Search

The development of effective chemotherapy for tumors of the central nervous system (CNS) is complicated in that the blood-brain barrier (BBB) hampers the penetration of most drugs into the brain and cerebrospinal fluid (CSF). This review summarizes the main reports on the distribution to CNS tumors and peritumoral normal brain of antitumor agents such as epipodophyllotoxins, cis-diamminedichloroplatinum(II), some nitrosoureas, bleomycin, vinblastine, and other clinically used antitumor agents as well as that of some experimental compounds with specific physicochemical properties. Drug levels were measured at surgical resection or in autopsy samples taken from patients who presented with different primary brain tumors or with brain metastases from extracerebral tumors. The observations made in each study were summarized in some detail, and the main points were then evaluated comparatively so as to highlight common aspects in the pharmacokinetic patterns of antitumor agents in human CNS tumors. Independently of their physicochemical properties, most antitumor agents appear to accumulate to a greater extent and to persist longer in intracerebral tumors than in the normal peritumoral brain. From in vitro cytotoxicity assays, it appears that epipodophyllotoxins, platinum compounds, bleomycin, and nitrosoureas reach potentially active therapeutic concentrations at the tumor target. However, all drugs have difficulty in reaching brain tissue adjacent to the tumor, as the intact BBB hampers their penetration. Plasma and CSF drug concentrations usually give little useful indication of the absolute quantity of drugs in brain tumors. To obtain a clear understanding of the CNS distribution of antitumor agents, one must determine whether the compound being measured is actually responsible for the observed activity and must consider the role of metabolites in the effect of the parent drug.

The differential distribution of antitumoral agents such as adriamycin, methylnitrosourea, cyclophosphamide and methotrexate in the vegetating and necrotic part of the intramuscular Lewis Lung carcinoma of C57B1 mice is investigated. As expected, these drugs are found to be mostly absent in the necrotic, while accumulating in the viable part of the tumor. Our findings suggest that measurement of antitumoral agents at the tumor level should be performed in the viable part of the tumor in order to get a proper estimate of the actual drug concentrations reached.

Two closely related lines of the same Walker 256 carcinoma in Crl-CD/COBS rats, described as behaving differently as regards tumor growth and host reaction, show different chemotherapeutic sensitivity to cyclophosphamide (CPA). Line B, which induces early cachexia with marked anorexia, is only moderately sensitive to CPA, while line A, which causes mild anorexia and only terminal cachexia, shows marked responsiveness to CPA, cure being attained in 75% of animals treated with a single dose of 120 mg/kg and in 90-100% of those given 20 mg/kg every other day. Comparative studies in both tumor lines on the distribution of CPA in vivo and on its metabolism by the liver perfusion technique showed no appreciable differences between the two lines in the pharmacokinetics of the compound, but indicate a much greater metabolizing capacity of CPA in the Walker 256/A animals. In vitro metabolic and covalent binding studies confirm that the liver of the Walker 256/A group metabolizes and covalently binds twice as much CPA as the liver of the Walker 256/B group. Conversely to Walker B, microsomal preparations of the Walker tumor line A are able to metabolize CPA to intermediates which irreversibly bind to tissue macromolecules, suggesting an in situ activation of the compound in the sensitive Walker tumor.

The interaction between cyclophosphamide (CPA) and phenobarbital (PB) was investigated in B6D2F, mice, checking both the antileukemic and immunosuppressive activity together with the serum levels of CPA and its metabolites. A reduced cytotoxic activity of CPA has been observed when PB is given for 2 days before CPA and an interval of at least 6 hours elapses between the last treatment of PB and the administration of CPA. On the contrary, when PB is given simultaneously with CPA for 2 or 4 consecutive days, an increased antileukemic activity of CPA occurs. In the experimental condition where PB decreases the activity of CPA, serum levels of CPA, assayed by means of a new specific gas-chromatographic method, and of its NBP-alkylating metabolites, indicate that this effect may be explained on a pure pharmacokinetic basis. However, for the situation where an increased effect of CPA was observed under the influence of PB, pharmacokinetic data did not provide a clear explanation.

Previous findings show that experimental tumors of the mouse, such as Ehrlich carcinoma and Sarcoma 180, transplanted in different sites, elicit a different sensitivity to the same drug according to the site of tumor implantation. Cyclophosphamide is highly active on intraperitoneal or subcutaneous tumors, but is ineffective on intracerebral tumors. However, the opposite is true for drugs such as nitroso-derivatives, which are active only on intracerebral tumors. The studies presented in this report were performed in order to elucidate whether different concentrations of drugs at the three tumor sites are related to the different antitumoral effects. It was found that measurements of blood and tissue concentrations explain fairly well the different activities of the drugs being tested.

