Metastatic involvement of skeletal muscle from gastric adenocarcinoma.

Gastric cancer represents the fifth most common cancer diagnosis worldwide and the third leading cause of cancer-related mortality. In the USA, the overall 5-year survival rate is 31%, with distant disease nearing 5%. The most common sites of metastasis are the liver and peritoneum. Skeletal muscle involvement has been rarely reported. Since clinical and imaging findings overlap with primary sarcomas, a confirmatory biopsy is required for diagnosis. Prognosis remains poor with treatment options including palliative chemotherapy, radiotherapy and surgical resection. We report the case of a 57-year-old female presenting with extensive involvement of skeletal muscle 10 years after achieving remission. In addition to illustrating the refractoriness and poor outcomes associated with muscle involvement, this case and comprehensive review of the literature highlights important characteristics of disease biology and tumor genomics that warrant detailed discussion and exposition to a wider audience.

Mechanism of action of a new antitumor agent, carbetimer.

Carbetimer, an intermediate molecular-weight-derivatized copolymer of maleic anhydride and ethylene, has been shown to possess significant antineoplastic activity in the stem cell assay. We have examined the antitumor activity of carbetimer in vivo and in vitro against HM5-Carb/S and M21, both primary human melanoma cell lines sensitive and resistant to carbetimer, respectively. The mechanism of action of carbetimer in HM5-Carb/S has been determined. Mice bearing palpable sensitive tumors were treated with 10% lethal doses of carbetimer (1500 mg/kg i.p.). The tumor nucleotide profile was determined 4 hours later. Uridine and cytidine nucleoside triphosphates were reduced by 36.6 and 58.2%, respectively. In a similar experiment using carbetimer-resistant tumor, there was no change in the tumor pool sizes of uridine and cytidine nucleoside triphosphate pools in carbetimer- or saline-treated animals. Following 24-h exposure of the cells to 1000 microM concentration of carbetimer, the carbetimer-sensitive cells were pulsed with [14C]uridine, cytidine, or thymidine for 30 min. Pyrimidine nucleotides, in particular triphosphates, were reduced significantly as compared to the saline-treated control. Similar treatment of carbetimer-resistant cells resulted in no change in the pool sizes of the nucleotides. [14C]Bicarbonate flux studies demonstrated that [14C]CO2 conversion into UMP and CMP was increased 200 and 140% of control in the carbetimer-sensitive cells treated with 1000 microM carbetimer; however, a similar treatment of the resistant cells showed no change in the pool sizes of the nucleotide. Examination of pyrimidine salvage enzymes demonstrated that, in the sensitive cells, carbetimer treatment reduced the specific activity of uridine, cytidine, and thymidine kinase by 46, 37, and 60%. In a similar study using resistant cells, the specific activities were reduced 7 and 0%, respectively. In the restitution studies coincubation of carbetimer-sensitive cells with carbetimer and uridine resulted in essentially the reversal of carbetimer cytotoxicity. Thus, carbetimer inhibits the growth of the sensitive cells by inhibiting the uptake and metabolism of performed nucleosides both in vivo and in vitro.

Collateral sensitivity to N-(phosphonacetyl)-L-aspartic acid in a line of P388 leukemia cells selected for resistance to L-(alpha S, 5S)-alpha-amino-3- chloro-4,5-dihydro-5-isoxazoleacetic acid (acivicin).

Administration of N-(phosphonacetyl)-L-aspartic acid (PALA) is ineffective in treating mice bearing the parent P388 leukemia line; however, such treatment becomes highly effective when a cell line, P388/ACIA, derived from P388/0 was selected for resistance to another antimetabolite, acivicin. The observed phenomenon of collateral sensitivity is associated with a significantly higher inhibition of the specific activity of carbamyl phosphate synthetase II, pyrimidine nucleoside kinases, adenine phosphoribosyl transferase, and hypoxanthine phosphoribosyl transferase in the PALA-sensitive line, P388/ACIA. Twenty-four hr following administration of PALA, 200 mg/kg, the 10% lethal dose i.p. to tumor-bearing mice, the intracellular concentrations of uridine triphosphate and cytidine triphosphate were decreased in the P388/ACIA, PALA-sensitive cells, whereas no significant change in the corresponding nucleotide pool sizes was observed in P388/0, PALA-resistant line. Moreover, the purine nucleotide pool demonstrated a significant expansion of adenosine triphosphate and guanosine triphosphate only in the P388/ACIA line following a similar treatment with PALA. It is proposed that the imbalance in the generation of pyrimidine and purine nucleoside triphosphate pools may explain the observed collateral sensitivity to PALA in P388/ACIA leukemia line.

