation. over a wide range of doses, producing long-term survivors when admin- d istered as a single i.p. bolus dose of 10, 20, 40, 60, 80, or 100 mg/kg nd, not determined. (Table 1). The therapeutic utility of 101M was also superior to the clinically useful CENU, BCNU, in this model of acute leukemia. Both Table 2 Effectiveness of a single oral treatment of 101M against i.p. implanted L1210 BCNU and 101M produced highly significant increases in life span. Note leukemia in mice ϫ 5 that treated (T)/control (C) values greater than 150% were produced at Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. A single oral treatment was administered 24 h later. 101M was suspended in corn oil and delivered at several dosages of both agents and both compounds were capable of a rate of 0.01 ml/g of mouse body weight. Data represent five mice per treatment group. producing long-term survivors (i.e., cures; mice alive and tumor-free 60 Dosage ⌬ wt.a T/Cb Long-termc days after tumor inoculation). The data presented in Table 1 also indicate (mg/kg) (%) (%) survivors (%) that 101M is therapeutically superior to BCNU. At least some long-term 40 ϩ2.6 214 0 survivors were produced with each of the six doses of 101M from 10 to 60 ϩ1.8 100 80 ϩ0.9 100 100 mg/kg and 100% of the mice were cured at doses of 20, 40, and 60 100 Ϫ3.9 100 mg/kg. In contrast, BCNU produced long-term survivors only at doses of 200 Ϫ1.9 457 80 40 and 60 mg/kg, with neither regimen producing 100% cures. Moreover, 300 Ϫ2.7 314 60 mice died of acute drug toxicity at the next higher dose of 80 mg/kg of a ⌬wt, change in weight from onset of therapy to day 6. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. BCNU. Thus, 101M exhibited a considerably wider therapeutic safety T/C calculations do not include long-term survivors. range than BCNU. c Long-term survivors, mice alive and tumor-free 60 days after inoculation. 3034

101M, were effective in crossing the BBB and eradicating leukemia cells in the CNS (Ͼ6 log kill), rivaling the efficacy of BCNU, which is one of the most effective agents for the treatment of experimental neoplasms in the CNS. Because curative therapy of the L1210 leukemia was possible with doses of 101M that did not produce severe depression of the bone marrow (Fig. 2), multidose scheduling of 101M should allow effective treatment of tumors with various growth characteristics. In addition, because 101M was effective at crossing the BBB, we were interested in whether it would be effective against a primary brain tumor. This was demonstrated with the U251 human glioma xenograft model. When administered to athymic nude mice implanted with this tumor using two dosage regimens that were well tolerated by the host, 101M caused complete regression of tumors in 16 of 16 treated mice (Fig. Fig. 2. Effects of a single i.p. bolus dose of 101M on the peripheral WBC count of 3). Both treatment regimens of 101M resulted in cumulative doses female CD2F1 mice. Data are average values of three mice per dosage level. exceeding 100 mg/kg yet produced no manifestations of the delayed wasting condition. Thus, it appears that treatment schedules can be Table 3 Occurrence of delayed toxicity in long-term survivors of L1210-inoculated designed that provide effective therapy for experimental leukemias as 101M-treated mice well as for solid tumors without the imposition of excessive host Delayed toxicity. The relatively wide range of safety in dosing makes this Dosage Long-term toxic Time of death (days agent particularly attractive for further development. a Route Schedule (mg/kg) survivors deaths after treatment) 101M was also highly active against an L1210 variant that is i.p. qdϫ1; d1 10 3/5 0/5 ndb resistant to BCNU (L1210/BCNU; Table 6). Although BCNU was 20 5/5 0/5 nd 40 5/5 0/5 nd marginally effective against L1210/BCNU, 101M was considerably 60 5/5 0/5 nd more active, producing 100% long-term survivors with single treat- 80 8/8 3/8 55, 56, 84 ments at two dosage levels. BCNU also had a low therapeutic index 100 6/6 5/6 58, 