Extracellularly tumor-activated prodrugs for the selective chemotherapy of cancer: application to doxorubicin and preliminary in vitro and in vivo studies.

Oligopeptidic derivatives of anthracyclines unable to penetrate cells were prepared and screened for their stability in human blood and their reactivation by peptidases secreted by cancer cells. N-beta-alanyl-L-leucyl-L-alanyl-L-leucyl-doxorubicin was selected as a new candidate prodrug. The NH2-terminal beta-alanine allows a very good blood stability. A two-step activation by peptidases found in conditioned media of cancer cells ultimately yields N-L-leucyl-doxorubicin. In vitro, when MCF-7/6 cancer cells are exposed to the prodrug, they accumulate about 14 times more doxorubicin than MRC-5 normal fibroblasts, whereas when exposed to doxorubicin the uptake is slightly higher in fibroblasts than in MCF-7/6 cells. This increased specificity of the prodrug over doxorubicin was confirmed in cytotoxicity assays using the same cell types. In vivo, the prodrug proved about nine times less toxic than doxorubicin in the normal mouse and also much more efficient in two different experimental chemotherapy models of human breast tumors.

Effect of MRC-5 fibroblast conditioned medium on breast cancer cell motility and invasion in vitro.

We used Transwell chambers to study separately cellular motility and invasion. In order to assess the cellular motility, polycarbonate microporous filters were coated with extracellular matrix proteins which adsorbed on the filters without clogging the pores. To investigate the invasive behavior of tumor cells, filters were covered with a layer of Matrigel which clogged the pores. The motility and the invasion of breast cancer cell lines (MDA-MB-231, MCF-7/6 and MCF-7/AZ cells) were assessed quantitatively in different culture media: defined (serum-free), serum-containing and normal human fibroblast MRC-5 conditioned media. In serum-containing medium, tumor cells migrated and invaded through the coated and covered filters. Their motility and invasion potentials were considerably lower in defined medium, whereas medium conditioned by MRC-5 fibroblasts stimulated both motility and invasion but not growth. The MRC-5 conditioned medium induced also the spreading of clusters of MCF-7/6 cells grown on Matrigel-coated plates.

Human pharmacokinetics of N-L-leucyl-doxorubicin, a new anthracycline derivative, and its correlation with clinical toxicities.

A pharmacokinetic study of N-L-leucyl-doxorubicin, a new derivative of doxorubicin, has been undertaken during a phase I trial in 19 patients with advanced cancer after intravenous bolus administration at doses ranging from 30 to 240 mg/m2. The pharmacokinetics of N-L-leucyl-doxorubicin was linear with a total body clearance of 41.3 +/- 25.7 L/hr/m2. N-L-leucyl-doxorubicin was extensively metabolized into doxorubicin, which appeared in plasma immediately after N-L-leucyl-doxorubicin infusion. The mean molar doxorubicin/N-L-leucyl-doxorubicin area under the curve (AUC) ratio was 0.49 +/- 0.22 and was independent of the administered dose. A relationship has been established between the doxorubicin AUC (r = 0.74; p less than 0.001) and the surviving factor in white blood cell counts. Other toxic side effects (thrombocytopenia or stomatitis) did not correlate with any pharmacokinetic parameter. These findings suggest that the degree of metabolization of N-L-leucyl-doxorubicin into doxorubicin may be responsible for the toxicity, that is, N-L-leucyl-doxorubicin may simply represent a pro-drug for doxorubicin.

Biodetermination of N-(deacetyl-O-4-vinblastoyl-23)-L-tryptophan, a metabolite of vintriptol, by high-performance liquid chromatography with fluorescence detection.

The determination of N-(deacetyl-O-4-vinblastoyl-23)-L-tryptophan (vintriptol acid, VtrpA), a metabolite of the investigational semi-synthetic vinca alkaloid vintriptol [N-(deacetyl-O-4-vinblastoyl-23)-L-ethyltryptophan, VtrpE], in plasma and urine samples is described. Sample pretreatment included liquid-liquid extraction of the buffered (pH 5.0) biological samples with chloroform-2-propanol (95:5, v/v). The analyses were performed by ion-exchange high-performance liquid chromatography on normal-phase silica with fluorescence detection. The assay was applied to the analysis of samples from cancer patients who had been treated with VtrpE in a phase I clinical study. VtrpA was found to be a principal metabolite of VtrpE with up to 1.2% of the administered dose excreted in the urine.

