A phase 1 study of pralatrexate in combination with paclitaxel or docetaxel in patients with advanced solid tumors.

PURPOSE: Pralatrexate is a rationally designed antifolate with greater preclinical antitumor activity than methotrexate. Pralatrexate was synergistic with paclitaxel and with docetaxel in mouse xenograft experiments. This phase 1 study was designed to determine the maximum tolerated dose and toxicity of pralatrexate plus paclitaxel or docetaxel in patients with advanced cancer. EXPERIMENTAL DESIGN: Pralatrexate was administered i.v. every 2 weeks (days 1 and 15) in a 4-week cycle. Depending on the taxane used and dose being tested, the taxane was administered on days 1 and 15; days 2 and 16; or days 1, 8, and 15. In the latter part of the study, patients in the docetaxel arm were treated with vitamin B(12) and folic acid supplementation to mitigate toxicity and allow pralatrexate dose escalation. RESULTS: For the combination of pralatrexate plus paclitaxel without vitamin supplementation, dose-limiting stomatitis and peripheral neuropathy were encountered at the lowest dose levels tested. For pralatrexate plus docetaxel plus vitamin supplementation, pralatrexate 120 mg/m(2) plus docetaxel 35 mg/m(2) administered on the same day every other week was defined as the maximum tolerated dose and schedule, with dose-limiting toxicities at higher dose combinations including stomatitis and asthenia. Significant antitumor activity was observed for this combination in patients with non-small-cell lung cancer. CONCLUSIONS: Pralatrexate (120 mg/m(2)) plus docetaxel (35 mg/m(2)) plus vitamin supplementation is well tolerated with signs of efficacy against non-small-cell lung cancer that merit phase 2 testing.

A phase I trial of intermittent high-dose gefitinib and fixed-dose docetaxel in patients with advanced solid tumors.

PURPOSE: Based on our mouse xenograft model demonstrating that intermittent high-dose gefitinib sensitizes tumors to subsequent treatment with taxanes, we initiated this phase I trial to explore docetaxel in combination with escalating doses of intermittent gefitinib (Iressa) given prior to docetaxel. METHODS: This was a phase I study where patients with advanced cancer were treated with escalating doses of gefitinib (1,000, 1,500, 2,250, 3,000 mg) on days 1 and 2 followed by docetaxel (75 mg/m2) on day 3 of a 21 day cycle. Gefitinib pharmacokinetic data were obtained on days 1, 2, and 3 of cycles 1 and 2 at each dose level. RESULTS: 18 patients were enrolled in this study with the most frequent tumor types being non-small cell lung cancer and head and neck squamous cell cancer. The dose-limiting toxicity was neutropenia (n=1 at dose level 2, n=2 at dose level 4). Rash, diarrhea, and fatigue were the most common grade 1-2 toxicities. Pharmacokinetic data indicated no accumulation of gefitinib between cycles 1 and 2 and no clear correlation between gefitinib plasma levels and toxicity. Partial responses were observed in one patient with head and neck squamous cell carcinoma and one patient with anaplastic thyroid cancer. CONCLUSION: The recommended dose for phase II studies is gefitinib 2,250 mg on days 1 and 2, followed by docetaxel 75 mg/m2 on day 3.

Pharmacokinetic basis for differences in methotrexate sensitivity of normal proliferative tissues in the mouse.

Following s.c. administration of varying doses of methotrexate (12 to 400 mg/kg) to mice, the drug accumulated more rapidly and to much higher levels in the small intestine in comparison to bone marrow. The persistence of exchangeable levels of drug (i.e. above that presumed to be equal to the dihydrofolate reductase-binding capacity) was also much greater in the small intestine. In addition, the more prolonged persistence of exchangeable drug in the small intestine compared to marrow correlated with a substantially longer duration of inhibition of DNA synthesis ([6-3H]deoxyuridine incorporation) in the former. Earlier recovery of DNA synthesis as a consequence of more rapid loss of drug appears to explain the lower sensitivity of marrow compared to small intestine to the effects of this agent in mice. These studies extend prior studies in our laboratory to the two major host proliferative populations in mice and allow us to propose that the property for accumulating and maintaining pharmacologically effective intracellular levels of folate analogs is differential among all proliferative tissues (tumor and normal) of this animal and probably in higher mammals as well.

Further evidence for a basis of selective activity and relative responsiveness during antifolate therapy of murine tumors.