59, 96, 100, 100 against this tumor. It is important to note that L1210/BCNU has p.o. qdϫ1; d1 40 5/5 0/5 nd Ͼ 60 5/5 0/5 nd multiple mechanisms of resistance, possessing a 3-fold increase in 80 5/5 0/5 nd the levels of the DNA repair enzyme, AGT, as well as increases in the 100 5/5 3/5 93, 102, 119 levels of glutathione S-transferase, glutathione reductase activities, 200 4/5 1/4 97 300 3/5 3/3 87, 100, 101 and increased reduced glutathione levels compared with the parental a qdϫ1, every day for one dose; d1, day 1. b nd, no delayed toxic deaths occurred. Table 4 Effectiveness of aqueous 101M against i.p. implanted L1210 leukemia in mice ϫ 5 Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. Treatment was initiated 24 h later. 101M was dissolved in water at its maximum solubility (0.6 mg/ml) To document the occurrence of “late nontumor deaths” in the and delivered at 0.01, 0.02, or 0.03 ml/g of mouse body weight to achieve doses of 6, 12, leukemia studies, mice that were “cured” were observed for extended or 18 mg/kg, respectively. Data represent five mice per treatment group. periods of time (i.e., longer than 60 days). This information is sum- Schedule of Dosage ⌬ wt.a T/Cb Long-termc marized in Table 3. Late deaths occurred only in mice treated with administration (mg/kg) (%) (%) survivors (%) relatively high bolus doses of 101M (Ն80 mg/kg i.p. and Ն100 mg/kg qd ϫ 1; d1 6 ϩ4.0 203 0 18 ϩ0.0 219 0 p.o.). It is important to note that these doses far exceed those required qd ϫ 6; d1–6 6 ϩ0.0 333 40 to cure L1210 leukemia-bearing mice. In an experiment designed to 12 ϩ2.9 100 test the efficacy of repeated administration of aqueous formulations of 18 Ϫ4.6 100 101M against the L1210 leukemia (Table 4), no late toxic deaths a ⌬wt, change in weight from onset of therapy to day 6; qd, every day; d1, day 1; d1–6, occurred when mice were treated at a dosage of once every day for 6 days 1–6. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. doses, days 1–6, with 18 mg/kg/day (a cumulative dosage of 108 T/C calculations do not include long-term survivors. mg/kg). This result suggests that the apparent delayed toxicity was c Long-term survivors, mice alive and tumor-free 60 days after inoculation. attributable to excessively large acute bolus exposure (Ն80 mg/kg) and that toxicity is probably not cumulative. Table 5 Effects of various anticancer drugs on the intracranially implanted L1210 Because the CENUs such as BCNU, largely because of their leukemia in mice capacity to traverse the BBB, are among the most effective chemo- CD2F1 mice were given a single i.p. injection of the drug 1 day after intracranial ϫ 5 therapeutic agents available for the treatment of brain tumors, we were inoculation with 4 10 L1210 cells. Viable L1210 cells per brain were estimated from inoculum response data that defined the relationship between life span and inoculum size. interested in the potential of the SHPs against tumors of the CNS. We Doxorubicin was dissolved in water and delivered to mice at a rate of 0.01 ml/g of body have used a quantitative bioassay system (25, 26) to compare the weight. BCNU and 101M were dissolved in DMSO and delivered to mice at a rate of 0.001 ml/g of body weight. Data for Doxorubicin, BCNU, and 101M represent five mice ability of chemotherapeutic agents to cross the BBB and kill tumor per treatment group. Data for methotrexate, 6-mercaptopurine, and cyclophosphamide cells in the brain. Comparative results of 101M and clinically estab- were taken from Ref. 27. lished agents (27) in the brain bioassay system are shown in Table 5. Dosage Log kill of leukemia Drugs such as methotrexate are unable to cross the BBB as evidenced Compound (mg/kg) cells in the brain by the lack of a reduction in the tumor cell burden in the brain tissue. Methotrexate 12 0.01 Other drugs such as 6-mercaptopurine and doxorubicin are marginally 6-Mercaptopurine 160 0.43 Cyclophosphamide 140 0.94 effective against tumor cells in the CNS. Cyclophosphamide is able to Doxorubicin 5 0.23 reach and kill tumor cells in the brain, producing nearly a one log BCNU 60 Ͼ5.97 reduction in the number of tumor cells. Several of the SHPs, including 101M 80 Ͼ6.54 3035