Mechanism of resistance of a variant of P388 leukemia to L-(alpha S,5S)-alpha-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid (acivicin).

Acivicin [L-(alpha S,5S)-alpha-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid; NSC 163501] is a fermentation-derived amino acid antibiotic antagonistic to L-glutamine which exhibits potent oncolytic properties. We have developed a variant of P388 leukemia resistant to acivicin (P388/ACIA) and compared its properties with those of the parent line (P388/S). An examination of the enzymes utilizing L-glutamine revealed that the basal specific activities of L-asparagine synthetase and L-glutaminase were 1-to 3-fold higher in the parent line. The activities of carbamoyl phosphate synthetase II, L-asparagine synthetase, formylglycinamide ribonucleotide amidotransferase, and guanosine monophosphate synthetase were about equally inhibited in the two cell lines, while there was a partial inhibition of 5-phosphoribosyl-1-pyrophosphate amidotransferase, fructose-6-phosphate amidotransferase, and L-glutaminase activities, found only in the sensitive line. Cytidine triphosphate synthetase activity was not inhibited in either line. There was no difference in the dose response or restitution of L-glutamine utilizing enzyme activities between the two lines. Acivicin treatment produced a 2- to 3-fold augmentation of the L-glutamine pools only in the sensitive line. Drug injection induced increased 5-phosphoribosyl-1-pyrophosphate levels in both lines. Acivicin perturbed guanosine nucleotide pools only in the sensitive line, indicating that the primary mechanism of action of acivicin in P388 leukemia may be directed at guanosine monophosphate synthetase. Transport studies demonstrated a restricted uptake of acivicin by the resistant cells. These studies suggest that the transport of acivicin and L-glutamine plays an important role in determining the sensitivity or resistance to acivicin in these tumors.

Reevaluation of the maximum tolerated dose of continuous venous infusion of 5-fluorouracil with pharmacokinetics.

5-Fluorouracil (5-FU) was administered as a continuous ambulatory venous infusion to 25 patients in a Phase I trial. The principal dose limiting toxic effect observed was mucositis. Skin rash and diarrhea occurred less frequently. Hematological toxicity was modest, and no hepatic toxicity was seen. One partial remission of 138 days duration was seen in a patient with metastatic breast carcinoma who was previously refractory to a 5-FU combination regimen. Patient tolerance of 5-FU delivered in this manner appeared highly variable. On the basis of this trial, we recommend that future studies evaluating the efficacy of long-term venous infusion of 5-FU should utilize a dosage of 450 mg/m2/day.

Mechanisms of sensitivity and resistance of murine tumors to 5-fluorouracil.

The biochemical basis for the resistance of murine leukemia P388 to 5-fluorouracil (FUra) was systematically investigated by examining the transport and metabolism of FUra, or its anabolites, as well as the inhibition of enzymes and processes known to be affected by the drug. Of these parameters, only three were found to be altered significantly in the resistant line: (a) the enzyme required for the phosphorylation of uridine 5'-monophosphate to uridine 5'-diphosphate was present at a significantly lower specific activity in the resistant line than in its sensitive counterpart; (b) the rates of generation and persistance of 5-fluoro-2'-deoxyuridine 5'-monophosphate were significantly lower and shorter in the variant; and (c) there was a 1.6- and 3-fold decrease in the incorporation of FUra into polyadenylic acid-containing RNA and polyadenylic acid-lacking RNA, respectively, in resistant versus sensitive cells. Taken together, these findings suggest a dual mechanism for resistance to FUra in these leukemic cells, namely, a depressed capacity to generate di- and triphosphates of the riboside and deoxyriboside of the drug leading to lower pools of the proximate antimetabolite, fluorouridine 5'-triphosphate, and accelerated excretion of 5-fluoro-2'-deoxyuridine 5'-monophosphate, so that thymidylate synthetase is perturbed in a less than lethal way.

