Modulation of thalidomide pharmacokinetics by cyclophosphamide or 5,6-dimethylxanthenone-4-acetic acid (DMXAA) in mice: the role of tumour necrosis factor

PURPOSE

There is considerable current interest in the use of thalidomide as a single agent or in combination with drugs such as cyclophosphamide in the treatment of multiple myeloma and other cancers. Our previous work has shown that thalidomide potentiates the antitumour activity of both cyclophosphamide and 5,6-dimethylxanthenone-4-acetic acid (DMXAA) against murine Colon 38 tumours. In both of these cases, thalidomide extends the half-life (t(1/2)) of the other drug. We wished to determine whether cyclophosphamide and DMXAA altered the t(1/2) of thalidomide. Since both thalidomide and DMXAA modulate tumour necrosis factor (TNF), we also wished to determine the role of TNF in this interaction.

METHODS

Mice with Colon 38 tumours were treated with cyclophosphamide (220 mg/kg) and/or thalidomide (20 mg/kg) or DMXAA (25 mg/kg) and thalidomide (100 mg/kg), combinations that have previously demonstrated synergistic activity. Plasma and tumour tissue drug concentrations were analysed by high-performance liquid chromatography. To determine the role of TNF, similar experiments were performed using mice defective in the TNF gene (TNF(-/-)) or the TNF receptor-1 gene (TNFR1(-/-)).

RESULTS

Coadministration of cyclophosphamide increased the thalidomide t(1/2) by 3.9- and 3.6-fold, respectively, in plasma and tumour tissue, with a corresponding increase in the concentration-time curve (AUC). The corresponding values following coadministration of DMXAA were 3.0- and 4.6-fold, respectively. Coadministration of cyclophosphamide had similar effects on thalidomide t(1/2) in C57Bl/6, TNF(-/-) and TNFR1(-/-) mice, while coadministration of DMXAA did not alter the t(1/2) or AUC in TNF(-/-) and TNFR1(-/-) mice.

CONCLUSIONS

Both cyclophosphamide and DMXAA have a pharmacokinetic interaction with thalidomide, increasing t(1/2) and AUC. TNF mediates the effect of DMXAA on thalidomide pharmacokinetics but not that of cyclophosphamide.

Metabolism of Thalidomide in Liver Microsomes of Mice, Rabbits, and Humans

Thalidomide is increasingly important in clinical treatment, not only of various inflammatory conditions but also in multiple myeloma and other malignancies. Moreover, the metabolism of thalidomide varies considerably among different species, indicating a need to understand its mechanistic basis. Our previous in vivo studies showed the plasma half-life of thalidomide to be much shorter in mice than in humans, with rabbits showing intermediate values. We were unable to detect hydroxylated thalidomide metabolites in humans and suggested that interspecies differences in thalidomide hydroxylation might account for the differences in plasma half-life. We sought here to establish whether these species differences in the formation of hydroxylated thalidomide metabolites could be discerned from in vitro studies. Liver microsomes of mice, rabbit, and human donors were incubated with thalidomide and analyzed using liquid chromatography-mass spectrometry. Hydrolysis products were detected for all three species, and the rates of formation were similar to those for spontaneous hydrolysis, except in rabbits where phthaloylisoglutamine formation increased linearly with microsomal enzyme concentration. Multiple hydroxylation products were detected, including three dihydroxylated metabolites not observed in vivo. Thalidomide-5-O-glucuronide, detected in vivo, was absent in vitro. The amount of 5-hydroxythalidomide formed was high in mice, lower in rabbits, and barely detectable in humans. We conclude that major interspecies differences in hepatic metabolism of thalidomide relate closely to the rate of in vivo metabolite formation. The very low rate of in vitro and in vivo hydroxylation in humans strongly suggests that thalidomide hydroxylation is not a requirement for clinical anticancer activity.

Relationship between tumour endothelial cell apoptosis and tumour blood flow shutdown following treatment with the antivascular agent DMXAA in mice