High-performance liquid chromatographic determination of vinblastine, 4-O-deacetylvinblastine and the potential metabolite 4-O-deacetylvinblastine-3-oic acid in biological fluids.

Procedures for the determination of vinblastine (VBL), 4-O-deacetylvinblastine (DVBL) and 4-O-deacetylvinblastine-3-oic acid (DVBLA) in biological samples using high-performance liquid chromatography (HPLC) combined with selective sample clean-up are presented. VBL and DVBL were determined in plasma and urine using ion-exchange normal-phase HPLC with fluorescence detection. The limit of detection was 1 microgram/l for both compounds using a 500-microliter sample. Successful chromatographic analyses of DVBLA were achieved by using a glass column packed with 5-microns Hypersil ODS and acetonitrile-0.05 M phosphate buffer (pH 2.7) (23:77, v/v). Positive identification was supported by the use of diode-array detection. The limit of detection (at 270 nm) was 10 micrograms/l using 1-ml samples.

Cytotoxic activity of daunorubicin or vindesin conjugated to a monoclonal antibody on cultured MCF-7 breast carcinoma cells.

Conjugates were constructed between daunorubicin or vindesin and a monoclonal antibody to human milk fat globule membrane associated antigen. This antibody recognizes a high molecular weight glycoprotein present at the cell surface of human normal and tumour epithelial cells; after specific binding to plasma membrane of cultured MCF-7 human breast carcinoma cells, it is endocytosed and gains access to lysosomes, wherein it is broken down (Aboud-Pirak et al., Cancer Res 48: 3188-3196, 1988). Covalent linkage of daunorubicin (through a succinylated tetrapeptide arm) or of vindesin (through a hemisuccinate arm) yields conjugates with maximal molar ratios (drug molecule/specific IgG under monomeric form, i.e. unaggregated) or 2.0 and 4.5 respectively. The conjugate with daunorubicin inhibits the binding of the 3H labelled antibody to MCF-7 cells as efficiently as the native unconjugated antibody, whereas the conjugate with vindesin inhibits it only by 56%. Both conjugates are entirely stable in plasma and serum; after 24 hr incubation at pH 4.8 in the presence of rat liver lysosomal enzymes, 60 and 33% of daunorubicin and vindesin respectively are released from the conjugates. Adherent non-confluent cultures of cells recognized (MCF-7) or not (Hep-G2, human hepatocarcinoma cells) by the antibody were incubated from 1 hr to 6 days with different concentrations of daunorubicin or vindesin, free or conjugated to the specific or to a control monoclonal antibody. LD50, defined as the drug concentration required to reach 50% of the amount of cell associated protein obtained in the absence of drug were determined at the end of 6 days continuous incubation or after shorter incubation followed by reincubation in drug free medium up to 6 days. Both cell lines are almost equally susceptible to the free drugs. The conjugate between daunorubicin and the antibody appears inactive, even at saturating concentrations of antibody. This could result from the extrusion out of the cells of daunorubicin molecules released from the conjugate, impairing the drug to reach the intracellular concentration required for cytotoxicity. In contrast, conjugation of vindesin to the specific but not to a control antibody restricts the activity of the drug to cells selectively recognized by the specific antibody. However, even after corrections for the loss of immunoreactivity and for the incomplete release of vindesin from the conjugate, cytotoxicity is achieved at higher concentrations or requires longer exposure to the conjugated than to the free drug.(ABSTRACT TRUNCATED AT 400 WORDS)

Alkylating activity in serum, urine, and CSF following high-dose ifosfamide in children.

