Brain and plasma pharmacokinetics and anticancer activities of cyclophosphamide and phosphoramide mustard in the rat

Summary By a sensitive and quantitative fluorometric assay, brain and plasma time-dependent concentration profiles were generated for phosphoramide mustard (PM) and active alkylating metabolites derived from cyclophosphamide (CPA) administration to rats. Whereas PM rapidly disappeared from plasma, with a monophasic half-life of 15.1 min, equimolar administration of CPA generated active metabolites in plasma that disappeared monoexponentially, with a composite half-life of 63 min. As a consequence, the time-dependent concentration integral of active alkylating metabolites derived from CPA administration, calculated between 5 min and infinity, was 3-fold that of PM. Pharmacokinetic parameters were calculated for each compound. The brain/plasma concentration-integral ratios of PM and active alkylating metabolites derived from CPA were 0.18 and 0.20, respectively. The cerebrovascular permeability-surface area product of PM was 7.5×10^–5s^–1, which is similar to that of other watersoluble anticancer agents that are restricted from entering the brain. The activities of a range of daily doses of PM and CPA were assessed against subcutaneous and intracerebral implants of Walker 256 carcinosarcoma tumor in rats. Inhibition of subcutaneous tumor growth by 50% was caused by CPA and PM doses of 6.6 and 12.0 mg/kg (daily for 5 consecutive days, starting 36 h after tumor implantation), respectively. However, administration of daily doses of up to 40 mg/kg did not significantly increase the survival of animals with intracerebral tumor implants. These studies indicate that active metabolites of CPA are restricted from entering the brain and that only subtherapeutic concentrations are achieved in brain tissue after systemic administration of CPA or PM.Abbreviations CPA cyclophosphamide - PM phosphoramide mustard - 4-HC 4-hydroxycyclophosphamide - AP aldophosphamide - PA cerebrovascular permeability-surface area product     

Validation of a novel procedure for quantification of the formation of phosphoramide mustard by individuals treated with cyclophosphamide

Purpose

Use of the patient’s body surface area (mg m^?2) as a basis for dosing does not take individual variation in metabolic capacity and rate of clearance into account. Here, we evaluated a novel approach for individual monitoring of short-lived cytotoxic agents formed from cytostatic drugs such as cyclophosphamide (CP).

Methods

The accumulated blood dose of the cytotoxic active agent phosphoramide mustard (PAM) formed from CP was measured as a reaction product with hemoglobin (Hb adduct). This adduct, N-[2-(2-oxazolidonyl)ethyl]-valyl Hb (OzVal-Hb), was detached from Hb with the adduct FIRE procedure?, and the formed analyte was quantified using LC-MS/MS. This dose biomarker for PAM and the analytical procedure was evaluated in accordance with the guidelines on bioanalytical method validation formulated by the European Medicine Agency. The evaluated method was applied to quantify blood dose levels of PAM in female breast cancer patients (n = 12) before and after three cycles of polychemotherapy regimes containing CP.

Results

OzVal-Hb, a specific and stable biomarker, could be measured with great sensitivity (lower limit of quantification = 33 pmol g^?1 Hb), high accuracy (within ±20 %) and good repeatability (CV < 20 %). The inter-individual variability in the blood level of this adduct in women with breast cancer (n = 12) who received three doses of CP in combination with one or two other cytostatic drugs was 250 % following the first dose and approximately 150 % after each subsequent dose.

Conclusions

Measurement of the biomarker OzVal-Hb can be used to quantify the short-lived cytotoxic agent PAM in a single blood sample drawn several days after therapy. This procedure may aid in individualizing doses of CP, thereby improving efficacy while both reducing the risk of and increasing the predictability of side-effects.     

Comparative effects of cyclophosphamide, isophosphamide, 4-methylcyclophosphamide, and phosphoramide mustard on murine hematopoietic and immunocompetent cells.