5,6-Dimethylxanthenone-4-acetic acid (DMXAA) is currently undergoing clinical evaluation as an antivascular agent for the treatment of cancer. We have previously demonstrated that DMXAA induces apoptosis of vascular endothelial cells in murine tumour sections and in a breast carcinoma biopsy from one patient in a Phase I trial. We wished to determine the tissue selectivity of this effect and its relationship to induced blood flow changes. Mice with Colon 38 tumours were treated with DMXAA and tissues were examined for apoptosis by TdT-mediated dUTP nick-end labelling (TUNEL). Hoechst 33342 was used to stain functional vessels, with the loss of stained vessels used as a measure of tumour vascular collapse. Treatment with DMXAA at 25 mg kg⁻¹, its maximum tolerated dose (MTD), showed, after 3 h, a 12-fold increase in TUNEL staining of tumour vascular endothelial cells. In contrast, tissue from the heart, brain, liver and spleen showed no increase. Induction of apoptosis in tumour tissue was both dose-dependent, observable at doses as low as 5 mg kg⁻¹, and time-dependent. Apoptosis was significantly lower in Colon 38 tumours of mice, with a targeted disruption in the TNF gene (TNF^−/−), or in the TNF receptor 1 gene (TNFR^−/−), as compared with that in wild-type mice. Increasing the DMXAA dose to 50 mg kg⁻¹ in these knockout mice raised tumour apoptosis to a level comparable to that induced in wild-type mice given DMXAA at the MTD. For all the data, a significant correlation (r=0.94; P<0.001) was found between logarithmic percentage apoptosis induction and the logarithmic density of Hoechst-stained vessels. These results suggest that blood flow inhibition caused by DMXAA is tumour tissue-specific and is a consequence of induction of apoptosis in tumour vascular endothelial cells.

Induction of tumour necrosis factor and interferon-γ in cultured murine splenocytes by the antivascular agent DMXAA and its metabolites

The induction of haemorrhagic necrosis by 5,6-dimethylxanthenone-4-acetic acid (DMXAA) in transplantable murine tumours depends on the in situ synthesis of cytokines, particularly tumour necrosis factor (TNF). Since the in vivo action of DMXAA would be greatly clarified by the development of an in vitro model, we investigated whether DMXAA could induce cytokines in cultured murine splenocytes. DMXAA alone induced low amounts of TNF with an optimal concentration of 10 microg/mL and an optimal time of 4 hr. When combined with low concentrations of lipopolysaccharide, deactivated-lipopolysaccharide (dLPS) or phorbol-12-myristate-13-acetate that did not elicit TNF production alone, synergistic TNF production was obtained. DMXAA also induced interferon-gamma at an optimal dose of 300 microg/mL, but the addition of dLPS had no further effect. Decreasing culture pH, although not changing the optimal concentrations for stimulation, increased both TNF and interferon-gamma production in response to DMXAA. The major DMXAA metabolites, DMXAA-glucuronide and 6-hydroxy-5-methylxanthenone-4-acetic acid, did not induce either cytokine alone, in combination with dLPS or at low pH. The results indicate that DMXAA rather than a metabolite is responsible for cytokine induction and suggest that the microenvironment of the tumour may be responsible for the observed selective induction of cytokines in tumour tissue.

Modelling cell death in human tumour cell lines exposed to the anticancer drug paclitaxel

Most anti-cancer drugs in use today exert their effects by inducing a programmed cell death mechanism. This process, termed apoptosis, is accompanied by degradation of the DNA and produces cells with a range of DNA contents. We have previously developed a phase transition mathematical model to describe the mammalian cell division cycle in terms of cell cycle phases and the transition rates between these phases. We now extend this model here to incorporate a transition to a programmed cell death phase whereby cellular DNA is progressively degraded with time. We have utilised the technique of flow cytometry to analyse the behaviour of a melanoma cell line (NZM13) that was exposed to paclitaxel, a drug used frequently in the treatment of cancer. The flow cytometry profiles included a complex mixture of living cells whose DNA content was increasing with time and dying cells whose DNA content was decreasing with time. Application of the mathematical model enabled estimation of the rate constant for entry of mitotic cells into apoptosis (0.035 per hour) and the duration of the period of DNA degradation (51 hours). These results provide a dynamic model of the action of an anticancer drug that can be extended to improve the clinical outcome in individual cancer patients.

Improvement of the antitumor activity of intraperitoneally and orally administered 5,6-dimethylxanthenone-4-acetic acid by optimal scheduling

PURPOSE

5,6-Dimethylxanthenone-4-acetic acid (DMXAA), a new anticancer drug that has recently completed Phase I clinical trial, is effective against transplantable murine tumors with established vasculature. We wished to determine the relationship between administration schedule and antitumor activity.

EXPERIMENTAL DESIGN

C57Bl/6 mice with s.c. implanted Colon 38 tumors were used for determination of maximal tolerated doses and tumor growth delay. Plasma and tissue DMXAA concentrations were measured by high-performance liquid chromatography.