A greater persistence of unbound (exchangeable) drug in tumor cells versus drug-limiting normal tissue (proliferating epithelium of small intestine) correlates with the therapeutic effects of various antifolates against a group of murine tumors. After approsimate equimolar doses (3 mg/kg i.p.) of methotrexate (MTX) methasquin (MQ), aminopterin, and N-([2,4-diamino-5-chloro-6-quinazolinyl) methyl]-amino)benzol)-L-glutamate (5-Cl-deaza-AM), total accumulation in small intestine was vie- to eight-fold greater than the dihydrofolate reductase content. Free drug persisted for less than 4 hr (MTX), 16 hr (MQ), 30 hr (aminopterin),and 48 hr (5-Cl-deaza-AM). Overall drug accumulation in L121O cells was greater (12- to 40-fold enzyme level), and drug persistence above enzyme level was more prolonged in the case of MTX (24 hr),MQ (32 hr), and 5-Cl-deaza-AM (greater than 48 hr). Persistence of aminopterin was similar to that seen in small intestine. After the same dose of each drug s.c., the results were similar in small intestine. In L121o cells, however, the total drug accumulation was much lower, but the relative persistence of each was similar to that seen after an i.p. dose. After a single optimal therapeutic dose (3, 0.75, 0.3, and 0.1 mg/kg i.p. for MTX, MQ, aminopterin, and 5-Cl-deaza-AM every other day), accumulation of each drug in for no more than 4 hr. In L121O cells, maximal accumulation of each drug also varied, but persistence differed in accordance with the relative therapeutic effectiveness of each (9 hr for 5-Cl-deaza-AM, 12 hr for aminopterin, and more than 20 hr for MTX amd MQ).

Kinetic correlates of methotrexate transport and therapeutic responsiveness in murine tumors.

Net accumulation of methotrexate by carrier-mediated transport in different murine tumor cells in vitro exhibits a positive correlation with the relative drug pharmacokinetics and therapeutic responsiveness in these tumors in vivo. The transport of methotrexate by Sarcoma 180, Ehrlich carcinoma, P388, P288, and L1210 leukemia cells is qualitatively similar. Influx of drug exhibits saturation kinetics and is highly temperature dependent (Q10, 6.1 to 9.4). Efflux of exchangeable methotrexate from all of the different tumor cells exhibited first-order kinetics and the same high temperature dependence seen for influx (Q10, 6.1 to 8.0). The major kinetic determinant of responsiveness is the Km for influx. Values vary from 3.1 to 11.2 X 10(-6) M and are highest in cells from a nonresponsive Sarcoma 180 tumor, somewhat lower in the poorly responsive Ehrlich tumor, lower in moderately responsive P388 and P288 leukemias, and lowest in the highly responsive L1210 leukemia. Values for the influx Vmax differ to some extent, but in a manner not correlatable with responsiveness. The level of responsiveness of the P388 leukemia in vivo can also be partially attributed to an efflux rate that is lower than that measured for the other tumor cells. Steady-state levels of drug accumulation in vitro reflected influx and efflux rates and were consistently correlatable with therapeutic responsiveness. There was no significant difference in the extent to which folate and reduced 5-substituted folate derivatives compete with methotrexate for uptake in cells from all five tumors. The average value for Ki measured with folate for each tumor cell type was 50- and 80-fold higher than for 5-formyltetrahydrofolate and 5-methyltetrahydrofolate.

The issues of transport multiplicity and energetics pertaining to methotrexate efflux in L1210 cells addressed by an analysis of cis and trans effects of inhibitors.