The pharmacokinetics of alkylating activity were studied in 17 children treated i.v. with ifosfamide (IF) at 3 g/m2 as a 1-h infusion for 2 consecutive days every 3 weeks, with mesna as a uroprotector. Two patients were treated for a newly diagnosed rhabdomyosarcoma according to the current SIOP (International Society of Pediatric Oncology) protocol. The other 15 patients were treated in a phase II study and presented with one of the following malignancies in relapse: neuroblastoma (7), osteosarcoma (3), soft tissue sarcoma (2), Wilms' tumor (1), non-Hodgkin's lymphoma (1), and acute lymphoblastic leukemia (1). Plasma alkylating activity levels determined by using 4(4'-nitro-benzyl)-pyridine showed considerable inter-individual and intercyclic variations and decreased biphasically, with mean alpha and beta half-lives of 60 min and 6-7 h, respectively. Probably as a result of liver mixed-function oxidase induction, on the 2nd day of treatment the terminal half-lives were shorter, the plasma exposures were lower, and the mean plasma clearances were higher. Renal excretion was almost complete after 24 h, accounting for a mean of 19% of the injected dose. The CSF alkylating activity levels, obtained in four children, were always lower than the plasma levels and ranged from 8 to 51 micrograms/ml, with a mean CSF/plasma ratio of 0.53 +/- 0.23 during the first 12 h. We conclude that IF alkylating activity was biphasically cleared from the plasma, with significant interindividual and intercyclic variability, that the renal contribution to the clearance was low, and that high levels of CSF alkylating activity could possibly contribute to the CNS toxic side effects observed in pediatric patients treated with high-dose IF/mesna.

Immunological evaluation of blood contamination in tissue distribution studies.

An original, simple and economical method allowing the evaluation of blood contamination in tissue distribution studies is described. This method is based on the immunochemical determination of the amount of serum albumin present in tissue samples as a quantitative measurement of blood contamination, using a single radial immunodiffusion technique according to Mancina et al. Two pharmacokinetic studies we have performed in mice for different research purposes illustrate the crucial importance of the correction for blood contamination in the exact interpretation of the tissue uptake of a given compound. We have indeed clearly demonstrated first in the development of drug-carrier complexes in targeted cancer chemotherapy and second in the study of receptor-mediated clearance of glycoproteins from the plasma that the blood contamination, if not taken into account, can greatly alter the data of uptake by target organs. Our method is thus particularly useful in in vivo pharmacokinetic studies and is moreover not restricted to one animal species.

Coupling products of amino acids to penicillin V and cephalothin: synthesis and susceptibility to carboxypeptidases and lysosomal enzymes.

Amino acids have been coupled to the carboxyl group of penicillin V and cephalothin by methods that keep the beta-lactam ring intact. Derivatives were successfully obtained with both neutral (Leu, Val, Ala, Ile, Trp, Tyr, Gly) and one acidic (Glu) amino acids. The new compounds were inactive in vitro against Staphylococcus aureus or Micrococcus luteus. Incubation in the presence of purified carboxypeptidases (A, B), soluble lysosomal fractions from liver, or cellular homogenates from liver, kidney, fibroblasts, and macrophages did not allow recovery of the antibacterial activity. Injection in mice also failed to cause liberation of microbiologically active compounds. HPLC studies confirmed that the amide linkage between the antibiotic and the amino acid was not hydrolyzed in the presence of soluble lysosomal fractions from liver. However, conversion of cephalothin and cephalothin-leucine to desacetyl derivatives was observed in the presence of soluble lysosomal fractions and extracts from liver and semipurified orange peel acetylesterase(s). It is concluded that amino acid derivatives of beta-lactam antibiotics do not offer potential chemotherapeutic use as prodrugs.

[Targeting of anthracyclines and hepatomas]. / Pilotage des anthracyclines et hépatomes.

DNR can be targeted to the liver by linking to galactosylated human serum albumin (AG). The linkage is stable in the blood stream and allows the release of active DNR after endocytosis of the conjugate by the target cells. Receptors for AG are present at the cell membrane of hepatocytes, primary human hepatoma cells and lung metastases. After i.v. administration of AG-DNR more than 70% of the drug is taken up by the liver and in rats 50% of DNR is eliminated in the bile after 16 h. Nude mice bearing human hepatoblastoma cells implanted s.c. were treated twice a week with a dose equivalent to 6.6 mg/kg of DNR either for the free drug or the conjugate AG-DNR. After 7 injections, tumor growth is inhibited in both case, however, the conjugate seems more active and is at least twice less toxic in terms of LD50. A phase I clinical trial of AG-DNR on 11 patients bearing hepatocarcinoma revealed that despite a transitory hyperthermia (37-38 degrees C) during the first day of the 4 day-perfusion, and modifications of hepatic enzymes in 3 cirrhotic patients, no hematologic and cardiac toxicity could be detected. A subjective response has been obtained in half of the patients with a decrease of plasmatic alpha-foetoprotein levels by more than 50% in 4 patients and one complete remission of more than 23 months with disparition of pulmonary metastases.

Doxorubicin and daunorubicin plasmatic, hepatic and renal disposition in the rabbit with or without enterohepatic circulation.