RESULTS

Continuous infusion (30 mg/kg/day for 3 days) and daily i.p. administration schedules (7.5 mg/kg) were ineffective. A pharmacokinetically guided schedule was developed to increase tumor tissue drug concentrations without increasing the maximal plasma concentration. A schedule comprising a loading dose (25 mg/kg, i.p.) followed by supplementary doses (5 mg/kg after 4 and 8 h) provided a 1.6-fold increase in tumor tissue area under the concentration-time curve, no increased toxicity, and superior antitumor activity (100% cure rate, as compared with 55% for a single i.p. dose of 25 mg/kg). A similar strategy was developed for oral administration with a loading dose (30 mg/kg) and supplementary doses (15 mg/kg after 4 and 8 h). It provided a 90% cure rate, in contrast to a single oral dose (0% cure rate).

CONCLUSIONS

The antitumor action of DMXAA is schedule dependent, and the achievement of an adequate tumor tissue DMXAA concentration above a threshold value appears to be critical for activity. The use of a pharmacokinetically guided schedule provides excellent oral activity against Colon 38 tumors and provides a basis for developing more effective administration schedules in clinical trials.

Mechanisms of Action of DNA Intercalating Acridine-based Drugs: How Important are Contributions from Electron Transfer and Oxidative Stress?

Reactive oxygen species (ROS) are produced continuously in living cells as a by-product of respiration and other metabolic activity. Some ROS may react with DNA, and in some cases may abstract an electron from the double helix, leading to long range electron transfer (ET) reactions. Thus, the DNA of living cells may be in a continuous state of ET. We consider here whether acridine-based anticancer or antimicrobial drugs, which bind to DNA by intercalation, might either donate electrons to, or accept electrons from, the double helix, thus actively participating in ET reactions. We focus in particular on two acridine-based drugs that have been tested against human cancer in the clinic. Amsacrine is a 9-anilinoacridine derivative that appears to act as an electron donor in ET reactions on DNA, while N-[2-(dimethylamino)ethyl]acridine-4-carboxamide (DACA) may act as an electron acceptor. Such reactions may make important contributions to the antitumor activity of these drugs.

A mathematical model for analysis of the cell cycle in cell lines derived from human tumors

The growth of human cancers is characterised by long and variable cell cycle times that are controlled by stochastic events prior to DNA replication and cell division. Treatment with radiotherapy or chemotherapy induces a complex chain of events involving reversible cell cycle arrest and cell death. In this paper we have developed a mathematical model that has the potential to describe the growth of human tumour cells and their responses to therapy. We have used the model to predict the response of cells to mitotic arrest, and have compared the results to experimental data using a human melanoma cell line exposed to the anticancer drug paclitaxel. Cells were analysed for DNA content at multiple time points by flow cytometry. An excellent correspondence was obtained between predicted and experimental data. We discuss possible extensions to the model to describe the behaviour of cell populations in vivo.

Dual topoisomerase I/II inhibitors in cancer therapy

While the majority of topoisomerase (topo) inhibitors show selectivity against either topo I or topo II, a small class of compounds can act against both enzymes. These can be divided into three classes. The first and largest class comprise drugs that bind to DNA by intercalation and include the clinically-evaluated acridine DACA, the benzopyridoindole intoplicine, the indenoquinolinone TAS-103, the benzophenazine XR11576, and the pyrazoloacridine NSC 366140. The second category comprises hybrid molecules, prepared by physically linking separate inhibitors of topo I and topo II, or by linking pure topo inhibitors to other DNA-interactive carriers. While several derivatives (e.g., camptothecin-epipodophyllotoxin and ellipticine-distamycin hybrids) have been prepared, there have been no detailed studies. The third category are less well defined as a structural class, but apparently recognize structural motifs that are present in both topo I and II enzymes. These include a series of benzoisoquinolinium quaternary salts such as NK 109, and more interestingly modified versions of classical topo I or topo II inhibitors; e.g., the modified camptothecin BN 80927 and the modified epipodophyllotoxin tafluposide (F-11782). There is as yet no detailed understanding of the factors that result in selective or dual inhibition, but structure-activity studies in several classes show that structural changes can influence topo I/II selectivity. DNA intercalation mode also appears to play a part. The basis for the high antitumor activity of some topo inhibitors is not yet understood but may depend on the complex pattern of activities that include both inhibition and poisoning of the two enzymes.