Studies are described addressing the controversial issue of the multiplicity of efflux routes for [3H]MTX in L1210 cells. We examined efflux multiplicity under conditions that do and do not maintain cellular ATP at the physiological level. In ATP-replete cells, the results delineate a probenecid-, bromosulfophthalein-, and verapamil-inhibitable route that accounts for nearly 90% of [3H]MTX efflux. Efflux of [3H]MTX by this route is inhibited by bromosulfophthalein in the trans orientation only, inhibited by probenecid only when present simultaneously in the cis and trans orientation and inhibited by verapamil only in the cis orientation. The remaining efflux of this folate analogue in ATP-replete cells appears to be mediated by the one-carbon, reduced folate system (MTX influx route), in that it is not inhibited by bromosulfophthalein or verapamil but is inhibited by the N-hydroxysuccinimide ester of MTX, a specific inhibitor of MTX influx, and a 10-fold higher concentration of probenecid than that required to inhibit ATP-dependent efflux. Under these conditions, MTX did not trans-stimulate [3H]MTX efflux. Also, no evidence was obtained for a putative bromosulfophthalein-insensitive, probenecid-inhibitable route for [3H]MTX efflux. In cells depleted to the extent of 90-95% of their ATP by 60-min incubation in medium in the absence of D-glucose and with 10 mM sodium azide, overall efflux of [3H]MTX was markedly reduced and appears to be mediated solely by the MTX influx route. Influx of [3H]MTX was both cis and trans inhibited by probenecid, and efflux under these conditions was markedly inhibited by the N-hydroxysuccinimide ester of MTX and trans-stimulated by MTX. Overall, the results of these studies are consistent with a model for methotrexate transport in L1210 cells derived [Dembo et al., J. Membrane Biol., 78: 9-17, 1984] in the authors' laboratory based solely upon a kinetic analysis of [3H]MTX influx and efflux in ATP-replete and depleted L1210 cells. As such, these new results identify a single ATP-dependent efflux route as the bromosulfophthalein-, probenecid-, and verapamil-inhibitable route in L1210 cells under conditions that maintain ATP levels at maximum.

Markedly improved efficacy of edatrexate compared to methotrexate in a high-dose regimen with leucovorin rescue against metastatic murine solid tumors.

10-Ethyl-10-deazaaminopterin (EDX, edatrexate) exhibits therapeutic activity against methotrexate (MTX)-resistant tumors in animals and patients. In an effort to improve its efficacy among more chemoresistant tumors, studies were initiated in murine models of advanced metastatic disease comparing EDX and MTX at their maximum tolerated dose alone and in a high-dose regimen incorporating low-dose, delayed Ca leucovorin (LCV) rescue. Both twice-weekly x 3 and weekly x 3 schedules of administration were used with LCV given 16, 20, and 24 h after EDX. The LCV dose required to protect mice was 1/40 and 1/20 of the EDX or MTX dose, respectively, on either schedule. Therapy was initiated 5 or 6 days following i.v. implant of 5 x 10(5) cells of the E0771 mammary adenocarcinoma, T241 fibrosarcoma, Lewis lung carcinoma, B16 melanoma, or C38 colon carcinoma. MTX was essentially ineffective (increase in life span = < 30%) when given alone and either ineffective or only modestly effective (increase in life span = 20-80%) in increasing survival when given in the high-dose regimen to tumor-bearing mice. EDX alone was more effective than MTX when it was given in either regimen of therapy. Also, EDX given in the high-dose regimen (either twice-weekly or weekly x 3) was markedly more effective than EDX alone. Increased survival with this regimen was 2-3-fold greater than EDX alone against all 5 tumors, and long-term survivors were obtained with E0771 (20%), T241 (30-40%), Lewis lung (10-15%), B16 (20%), and C38 (40%) tumors. The administration of 6 doses rather than 3 doses on the twice-weekly schedule against T241 and Lewis lung tumors required a modest increase in the LCV dose but substantially improved efficacy, with as much as 70% long-term survivors (T241 tumor). We conclude that the use of a high-dose regimen with delayed LCV rescue markedly improved the therapeutic effectiveness of EDX against advanced metastatic disease in tumor-bearing mice. These studies should provide a framework for further clinical work with EDX, using this modality of therapy.

Enhancement of folate analogue transport inward in L1210 cells during methotrexate therapy of leukemic mice: evidence of the nature of the effect, possible host mediation, and pharmacokinetic significance.