The pharmacokinetics, metabolism and disposition of doxorubicin and daunorubicin were studied for periods up to 100 hr in rabbits with (group II) or without a biliary fistula (groups I and III) and with (group I) or without (groups II and III) ligatured ureters using high-performance liquid chromatography to separate parent drug and metabolites. The plasma decay of doxorubicin and daunorubicin was triexponential. Metabolites appearing in the plasma after doxorubicin and daunorubicin bolus i.v. injection were respectively doxorubicinol and daunorubicinol, the latter being the major compound after daunorubicin injection. The elimination of daunorubicin was faster than that of doxorubicin. No differences in the elimination were observed between the 3 groups. In bile, 21% of the injected dose of doxorubicin were excreted mainly as the parent drug and 60% of the injected dose of daunorubicin were excreted, mainly as daunorubicinol. Enterohepatic circulation did not affect the biliary excretion of both doxorubicin and daunorubicin. Ligature of ureters increased slightly the biliary excretion of doxorubicin. The hepatic clearance of daunorubicin was greater than that of doxorubicin. The total urinary excretion was not different between the II and III groups and amounted to 11.6 and 12.8% of the injected dose of doxorubicin and daunorubicin, respectively. Metabolic ratios of doxorubicinol/doxorubicin and daunorubicinol/daunorubicin were similar in bile and urine.

Pharmacokinetics of daunorubicinol in the rabbit: comparison with daunorubicin.

The pharmacokinetics of daunorubicinol (DOL), the main metabolite of daunorubicin (DNR), was studied in rabbits and compared to that of daunorubicin after an 8 mg/kg dose. High-performance liquid chromatography was used to separate parent drug and metabolites. The plasma disappearance of DNR and DOL was triexponential. DOL was the major species detected in plasma and urine. Both drugs had large volumes of distribution. About 70% of DNR or DOL were bound to plasma proteins and mainly to albumin. Pharmacokinetic parameters of DOL obtained after injection of DOL were different from those calculated for DNR and those calculated for DOL after injection of DNR. The total urinary excretions of DNR or DOL were similar and amounted to 25% of the dose. No conjugates were identified in urine after enzymatic treatment. No fluorescent drug was identified in the feces. Anthracyclines were degraded in vitro in rabbit feces. The rabbit seems to be a good model for the study of anthracycline pharmacokinetics as our results in rabbits after DNR injection were similar to those in human studies.

Degradation of doxorubicin and daunorubicin in human and rabbit biological fluids.

The stability of doxorubicin (DOX) and daunorubicin (DNR) in rabbit and human plasma, bile and urine and in rabbit faeces was studied in the presence or absence of light, and at body, room and cold room temperatures. Fluorescence was determined by spectrofluorimetry after normal and reversed phase HPLC. Under each set of conditions, DOX and DNR fluorescence decreased with time; the decrease was more rapid with DOX. As the parent drugs were degraded, apolar compounds were formed which behaved like 7-deoxyaglycones and generally did not compensate for the loss in fluorescence of the parent drug. The degradation of anthracyclines occurred even in the absence of light, was not due to bacterial contamination and was faster at higher pH or temperature. The rapid degradation of DOX and DNR in biological fluids at body temperature may have implications on the disposition of anthracyclines in vivo. Prior to analysis, biological fluids and solutions containing anthracyclines should be processed quickly at 4 degrees C, in the absence of light, and at a pH no greater than 6 to avoid degradation.

Doxorubicin pharmacokinetics in the rabbit.

The levels of doxorubicin, doxorubicinol and DOX aglycone achieved in plasma, urine and feces after i.v. administration of doxorubicin at 7.9 mg/kg into five New Zealand White rabbits were determined by high-performance liquid chromatography and fluorometry. The parent drug was the major compound found in the plasma. Doxorubicinol and DOX aglycone only appeared in trace amounts in the plasma. The plasma disappearance curves were triphasic with mean half-lives of 2 min, 18 min and 15 h for the alpha-, beta- and gamma-phases, respectively. Plasma protein binding was studied by ultracentrifugation and electrophoresis. Doxorubicin was mainly bound to albumin to an extent of 73.9%. The mean percentage of the administered dose excreted in the urine was 8.0%. Doxorubicinol and doxorubicin were the major compounds present in the urine with small amounts of DOX aglycone. Neither the parent drug nor any fluorescent metabolites were identified in the feces. The rabbit seems to be a good experimental model for the pharmacokinetics study of anthracyclines since the plasma and urine pharmacokinetics parameters and the extent of metabolism are similar in both humans and rabbits.