Clinical aspects of a phase I trial of 5,6-dimethylxanthenone-4-acetic acid (DMXAA), a novel antivascular agent

The antitumour action of 5,6-dimethylxanthenone-4-acetic acid (DMXAA) is mediated through tumour-selective antivascular effects and cytokine induction. This clinical phase I trial was conducted to examine its toxicity, maximum tolerated dose, pharmacokinetics (PK) and pharmacodynamics (PD). A secondary objective was to assess its antitumour efficacy. DMXAA was administered every 3 weeks as a 20-min i.v. infusion. Dose escalation initially followed a modified Fibonacci schema but was also guided by PK and toxicity. A total of 63 patients received 161 courses of DMXAA over 19 dose levels ranging from 6 to 4900 mg m(-2). DMXAA was well tolerated at lower doses and no drug-related myelosuppression was seen. Rapidly reversible dose-limiting toxicities were observed at 4900 mg m(-2), including confusion, tremor, slurred speech, visual disturbance, anxiety, urinary incontinence and possible left ventricular failure. Transient prolongation of the corrected cardiac QT interval was seen in 13 patients evaluated at doses of 2000 mg m(-2) and above. A patient with metastatic cervical carcinoma achieved an unconfirmed partial response at 1100 mg m(-2), progressing after eight courses. The results of PK and PD studies are reported separately. DMXAA has antitumour activity at well-tolerated doses.

Paclitaxel induces nucleolar enlargement in dorsal root ganglion neurons in vivo reducing oxaliplatin toxicity

Paclitaxel and oxaliplatin are promising drugs for combination trials but both induce peripheral neurotoxicity. To investigate this toxicity, 10-week-old female Wistar rats were given single intraperitoneal doses of paclitaxel and oxaliplatin, alone or in combination. Neurotoxicity was assessed by L5 dorsal root ganglion morphometry and H-reflex-related sensory nerve conduction velocity. Platinum concentrations in dorsal root ganglia and plasma were measured by inductively coupled plasma mass spectrometry. Dorsal root ganglion nucleolus size was significantly increased following single doses of paclitaxel of 10 and 20 mg kg(-1) at 24 h and 6 days (P<0.02). In contrast, dorsal root ganglion nucleolus size was significantly decreased following single doses of oxaliplatin ranging from 3 to 30 mg kg(-1) at time points ranging from 2 h to 14 days. Sensory nerve conduction velocity was altered after a single dose of oxaliplatin but not after paclitaxel. In combination with oxaliplatin, paclitaxel did not alter the plasma pharmacokinetics or dorsal root ganglion accumulation of oxaliplatin-derived platinum. However, prior paclitaxel inhibited oxaliplatin-induced reductions of dorsal root ganglion nucleolar diameter (P<0.02). Sensory nerve conduction velocity was reduced after oxaliplatin alone (P&<0.05) but unchanged when paclitaxel was given before oxaliplatin. In conclusion, paclitaxel induces nucleolar enlargement in dorsal root ganglion neurons after pharmacologically relevant doses in vivo and reduces oxaliplatin nucleolar damage and neurotoxicity.

Preclinical factors affecting the interindividual variability in the clearance of the investigational anti-cancer drug 5,6-dimethylxanthenone-4-acetic acid

Cancer chemotherapy is characterized by significant interindividual variations in systemic clearance, therapeutic response, and toxicity. These variations are due mainly to genetic factors, leading to alterations in drug metabolism and/or target proteins. The aim of this study was to determine, using a human liver bank (N=14), the interindividual variations in the expression and activity of liver enzymes that metabolize the investigational anticancer drug 5,6-dimethylxanthenone-4-acetic acid (DMXAA), i.e cytochrome P450 (CYP1A2) and uridine diphosphate glucuronosyltransferase (UGT1A9/2B7). In addition, interindividual variations in enzyme inhibition, hydrolysis of DMXAA acyl glucuronide (DMXAA-G) by plasma and hepatic microsomes, and the binding of DMXAA by plasma proteins also were examined. The results indicated that there was approximately one order of magnitude of interindividual variation in the expression of CYP1A2 and UGT2B7, activity of the enzymes toward DMXAA, and inhibition potency (IC(50)) by diclofenac, cyproheptadine, and alpha-naphthoflavone. The enzyme activities toward DMXAA and IC(50) values were closely correlated with enzyme expression. There was a smaller (2- to 3-fold) variation in the enzyme-catalyzed hydrolysis of DMXAA acyl glucuronide in human plasma and liver microsomes (N=6) and in the binding of DMXAA by plasma proteins in humans. In conclusion, the interindividual variability of DMXAA disposition observed in vitro might reflect the greater elimination variability (>one order of magnitude) in Phase I cancer patients. The variability in DMXAA clearance in these cancer patients would be due mainly to differences in its metabolism and its metabolic inhibition by co-administered drugs. To a lesser extent, variability in the clearance of DMXAA could be due to the hydrolysis of its acyl glucuronide and/or its binding to plasma proteins. Further study is needed to examine the genotype-phenotype relationship, and the result, together with therapeutic drug monitoring may provide a useful strategy for optimizing DMXAA treatment.