Pharmacokinetic and biochemical studies on acivicin in phase I clinical trials.

Acivicin pharmacokinetics were studied in Phase I patients receiving i.v. treatment on single-dose or daily x5 (daily times five doses) regimens repeated every 3 weeks. In 14 patients, the time course of plasma concentrations was characterized by a biexponential equation with a terminal (elimination-phase) half-life of 9.92 +/- 3.91 h (mean +/- SD), distribution phase half-life of 0.32 +/- 0.28 h, total body clearance of 1.69 +/- 0.48 liters/h/m2, and volume of distribution of 21.79 +/- 2.94 liters/m2. Acivicin kinetics appeared to be dose-independent over the range of 8.5-150 mg/m2/day. Urinary excretion of intact acivicin in nine patients ranged from 2-42% in the first 24 h following administration; interpatient variability in urinary excretion was large, but daily urinary recovery within patients on the daily x5 schedule was quite consistent. Measurements of acivicin effects on the activity of carbamyl phosphate synthetase II (CPS II) were conducted using leukocytes and/or malignant ascites of three colon cancer patients. Acivicin given to one patient at 8.5 mg/m2/day on the daily x5 schedule caused a 70% reduction in leukocyte CPS II activity within 5 h after therapy was initiated. Leukocyte CPS II activity remained suppressed at this level over the 5-day dosing regimen. In this patient, CPS II activity in malignant ascitic cells had decreased by 75% on day 4 of the daily x5 regimen. On the single dose schedule, treatment of two patients with 100 mg/m2 caused leukocyte CPS II activity to decrease by greater than 90% within 4 h of treatment with gradual recovery over the next 2 days.

Long-term association of N-(phosphonacetyl)-L-aspartate with bone.

By means of enzymatic and autoradiographic techniques, it has been demonstrated that, 24 hr after a single dose of the antitumor amino acid N-phosphonacetyl-L-aspartic acid (PALA), (400 mg/kg i.p.; 1.15 mmol/kg) to C57BL x DBA/2 F1 mice, the agent accumulates in bone to a concentration of approximately 400 microM; this is 3000 times greater than the Ki of PALA for its target enzyme, aspartate carbamoyltransferase. However, disproportionately low inhibition of enzyme activity was demonstrated in homogenates of bone from these recipients, suggesting that the drug was sequestered from its target in this tissue. Autoradiography of sections of femoral shafts from mice treated with 14C-labeled drug demonstrated that autoradiogram density due to [14C]PALA equivalents was confined to the bony matrix, with no label above background resolvable in bone marrow. Following in vivo administration of PALA (400 mg/kg i.p.), the half-life of the drug in the bone was approximately 23 days. In vitro, with equilibrium dialysis at pH 7.4, it was demonstrated that: (a) normal pulverized and decalcified bone bound PALA with capacities of 3.5 nmol/mg and 0.1 nmol/mg bone, respectively, at a PALA concentration of 5 mM; (b) binding of PALA to normal bone reached saturation at a concentration of 200 mM; and (c) PALA functions as a solubilizer of bone at concentrations above this. Since administration of PALA was shown to produce long-lasting inhibition of aspartate carbamoyltransferase in liver and tumor and since its ultimate half-life in the plasma of mice, following a single 400-mg/kg administration of the drug, is 8 days, it is suggested that bone serves as a reservoir from which PALA is released at a slow rate into plasma and other tissues.

Mechanism of resistance of variants of the Lewis lung carcinoma to N-(phosphonacetyl)-L-aspartic acid.