High dose melphalan in children with advanced malignant disease. A pharmacokinetic study.

Nine children with poor-prognosis malignancies--seven with advanced neuroblastoma and two with metastatic Ewing's sarcoma--were given high doses of melphalan (HDM), 150 mg/m2 (3 patients) and 180 mg/m2 (6 patients), as a 'late intensification' agent combined with noncryopreserved autologous bone marrow transplants. Melphalan levels in the plasma decreased biphasically, with mean half-lives of 6.6 min and 3.0 h. At the time of marrow reinfusion (12-21 h after HDM) the melphalan plasma level was generally below 0.1 microgram/ml. The renal contribution to melphalan clearance was low, a mean of 5.8% of the injected dose being found in patients' urine over the 12 h following HDM administration. No significant difference was seen in pharmacokinetic parameters between patients undergoing and not undergoing forced diuresis.

Cellular pharmacology of amino acid derivatives of daunorubicin and doxorubicin in suspension of renal proximal tubules.

One approach currently being used to target drug action selectively to specific cells and tissues is to link active drugs to proteins and peptides that are preferentially recognized by, distributed to or activated by the target cells. Inasmuch as the kidney proximal tubule cells are very active in recapturing and catabolizing peptides and proteins which appear in the glomerular filtrate, we have examined the cellular pharmacology of daunorubicin (DNR) and doxorubicin (DOX) and selected amino acid and dipeptide derivatives in suspensions of rabbit renal proximal proximal tubules. The tubules accumulated the DNR series of drugs and their metabolites to a greater extent than the DOX series. Although all of the amino acid derivatives of these drugs entered the cells for sequestration and metabolism, the dipeptide derivative, alanyl-leucyl-DNR, was not detected within the cells. Using high-performance liquid chromatography to quantify the various metabolites and isopycnic centrifugation of tubule-derived post-nuclear supernates in linear sucrose gradients to resolve various subcellular organelles, the subcellular sites of metabolism of these drugs were examined. A NADPH-dependent reduction of the C-13 carbonyl group of the parent drugs, which was the primary route of metabolism, was localized to the cytoplasm. Formation of aglycones generated by the cleavage of the daunosamine moiety from the anthracycline core followed the microsomal marker, NADPH-cytochrome reductase, in the sucrose gradients. Removal of the terminal amino acid from alanyl-leucyl-DNR was tentatively assigned to a cysteine-requiring enzyme on the plasma membrane.(ABSTRACT TRUNCATED AT 250 WORDS)

Cellular pharmacology of detorubicin and doxorubicin in L1210 cells.

Detorubicin (DET), a semi-synthetic analog of daunorubicin, releases at neutral pH doxorubicin (DOX) upon hydrolysis. DET enters faster than DOX into the cultured L1210 cells and reaches higher intracellular levels. When the cells are incubated for 120 min at pH 6.5, in spite of its rapid hydrolysis, one third of the intracellular fluorescence was due to undegraded DET. DET, like the other anthracyclines studied, is found associated intracellularly only to the lysosomes and to the nuclei. Unchanged DET is found mainly inside the lysosomes where the drug is stabilized, while DOX is found essentially associated to the nuclear DNA of L1210 cells. DET can therefore be viewed as an hydrophobic prodrug of DOX characterized, however, by a distinct subcellular localization in L1210 cells.

Metabolism of daunorubicin in sensitive and resistant Ehrlich ascites tumor cells. Determination by high pressure liquid chromatography.

The intracellular metabolism of daunorubicin (DNR) has been studied in sensitive and resistant Ehrlich ascites tumor (EAT) cells. The subcellular localization of metabolites has been followed by normal-phase and reverse-phase high-pressure liquid chromatography (HPLC). The metabolism of DNR by either sensitive or resistant EAT cells is not significant; unmetabolized DNR is always the main intracellular compound. Daunorubicinol (DOL) accounts for less than 5% after 24 h and an unidentified product is also observed. This highly apolar compound, having an intrinsic fluorescence one order of magnitude greater than that of DNR is formed in acellular conditions and could be a chemical artifact. DNR and DOL are mainly associated with DNA-containing fractions. No significant differences can be observed in the metabolism of DNR in sensitive and resistant EAT cells.