reach a mass of 1 gram) of 8.8 days (23). In the current study, the s.c. tumor implant was allowed to grow for nine days, at which time tumors were easily palpable, before initiating treatment with 101M. With the average starting size of these established tumors at ϳ150 mg, 101M produced significant growth delays of 8 and 14 days for the 10 and 20 mg/kg/injection regimens, respectively. Thus, the growth delay produced by the 20 mg/kg/injection was more pronounced than that produced by cyclophosphamide in previous studies with far-less-well- established tumors. Because expression of the DNA repair protein, AGT, by tumor cells limits the effectiveness of the CENUs, we were interested in whether the activity of 101M would be affected by the AGT status of the tumor. We used two human colon tumor cell lines with different levels of AGT expression in cell culture studies to address this question (30). Fig. 3. Effects of multiple dose regimens of 101M on the growth of the U251 human Cells that express AGT activity and, therefore, have the capacity to glioma in athymic nude mice. Treatment on the indicated schedule was initiated 21 days repair O6-alkylguanine lesions in the DNA are termed Merϩ, and after s.c. implantation of tumor fragments near the flank. The average tumor volume was Ϫ ϳ300 mm3 at the time of the first treatment. Tumor size was estimated from caliper those with very low or negligible AGT expression are termed Mer . measurements made every 4 days using the formula for the volume of a prolate ellipsoid. As shown in Fig. 5, 101M was found to be nearly as effective in Data are average values of eight mice per treatment group. q2dϫ11, every 2 days for 11 ϩ Ϫ doses; q4dϫ6, every 4 days for 6 doses. inhibiting the growth of Mer HT-29 colon carcinoma cells as Mer BE colon carcinoma cells, which express very low levels of AGT. This is in sharp contrast to the differential sensitivity of these two Table 6 Effectiveness of a single treatment of 101M or BCNU against i.p. implanted tumor lines to 90CE, a SHP that generates the identical alkylating BCNU-resistant L1210 leukemia ϫ 6 species as 101M but that does not generate methyl isocyanate. This Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. A single i.p. treatment was administered 24 h later. Drugs were dissolved in DMSO and delivered at result implicates methyl isocyanate in overcoming the AGT-depend- a rate of 0.001 ml/g of mouse body weight. Data represent five mice per treatment group. ent resistance to chloroethylation-induced cytotoxicity. This concept Dosage ⌬ wt.a T/Cb Long-termc has been strengthened by recent studies in our laboratory that have Compound (mg/kg) (%) (%) survivors (%) 101M 20 ϩ14.8 133 0 40 ϩ2.0 176 0 60 ϩ0.0 100 Table 7 Effectiveness of a single treatment of 101M or cyclophosphamide (CTX) 80 Ϫ8.2 100 against i.p. implanted CTX-resistant L1210 leukemia ϫ 6 BCNU 20 ϩ14.7 129 0 Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. A single i.p. 40 Ϫ14.5 157 0 treatment was administered 24 h later. 101M was dissolved in DMSO and delivered at a 60 Toxic 105 0 rate of 0.001 ml/g of mouse body weight. CTX was dissolved in water and delivered at a rate of 0.01 ml/g of mouse body weight. Data represent five mice per treatment group. a ⌬wt, change in weight from onset of therapy to day 6. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. Dosage ⌬ wt.a T/Cb Long-termc T/C calculations do not include long-term survivors. Compound (mg/kg) (%) (%) survivors (%) c Long-term survivors, mice alive and tumor-free 60 days after inoculation. 101M 20 ϩ1.9 343 60 40 Ϫ1.9 100 60 Ϫ7.0 100 Ϫ L1210 leukemia (28, 29). Pharmacological doses of 101M do not 80 10.1 557 60 impact the pool of reduced glutathione and do not inhibit glutathione CTX 140 ϩ7.0 114 0 280 ϩ5.0 114 0 reductase.4 Thus, 101M may be an attractive alternative to BCNU, not a ⌬wt, change in weight from onset of therapy to day 6. only because of its greater therapeutic index, but also because it b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. possesses superior activity against nitrosourea-resistant tumors. 