The effects of equimolal doses of cyclophosphamide (CY), isophosphamide (IP), 4-methylcyclophosphamide (4-MCY), and phosphoramide mustard (PM) on murine hematopoietic spleen colonies and adoptively transferred antibody-forming cells in vivo were compared. Equimolal doses of the drugs produced significantly different effects. All the drugs exerted an increasing effect against the ability of adoptively transferred immunocompetent cells to produce a significant anti-sheep red blood cell titer as the length of time between cell transfer and drug administration was increased. The maximum effect was seen when a drug was given 48--72 hours after antigen and spleen cell transfer. CY and IP produced significantly greater immunosuppressive effects than did the other drugs at all times after cell transfer and at all doses administered. PM had the least immunosuppressive effect at each dose evaluated. Against hematopoietic spleen colonies, the cytotoxic effects of 4-MCY and PM were similar and, at most doses studied, significantly greater than the effect of either CY or IP. Inasmuch as PM is an active metabolite of CY, it appeared either that one of the prior metabolites of CY was responsible for this marked immunosuppressive effect or that due to differences in polarity, PM was differentially distributed within the two cell systems as compared to CY. The differences in hematopoietic effects among all drugs were much less than those seen against immunocompetent cells and were not dependent on time of drug administration.     

Specific p53 mutations detected in plasma and tumors of hepatocellular carcinoma patients by electrospray ionization mass spectrometry

Hepatocellular carcinoma (HCC), a common cause of cancer deaths worldwide, has several major etiological risk factors, including infection with the hepatitis viruses and exposure to aflatoxin B1. A specific missense mutation resulting from a guanine to thymine transversion at the third position of codon 249 in the p53 tumor suppressor gene has been reported in 10-70% of HCCs from areas of high dietary exposure to aflatoxin B1. Short oligonucleotide mass analysis was compared with DNA sequencing in 25 HCC samples for specific p53 mutations. Mutations were detected in 10 samples by short oligonucleotide mass analysis in agreement with DNA sequencing. Analysis of another 20 plasma and tumor pairs showed 11 tumors containing the specific mutation, and this change was detected in six of the paired plasma samples. Four of the plasma samples had detectable levels of the mutation; however, the tumors were negative, suggesting possible multiple independent HCCs. Ten plasma samples from healthy individuals were all negative. This molecular diagnostic technique has implications for prevention trials and for the early diagnosis of HCC.     

A proposed mechanism of resistance to cyclophosphamide and phosphoramide mustard in a Yoshida cell line in vitro

Summary A Yoshida sarcoma cell line (Y_R/cyclo) showing decreased sensitivity to metabolically activated cyclophosphamide in vitro has been shown to be cross-resistant to phosphoramide mustard, the ultimate alkylating agent formed from cyclophosphamide. Resistance to these alkylating agents has been shown to be associated with increased activity of the glutathione S-transferase group of enzymes, and with elevated levels of glutathione, the cosubstrate of the enzyme. The resistant cell line shows lower levels of cellular damage, as measured by alkaline elution following treatment with phosphoramide mustard, than the parental (Ys) line. The mechanism of resistance is ascribed to increased deactivation of potentially damaging metabolites of cyclophosphamide by the glutathione S-transferase enzymes, resulting in decreased cellular damage in the resistant cell line.This work was supported by a grant from the Cancer Research Campaign     

Pharmacokinetics of cyclophosphamide and its metabolites in paediatric patients receiving high-dose myeloablative therapy

Introduction

In order to better understand the impact of high-dose on the pharmacokinetics and metabolism of cyclophosphamide, a pharmacological study was performed in children with malignant mesenchymal tumours with metastatic disease.

Methods

Patients received four courses of chemotherapy including two courses of cyclophosphamide. Plasma concentrations of cyclophosphamide and the metabolites 4-ketocyclophosphamide, dechloroethylcyclophosphamide and carboxyphosphamide were determined on days 1, 2 and 3 of each course. A population pharmacokinetic model for cyclophosphamide was developed using non-linear mixed effects modelling and metabolite AUC values were compared between days and courses.