Studies are described that sought the basis for a discrepancy in values for a key kinetic parameter of methotrexate transport (influx Vmax) in L1210 cells derived alternately from biochemical or pharmacokinetic measurements. Our results show that, within a short period of time following administration of a therapeutic dose of methotrexate to leukemic mice, influx of this folate analogue measured in L1210 cells removed from these mice was markedly stimulated. Enhancement of [3H]methotrexate influx in these cells was observed within 15 min of drug administration, was maximum (up to 3-fold) within 2 to 3 h, then decreased with time until 24 h when influx was at the control level. Measurements of [3H]methotrexate influx in cells removed from drug-treated mice were made after a period of incubation in drug-free medium to allow for efflux of exchangeable drug. Enhanced influx of [3H]methotrexate was accounted for by an increase in influx Vmax (influx Km was unchanged) and was further enhanced (to a total of 5-fold) by coadministration of leucovorin. Also, enhancement of influx of [3H]methotrexate in L1210 cells did not occur following administration of 1-beta-D-arabinofuranosylcytidine at a therapeutically equivalent dose to leukemic mice or following exposure of these cells to methotrexate or methotrexate with leucovorin during growth in culture. Methotrexate therapy did not affect all transport systems, since the same therapy of leukemic mice had no effect on influx of the purine nucleoside analogue, 9-beta-D-arabinofuranosyl-2-fluoroadenine, in these same L1210 cells. These findings suggest that stimulation of [3H]methotrexate influx in L1210 cells during therapy with this folate analogue was not due to transstimulation during exchange between folate compounds and was not related to the antiproliferative effect of methotrexate on these tumor cells. The coadministration of cycloheximide with methotrexate to leukemic mice at a dose which markedly inhibited 3H-leucine incorporation into L1210 cell protein severely diminished the stimulation of [3H]methotrexate influx. However, in L1210 cells removed from leukemic mice treated with methotrexate, there was no increase compared to control cells in affinity labeling with the N-hydroxysuccinimide ester of [3H]methotrexate. This suggested that the effect of cycloheximide was not on increased synthesis of folate transporter and that increased rate of translocation of folate transporter, rather than increased amount of transporter, accounted for the increase in [3H]methotrexate influx.(ABSTRACT TRUNCATED AT 400 WORDS)

Relationship of reduced folate changes to inhibition of DNA synthesis induced by methotrexate in L1210 cells in vivo.

Reduced folate levels and DNA synthesis were examined in L1210 cells in mice after exposure to a wide range of methotrexate doses. A radioenzymatic assay based upon entrapment of tissue 5,10-methylenetetrahydrofolate (CH2FH4), and other reduced folates after cycling to this form, into a stable ternary complex with thymidylate synthase and [3H]-5-fluoro-2'-deoxyuridine 5'-monophosphate was used to estimate reduced folates. DNA synthesis was estimated from incorporation of [3H]-2'-deoxyuridine into DNA. The predominant reduced folate in cells from untreated animals was 10-formyltetrahydrofolate (10-CHOFH4; 3.59 pmol/10(6) cells). The four other reduced folates measured, tetrahydrofolate (FH4), CH2FH4, dihydrofolate (FH2), and 5-methyltetrahydrofolate (5-CH3FH4), were present in nearly equal amounts yielding a total reduced folate level of 6.24 pmol/10(6) cells. When methotrexate was administered s.c. at doses of 1.5, 12, and 400 mg/kg, the level of FH2 increased dramatically and total tetrahydrofolates measured decreased extensively within 1 h. DNA synthesis was completely inhibited during the first 1-2 h after administration of each dose of methotrexate with onset of recovery after 4 and 20 h at 1.5 and 12 mg/kg and not at all after the highest dose. Both FH4 and CH2FH4 were extensively depleted at 12 and 400 mg/kg methotrexate but considerably less depletion of CH2FH4, and none of FH4, was observed at 1.5 mg/kg during a period when DNA synthesis was essentially abolished. The metabolically linked 5-CH3FH4 and 10-CHOFH4 pools were also extensively depleted following treatment with methotrexate. While FH2 levels expanded extensively after drug treatment, the total increase did not account for the extent of depletion of the combined tetrahydrofolate pools. The change in concentration with time of any one folate pool was apparently not sufficient to explain completely the duration of inhibition of DNA synthesis; however, sustained inhibition of DNA synthesis was generally associated with maintenance of an expanded FH2 pool and delay in repletion of the combined tetrahydrofolate pools. Discussion is presented with respect to the impact of these results on changing notions of the mode of action of classical antifolates.

Similar characteristics of folate analogue transport in vitro in contrast to varying dihydrofolate reductase levels in epithelial cells at different stages of maturation in mouse small intestine.