101M T/C calculations do not include long-term survivors. c also exhibited pronounced activity in mice against cyclophosphamide- Long-term survivors, mice alive and tumor-free 60 days after inoculation. resistant (Table 7) and melphalan-resistant (Table 8) L1210 sublines, making this agent an attractive alternative to these drugs. Table 8 Effectiveness of a single treatment of 101M or melphalan against i.p. Because one of the most valuable predictors of clinical activity of implanted melphalan-resistant L1210 leukemia in mice ϫ 6 a potential anticancer drug is broad-spectrum activity in preclinical Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. A single i.p. treatment was administered 24 h later. 101M was dissolved in DMSO and then further model tumor systems, we examined the efficacy of 101M in several diluted with water for a final DMSO concentration of 10%. Melphalan was suspended in additional model systems. Table 9 shows the results of an experiment corn oil. Both drugs were delivered at a rate of 0.01 ml/g of mouse body weight. Data with the murine P388 lymphoid leukemia. In this dose-ranging study, represent five mice per treatment group. 101M was highly effective, producing long-term survivors at all of the Dosage ⌬ wt.a T/Cb Long-termc six dose levels used between 10 and 100 mg/kg. Compound (mg/kg) (%) (%) survivors (%) Mice implanted with the murine C26 colon carcinoma were cured 101M 20 ϩ8.1 183 0 40 ϩ2.1 100 by two injections of 10, 20, 40, or 60 mg/kg of 101M (Table 10). 60 Ϫ7.9 100 Although BCNU was also effective in this model system, the CENU 80 Ϫ12.4 100 demonstrated a narrower safety range than 101M. Significant activity Melphalan 5 100 0 by 101M was also demonstrated against the advanced M109 murine 10 100 0 20 Ϫ2.1 123 0 lung carcinoma in mice (Fig. 4). In previously published studies from 40 Toxic 91 0 our laboratory, when treatment was initiated 24 h after tumor implan- 60 Toxic 100 0 tation, cyclophosphamide produced a tumor growth delay (time to a ⌬wt, change in weight from onset of therapy to day 6. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. T/C calculations do not include long-term survivors. 4 P. G. Penketh et al., unpublished observations. c Long-term survivors, mice alive and tumor-free 30 days after inoculation. 3036

Table 9 Effectiveness of a single treatment of 101M against i.p. implanted P388 peutic utility of compounds of interest. Of the many SHPs with activity leukemia in mice in this system (examples can be found in Refs. 23 and 24), 101M ϫ 6 Female CD2F1 mice were inoculated i.p. with 1 10 leukemia cells. A single i.p. treatment was administered 24 h later. 101M was dissolved in DMSO and then further demonstrated the widest margin of safety, curing mice with a single diluted with water for a final DMSO concentration of 10%. The drug was delivered at a treatment over a wide range of doses. Moreover, the ability of 101M to rate of 0.01 ml/g of mouse body weight. Data represent five mice per treatment group. cure mice at levels that did not produce severe weight loss or depression b Dosage ⌬ wt.a T/C Long-termc of the bone marrow indicates that this agent may be used effectively on (mg/kg) (%) (%) survivors (%) a variety of scheduling regimens, thereby being applicable to the treat- 5 ϩ0.0 158 0 ment of tumors with diverse growth characteristics. 10 Ϫ2.7 188 80 20 ϩ0.0 100 Although 101M has limited solubility in water, effective scheduling 40 Ϫ3.6 100 regimens using aqueous solutions of 101M can be designed to pro- Ϫ 60 2.8 100 duce curative therapy for L1210 bearing mice (Table 4). Improved 80 Ϫ3.9 100 100 Ϫ19.1 125 20 formulations containing polyethylene glycol, ethanol, and water are a ⌬wt, change in weight from onset of therapy to day 6. planned for use in future clinical trials. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. The ability of 101M to penetrate the BBB of mice, with its activity T/C calculations do not include long-term survivors. c Long-term survivors, mice alive and tumor-free 60 days after inoculation. against human tumors of CNS origin, coupled with its exceptional therapeutic properties, suggests that this agent may be clinically superior to the CENUs in the treatment of brain tumors. The importance of the Table 10 Comparison of the effectiveness of 101M and BCNU against the C26 colon potential role of 101M in the treatment of these tumors can be appreciated carcinoma in mice by an understanding of the immensity of the CNS tumor problem. The Female Balb/c mice were inoculated i.p. with brei of C26 colon carcinoma. On postinoculation days 1 and 5, mice were injected with the indicated dosage of either 101M combined incidence of primary intracranial and spinal axis tumors is or BCNU. Drugs were dissolved in DMSO and delivered at a rate of 0.001 ml/g of mouse between 2 and 19 in 100,000 persons per year, depending on age, with body weight. Data represent five mice per treatment group. CNS neoplasms being the most common solid tumors of childhood (31). 101M BCNU Brain metastases from primary tumors originating in other tissues are the

Dosage ⌬ wt.a T/Cb Long-termc ⌬ wt.a T/Cb Long-termc most prevalent intracranial tumors among adults, with as many as (mg/kg) (%) (%) survivors (%) (%) (%) survivors (%) 170,000 new cases of brain metastases occurring each year in the United 10 Ϫ4.6 100 Ϫ3.8 166 20 20 Ϫ4.1 100 Ϫ12.2 100 40 Ϫ6.8 100 Ϫ13.4 81 20 60 Ϫ9.7 84 60 Toxic 25 0 80 Toxic 39 0 ndd nd nd a ⌬wt, change in weight from onset of therapy to day 6. b T/C (%), mean life span of treated (T) mice/mean life span of control (C) mice ϫ 100. T/C calculations do not include long-term survivors. c Long-term survivors, mice alive and tumor-free 100 days after inoculation. Control mice lived 22.0 Ϯ 2.0 days. d nd, not determined. shown that methyl isocyanate can inhibit AGT activity in simple in vitro chemical systems.5

DISCUSSION The SHPs represent a new class of agents that possess broad spectrum Fig. 4. Growth of the murine M109 lung carcinoma in mice treated with 101M. Treatment on the indicated schedule was initiated 9 days after s.c. implantation of tumor fragments. The antineoplastic activity against a variety of experimental murine and average tumor volume was ϳ150 mm3 at the time of the first treatment. Tumor size was human tumors (22–24). The discovery of numerous agents in this class estimated from caliper measurements made every 2 days. Data are average values of five mice ϫ ϫ with significant activity in the preclinical setting has necessitated addi- per treatment group. q2d 10, every 2 days for 10 doses; q4d 5, every 4 days for 5 doses. tional studies aimed at the identification of an agent or agents most likely to be of benefit in the clinic and, therefore, to merit further development. We have used the i.p. implanted L1210 lymphocytic leukemia in mice as an initial test system to provide quantitative comparisons of the therapeutic potential of the various SHPs among themselves, as well as comparisons with proven clinically useful agents. This system was selected because the results obtained are highly reproducible, readily quan- tifiable, and rapidly generated. Furthermore, a single L1210 cell has the malignant potential to replicate and kill the host mouse in about 18 days, and a control mouse inoculated i.p. with 1 ϫ 105 cells will die about 7 days after inoculation. In addition, because i.p. inoculated cells dissem- inate rapidly throughout the body and are found in tissues within 3 h after implantation, candidate drugs must reach and eradicate every leukemia cell in tissues to effect a cure (25). The use of an animal tumor:host system also provides valuable information about the comparative thera- Fig. 5. Growth inhibitory capacity of 90CE and 101M for cultured colon carcinoma cell lines with differential expression of AGT activity. Colon carcinoma cells were subcultured at 5 P. G. Penketh, K. Shyam, R. P. Baumann, and A. C. Sartorelli. Role of O6- a density of 200 cells per well in 96-well plates. Twenty-four h later, various concentrations alkylguanine-DNA alkyltransferase in the inhibition of the cross-linking of T7 DNA by of either 90CE or 101M were added to the wells. Five days thereafter, the number of viable 1,2-bis(sulfonyl)hydrazines, manuscript in preparation. cells was determined using the CellTiter 96 AQ cell proliferation assay (Promega). 3037