Results

Data were available on 21 cyclophosphamide courses from 15 patients. A one compartment model, incorporating a term in surface area for both CL and V, best described cyclophosphamide pharmacokinetics. Typical CL and V on day 1 of treatment for a patient with a SA of 1.4 m² were 4.3 L/h and 28.5 L, respectively. On days 2 and 3 CL increased by 88% (95% CI, 72-105%) and 125% (95% CI, 108-145%) over day 1 levels; V increased by 14% (95% CI, 5-23%) on days 2 and 3. V tended to be larger for males than similarly sized females but no effect of age was found upon CL or V. Significant increases in metabolite AUCs were observed on days 2 and 3 compared to day 1 and a significant increase in CXCP AUC from course 1 to course 3.

Conclusion

Administration of high-dose cyclophosphamide over several days results in an increase in metabolism, possibly by induction of the activation pathway. This induction is effectively reversed following a four week period between cyclophosphamide doses. The degree of intersubject variation in cyclophosphamide elimination is largely accounted for by body surface area and is less than previously reported.     

Quantitative analysis of plasma TP53 249Ser-mutated DNA by electrospray ionization mass spectrometry.

A mutation in codon 249 of the TP53 gene (249(Ser)), related to aflatoxin B(1) exposure, has previously been associated with hepatocellular carcinoma risk. Using a novel internal standard plasmid, plasma concentrations of 249(Ser)-mutated DNA were quantified by electrospray ionization mass spectrometry in 89 hepatocellular carcinoma cases, 42 cirrhotic patients, and 131 nonliver diseased control subjects, all from highly aflatoxin-exposed regions of The Gambia. The hepatocellular carcinoma cases had higher median plasma concentrations of 249(Ser) (2,800 copies/mL; interquartile range: 500-11,000) compared with either cirrhotic (500 copies/mL; interquartile range: 500-2,600) or control subjects (500 copies/mL; interquartile range: 500-2,000; P < 0.05). About half (52%) of the hepatocellular carcinoma cases had >2,500 copies of 249(Ser)/mL plasma, corresponding to the prevalence of this mutation in liver tumors in The Gambia. In comparison, only 15% of control group and 26% of cirrhotic participants exceeded this level (P < 0.05). Further subset analysis revealed a statistically significant, quantitative relation between diagnosis of hepatocellular carcinoma and levels of 249(Ser) detected at 2,501 to 10,000 copies/mL plasma (odds ratio, 3.8; 95% confidence interval, 1.3-10.9) and at >10,000 copies/mL plasma (odds ratio, 62; 95% confidence interval, 4.7-820) when compared with control subjects and after adjusting for age, gender, recruitment site, hepatitis B and C serologic status, and total DNA concentration. Levels of >10,000 copies of 249(Ser)/mL plasma were also significantly associated with the diagnosis of hepatocellular carcinoma (odds ratio, 15; 95% confidence interval, 1.6-140) when compared with cirrhotic patients. Potential applications for the quantification of 249(Ser) DNA in plasma include estimation of long-term, cumulative aflatoxin exposure and selection of appropriate high-risk individuals for targeted intervention.     

Determination of vinca alkaloids in human plasma by liquid chromatography/atmospheric pressure chemical ionization mass spectrometry

 A sensitive assay was developed for the quantitation of vinblastine, desacetylvinblastine and vincristine using liquid chromatography-atmospheric pressure chemical ionization mass spectrometry (LC-APCI-MS). Analyses were performed on an Ultrasphere C₁₈ microbore column using ammonium acetate as mobile phase. The calibration curves were linear across the range of 0.51–4.00 ng/ml (0.63–4.93 nM) for vinblastine, 0.74–3.93 ng/ml (0.96–5.11 nM) for desacetylvinblastine and 0.30–3.95 ng/ml (0.36–4.79 nM) for vincristine. Vinca alkaloid concentrations were measured with an accuracy and precision within 11%. This assay could be implemented to determine the plasma concentrations for pharmacokinetic studies of vinblastine, desacetylvinblastine and vincristine in conjunction with clinical trials. Received: 15 December 1995 / Accepted: 16 June 1996     