We describe studies of folate analogue transport in purified epithelial cell fractions isolated from mouse small intestine. Fractionation of these cells into immature proliferative and mature absorptive components and two components representative of intermediate stages of maturation was carried out by stepwise, nonenzymatic stripping of the everted organ. In contrast to the proliferative-specific enzyme markers, thymidine kinase and dihydrofolate reductase, folate analogue transport did not vary with the alteration in proliferative potential of these cells during maturation. The brush-border enzyme, alkaline phosphatase, was used as a positive marker for maturation. Initial influx of [3H]-aminopterin into both mature and immature cell fractions showed the same kinetics and did not exhibit pH dependence within the range of 6.0 to 7.8. A single saturable component (Km = 16 +/- 3 microM; V37 = 57 +/- 8 pmol/min/10(7) cells) was delineated, with the same temperature dependence (Q10 27-37 degrees = 3.2 +/- 0.4; Arrenhius constant = 11.1 +/- 3 kcal/mol) and same specificity for various folate compounds. Initial efflux of [3H]aminopterin was also similar in both cell types. Efflux was first-order (t1/2 37 degrees = 1.1 to 1.3 min; K37 = 0.53 +/- 0.04 min-1) and equal to the decay-time constant for approach to steady-state during accumulation of [3H]aminopterin, but showed higher-temperature dependence (Q10 27-37 degrees = 6.7 +/- 0.8; Arrenhius constant = 25.3 +/- 4 kcal/mol). Under the conditions used which do not allow polyglutamylation of [3H]aminopterin, steady-state levels of accumulation of exchangeable drug at 37 degrees in each cell fraction were accounted for by the various kinetic parameters for each flux. At all concentrations of [3H]aminopterin examined, both types of epithelial cells appeared to maintain a negative electrochemical gradient under physiological conditions. Overall, the data conform to a two-carrier model for folate analogue transport in which each flux is mediated by a separate system. However, specificity and saturability of influx for folate compounds, and inhibition of this flux by various agents was markedly different from that reported for various tumor cells.

Increased accumulation of methotrexate by murine tumor cells in vitro in the presence of probenecid which is mediated by a preferential inhibition of efflux.

Carrier-mediated influx and efflux of [3H]methotrexate by L1210 leukemia cells was inhibited by probenecid. The concentration of probenecid required for inhibition of influx was markedly greater than that required to inhibit efflux. The concentration determined for 50% inhibition of influx was 1.35 +/- 0.15 (S.D.) mM. Inhibition of this flux was competitive (Ki = 1.23 +/- 0.2 mM) and was reversed after removal of probenecid. In contrast, the concentration determined for 50% inhibition of efflux was only 0.12 +/- 0.016 mM, and inhibition was also reversed after removal of probenecid. As a consequence of the different extent of inhibition of each flux by probenecid, the level of intracellular [3H]methotrexate at steady state was markedly increased. At 0.1 and 1 mM probenecid, the steady state level was increased 2- and 2.6-fold, respectively. These observed increases are in close agreement with that expected from the effect on each flux at these concentrations. From other data on the inhibition of each flux at higher concentrations of probenecid, a maximum effect (3- to 4-fold increase) in steady-state level would be expected at a probenecid concentration between 2 and 3 mM. A similar relationship between the inhibition by probenecid of influx and efflux of [3H]methotrexate was also shown for Sarcoma 180 and Ehrlich carcinoma cells. These results have pharmacological implications with respect to the adjuvant use of probenecid or related organic ions during folate analog therapy of human cancer.

Differential synthesis of methotrexate polyglutamates in normal proliferative and neoplastic mouse tissues in vivo.

Synthesis of poly-gamma-glutamyl metabolites of methotrexate was demonstrated in mouse small intestine, liver, and bone marrow and in L1210 leukemia, Sarcoma 180, and Ehrlich tumor cells after s.c. injections of [3H]methotrexate to tumor-bearing mice. Ion-exchange chromatography of tissue extracts resolved six peaks of radioactivity believed to represent methotrexate and metabolites with up to five additional glutamyl residues. Polyglutamate formation in L1210 cells and small intestine was shown to be independent of dose at least to 400 mg/kg as long as intracellular levels of drug in excess of the dihydrofolate reductase-binding capacity (exchangeable) were maintained. Both the total amount of polyglutamates and the average length of the polyglutamyl chain increased with time as long as exchangeable level of drug was present intracellularly. The results also showed differences in the extent of metabolism of methotrexate polyglutamates among the tissues examined. Although these differences were at times very large, there was no consistent correlation between these differences and other pharmacological parameters or cytotoxicity. Tumor cells appeared to synthesize more polyglutamates than did the normal tissues examined. However, differences in total drug persistence and sensitivity to drug among tumor cells and among normal tissues did not reflect the relative extent of polyglutamate synthesis in each group. It is concluded that the extent of polyglutamate synthesis per se may not be a determinant of drug sensitivity in murine tissues. However, the accumulation of these metabolites may contribute in some way to overall therapeutic response or relative cytotoxicity.