The distribution of adriamycin (AM) in C57Bl/6 mice bearing intramuscular Lewis lung carcinoma under the influence of combined treatment with warfarin (W) was investigated by a fluorimetric procedure. AM was injected IV at the dose of 7.5 mg/kg 14 days after tumor transplantation and W was given in the drinking water for 96 h, starting 24 h before AM.. No substantial modifications in the serum and tissue distribution of AM fluorescence were observed under combined short-term treatment with W.

The acute cardiac toxicity of three second-generation anthracycline analogues and doxorubicin was compared in a model of the rat isolated Langendorff perfused heart. The drugs, doxorubicin (DX), 4-epi-doxorubicin (4′EDX), 4-demethoxy-daunorubicin (4DMDR) and 4′-deoxy-doxorubicin (4′dxDX) were infused for 40 min at a concentration of 26 μM into the isolated hearts. All four compounds significantly reduced cardiac work and its first derivative. The time to 50% decrease in work (TW50) was respectively 36, 23, 9 and 7 min for DX, 4′EDX, 4′dxDx and 4DMDR. The three anthracycline derivatives, but not DX, significantly increased coronary resistance. Heart rate was reduced by all compounds compared with baseline, but not compared with controls. Rhythm disturbances were seen with all five hearts perfused with 4DMDR, which stopped beating before 40 min; 2/5 hearts in the 4′EDX group and 1/5 hearts in the 4′dxDX group also stopped before the end of perfusion. All the compounds reached concentrations in the myocardium 8 to 50 times higher than in the perfusing medium. The more cardiotoxic the compound, the higher was its myocardial concentration; a significant correlation was found for all four agents. Noradrenaline was never measurable in the perfusate of control and DX hearts; perfusion with the three anthracycline derivatives caused some release, but the pattern was not clearcut and the maximum concentrations attained in the perfusate were relatively low (< 1ṁ6 times 10−9 M). In conclusion, in the rat isolated perfused heart, the early cardiotoxicity induced by equimolar concentrations of the three anthracycline compounds was greater than that induced by DX and was directly related to drug accumulation in the myocardium. Catecholamines do not seem to have a major role in the development of toxicity in this model.

Cyclophosphamide (CP) injected at different treatment schedules into CD male rats, prolongs narcosis after either parenteral or oral administration of pentobarbital. However, it takes 7 days before any change appears after CP pretreatment. The level of pentobarbital in blood and brain at different times after injection was higher in CP pretreated than in untreated animals. Since previous studies reported a decreased rate of inactivation of drugs after treatment with cytotoxic agents, experiments were performed to ascertain whether, beside the impairment of the metabolism, the animals respond differently to the same brain concentration of pentobarbital.

A “tissue culture” method for testing the cytotoxic activity of compounds active through the formation of metabolites is described and cyclophosphamide (CPA), which is known to exert its effect only after hepatic biotransformation, has been chosen as “model” substance. A monolayer of cultured KB cells is treated for one hour with CPA at 37°C, in the presence of different amounts of microsomal enzymes (105,000 × g fraction) and their co-factors and at the end of the incubation the medium is removed and changed. A dose-effect relation is observed by measuring the inhibition of growth or the destruction of the KB cells after 24 hr. The significance of this procedure for screening purposes or for studying the possible cytotoxicity of unstable metabolites is outlined.

The rate of pentobarbital disappearance from the medium was followed during the perfusions of livers isolated from normal and Walker 256 tumor-bearing rats. The half life of pentobarbital found when tumor liver was perfused with tumor blood was ten times higher (323 min) than that found in normal conditions. The cross experiments—normal liver perfused with tumor blood and vice versa—demonstrated that blood of tumor-bearing animals decreased the rate of pentobarbital metabolism by liver tissue.

SUMMARY A comparative study of the immunosuppressive activities of 6-mercaptopurine and its derivative azathioprine has been performed. It has been shown that in mice and rats, 6-mercaptopurine is markedly superior to azathioprine in reducing the number of spleen hemolytic plaque-forming cells when injected by parenteral routes, whereas azathioprine is more effective by the oral route. Determinations of the levels of 6-mercaptopurine in serum and spleen after injection of both agents did not reveal substantial differences in the concentration of this compound. The possible role of the imidazole moiety released from the in vivo splitting of azathioprine in the reduced activity of the latter drug given parenterally is advanced.