Quantitation of methylglyoxal bis(guanylhydrazone) in blood plasma and leukemia cells of patients receiving the drug

Methylglyoxal bis(guanylhydrazone), a cytostatic compound which apparently interferes with the metabolism and/or functions of the natural polyamines (spermidine and spermine), was effectively taken up by cultured human lymphocytic leukemia cells, rapidly resulting in the formation of a concentration gradient of up to 1,000-fold across the cell membrane in cells grown in the presence of micromolar concentrations of the drug. For an anti-proliferative effect on the leukemia cells, an intracellular concentration of more than 0.5 mm was required. The uptake of methylglyoxal bis-(guanylhydrazone) was critically dependent on the growth rate of the leukemia cells. Low intracellular concentrations of the drug were present in cells growing slowly, whereas in rapidly dividing cells the intracellular concentration of the drug approached 5 mm. When given as repeated intravenous infusions to two leukemic children, methylglyoxal bis(guanylhydrazone) exhibited sharp and transient peaks of plasma concentration, the drug having an apparent half-life in plasma of only 1-2 h. However, as in cultured cells, the drug was rapidly concentrated in the leukemia cells, reaching concentrations that were distinctly anti-proliferative. In contrast to the rapid disappearance of methylglyoxal bis(guanylhydrazone) from plasma, the circulation leukemia cells retained the drug for a period of several days with only minimal decrease in the initial concentrations. Methylglyoxal bis(guanylhydrazone) was given to the patients for 1 to 2 months as intravenous infusions, the timing of which was determined by regular assays of the drug concentrations in the leukemia cells. In agreement with the results obtained with the cultured cells, an intracellular concentration of about 0.5 to I mm was apparently required for growth-inhibitory action to occur. Regular determination of the cellular drug concentrations indicated that methylglyoxal bis(guanylhydrazone) could be given as weekly infusions. This treatment schedule represents much lower dosing of the drug than the earlier daily regimens which were commonly associated with unacceptable toxicity.     

Quantitation of dolastatin-10 using HPLC/electrospray ionization mass spectrometry: application in a phase I clinical trial

A highly sensitive and specific assay for the quantitation of the anticancer agent dolastatin-10 (DOL-10) in human plasma is described. The method was based on the use of electrospray ionization-high-performance liquid chromatography/mass spectrometry (ESP-LC/MS). The analytical procedure involved extraction of plasma samples containing DOL-10 and the internal standard (DOL-15) with n-butyl chloride, which was then evaporated under nitrogen. The residue was dissolved in 50 μl mobile phase and 10 μl was subjected to ESP-LC/MS analysis using a C₁₈ microbore column. A linear gradient using water/acetonitrile was used to keep the retention times of the analytes of interest under 5 min. The method exhibited a linear range from 0.005 to 50 ng/ml with a lower limit of quantitation (LLQ) at 0.005 ng/ml. Absolute recoveries of extracted samples in the 85–90% range were obtained. The method's accuracy (≤5% relative error) and precision (≤10% CV) were well within industry standards. The analytical procedure was applied to extract DOL-10 metabolites from samples obtained following incubation of the drug with an activated S9 rat liver preparation. Two metabolic products were detected and were tentatively identified as a N-demethyl-DOL-10 and hydroxy-DOL-10. Structural assignments were made based on the fragmentation patterns obtained using the electrospray source to produce collision-induced dissociation (CID). The method was also applied to the measurement of DOL-10 in the plasma of patients treated with this drug. Preliminary investigation of the pharmacokinetics suggested that drug distribution and elimination may be best described by a three-compartment model with t_1/2α = 0.087 h, t_1/2β = 0.69 h and t_1/2γ = 8.0 h. Plasma clearance was 3.7 l/h per m². Received: 17 March 1997 / Accepted: 13 June 1997     