Differing specificities for 4-aminofolate analogues of folylpolyglutamyl synthetase from tumors and proliferative intestinal epithelium of the mouse with significance for selective antitumor action.

Folylpolyglutamyl synthetase (FPGS), partially purified from murine L1210 leukemia and Sarcoma 180 cells and the proliferative fraction of luminal epithelium from mouse small intestine (the site of limiting toxicity to folate analogues), was examined for its ability to utilize various 4-aminofolates as substrates. For tumor-derived FPGS, aminopterin was the most preferred substrate overall, exhibiting the lowest value for apparent Km and highest Vmax. The other analogues and folic acid exhibited nearly 2-fold lower Vmax. Folic acid exhibited a 3-fold higher Km than aminopterin. Alkylation of aminopterin (methotrexate) or carbon for nitrogen substitution (10-deazaaminopterin) at N-10 increased Km 3- to 6-fold, while alkylation at C10 (10-ethyl-10-deazaaminopterin) restored Km to near equivalency with aminopterin. For FPGS derived from proliferative intestinal epithelium, aminopterin was also the preferred substrate, but the value for Vmax (derived with crude cell-free extract) was 6-fold lower than for tumor cell FPGS. Values for Vmax (derived with partially purified FPGS) for the other 4-aminofolate analogues and folic acid were similar (methotrexate) or 2-fold (10-ethyl-10-deazaaminopterin) and 5-fold (folic acid) lower than for aminopterin. The value for Km derived with aminopterin was similar to that derived for either tumor cell FPGS. The value for folic acid was 2-fold higher, and alkylation of aminopterin (methotrexate) or carbon to nitrogen substitution (10-deazaaminopterin) at N-10 with (10-ethyl-10-deazaaminopterin) or without alkylation markedly increased Km (27-, 90-, and greater than 100-fold, respectively, for methotrexate, 10-ethyl-10-deazaaminopterin, and 10-deazaaminopterin). In other studies, it was found that the diglutamate of aminopterin (aminopterin +G1) was a relatively poor substrate for FPGS derived from all three sources compared with methotrexate diglutamate, both in respect to values for Km and Vmax that were measured in each case. Findings with FPGS derived from L1210 cells were confirmed by high-pressure liquid chromatography analysis of product formation during the reaction with the parent compounds. The significance of the results presented here to the question of relative toxicity and therapeutic activity of these analogues is discussed.

Optimization of high-dose methotrexate with leucovorin rescue therapy in the L1210 leukemia and sarcoma 180 murine tumor models.

An analysis of dose and schedule dependence of calcium leucovorin rescue during high-dose methotrexate therapy of ascitic forms of l1210 leukemia and Sarcoma 180 is reported. Schedules with very delayed "low-dose" leucovorin rescue following lethal doses of methotrexate were highly effective in preventing toxicity and achieved a pronounced antitumor effect in both ascites tumor models. Best results were obtained on a schedule of methotrexate (400 mg/kg s.c.) followed 16 to 20 hr later by calcium leucovorin (12 mg/kg s.c.) given once every 2 hr for a total of 5 doses. Progressive increases in the calcium leucovorin dosage on any schedule reduced both toxicity and the antitumor effect of methotrexate in each model. Following a single course of therapy, essentially no toxicity was observed, and the antitumor effects were 2-fold (L1210 leukemia) and 4-fold (Sarcoma 180) greater than a single, maximally tolerated dose (24/kg s.c.) methotrexate alone. An increase in the methotrexate dosage to 800 mg/kg s.c. with or without an increase in calcium leucovorin dosages on the same schedule did not appreciably increase the antitumor effect observed. Two courses of high-dose methotrexate (400 mg/kg s.c.) with leucovorin rescue (24 mg/kg s.c. 16, 20, and 24 hr after drug) given with an 8-day interval between courses doubled the total antitumor effect in each model with no substantial increase in toxicity and gave long-term survivors with Sarcoma 180. The results, overall, are in close agreement with prior prediction for schedule and dose dependence made on the basis of related pharmacokinetic and biochemical studies in murine tumor models reported from this laboratory.