States and between 20 and 40% of all cancer patients developing CNS Harley, J. B., Seligman, B., Cornell, C., Jr., Henry, P., Senn, H., Brunner, K., Martz, G., involvement (32). Because 101M was capable of readily crossing the Maurice, P., Bank, A., Shapiro, L., James, G. W., and Holland, J. F. Comparative study of a BCNU containing 4-drug program versus MOPP versus 3-drug combination in BBB and was active against BCNU-, cyclophosphamide-, and melpha- advanced Hodgkin’s disease. Cancer (Phila.), 43: 31–40, 1979. lan-resistant tumor cells, and was highly active against a human glioblas- 7. DeVita, V. T., Carbone, P. P., Owens, A. H., Jr., Gold, G. L., Krant, M. J., and Edmonson, J. Clinical trials with 1,3-bis(2-chloroethyl)-1-nitrosourea, NSC-409962. toma xenograft, as well as against tumors of various tissue types, this new Cancer Res., 25: 1876–1881, 1965. agent may have an immediate niche in the treatment of CNS tumors. 8. Montgomery, J. A., James, R., McCaleb, G. S., Kirk, G. S., and Johnston, T. P. An additional aspect favoring 101M as a clinically useful alternative to Decomposition of N-(2-chloroethyl)-N-nitrosoureas in aqueous media. J. Med. Chem., 18: 568–570, 1975. presently available alkylating agents is the limited responsiveness of 9. Montgomery, J. A., James, R., McCaleb, G. S., and Johnston, T. P. The modes of human tumors, especially brain tumors, to treatment with CENUs be- decomposition of 1,3-bis(2-chloroethyl)-1-nitrosourea and related compounds. cause of the expression of the DNA repair protein, AGT. This enzyme J. Med. Chem., 10: 668–674, 1967. 10. Weinkam, R. J., and Lin, H-S. Reactions of 1,3-bis(2-chloroethyl)-1-nitrosourea and protects the genome from the toxicity of various CENUs by removing the 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea in aqueous solution. J. Med. Chem., 22: alkyl group transferred by the alkylating agent onto the O6 position of 1193–1198, 1979. guanine prior to the formation of the lethal DNA cross-link (33, 34). AGT 11. Brundrett, R. B. Chemistry of nitrosoureas. Intermediacy of 4,5-dihydro-1,2,3- oxadiazole in 1,3-bis(2-chloroethyl)-1-nitrosourea decomposition. J. Med. Chem., 23: is significantly overexpressed in numerous human cancers including 1245–1247, 1980. many brain tumors (35) and, therefore, plays an important role in the 12. Kohn, K. W. Mechanistic approaches to new nitrosourea development. In: S. K. sensitivity of these tumors to alkylating agents such as the CENUs. The Carter, Y. Sakurai, and H. Umezawa (eds.), Recent Results in Cancer Research, Vol. 76, pp. 141–152. Berlin: Springer, 1981. carbamoylating species (i.e., the isocyanate) generated by the CENUs can 13. Pelfrene, A., Mirvish, S. S., and Gold, B. Induction of malignant bone tumors in rats by react with thiol and amine functionalities in proteins. We postulated that 1-(2-hydroxyethyl)-1-nitrosourea. J. Natl. Cancer Inst. (Bethesda), 56: 445–446, 1976. the methyl isocyanate generated by 101M could inhibit the DNA repair 14. Swenson, D. H., Frei, J. V., and Lawley, P. D. Synthesis of 1-(2-hydroxyethyl)-1- nitrosourea and comparison of its toxicity to that of 1-ethyl-1-nitrosourea. J. Natl. activity of AGT, an enzyme containing a cysteine thiol in its active site Cancer Inst. (Bethesda), 63: 1469–1473, 1979. (33, 34), and could, therefore, be beneficial in the treatment of AGT- 15. Baril, B. B., Baril, E. F., Lazlo, J., and Wheeler, G. P. Inhibition of rat liver DNA expressing tumors. This notion is supported by the work of Gibson and polymerase by nitrosoureas and isocyanates. Cancer Res., 35: 1–5, 1975. 