[CANCER RESEARCH54, 41)72-4076, August 1, 19941 murine tumors (4) and melanoma xenografts (5) and is significantly more effective than DTIC in inhibiting the growth of the Walker tumor in rats (6). Like DTIC, CB1O-277 requires activation by oxi dative N-demethylation, but the overall production of the putative active monomethyl metabolite in rats was iS-fold greater than that for DTIC, suggesting that species-dependent activation is less likely to be a problem in humans (6). Thus, due to the structural similarities to DTIC, the improved in vitro stability and solubility, and the possibil ity of improved metabolic activation, CB1O-277 was selected for clinical trial by the Cancer Research Campaign (UK) Clinical Trial Committee (7). With respect to the mechanisms of antitumor action, there is a considerable amount of evidence to indicate that the cell-killing effects of DTIC, CB1O-277, and related methylating agents are me diated mainly by the formation of 06-MedG in DNA and that the principal mechanism of resistance is via the DNA repair protein ATase. This protein acts by the stoichiometric transfer of the methyl group from the 06-position of guanine in alkylated DNA to a cysteine residue in the protein itself, in an autoinactivating reaction (5, 8—12). The strongest evidence for the cytotoxic effects of 06-alkylguanine in DNA comes from experiments which show that the expression of a transfected prokaryotic or eukaryotic ATase complementary DNA in mammalian cells protects them against the toxic effects of these agents (13—17). . We recently demonstrated a wide interindividual variation in the depletion of lymphocyte ATase during a 24-h continuous infusion of CB1O-277 (18). These findings can be attributed to the CBiO-277- mediated formation of 06-MedG in lymphocyte DNA, and the sub sequent repair of this adduct by ATase is manifested as a depletion of ATase activity in lymphocyte extracts. The relationship between the ability to activate CB1O-277, which is reflected in the formation of 06-MedG in DNA, and the wide range of clinical responses seen among patients with metastatic melanoma receiving this drug has not been explored. The aims of the present study are therefore to examine whether there is a correlation between 06-MedG accumulation in leukocyte DNA and lymphocyte ATase activity and also to assess whether there is any relationship between 06-MedG formation and clinical response. MATERIALS AND METhODS Patients and Blood Samples. CB1O-277 (sodium salt, Mr 215) was sup plied as a lyophilized, pyrogen- and preservative-free powder (in 1000-mg vials) by the Developmental Therapeutics Program, National Cancer Institute (Bethesda, MD). All nine patients had metastatic melanoma, and details of the individuals studied as reported previously (18) are shown in Table 1. Each patient received CB1O-277 (12 mg/m2) by continuous iv. infusion over 24 h and the treatment was repeated every 4 weeks. Serial blood samples (20 ml) were collected at various times during the infusion and for 24 h after comple tion of the first cycle of CB1O-277. Blood samples were dispensed into two universal containers (10 ml each) containing 0.5 ml of 0.5% EDTA, pH 8.0. One container was stored at —20°C prior to DNA extraction and radioimmu noassay for 06-MedG levels (see below). The second container was kept at 4072 Relationships between the Formation of 06-Methyldeoxyguanosine by 1-p-Carboxyl 3,3-dimethylphenyltriazene in DNA and 06-AlkylguanineDNA Alkyltransferase in Human Peripheral Leukocytes1 Slow Ming Lee,2 Peter J. O'Connor, Nicholas Thatcher, Derek Crowther, Geoffrey P. Margison, and Donald P. Cooper CRC Department of Carcinogenesis, Paterson Institute for Cancer Research (S. M. L., P. J. 0., G. P. M., D. P. C.J, and CRC Department ofMedical Oncology, Christie Hospital National Health Service Trust [S. M. L., N. T., D. C.], Manchester, M20 9BX, UK ABSTRACT There is increasing experimental evidence to indicate that 06-meth yldeoxyguanosine (06-MedG) formation in DNA is a critical cytotoxic event following exposure to certain antitumor alkylating agents and that the DNA repair protein O@-alkylguanine-DNA-alkyltransferase(ATase) can confer resistance to these agents. We recently demonstrated a wide interindividual variation in the depletion and subsequent regeneration of ATase in peripheral blood lymphocytes of patients treated with 24-h continuous infusion of 1-p-carboxyl-3,3-dimethylphenyltnazene (CB1O- 277) for metastatic melanoma. We have now measured the formation of O't-MedG in the DNA of peripheral leukocytes of nine patients receiving this treatment regimen. This lesion could be detected in DNA within 1h and a progressive increase in adduct levels occurred during the CB1O-277 infusion and for 24 h after completion. Considerable interindividual var iation was observed in the peak O'@-ModG levels, with values ranging from 3.0 to 23.8 @unol O'-MedG/mol deoxyguanosine (mean, 12.3 ±6.4 @.tmol 06-MedG/mol deoxyguanosine) following the first treatment cycle, possi bly as a consequence of differences in the capacity of patients to metab olize CB1O-277 to a methylating agent. There was, nevertheless, a clear temporal relationship between the progressive formation of leukocyte O@-MedGand lymphocyte ATase depletion. Repeated-measures regres sion showed that this was statisticaHy significant (P < 0.001) during the CB1O.277 infusion. A significant inverse correlation was also seen between pretreatment lymphocyte ATase activity and peak O'@-MedGlevels in leukocyte DNA (r = —0.73)and the area under the leukocyte O'@-MedG concentration-time curve (r —0.76).Metabolism of CB1O-277 to a methylating agent could be one factor that combines with DNA repair capacity to determine clinical response, because the two responses oh served in this series occurred in the two patients with the highest leukocyte O@-MedGlevels and also the lowest pretreatment ATase activity. Hema tological toxicity developed in the same two patients. INTRODUCTION The treatment of metastatic melanoma remains unsatisfactory. Al though DTIC3 is still regarded as the most effective chemotherapeutic agent and regularly produces a response rate of 20%, only 4% of patients (jrimarily those with soft tissue metastasis) show a complete response and responses are rare in cases with visceral metastasis (i, 2). It has been suggested that one possible reason for the relatively poor clinical activity of DTIC may be the relatively inefficient met abolic activation, compared to murine models. Thus, following equiv alent doses, the plasma levels of the active monomethyl metabolite 5-(3-methyl-1-triazeno)imidazole-4-carboxamide are much lower in rats and humans than in mice (3). Another triazene, CB1O-277 (Fig. 1), has been shown to have marked activity against experimental Received 1/10/94; accepted 5/24/94. I Supported by funds from the Christie Hospital (National Health Service) Trust Endowment Fund, North West Regional Health Authority, and the Cancer Research Campaign (UK). 2 To whom requests for reprints should be addressed. 3 The abbreviations used are: DTIC (dacarbazine), 5-(3,3-dimethyl-1-triazeno)imida zole-4-carboxamide; ATase, O―-alkylguanine-DNA alkyltransferase; AUC, area under the concentration-time curve; CB1O-277, 1-p-carboxyl-3,3-dimethylphenyltriazene; O―-MedG,O―-methyldeoxyguanosine. Research. on October 15, 2021. © 1994 American Association for Cancer cancerres.aacrjournals.org Downloaded from