Intracavitary therapy of murine ovarian cancer with cis-diamminedichloroplatinum(II) and 10-ethyl-10-deazaaminopterin incorporating systemic leucovorin protection.

Administration i.p. of 10-ethyl-10-deazaaminopterin (10EDAM) with cis-diamminedichloroplatinum(II) (cis-Pt) had significant antitumor activity against the murine ovarian tumor. This tumor is a teratoma originating in the ovary with pathogenesis and metastatic properties similar to those of human ovarian cancer. Drug was given on a schedule of once every 3 days for 3 doses 1 or 2 days after i.p. implant of 10(7) tumor cells. Despite the 2-fold attenuation of dosage required, antitumor activity of the combination (increased life span, 161%) was approximately twice that obtained with maximum tolerated doses of either agent alone and tumor-free, long-term survivors were obtained. Incorporation of s.c. calcium leucovorin administration 16 h after each dose of 10EDAM and cis-Pt allowed a 4-fold increase in dosage of 10-EDAM without an increase in toxicity, increased median survival by an additional 120%, and quadrupled the number of tumor-free, long-term survivors to 40% of treated animals. By comparison, methotrexate was only modestly active against this tumor model either as a single agent, with cis-Pt, or with delayed s.c. calcium leucovorin administration. These results appear to suggest that 10EDAM with cis-Pt may have considerable potential for intracavitary therapy of human cancer, including ovarian carcinoma, particularly when incorporating delayed systemic calcium leucovorin administration.

Extent of the requirement for folate transport by L1210 cells for growth and leukemogenesis in vivo.

In these studies the extent of the requirement for 5-methyltetrahydrofolate by L1210 cells for growth and leukemogenesis in vivo was addressed from the aspect of its cellular membrane transport. Growth characteristics and leukemogenesis in vivo were determined for parental and methotrexate-resistant L1210 cell variants with reduced capacity for folate coenzyme transport inward. These variants exhibited 6-, 16-, and 100-fold reductions compared to parental cells in influx Vmax for the high-affinity system transporting 5-substituted reduced folates and methotrexate. They also exhibited reduced saturability for methotrexate influx (3-fold higher Km), but not for influx of 5-formyltetrahydrofolate or 5-methyltetrahydrofolate. The reduced influx capacity in these variants correlated with their increased requirement for reduced folates during growth in vitro and with the ability of the variants to proliferate and develop leukemia in vivo. Lack of growth potential in vivo for one variant appears to reflect the inability for net intracellular accumulation of reduced folate per se, since growth of this variant could be restored by treatment of mice with folic acid, but not with 5-methyltetrahydrofolate or 5-formyltetrahydrofolate, and following reversion to a more transport-proficient phenotype.

Similar differential for total polyglutamylation and cytotoxicity among various folate analogues in human and murine tumor cells in vitro.

Four folate analogues, methotrexate, aminopterin, 10-deazaminopterin, and 10-ethyl-10-deazaaminopterin were assessed for their ability to be metabolized to poly-gamma-glutamyl derivatives in three tumor lines which vary in their sensitivity to these agents. Cytotoxicity of the four analogues against the murine L1210 leukemia and the human Manca B cell leukemia, as determined by a 3-h clonogenic assay, showed aminopterin and the two 10-deazaaminopterin compounds to be approximately equivalent for each cell type and were 3- to 10- (L1210) and 7- to 14-fold (Manca) more potent than methotrexate. In murine Sarcoma 180 cells, 10-ethyl-10-deazaaminopterin and aminopterin were similarly potent but were 5- to 10-fold more potent than 10-deazaaminopterin and 40- to 80-fold more potent than methotrexate. These results could be explained in part by the differences in transport properties and substrate activities for polyglutamylation for each analogue in these cell types. Initial rates of polyglutamate accumulation of the four analogues, which were determined under conditions of comparable rates of drug entry into the three tumor cell lines, were 7- to 18-fold less than drug entry rates. In L1210 and Sarcoma 180 cells, the relative rates of polyglutamylation were in the order aminopterin greater than 10-ethyl-10-deazaaminopterin greater than methotrexate greater than 10-deazaaminopterin. In contrast, the relative rates of polyglutamylation in Manca cells were in the order 10-ethyl-10-deazaaminopterin approximately equal to aminopterin greater than 10-deazaaminopterin greater than methotrexate, suggesting that folylpolyglutamyl synthetase may have varying substrate preferences in different cell types. The maximum relative extents of total polyglutamate accumulation in L1210 cells were 85 to 95% of the total drug at 24 h. In Manca cells, the maximum polyglutamate accumulation was also 85 to 95%, but this was obtained by 6 h. However, in Sarcoma 180 cells, only aminopterin polyglutamates reached a similar maximum percentage of accumulation, while lower relative polyglutamate levels were achieved with the other analogues. Accumulation of individual polyglutamates in each cell line was similar for all analogues except aminopterin. For methotrexate and the two 10-deazaaminopterins, accumulation occurred mainly as the tetraglutamate or as higher polyglutamates. Aminopterin was accumulated mainly as the diglutamate, particularly in Manca cells where 70% of total drug was in the diglutamate form within the first 3 h and remained the predominant form for 24 h.(ABSTRACT TRUNCATED AT 400 WORDS)