NF-kappa B activation in vivo in both host and tumour cells by the antivascular agent 5,6-dimethylxanthenone-4-acetic acid (DMXAA)

5,6-Dimethylxanthenone-4-acetic acid (DMXAA), a new anticancer agent developed in this centre, has an antivascular action and causes regression of transplantable murine tumours that is mediated partially by the intratumoral production of tumour necrosis factor (TNF). DMXAA activates the nuclear factor-kappaB (NF-kappaB) transcription factor, which is involved in TNF synthesis and has also been suggested to mediate resistance to TNF. We wished to determine whether tumour cell NF-kappaB activation modulated the in vitro and in vivo effects of DMXAA. We compared the response of the 70Z/3 pre-B lymphoma cell line with that of its mutant 1.3E2 sub-line, which has a defective gamma-subunit of IKK, the kinase that phosphorylates IkappaB leading to NF-kappaB activation. As shown by electrophoretic mobility shift assays (EMSAs), DMXAA induced in vitro translocation of NF-kappaB (p50 and p65 subunits) into the nucleus of 70Z/3 cells, but not of 1.3E2 cells. However, when the cell lines were then grown as subcutaneous tumours in mice and treated with DMXAA (25 mg/kg), activation of NF-kappaB was found in nuclear extracts prepared from both 70/Z3 and 1.3E2 tumours, as well as from Colon 38 tumours that were used for comparison. This suggests that DMXAA induces NF-kappaB responses in host components of the tumour. Tumours grown from both 70Z/3 and 1.3E2 cells were found to regress completely following DMXAA treatment. Thus, the antitumour action of DMXAA appears to be independent of the ability of the target tumour cell population to induce NF-kappaB expression. Moreover, activation of NF-kappaB in the tumour cell did not confer resistance to DMXAA-induced therapy.

Thalidomide metabolites in mice and patients with multiple myeloma

PURPOSE

This research examines the profile of metabolites of thalidomide that are formed in refractory multiple myeloma patients undergoing thalidomide therapy in comparison with those that are detected in healthy mice.

EXPERIMENTAL DESIGN

Urine or plasma samples from patients during thalidomide therapy (100-400 mg daily), or from mice treated i.p. (100 mg/kg) or p.o. with thalidomide (50 mg/kg) were analyzed using liquid chromatography-mass spectrometry. Metabolites in each of the peaks observed in the UV- and mass spectrometry-detected high-performance liquid chromatography traces were identified by comparison of retention times and spectra with those of authentic standards.

RESULTS

Plasma and urine samples from mice 4 h after treatment with thalidomide contained eight major metabolites formed by hydroxylation and/or hydrolysis of thalidomide. In contrast, urine samples from seven multiple myeloma patients at steady state levels of thalidomide therapy showed the presence of only three hydrolysis breakdown products and no hydroxylated metabolites.

CONCLUSIONS

Our results show that thalidomide metabolite profiles in multiple myeloma patients differ considerably from those in mice. The lack of measurable hydroxylated metabolites in urine and in 1 case plasma of these patients suggests that such metabolites are not responsible for the therapeutic effects of thalidomide in multiple myeloma. We suggest that thalidomide may act directly, down-regulating growth factors essential for multiple myeloma growth.

Uptake of the antivascular agent 5,6-dimethylxanthenone-4-acetic acid (DMXAA) and activation of NF-kappaB in human tumor cell lines