Variants of the Lewis lung carcinoma were selected for resistance to N-(phosphonacetyl)-L-aspartic acid (PALA) by treatment of tumor-bearing mice with repetitive subcurative doses of PALA. The specific activity of the target enzyme, L-aspartic acid transcarbamylase (ATCase), was measured in the four variants developed. Three had markedly elevated ATCase activities; however, the fourth line, LL/PALA-C, had an ATCase activity identical to that of the parent, PALA-sensitive line (LL/O). One high-ATCase variant, LL/PALA-J, and LL/PALA-C were compared with LL/O in subsequent biochemical studies on the mechanism of resistance to PALA. Enzyme activities in the salvage pathways which phosphorylate pyrimidine nucleosides and deoxynucleosides were found to be similar in all three lines. ATCase in these lines exhibits closely comparable kinetics with its natural substrates as well as with PALA. The time courses of restitution of ATCase after a single therapeutic dose of PALA show that both resistant variants recover full activity more rapidly than the parent. Additionally, inhibition of ATCase 24 hr following graded doses of PALA is lower in the resistant lines. The uptake of [14C]PALA in vitro into cell lines derived from the three Lewis lung carcinomas apparently occurs by passive diffusion and at comparable rates in both sensitive and resistant cells. Analysis of the nucleotide content of tumors reveals comparable spectrums of purine and pyrimidine nucleotide levels in the LL/O and LL/PALA-C lines, whereas the LL/PALA-J line has augmented nucleotide pools. In all three lines, 24 hr after treatment with PALA (400 mg/kg), uridine and cytidine nucleotide levels were substantially diminished (70 to 80%) while adenosine 5'-triphosphate and guanosine 5'-triphosphate levels were elevated (50 to 100%). Estimations of precursor flux through the de novo pyrimidine pathway by measuring orotate and orotidine levels in tumors of mice treated with pyrazofurin (an inhibitor of orotidine-5'-monophosphate decarboxylase) and either 0.9% NaCl solution or PALA shows that PALA treatment eliminates orotate and orotidine accumulation in LL/O but reduces it by only 75 and 50% in LL/PALA-C and LL/PALA-J, respectively. Similarly, PALA treatment (20 microM) of tumor lines in culture provokes a dramatic decrease in the incorporation of NaH14CO3 into pyrimidine intermediates and nucleotides in the LL/O cell line only. Determinations of specific activities of the other enzymes in this pathway reveal that the activity of carbamyl phosphate synthetase II, the rate-limiting step, is elevated 2- to 3-fold in both resistant lines. Since carbamyl phosphate synthetase II exists as a complex with ATCase, the suggestion is made that levels of carbamyl phosphate synthetase II are collaterally important determinants of PALA activity. An augmented pool of carbamyl phosphate in the resistant variants may serve to competitively displace PALA from ATCase, diminish enzyme inhibition, and allow pyrimidine biosynthesis to proceed despite therapy.

A randomized phase I and II study of short-term infusion of high-dose fluorouracil with or without N-(phosphonacetyl)-L-aspartic acid in patients with advanced pancreatic and colorectal cancers.

Fifty-two patients with advanced gastrointestinal (GI) malignancies who had not received previous chemotherapy or radiation therapy were randomized to be treated either with 24-hour infusion of weekly fluorouracil (5-FU) or the same plus N-(phosphonacetyl)-L-aspartic acid (PALA). Forty-seven patients were evaluable for the assessment of toxicity and antitumor activity. PALA was administered as an intravenous (IV) bolus over 15 minutes at a fixed dose, 250 mg/m2. The latter agent was administered 24 hours before the start of 5-FU infusion. 5-FU was initially administered at 750 mg/m2 and was incrementally increased to 3,400 mg/m2. In both arms of the randomized study, the courses were repeated every week. In both arms of the study, ataxia and myelosuppression were the dose-limiting toxic effects. At 5-FU dose of 3,400 mg/m2, one patient in each arm developed grade 3 hematologic toxicity. Other reversible side effects included grade 2 skin changes, nausea, and vomiting. During the administration of 2,600 mg/m2 of 5-FU over 24 hours, the steady state plasma 5-FU concentration was approximately 20 mumol/L. The maximum tolerated dose (MTD) for 5-FU for protracted treatment is 2,600 mg/m2 in either arm of the study. Therapeutic response was predominantly seen in the combination arm: there were two patients with complete response (CR) and 11 patients with partial response (PR) of 28 patients in the study. In the 5-FU alone arm there were four PR and 19 patients in the study.