16. Kann, H. H., Kohn, K. W., and Lyles, J. M. Inhibition of DNA repair by 1,3-bis(2- Hickman (21) with the TLX5 lymphoma, which is naturally resistant to chloroethyl)-1-nitrosourea breakdown product, 2-chloroethyl isocyanate. Cancer alkylating agents of the 2-chloroethylamine type, yet is sensitive to the Res., 34: 398–402, 1974. 17. Schallreuter, K. U., Gleason, F. K., and Wood, J. M. The mechanism of action of the CENUs; these authors postulated that intracellular release of isocyanates nitrosourea antitumor drugs on thioredoxin reductase, glutathione reductase and from the CENUs may overcome the resistance mechanism. We indirectly ribonucleotide reductase. Biochim. Biophys. Acta, 1054: 14–20, 1990. addressed this possibility using two human colon tumor cell lines with 18. Colvin, M., Brundrett, R. B., Cowens, J. W., Jardine, I., and Ludlum, D. B. A chemical basis for the antitumor activity of the chloroethylnitrosoureas. Biochem. very different levels of AGT expression. 101M was found to be nearly as ϩ Pharmacol., 25: 695–699, 1976. effective in inhibiting the growth of Mer HT-29 colon carcinoma cells 19. Wheeler, G. P., Bowdon, B. J., Grimsley, J., and Lloyd, H. H. Interrelationships of as it was with MerϪ BE colon carcinoma cells, which express very low some chemical, physiochemical and biological activities of several 1-(2-haloethyl)- 1-nitrosoureas. Cancer Res., 34: 194–200, 1974. levels of AGT. This is in sharp contrast to the differential sensitivity of 20. Panasci, L. C., Green, D., Nagourney, R., Fox, P., and Schein, P. S. A structure- these two tumor lines to 90CE, a SHP that generates the identical activity analysis of chemical and biological parameters of chloroethylnitrosoureas in alkylating species as 101M but that does not generate methyl isocyanate. mice. Cancer Res., 37: 2615–2618, 1977. 21. Gibson, N. W., and Hickman, J. A. The role of isocyanates in the toxicity of antitumor Thus, when the DNA alkylating/cross-linking capacity of the sulfonyl- haloalkylnitrosoureas. Biochem. Pharmacol., 31: 2795–2800, 1982. hydrazines is coupled with the capacity to generate an isocyanate to 22. Penketh, P. G., Shyam, K., and Sartorelli, A. C. Comparison of DNA lesions inhibit AGT, the spectrum of activity of the SHPs is extended to Merϩ produced by tumor inhibitory 1,2-bis(sulfonyl)hydrazines and chloroethylnitrosoureas. Biochem. Pharmacol., 59: 283–291, 2000. tumor cells. 23. Shyam, K., Penketh, P. G., Loomis, R. H., Rose, W. C., and Sartorelli, A. C. Broad spectrum activity in transplantable tumor models is one of the Antitumor 2-(aminocarbonyl)-1,2-bis(methylsulfonyl)-1-(2-chloroethyl)hydrazines. most valuable predictors of clinical potential. In these studies, 101M J. Med. Chem., 39: 796–801, 1996. 24. Shyam, K., Penketh, P. G., Divo, A. A., Loomis, R. H., Rose, W. C., and Sartorelli, demonstrated significant activity in each of nine different model tumor A. C. Synthesis and evaluation of 1,2,2-tris(sulfonyl)hydrazines as antineoplastic and systems, including several leukemias and solid tumors originating from trypanocidal agents. J. Med. Chem., 33: 2259–2264, 1990. 25. Skipper, H. E., Schabel, F. M., Trader, M. W., and Thompson, J. R. Experimental colon, lung, and brain. In previous evaluations, 101M was also active evaluation of potential anticancer agents. XI. Anatomical distribution of leukemic against the murine B16 melanoma (23). Taken together, with the supe- cells and failure of chemotherapy. Cancer Chemother. Rep., 21: 1154–1164, 1961. riority of 101M over BCNU, its broad-spectrum antitumor activity in 26. Skipper, H. E., Schabel, F. M., Wilcox, W. S., Laster, W. R., Trader, M. W., and Thompson, S. A. Experimental evaluation of potential anticancer agents. XVIII. preclinical models, the avoidance of the mutagenic/carcinogenic activities Effects of therapy on viability and rate of proliferation of leukemic cells in various of hydroxyethylation, the demonstrated activity against tumors resistant anatomic