5,6-Dimethylxanthenone-4-acetic acid (DMXAA), a new anticancer drug synthesized in this laboratory and currently in clinical trial, induces tumor vascular damage in vivo that is mediated primarily by cytokine synthesis by host cells. Although its pharmacology and antitumor activity have been extensively studied, little is known of its action on tumor cell lines. We measured [3H]DMXAA uptake in the Raji, Daudi, Jurkat, ECV304, NZM12, HL60, and K562 human tumor lines using velocity centrifugation through silicon oil layers, and also measured NF-kappaB activation by electrophoretic mobility shift assays. All lines accumulated [3H]DMXAA, and uptake by ECV304 cells was rapid, pH dependent (greater uptake at pH 6.5), similar at 4 degrees C and 37 degrees C, and unaffected by the addition of 5 mM sodium azide. The uptake ratio was 4.5-fold at a low drug concentration (4 microM) and decreased significantly (P < 0.01) to 4.0 as the external drug concentration was increased to 0.7 mM, providing evidence of saturability. [3H]DMXAA interacted weakly with isolated cytoplasmic proteins, as measured by equilibrium dialysis, providing a basis for the observed cellular uptake. Uptake was slightly reduced by addition of a less potent analogue, flavone acetic acid, or of an inactive analogue, 8-methylxanthenone-4-acetic acid, suggesting competition for binding sites. The Raji, Daudi, Jurkat. and ECV304 lines showed evidence of activation of the NF-kappaB transcription factor in response to DMXAA, but the identity of the NF-kappaB subunits translocated to the nucleus varied according to the line. The results are consistent with the hypothesis that DMXAA is taken up rapidly into cells by passive diffusion and binds to cellular proteins. The observed activation of NF-kappaB in some lines suggests that the effects of DMXAA on tumor cells, as well as host cells, must be considered in understanding its antitumoraction.

Synthesis and cytotoxic activity of carboxamide derivatives of benzo[b][1,6]naphthyridines

The reaction of 4-dimethylaminomethylene-6-methyl-4H-pyrano[4,3-b]quinoline-1,3-dione with a range of primary amines gave rise to a series of 2-substituted 6-methyl-1-oxo-1,2-dihydrobenzo[b][1,6]naphthyridine-4-carboxylic acids. The derived 4-N-[2-(dimethylamino)ethyl]carboxamides were tested for growth inhibitory properties against murine P388 leukemia, Lewis lung carcinoma (LLTC), and human Jurkat leukemia cell lines. Most compounds were potent cytotoxins, with some having IC(50) values less than 10 nM. Five were tested in vivo against subcutaneous colon 38 tumors in mice, and a single dose (3.9 mg/kg) proved to be curative for the 2-methyl and 2-(3,4-dimethoxyphenyl) derivatives in this refractory model.

Antivascular therapy of cancer: DMXAA

The vascular endothelium of tumour tissue, which differs in several ways from that of normal tissues, is a potential target for selective anticancer therapy. By contrast with antiangiogenic agents, antivascular agents target the endothelial cells of existing tumour blood vessels, causing distortion or damage and consequently decreasing tumour blood flow. DMXAA (5,6-dimethylxanthenone-4-acetic acid), a low-molecular-weight drug, has a striking antivascular and in some cases curative effect in experimental tumours. Its action on vascular endothelial cells seems to involve a cascade of events leading to induction of tumour haemorrhagic necrosis. These events include both direct and indirect effects, the latter involving the release of further vasoactive agents, such as serotonin, tumour necrosis factor, other cytokines, and nitric oxide from host cells. Phase I clinical trials of DMXAA have been completed and the next challenge to face is how the antivascular effect of this drug should be exploited for the treatment of human cancer.

Estimation of Radiation-Induced Interphase Cell Death in Cultures of Human Tumor Material and in Cell Lines

A short-term assay method able to estimate the radiation response of human cancer tissue samples would be of great advantage to the individualization of radiotherapy in cancer patients. However, the effect of radiation on [3H]thymidine incorporation by proliferating cells reflects a composite of cell cycle arrest and induced cell death pathways. Here we consider whether it is feasible to correct for cell cycle effects based on comparison of the effects of radiation and the mitotic inhibitor paclitaxel on [3H]thymidine incorporation. Sixty-two short-term (7-day) cultures of human tumor tissue from 61 patients with melanoma, gynecological cancer, brain cancer, and head and neck cancer, as well as 18 5-day cultures of low passage human tumor cell lines, were irradiated at doses from 2 to 9 Gy, or exposed to paclitaxel (200 nM). [3H]Thymidine incorporation was measured at the end of the incubation. Cell cycle times could be estimated from the paclitaxel data and were 2.7 to 18.6 days for melanomas, 2.5 to >40 days for carcinomas, 3.9 to 39 days for brain tumors, and 1.1 to 3.8 days for cell lines. The effects of radiation on [3H]thymidine incorporation varied widely (0-97% and 0-99% inhibition for 2 and 9 Gy, respectively), and in 23 of the clinical samples, but in none of the cell lines, radiation caused significantly greater inhibition of [3H]thymidine incorporation than paclitaxel (p < 0.05). We argue that that these differences reflect radiation-induced cell loss from G1 phase and/or S phase. Responses of short-term cultures of clinical tumor material to radiation, with appropriate correction for cell cycle effects, might have the potential to provide information on radiation-induced cell death in individual patients.