A phase II study of weekly 24-hour infusion with high-dose fluorouracil with leucovorin in colorectal carcinoma.

Twenty-two patients with advanced colorectal carcinoma were enrolled in this study. Ten patients had received prior chemotherapy that included the combination of fluorouracil (5-FU) and leucovorin (LV). All patients required subcutaneous port insertion and portable external infusion pumps to allow outpatient treatment. 5-FU (2,600 mg/m2) was administered concurrently with LV (500 mg/m2) over 24 hours of continuous infusion. The mean steady-state plasma concentration of 5-FU was 10 mumol/L (range, 7 to 14 mumol/L). The 5-FU dose was based on our previous phase I study, in which maximum-tolerated dose (MTD) of 5-FU was determined to be 2,600 mg/m2 in combination with a fixed dose of LV at 500 mg/m2. The treatment was repeated weekly. Twenty-two patients received a total of 560 courses of treatment. Eleven instances of grade 2-3 toxicity were observed: diarrhea (five), stomatitis (three), hand/foot syndrome (three). The overall objective response was 45% (10 of 22) and among previously untreated patients was 58%. Three of the responders achieved complete response (CR), with lung and liver as the metastatic sites. The median duration of survival for the previously untreated patients was not reached at 22 months, and was 10 months for the previously treated patients. These results suggest that short-term infusional therapy of 5-FU and LV in patients with advanced metastatic colorectal cancer generates acceptable toxicity, with equivalent or superior survivability in previously treated and untreated patients versus alternative methods of administration of the two agents.

Phase II study of fluorouracil and its modulation in advanced colorectal cancer: a Southwest Oncology Group study.

PURPOSE: A variety of fluorinated pyrimidine-based regimens for the treatment of disseminated colorectal cancer have been presented in the medical literature. The Southwest Oncology Group designed a screening trial of seven regimens of fluorouracil (5-FU) to assess efficacy and toxicity afforded by biochemical modulation or schedule variations. PATIENTS AND METHODS: Six hundred twenty patients were entered into this trial between August 1989 and January 1993. Eligible patients were classified as having recurrent or disseminated disease that was measurable or nonmeasurable. All eligible patients were evaluated for toxicity and survival; patients with measurable disease were evaluated for response according to standard criteria. RESULTS: No regimen achieved a higher response rate than single-agent bolus 5-FU. Eighty-four percent of patients have been monitored until death. The median survival time for the entire cohort is 14 months. Survival hazards ratios showed a positive trend in favor of the unmodulated infusion regimens, while high-grade toxicities occurred more frequently in the 5-FU bolus arms. The major high-grade toxicities were granulocytopenia and diarrhea. CONCLUSION: In this trial, no regimen provided substantial improvement relative to 5-FU bolus or single-agent therapy for either response or survival in the treatment of disseminated colorectal cancer. The single-agent infusion regimens demonstrated the most encouraging results with a favorable toxicity profile and a 2-month longer survival duration than 5-FU bolus therapy.

Phase I study of high dose 5-fluorouracil and high dose Leucovorin with low dose phosphonacetyl-L-aspartic acid in patients with advanced malignancies.