Modulation of plasma thiols and mixed disulfides by BNP7787 in patients receiving paclitaxel/cisplatin therapy

BACKGROUND: BNP7787 (disodium 2,2'-dithio-bis-ethane sulfonate) was evaluated in a phase I clinical trial with paclitaxel and cisplatin to assess the safety and potential efficacy for preventing or reducing cisplatin- and paclitaxel-induced toxicities. During this trial the effects of BNP7787 administration on the total concentrations (oxidized plus free) of cysteine, homocysteine and GSH in plasma, free and total GSH in WBC and rate of urinary excretion of cysteine were studied. The pharmacokinetics of ultrafilterable (free, non-protein bound) platinum were also determined after cisplatin (75 mg/m(2)) treatment which followed paclitaxel (175 mg/m(2)) and BNP7787 (8.2 to 27.6 g/m(2)). METHODS: Plasma thiols were measured by HPLC with fluorescence detection and platinum was measured by atomic absorption spectrophotometry. RESULTS: BNP7787 administration produced a significant depletion of all plasma thiols in all the patients studied. Differences were noted in the kinetics of BNP7787-induced depletion of cysteine and other thiols. A significant depletion of cysteine occurred with a time lag of about 2 h after the end of BNP7787 infusion, while a reversible depletion of GSH and homocysteine occurred immediately following the start of BNP7787 infusion, with the plasma thiol/disulfide nadir corresponding to the end of infusion. The mean half-life of cysteine depletion following BNP7787 administration was 2.2 h, significantly longer than for homocysteine (0.23 h), or GSH (0.18 h; P<0.05 for both). A several-fold increase in the urinary excretion of cysteine occurred following BNP7787 administration in all patients. The BNP7787-induced thiol/disulfide depletion in plasma was not affected by cisplatin administration ( P>0.05). BNP7787 administration had no effect on the ultrafilterable platinum pharmacokinetics. The 2-h lag in the depletion of cysteine, the most abundant thiol in plasma, suggests that the process may be related to the formation of free mesna from BNP7787 and that increased levels of mesna are not in circulation until after 2 h after BNP7787 administration. No effect of BNP7787 was seen on the GSH concentration in WBC, possibly reflecting the inability of these cells to take up BNP7787. CONCLUSION: The results suggest that BNP7787 has the potential to enhance cisplatin antitumor activity by depleting the reactive thiols in plasma.     

Quantitation of 1-beta-D-arabinofuranosylcytosine 5'-triphosphate in the leukemic cells from bone marrow and peripheral blood of patients receiving 1-beta-D-arabinofuranosylcytosine therapy

A method for the detection and quantitation of 1-beta-D-arabinofuranosylcytosine 5'-triphosphate (ara-CTP), the active metabolite cells and leukemic cells of the peripheral blood from patients receiving ara-C therapy is described. ara-CTP is separated from normal cellular nucleotides by high-pressure liquid chromatography and is quantitated by its absorbance of ultraviolet light at 280 nm with a lower limit of sensitivity of 25 pmol/2 x 10(7) cell equivalents. During separate courses of continuous infusion of different therapeutic doses of ara-C, ara-CTP accumulated in the leukemic bone marrow cells of a patient with acute myelogenous leukemia in proportion to the dose of ara-C. Continuous infusion of ara-C (90 mg/sq m/day) resulted in plateau levels of ara-CPT in peripheral blast cells after 24 hr (115 pmol/1 x 10(7) cell equivalents). A priming dose of ara-C(125 to 250 mg/sq m) followed by a 1-hr infusion of an equal dose of ara-C to patients with acute myelogenous leukemia facilitated the determination of ara-CTP retention in bone marrow and peripheral blood leukemic cells in vivo. This procedure should be useful for extended studies of the biochemical pharmacology of ara-CTP in vivo.     