Rats and mice transplanted with different solid or ascites tumors are more sensitive than controls to the paralyzing effect of zoxazolamine and to the hypnotic action of pentobarbital. This increased effectiveness of pentobarbital can be partly explained by the fact that tumor-bearing animals show, in relation to controls, higher serum and brain concentrations of these two drugs. Other experiments indicate that zoxazolamine is less metabolized “ in vitro ” when incubated with liver homogenates obtained from tumor-bearing animals relative to controls. Also other substrates of liver microsomal enzymes, such as aniline, p -nitro-anisol and amidopyrine, are metabolized to a lesser extent, per unit of liver proteins, in tumor-bearing vs. control animals. In the case of pentobarbital, it is demonstrated that the impairment of metabolism is not the only factor responsible for the increased effect of pentobarbital in tumor-bearing animals because the onset of the narcosis is present at lower concentrations of brain pentobarbital as compared to controls. The surgical removal of the tumor abolishes the impairment of drug metabolism. Preliminary experiments seem to indicate that the serum of tumor-bearing animals contains principle(s) able to inhibit the activity of the microsomal liver enzymes.

A combined treatment with warfarin (W) and the alkylating agent nitrogen mustard (HN 2 ) was performed from the human heteroploid cell line (EUE). Warfarin was added to the cultures and after 24 hr contact the W treatment was repeated together with different HN 2 doses. Cytotoxicity was measured as percentage reduction in cell density and the chromosomal effects were evaluated by determining the frequency of chromosomal aberrations. As compared to EUE cultures treated only with HN 2 a significant increase in the toxicity values and in the amount of cytogenetic damage was found in cultures submitted to a combined treatment of HN 2 and W, which does not display toxicity in the experimental conditions utilized.

Adriamycin (NSC-123, 127) is an anticancer drug of the anthracycline type widely used in the clinical practice for the treatment of different malignancies (1). Intercalation with DNA has been suggested as having a role in the mechanism of action of this drug (2, 3, 4). Recently, reports by Sinha (5) and Lucacchini et al. (6) suggested a covalent interaction of Adriamycin (AM) with DNA and proteins. These studies, however, were not conclusive since they were carried out by spectrophotometric and Chromatographie procedures. In the present study a covalent interaction between AM and microsomal proteins was demonstrated in a more “classical” covalent binding study with the 14 C-labelled compound.

Besides the liver, other organs such as lung, kidney, and placenta possess microsomal enzymes. Since several antitumoral drugs are metabolized by this enzyme system, it was of interest to investigate whether the microsomal fraction of the tumor tissue was able to metabolize foreign compounds. The present paper shows that within the sensitivity of the method, 9000 × g and 105,000 × g fractions of various experimental tumors of the rat and the mouse do not metabolize known substrates of the microsomal enzymes such as p-nitro-anisol and aniline. Lack of metabolism was also observed by increasing the amount of tumor tissue employed, the amount of cofactors and the incubation time. Neither pretreatment of the tumor-bearing animals with phenobarbital or 3-methylcholanthrene induces metabolization.

The antitumor activity of mafosfamide (MFA), and its parent compound cyclophosphamide (CTX), was investigated against an ovarian reticular cell sarcoma growing i.m. in C57BI/6 mice (M5), which is very sensitive to CTX, and against a subline of this tumor (R16) resistant to CTX. MFA is the prototype of a class of oxazaphosphorines which do not require metabolic activation since under physiologic conditions they undergo rapid spontaneous hydrolyzation to the activated 4-hydroxyoxazaphosphorine and retain a spectrum of activity very similar to the parent compound. After a single dose (300 mg/kg × 1) or repeated low doses (100 mg/kg × 6) the antitumoral activity of MFA on the M5 tumor appeared comparable to or only slightly lower than CTX; the highest T/C value for median survival times was 167% in MFA-treated mice vs. 176% in the CTX group. MFA showed no activity against the R16 subline, thus indicating cross-resistance between the two drugs. Marked thickening of the glissonian capsule with compression of the lobular area of the liver, observed on i.p. administration of MFA, did not result in histopathologic abnormalities of the hepatic parenchyma. The therapeutic efficacy of MFA was similar with the i.p. and the i.v. route. MFA may represent a good candidate to replace CTX in cases in which a compound acting per se, and not through metabolites, is preferred.

The differential distribution of doxorubicin (Adriamycin® = AM) and daunorubicin (Daunomycin = DM) within the blood components, after an i.v. injection of 10 or 15 mg/kg of body weight, was investigated in Cd-COBS rats. AM and DM were measured by a fluorimetric procedure, and the unchanged compound was separated from its metabolites and quantified by means of the TLC scanning fluorescence technique. AM accumulated in the following order (of decreasing percentages): plasma and red cells (RBC), white cells (WBC), and platelets (PT), but the absolute amount of drug that reached each cell type was related to its relative volume. In the presence of higher blood concentrations (after injection of 15 mg/kg of body weight) the RBCs accumulated much more AM than the plasma, WBC, and PT; suggesting that the RBC fraction has a greater capacity to concentrate the drug. However, if the concentration of AM is expressed per unit volume of each component, a markedly higher value was observed for PT, and this was confirmed byin vitro results obtained by incubating blood in the presence of AM. DM seemed to be distributed on a percent basis to a greater extent than AM in the RBC fraction. Both compounds were taken up by blood cells, particularly platelets, to levels in excess of the extracellular concentration. This work was supported by NCI (Contract No. NOI CM 67064) and CNR (Progetto Finalizzato: Control of Cancer Growth).