Twenty-eight patients with refractory advanced malignancies were treated with a 24 hr infusion of 5-fluorouracil (5-FU), Leucovorin (LV), and N-(phosphonacetyl)-L-aspartic acid (PALA) weekly. Twenty-seven patients were evaluable for the assessment of toxicity and anti-tumor activity. PALA was administered as intravenous bolus over 15 min at a fixed dose, 250 mg/m2 24 hr before the start of 5-FU and LV infusions. 5-FU was initially administered at 750 mg/m2 and was incrementally increased to 2600 mg/m2. LV was administered in a fixed dose of 500 mg/m2 concurrently with 5-FU over a 24-hr period. The course was repeated weekly. Diarrhea, stomatitis, nausea, and vomiting were among dose-limiting toxic effects. Other toxicities observed were hand-foot syndrome, hair loss of scalp/eyelashes, overall weakness, rhinitis, and chemical conjunctivitis. Maximum tolerated dose (MTD) of 5-FU in this combination and schedule was 2600 mg/m2. Seven of 14 patients treated at 2600 mg/m2 were able to tolerate the chemotherapy on a weekly basis without interruption. The other seven patients required dose de-escalation, a majority of whom contained 5-FU at a dose of 2100 mg/m2. Twenty-three of 27 patients had been previously treated. Eight patients achieved a partial response, all of whom were previously treated, except three patients. A complete response was observed in a patient with pancreatic carcinoma, previously untreated. Overall response rate for the patients who were treated at the 5-FU dose of 2100 mg/m2 or 2600 mg/m2 is 9 of 18 patients (50%).

Effect of L-glutamine antagonists on 5-phosphoribosyl 1-pyrophosphate levels in P388 leukemia and in murine colon adenocarcinomas in vivo.

The intratumoral content of 5-phosphoribosyl 1-pyrophosphate (PRPP) and the activity of the enzymes anabolizing and catabolizing the sugar phosphate were determined following i.p. administration of an LD10 dose of an L-glutamine antagonist or saline to tumor-bearing animals. Elevation of PRPP pool size following administration of L-[alpha S,5S]-alpha-amino-3-chloro-4,5-dihydro-5-isopazoleacetic acid (NSC-163501) (AT-125) was maximal at 8 hr and returned to pretreatment levels by 24 hr. In P388 leukemia, dose for dose, at 4 hr, 6-diazo-5-oxo-L-norleucine (NSC-7365) (DON) was the most potent of the L-glutamine antagonists in elevating basal PRPP pool size (50% above control) followed by AT-125 and azaserine, 300 and 100% above control respectively. Moreover, such augmentation in PRPP pool size preferentially affected P388 tumor rather than the small intestine. Following i.p. administration of LD10 doses of AT-125, DON and azaserine, the specific activities of PRPP anabolizing and catabolizing enzymes were determined. A significant inhibition of PRPP amidotransferase was demonstrated with DON and AT-125 (P less than 0.05), and no inhibition with azaserine. A similar modulation of PRPP pool size demonstrated in vivo following administration of 250 mg/kg of ART-125 in mice bearing colonic adenocarcinoma lines. It was suggested that a significant increase of PRPP pool size might cause the possible synergism of a selected L-glutamine antagonist and 5-fluorouracil as reported after the appropriately scheduled administration of methotrexate and 5-fluorouracil.

Biochemical mechanisms for the scheduled synergism of (alpha S, 5S)-2 amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid and 5-fluorouracil in P388 leukemia.

A study was made of the in vivo effects of equitoxic doses of AT-125 and 5-FU combination, being administered either simultaneously (% ILS 152) or with a 6-h pretreatment with AT-125 (% ILS 184). To examine the biochemical basis for the scheduled synergism, measurements were made of the concentration of PRPP, the specific activities of CPS II, cytidine, thymidine, uridine, deoxyuridine kinases, and fluorinated nucleotide formation in P388 tumors and the small intestine. Two hours after in vivo simultaneous treatment of mice bearing tumors the concentration of PRPP increased 9- and 6-fold above baseline in the tumor and the small intestine, respectively. In the AT-125 pretreatment arm the concentration of PRPP increased 18- and 7-fold above baseline in the tumor and the small intestine, respectively. CPS II activity was reduced to 28%-18% of control in the tumors in the simultaneous and pretreatment groups, respectively, whereas it remained unchanged in the small intestine. Specific activities of cytidine kinase (5.5 +/- 1), thymidine kinase (4.0 +/- 1.6), uridine kinase (35.6 +/- 6.5), and deoxyuridine kinase (2.4 +/- 1.1) nmol/mg protein/h remained unchanged with treatment. In concert with the increased intratumor concentration of PRPP, fluorinated nucleotide formation was proportionally increased in the treatment arms. These results indicate the importance of drug scheduling of the above two agents in treating P388 leukemia.