Effect of 3-Fluorothalidomide and 3-Methylthalidomide Enantiomers on Tumor Necrosis Factor Production and Antitumor Responses to the Antivascular Agent 5,6-Dimethylxanthenone-4-Acetic Acid (DMXAA)

5,6-Dimethylxanthenone-4-acetic acid (DMXAA) is an antivascular drug that induces tumor necrosis factor (TNF) in mice. Thalidomide inhibits TNF induction by DMXAA and also potentiates its antitumor activity. We investigated whether these effects were enantiomer specific, using the R- or S-enantiomers of two nonracemizable thalidomide analogues. Racemic 3-fluorothalidomide (3FThal) and racemic 3-methylthalidomide (3MeThal) were separated into enantiomers of greater than 98% optical purity using preparative chiral column chromatography. C57Bl/6 mice implanted with subcutaneous Colon 38 tumors were treated with DMXAA (25 mg/kg) alone or together with the pure R- or S-enantiomers by a single i.p. injection. TNF levels in the serum or tumor tissues 3 h after treatment were measured using ELISAs and tumor growth was also measured. 3FThal and 3MeThal, at their respective single maximum tolerated doses (MTD) of 15 and 50 mg/kg, were more toxic in mice than thalidomide (100 mg/kg). The R- and S-enantiomers of either 3FThal or 3MeThal, at their respective MTD, inhibited DMXAA-induced TNF activity in serum and tumor tissue, but no significant differences were observed between the enantiomers. Coadministration of racemic or enantiomers of 3FThal or 3MeThal at their respective MTD did not potentiate the antitumor responses above that obtained with DMXAA alone, and no enantioselectivity was apparent. We conclude that there is no advantage in using the nonracemizable thalidomide analogues to improve the antitumor activity of DMXAA.

Preclinical factors influencing the relative contributions of Phase I and II enzymes to the metabolism of the experimental anti-cancer drug 5,6-dimethylxanthenone-4-acetic acid

It is important to determine the relative contribution of each metabolic pathway (f(p)) and of enzymes to the net metabolism of a drug. The aim of this study was to investigate, using a human liver bank, the f(p) of the anti-cancer drug 5,6-dimethylxanthenone-4-acetic acid (DMXAA) and the effects of various inhibitors and inducers on f(p). The mean apparent K(m) and V(max) values (N=14) were 21+/-5 microM and 0.04+/-0.02 nmol/min/mg, respectively, for 6-methylhydroxylation, and 143+/-79 microM and 0.71+/-0.52 nmol/min/mg, respectively, for acyl glucuronidation in human liver microsomes. 6-Methylhydroxylation and acyl glucuronidation contributed 26 and 74%, respectively, to DMXAA metabolism at 5 microM; values were 7 and 93% at 350 microM DMXAA. There was a significant relationship between the ratio of metabolic activity by Phase II and I reactions (R(II/I)) and uridine diphosphate glucuronosyltransferase (UGT2B7) protein level (r=0.605, P=0.022), whereas a reverse correlation between R(II/I) and cytochrome P450 (CYP1A) protein level was observed (r=-0.540, P=0.046). Various compounds inhibited either DMXAA glucuronidation or 6-methylhydroxylation, or both pathways. Pretreatment of rats with beta-naphthoflavone, but not phenobarbitone and cimetidine, increased the percentage of the contribution by 6-methylhydroxylation to 17% from 4% of control at 5 microM DMXAA. Our results indicate that the f(p) of DMXAA is subject to substrate concentration, inhibition, induction, and the protein levels of enzymes that biotransform DMXAA. However, clinical studies are important to verify the conclusions drawn from in vitro data.

DMXAA: an antivascular agent with multiple host responses

PURPOSE

To measure host responses to the antivascular agent DMXAA (5,6-dimethylxanthenone-4-acetic acid) and to compare them with those of other antivascular agents.

METHODS

Induction of tumor necrosis was measured in s.c. murine Colon 38 carcinomas growing in normal or tumor necrosis factor (TNF) receptor-1 knockout mice. Plasma and tumor tissue TNF concentrations were measured by ELISA. Plasma concentrations of 5-hydroxyindoleacetic acid (as a measure of serotonin release) and nitrite (as a measure of nitric oxide release) were measured by high-performance liquid chromatography.