1. The pharmacokinetics of cyclophosphamide and its alkylating metabolites have been studied in rats whose liver microsomal enzymes had been induced by phenobarbital pre-treatment. 2. Serum levels of cyclophosphamide were determined using a new g.l.c. method. The half-life of cyclophosphamide in blood of rats pre-treated with phenobarbital was shorter than in control rats. This change is closely related to higher rates of production of p-nitrobenzylpyridine-positive alkylating metabolites of cyclophosphamide, which in turn is followed by their more rapid disappearance from the circulation. 3. Urinary excretion reflects this situation; lower amounts of cyclophosphamide and higher concentrations of its alkylating metabolites are present in the urine of phenobarbital-treated rats. 4. Perfusion of livers isolated from phenobarbital-pre-treated rats confirmed the results in vivo. With this preparation, too, disappearance of cyclophosphamide was more rapid and formation of its alkylating metabolites was accelerated after phenobarbital treatment.

The relative distribution of adriamycin to plasma and blood cells after IV injection of 10 mg/kg was investigated in CD rats bearing intramuscular 256 Walker carcinosarcomas 15 days old. The drug was measured by a fluorimetric procedure and the amount of unchanged compound was separated from metabolites and quantitated by means of a TLC scanning fluorescence technique. In the presence of a tumor much lower hematocrit values are found, with marked anemia and thrombocytopenia associated with leukocytosis. These modified hematologic parameters account for an altered pattern of drug distribution. The low number of blood cells per milliliter results in a smaller amount of drug being present in the cellular fraction, so that more of the compound (even twice as much) is made available in plasma. Changes in adriamycin concentrations per unit volume or cell of each cell type are inversely related to changes in their relative number per milliliter. The only cell fraction where the drug increase per cell or cubic micrometer does not compensate the marked reduction in cell count observed in the presence of tumor is the platelet fraction, in which adriamycin amounts are 25% or less of those observed in the blood of normal rats, indicating that these blood cells become saturated in tumor-bearing animals.

The covalent binding of hexamethylmelamine (HMM) and its metabolites was studied in liver, tumor, blood, kidney, spleen, lung, brain, heart, and small intestine after a single IP injection of 2,4,6-14C-hexamethylmelamine (50 mg/kg) to C57Bl/6J female mice bearing 20-day-old M5076/73A ovarian cancer. Covalent binding to tissue macromolecules was measured 2, 10, and 40 h after injection of the drug. At 2 h liver and small intestine showed the highest levels of irreversibly bound metabolites, the lowest being found in brain and heart. Except in the small intestine, where a decrease was observed between 2 and 10 h, the level of covalent binding was constant up to 40 h. HMM metabolism was also studied. Tissue distribution of pentamethylmelamine (PMM), 2,2,4,6-tetramethylmelamine (TMM), and 2,4,6-trimethylmelamine (TriMM) was determined at the three times considered. At 2 h the drug was already extensively metabolized, TriMM being the major metabolite among those determined.

6-Aminochrysene (6AC) and a series of derivatives, which are selective inhibitors of the growth of a spontaneous mammary carcinoma in mice and a rhabdomyosarcoma in rats, antagonize the muscular paralysis induced by zoxazolamine in Sprague-Dawley male rats. This effect is shown to be maximal 24 hr after treatment and decreases at longer intervals. The route of administration of both drugs does not appear to be important. Measurements of blood and brain zoxazolamine in rats pretreated with 6AC indicate lower levels of the compound as compared with controls. Furthermore in vitro experiments show that the 9000g liver fraction of 6AC pretreated rats metabolize zoxazolamine twice as rapidly as controls. On the contrary, 6AC pretreatment does not reduce the duration of paralysis induced by carisoprodol, the sleeping time by pentobarbital, the convulsions or the toxicity by strychnine. Moreover, the in vitro metabolism of two microsomal substrates, p-NO 2 anisol and aniline, is not affected by 6AC. These findings may well account for the strong specific inhibition of 6AC on zoxazolamine induced paralysis and are consistent with the hypothesis that 6AC resembles polycyclic hydrocarbons in its stimulating activity on hepatic microsomal enzymes.