Studies on the mechanism of 3-deazaguanine cytotoxicity in L1210-sensitive and -resistant cell lines.

3-Deazaguanine (3-DG), a purine analogue, has unusual antitumor activity against experimental mammary tumor models and a number of other solid tumors. Others have shown that mutant CHO cells deficient in hypoxanthine guanine phosphoribosyl transferase (HGPRTase) or adenine phosphoribosyl transferase (APRTase) are resistant to 3-DG. We developed a L1210 cell line resistant to 3-DG, L1210/3-DG, by subculturing the parent L1210/0 cells in the presence of increasing concentrations of 3-DG. The IC50 was 3.5 microM and 620 microM for L1210/0 and L1210/3-DG, respectively. Cytotoxicity studies proved the resistance to be stable. Examination of the baseline-specific activity of HGPRTase and APRTase showed that the former was 118-fold lower in L1210/3-DG than in L1210/0, and the latter demonstrated no difference. A 4-h treatment of the cell lines at IC50 doses showed 48% and 23% reductions in IMP dehydrogenase in L1210/0 and L1210/3-DG, respectively. The rate of de novo purine biosynthesis was studied by using [14C]formic acid. Formate flux increased 2-fold in L1210/3DG in concert with the observed deficiency of HGPRTase in the cell line. 3-DG uptake was studied with [14C]-labelled compound. The total radioactivity was 9-fold higher in L1210/0 than in L1210/3-DG at 2 h. Subsequent chromatographic separation of radioactivity showed the 3-DG and 3-deazaguanosine pools of the drug to be equal in both lines. However, 3-DG nucleotide pools at 1 min and 2 h were 2.5-fold and 16-fold lower, respectively, in L1210/3-DG than in L1210/0. 3-DG incorporation studies with radiolabelled drug demonstrated that 3-deazaguanine is incorporated in the acid-insoluble fraction of the cell. These studies conclude that HGPRTase, and not APRTase, is required for the activation of drug. Inhibition of IMP dehydrogenase is partially responsible for antitumor activity of the drug. The incorporation of drug into nucleic acids may be a major mechanism for its antitumor activity. Further studies using a cloned cDNA probe for hypoxanthine guanine phosphoribosyltransferase (HGPRT) demonstrated no change in the DNA arrangements of the L1210/3-DG cell line, and Northern blot analysis showed approximately equal expression of mRNA in both cell lines.

Prochlorperazine as a doxorubicin-efflux blocker: phase I clinical and pharmacokinetics studies.

Doxorubicin (DOX) efflux in drug-resistant cells is blocked by phenothiazines such as trifluoperazine (TFP) and prochlorperazine (PCZ) in vitro. The present phase I study was conducted in 13 patients with advanced, incurable, nonhematologic tumors to determine whether PCZ plasma levels high enough to block DOX efflux could be achieved in vivo. The treatment schedule consisted of prehydration and i.v. administration of 15, 30, 50, and 75 mg/m2 PCZ followed by a standard dose of 60 mg/m2 DOX. The hematologic toxicities attributable to DOX were as expected and independent of the PCZ dose used. Toxicities attributable to PCZ were sedation, dryness of the mouth, cramps, chills, and restlessness. The maximal tolerated dose (MTD) of PCZ in this schedule was 75 mg/m2. Pharmacokinetic analysis indicated a large interpatient variation in peak plasma PCZ levels that ranged from 95 to 1100 ng/ml. The three plasma half-lives of PCZ were: t1/2 alpha (+/- SE), 20.9 +/- 5.3 min; t1/2 beta, 1.8 +/- 0.3 h; and t1/2 gamma, 21.9 +/- 5.3 h. The volume of distribution (Vd), total clearance (ClT), and area under the curve (AUC) for PCZ were 2254 +/- 886 l/m2, 60.2 +/- 13.5 l m-2 h-1, and 1624 +/- 686 ng ml-1 h, respectively. DOX retention in tumor cells retrieved from patients during the course of therapy indicated the appearance of cells with enhanced DOX retention. The combination of DOX and high-dose i.v. PCZ appeared to be safe, well tolerated, and active in non-small-cell lung carcinoma.