DNA cross-linking and single-strand breaks induced by teratogenic concentrations of 4-hydroperoxycyclophosphamide and phosphoramide mustard in postimplantation rat embryos

Postimplantation rat embryos (Day 10) were exposed in vitro to teratogenic concentrations of 4-hydroperoxycyclophosphamide, an activated form of cyclophosphamide, and phosphoramide mustard, the major teratogenic metabolite of cyclophosphamide. Following a 5-h exposure to these agents, drug-induced DNA damage was assessed by alkaline elution. Both drugs induced detectable DNA cross-linking at teratogenic concentrations. Alkaline elution combined with proteinase K digestion indicated that approximately half of the DNA cross-linking was DNA-DNA cross-linking and the other half was DNA-protein cross-linking. In addition to DNA cross-linking, phosphoramide mustard produced DNA strand breaks and/or alkaline labile sites. However, 4-hydroperoxycyclophosphamide did not produce detectable DNA strand breaks or alkaline labile sites. Our data also indicate that the induction of abnormal morphogenesis by 4-hydroperoxycyclophosphamide and phosphoramide mustard is correlated with drug-induced DNA cross-linking.     

Identification of patients with nasopharyngeal carcinoma by serum protein profiling using surface-enhanced laser desorption/ionization time-of-flight mass spectrometry

Background

As diagnosis of nasopharyngeal carcinoma at an early disease stage is important, we attempted to distinguish between patients with nasopharyngeal carcinoma and noncancer controls by using serum protein profiles.

Methods

Surface-enhanced laser desorption/ionization time-of-flight mass spectrometry and CM10 protein chip were used to detect the serum proteomic patterns of 65 patients with nasopharyngeal carcinoma before radiotherapy and 93 noncancer controls. Proteomic spectra of serum samples from 50 nasopharyngeal carcinoma patients and 60 noncancer controls were used as a training set. The validity of the classification tree was then challenged with a blind test set which included another 15 patients with nasopharyngeal carcinoma and 33 noncancer controls. Biomarker Wizard 3.01 and Biomarker Pattern 5.01 were used in combination to analyze the data and to develop diagnostic models.

Results

21 protein peaks were significantly different between nasopharyngeal carcinoma and controls. 4 mass peaks (M4182, M5343, M5913 and M8702 mass/charge ratio) were chosen automatically to construct a classification tree. The classification tree correctly determined 93.8 % (45/48) of the test samples with 93.3 % (14/15) of the nasopharyngeal carcinoma samples and 93.9 % (31/33) of the noncancer samples. Using a combination of serum protein profiles and Epstein?Barr viral capsid antigen immunoglobulin A antibody tests, the diagnostic sensitivity and specificity were increased to 100 and 97 %, respectively.

Conclusions

Surface-enhanced laser desorption/ionization time-of-flight mass spectrometry could correctly distinguish nasopharyngeal carcinoma from noncancer individuals and showed great potential for the development of a screening test for the detection of nasopharyngeal carcinoma.     

Acquired protein C deficiency in patients with breast cancer receiving cyclophosphamide, methotrexate, and 5-fluorouracil

Recent reports have documented an increase of thrombotic complications in patients with carcinoma of the breast receiving chemotherapy regimens containing cyclophosphamide, methotrexate, and 5-fluorouracil. The authors studied blood from nine such patients screening for abnormalities that might predispose to thrombosis or indicate that the coagulation cascade had been activated. Six of the patients were in the adjuvant setting, and three had metastatic disease. Samples were collected from each patient before, during, and after completion of the chemotherapy in question. In each patient a statistically significant decline in functional protein C activity (P = 0.001) was demonstrated at midtherapy. In seven of nine patients functional protein C level normalized after the cessation of therapy. No other positive results were found. The authors conclude that the combination of cyclophosphamide, methotrexate, and 5-fluorouracil, when administered to patients with a diagnosis of carcinoma of the breast, causes a reversible decline in the activity of protein C.     