sites. Cancer Chemother. Rep., 47: 41–64, 1965. to other clinically useful alkylating agents, attractive therapeutic proper- 27. Avery, T. L., Finch, R. A., Vasquez, K. M., Radparvar, S., Hanna, N. B., Revankar, G. R., and Robins, R. K. Chemotherapeutic characterization in mice of 2-amino-9- ties including oral activity, and its effectiveness against tumor cells in the ⅐-D-ribofuranosylpurine-6-sulfinamide (sulfinosine), a novel purine nucleoside with CNS, suggest that 101M exhibits sufficient properties to merit consider- unique antitumor properties. Cancer Res., 50: 2625–2630, 1990. ation for further development and clinical trial. 28. Gorbacheva, L. B., Kukushkina, G. V., Elksne, S. J., Mikiner, S. J., Melnik, S. Y., and Preobrazhenskaya, M. N. O6-Methylguanine as a modulator of antitumor activity of N-alkyl-N-nitrosoureas in vivo. Anticancer Drugs, 2: 185–189, 1991. REFERENCES 29. Ribrag, V., Massaad, L., Janot, F., Bissery, M. C., Parise, O., Gouyette, A., and Chabot, G. G. Principal drug-metabolizing enzyme systems in L1210 leukemia 1. Teicher, B. A. Antitumor alkylating agents. In: V. T. DeVita, Jr., S. Hellman, and sensitive or resistant to BCNU in vivo. Leukemia Res., 18: 829–835, 1994. S. A. Rosenburg (eds.), Cancer: Principles and Practice of Oncology, Ed. 5, pp. 30. Pratviel, G., Shyam, K., and Sartorelli, A. C. Cytotoxic and DNA-damaging effects ϩ Ϫ 405–418. Philadelphia: Lippencott-Raven, 1997. of 1,2-bis(sulfonyl)hydrazines on human cells of the Mer and Mer phenotype. 2. Lawley, P. D., and Brooks, P. Interstrand cross-linking of DNA by bifunctional Cancer Biochem. Biophys., 10: 365–375: 1989. alkylating agents. J. Mol. Biol., 25: 143–160, 1967. 31. Levin, V. A., Leibel, S. A., and Gutin, P. H. Neoplasms of the central nervous system. 3. Ludlum, D. B., Kramer, B. S., Wang, J., and Fenselau, C. Reaction of 1,3-bis(2-chloroethyl)- In: V. T. DeVita, Jr., S. Hellman, and S. A. Rosenberg (eds.), Cancer: Principles and 1-nitrosourea with synthetic polynucleotides. Biochemistry, 14: 5480–5485, 1975. Practice of Oncology, Ed. 5, pp. 2022–2082. Philadelphia: Lippincott-Raven, 1997. 4. Kohn, K. W. Interstrand cross-linking of DNA by 1,3-bis(2-chloroethyl)-1-nitro- 32. Posner, J. B. Management of brain metastases. Rev. Neurol., 148: 477–487, 1992. sourea and other 1-(2-haloethyl)-1-nitrosoureas. Cancer Res., 37: 1450–1454, 1977. 33. Pegg, A. E., Dolan, M. E., and Moschel, R. C. Structure, function, and inhibition of O6- 5. Lown, J. W., McLaughlin, L. W., and Chang, Y-M. Mechanism of action of 2- alkylguanine-DNA alkyltransferase. Prog. Nucleic Acid Res. Mol. Biol., 51: 167–223, 1995. haloethylnitrosoureas on DNA and its relation to their antileukemic properties. 34. Mitra, S., and Kaina, B. Regulation and repair of alkylation damage in mammalian Bioorg. Chem., 7: 97–110, 1978. genomes. Prog. Nucleic Acid Res. Mol. Biol., 44: 109–142, 1995. 6. Nissen, N. I., Pajak, T. F., Glidewell, O., Pedersen-Bjergaard, J., Stutzman, L., Falkson, 35. Silber, J. R., Mueller, B. A., and Berger, M. S. Comparison of alkyltransferase G., Cuttner, J., Blom, J., Leone, L., Sawitsky, A., Coleman, M., Haurani, F., Spurr, C. L., activity in brain tumors and adjacent brain. Cancer Res., 53: 3416–3420, 1993. 3038

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2001 American Association for Cancer Research. 1,2-Bis(methylsulfonyl)-1-(2-chloroethyl)-2-(methylamino)carbo nylhydrazine (101M): A Novel Sulfonylhydrazine Prodrug with Broad-Spectrum Antineoplastic Activity

Rick A. Finch, Krishnamurthy Shyam, Philip G. Penketh, et al.

Cancer Res 2001;61:3033-3038.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/61/7/3033

Cited articles This article cites 31 articles, 8 of which you can access for free at: http://cancerres.aacrjournals.org/content/61/7/3033.full#ref-list-1

Citing articles This article has been cited by 12 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/61/7/3033.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/61/7/3033. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.