Prediction of clinical response to 5-fluorouracil-containing chemotherapy: preliminary results of in vitro assay in human breast cancer.

An in vitro assay system measuring the production of phosphorylated nucleotide products of 5-FU by human tumor cytosols is described. Samples from 11 patients with breast cancer treated with 5-FU-containing chemotherapy regimens were assayed. Significant differences were seen in quantitative and qualitative patterns of phosphorylation between samples from chemotherapy responding and nonresponding patients. The rate of 5-fluorodeoxyuridylate production and the amount of nucleotide triphosphate produced were significantly P < 0.05 increased in chemotherapy-responding patients. This in vitro assay system is a simple, reproducible procedure that requires further evaluation as a potential system for prospectively predicting clinical response to fluorinated pyrimidines.

Phase II trial of mitoxantrone in head and neck carcinoma.

Nineteen patients with measurable and incurable head and neck carcinoma (17 squamous cell and two adenoid cystic) received intravenous bolus doses of 14 mg/m2 mitoxantrone in the first course. The doses were escalated or de-escalated by 2 mg/m2 in subsequent courses, based on leukocyte nadir, to achieve mild (3,000-3,999/mm3) or moderate (2,000-2,999/mm3) toxicity and response. The courses were repeated every 3 weeks. All 60 courses were evaluated for toxicity. Leukopenia was mild, moderate, severe (1,000-1,999/mm3), and life-threatening (less than 1,000/mm3) in 17%, 23%, 42%, and 2% of courses, respectively. Mild thrombocytopenia (100,000-129,999/mm3) occurred in two courses. The median interval to nadir leukopenia was 14 days (range 7-36) with a median of 13 days (range 3-50) to recover to normal. After the first course, leukopenia was mild in 16%, moderate in 32%, severe in 26%, and life-threatening in 5%. One patient died of pulmonary embolism 8 days after the first course and had concomitant leukocyte count of 700/mm3. Eighteen patients had at least one course resulting in leukopenia. Three of six patients receiving greater than or equal to 4 courses (cumulative dose 56-102 mg/m2) had an asymptomatic decrement of 14%, 17%, and 29%, respectively, in radionuclide left ventricular ejection fraction. The other toxicities were mild. In the 16 patients with squamous cell carcinoma that could be evaluated for response, one had a partial response lasting 8 months, and six had stable disease. One of the two patients with parotid adenoid cystic carcinoma had a minor response lasting 16 months. Mitoxantrone on a bolus schedule has minimal activity and is not indicated in head and neck squamous cell carcinoma.

A phase II trial of recombinant leukocyte interferon plus doxorubicin in patients with hepatocellular carcinoma.

A Phase II trial of combination therapy with recombinant leukocyte interferon (alpha IFN) and doxorubicin was performed in patients with unresectable hepatocellular carcinoma. alpha IFN was administered at a starting dose of 20 x 10(6) U/m2 intramuscularly or subcutaneously with doxorubicin 20 mg/m2 intravenously weekly x 3 weeks followed by a 2-week period rest. There were 22 patients entered into the study. Among the 21 patients, there were 2 partial responses (10%), one minor response, and one patient had stable disease. Toxicity was generally tolerable, with fever, fatigue, and myelosuppression being the most common side effects. This combination of weekly recombinant leukocytic interferon and doxorubicin has modest and limited activity in hepatocellular carcinoma.