Antitumor activity of halogen analogs of phosphoramide,isophosphoramide, and triphosphoramide mustards,the cytotoxic metabolites of cyclophosphamide,ifosfamide, and trofosfamide

A series of halogen analogs of phosphoramide mustard, isophosphoramide mustard, and triphosphoramide mustard, the cytotoxic metabolites of the antitumor drugs cyclophosphamide, ifosfamide, and trofosfamide, respectively, was evaluated in vitro against human tumor cell lines and in vivo against experimental tumors to investigate the effect of replacement of chlorine with bromine or fluorine on the antitumor activity of the parent phosphoramide mustards. In the experimental tumors L1210 leukemia, B16 melanoma, mammary adenocarcinoma 16/C, and ovarian sarcoma M5076, the antitumor activity of the analogs was observed to be generally comparable with that of the parent mustards when chlorine was replaced by bromine but uniformly lower when chlorine was replaced by fluorine. Furthermore, the monobromo analog of isophosphoramide mustard displayed equal or somewhat greater activity in comparison with cyclophosphamide when evaluated against subcutaneously implanted L1210 leukemia with intraperitoneal drug treatment and against mammary adenocarcinoma 16/C.This work was financially supported by NIH, NCI grant PO1 CA34200     

Metabolites of 5-fluorouracil in plasma and urine, as monitored by 19F nuclear magnetic resonance spectroscopy, for patients receiving chemotherapy with or without methotrexate pretreatment

19F NMR spectroscopy at 470 MHz (11.7 Tesla) has been used to directly measure the levels of 5-fluorouracil (FU) and its fluorine-containing catabolites in plasma and urine of colon cancer patients after i.v. infusion (10 min) of 60-230 mumol (8-30 mg) FU/kg, either with or without pretreatment with methotrexate (5.1-12.5 mg/kg). With a 1.5-ml sample the minimum metabolite concentration that can be quantified is approximately 15 +/- 5 microM within 30 min and 3 +/- 1 microM within 12 h of data acquisition. The first and second catabolites of FU, dihydrofluorouracil and alpha-fluoro-beta-ureidopropanoic acid, exhibit steady-state behavior with dose-dependent plasma concentrations of 5-40 microM for approximately 10-90 min after infusion (12 patients, 16 treatments). The final catabolite alpha-fluoro-beta-alanine (FBAL) was detected in plasma after 5-15 min, and the rate at which its concentration increased was independent of FU dose, while the maximum concentration reached at about the time FU disappeared (FU less than 5 microM in 1-2 h) was dose-dependent. The area under the time curve for FU in plasma increased more than linearly with dose. Several patients showed elevated levels of free fluoride anion (F-) in plasma (63 samples: median, 5 microM; maximum, 33 microM). In urine all of the above catabolites and F- could be observed. In samples with pH greater than or equal to 7.3 (methotrexate patients, due to bicarbonate infusion) N-carboxy-FBAL was also found in significant amounts. Urinary excretion of FU and catabolites amounted to 2.6-30% of the dose within 2 h (14 patients, 18 treatments) and 60-66% within 24 h (three patients). The ratio FU/creatinine in 2-h urine increased more than linearly with FU dose. Urinary fluoride concentration reached a maximum during the first day after FU infusion and returned to normal background levels after 2-3 days (four patients). The pattern of FU catabolites observed in plasma or urine did not differ significantly between responders and nonresponders to therapy or between patients with FU monotherapy and patients with methotrexate pretreatment. Cytotoxic FU anabolites, i.e., nucleotides, were not detected in plasma or urine (i.e., are less than 3 microM). Their detection in tumor tissue will be required for an assessment of individual responsiveness to FU. Possible toxic metabolic products derivable from FBAL, e.g., 2-fluoroacetate or 2-fluorocitrate, were not detected (i.e., are less than 3 microM) in plasma or urine.(ABSTRACT TRUNCATED AT 400 WORDS)