RESULTS

Administration of DMXAA to tumor-bearing mice increased plasma and tumor tissue-associated TNF, in addition to increasing plasma nitric oxide, distinguishing its action from that of mitotic poisons that had an antivascular action. Results from TNF receptor-1 knockout mice showed that TNF played an important role in both its antitumor action and its host toxicity. Release of serotonin occurred in response to mitotic poisons, as well as to DMXAA. CONCLUCIONS: The antivascular action of DMXAA involves in situ production in tumor tissue of a cascade of vasoactive events, including a direct effect on vascular endothelial cells and indirect vascular effects involving TNF, other cytokines, serotonin, and nitric oxide. Now that Phase I clinical trials of DMXAA are completed, the optimization of this cascade in cancer patients is a major challenge. Plasma 5-hydroxyindoleacetic acid concentrations may provide a useful surrogate marker for the antivascular effects of DMXAA and other antivascular agents.

Potential of DMXAA combination therapy for solid tumors

DMXAA is one of the first examples of a new class of anticancer agents that attack existing tumor blood vessels and thus deprives tumor tissue of an adequate blood supply. Its mechanism of action appears to rely on the induction within tumor tissue of cytokines, such as tumor necrosis factor. In experimental tumors, DMXAA interacts productively with radiation, hyperthermia and a number of chemotherapeutic drugs. This review discusses the mechanisms underlying such interactions and how these might be exploited in clinical cancer treatment.

Novel strategies for overcoming multidrug resistance in cancer

The problems of why metastatic cancers develop pleiotropic resistance to all available therapies, and how this might be countered, are the most pressing in cancer chemotherapy. It is likely that such resistance involves a combination of mechanisms including changes in drug transport/drug targets, reduction in the degree of drug-induced apoptosis/cell loss, and increased rate of tumour repopulation following therapy. Current research must consider not only which mechanisms contribute, eventually relating this to individual patients with cancer, but also what strategies might be utilised to counter each of the important resistance mechanisms. A considerable amount of work has been devoted to the development of inhibitors of membrane-associated transport proteins such as P-glycoprotein, which mediate drug efflux. This work is now being complemented by approaches that target cell death pathways such as those mediated by release of mitochondrial proteins and by activation of surface receptors such as Fas. Rapid progress has been made in developing small-molecular-weight drugs that influence the rate of apoptosis, for instance by binding to the bcl-2 family of proteins regulating mitochondrial permeability. Antisense approaches aimed at reducing bcl-2 expression, and thus increasing the rate of cell death, are also showing promise. Modification of repopulation kinetics provides a further approach but has not received as much attention as other aspects of tumour resistance. New therapeutic approaches will have to be complemented by improved diagnostic tests to evaluate the contributions of different resistance mechanisms in individual patients with cancer.

Potentiation of the antitumour effect of cyclophosphamide in mice by thalidomide

PURPOSE

Thalidomide has recently shown significant promise in the treatment of some types of cancer, and trials in combination with conventional chemotherapy are being undertaken. We wished to determine whether thalidomide potentiated the effect of cyclophosphamide, a commonly used cytotoxic drug, in a murine tumour model.

METHODS

C57Bl/6 mice implanted with subcutaneous Colon 38 tumours were treated with cyclophosphamide alone or together with thalidomide as a single intraperitoneal injection and tumour growth was measured. Concentrations of cyclophosphamide, 4-hydroxycyclophosphamide, 4-ketocyclophosphamide and 2-dechloroethylcyclophosphamide were determined in plasma, liver and tumour tissue using coupled high-performance liquid chromatography-mass spectrometry at different times after treatment.

RESULTS

Cyclophosphamide alone (220 mg/kg) induced growth delays of 11-13 days with no cures, whereas cyclophosphamide together with thalidomide (100 mg/kg) cured mice of their tumours. Thalidomide at lower doses (1-20 mg/kg) also potentiated the antitumour effect. Coadministration of thalidomide (100 mg/kg) dramatically decreased the clearance of cyclophosphamide and its metabolites from plasma and tissue, with corresponding increases in the area under the concentration-time curves. The magnitude of the effect was dependent on the dose of thalidomide over the range 1-20 mg/kg with no further effect at a dose of 100 mg/kg.

CONCLUSIONS

Coadministration of thalidomide and cyclophosphamide gave markedly greater activity against Colon 38 tumour compared with either drug alone. Investigation of the reason for this effect revealed thalidomide to possess the novel property of dramatically decreasing the clearance of cyclophosphamide and its metabolites.