Biochemical and pharmacokinetic effects of leucovorin after high-dose methotrexate in a murine leukemia model.

The administration of calcium leucovorin, either concurrently or after high dosages of methotrexate in L1210 leukemic mice, has both pharmacokinetic and biochemical effects in tumor cells and drug-limiting proliferative normal tissue in small intestine. A reduction in the maximal level of accumulation and retention of exchangeable drug (unbound to dihydrofolate reductase) in tissue could be demonstrated when calcium leucovorin was given simultaneously with methotrexate at equal or greater dosages than the latter. The dose dependence for calcium leucovorin-introduced drug loss is similar in both tissues and showed the expected variation when the time interval between methotrexate and calcium leucovorin doses was increased. With 400 mg methotrexate per kg, greater than 96 mg calcium leucovorin per kg were required maximally to affect overall drug retention in tissue 2 hr after drug, whereas only 24 mg calcium leucovorin per kg were required 16 hr after drug. Calcium leucovorin, given after methotrexate, induced synchronous recovery of DNA synthesis (measured by labeled deoxyruridine incorporation) in both small intestine and L1210 cells. An initial cycle of synthesis was induced in the presence of exchangeable levels of drug. Two hr after methotrexate, 12 mg/kg, calcium leucovorin induced an immediate but only partial (20 to 25% of control rate) recovery of synthesis with dose dependence from 3 to 12 mg calcium leucovorin per kg. Less synthesis was induced after 96 mg/kg and almost none after methotrexate, 400 mg/kg. With calcium leucovorin, 24 mg/kg, given 2 hr after methotrexate, 12 or 96 mg/kg, a major cycle of synthesis occurred when total drug levels approached the equivalence of the dihydrofolate reductase content. The magnitude of this cycle of synthesis in both L1210 cells and small intestine was the same as that seen in animals recovering from methotrexate alone. However, this is based on the assumption that an approximately equivalent relationship between DNA synthesis and labeled doexyuridine incorporation in each tissue during the period of maximal incorporation within the cycle. The major effect of calcium leucovorin, then, was to induce an earlier resumption of DNA synthesis as a consequence of the pharmacokinetic effect in each tissue. With calcium leucovorin, 24 or 400 mg/kg, given 16 hr after methotrexate, an identical effect on drug retention was observed in both L1210 cells and small intestine. Although there was a difference in the time course for recovery in small intestine at each dosage of calcium leucovorin, the recovery of DNA synthesis as drug levels approached the dihydrofolate reductase content was similar in magnitude. In L1210 cells, however, substantial recovery of synthesis to a comparable level and with a similar time-course occurred only after leucovorin, 400 mg/kg. Little or no recovery of DNA synthesis was observed after calcium leucovorin, 24 mg/kg, during the same time period. This dosage schedule (methotrexate, 400 mg/kg, s.c...

Tissue pharmacokinetics, inhibition of DNA synthesis and tumor cell kill after high-dose methotrexate in murine tumor models.

In Sarcoma 180 and L1210 ascites tumor models, the initial rate of methotrexate accumulation in tumor cells in the peritoneal cavity and in small intestine (intracellularly) after s.c. doses up to 800 mg/kg, showed saturation kinetics. These results and the fact that initial uptake in these tissues within this dosage range was inhibited to the expected relative extent by the simultaneous administration of leucovorin suggest that carrier mediation and not passive diffusion is the major route of drug entry at these extremely high doses. Maximum accumulation of intracellular drug occurred within 2 hr and reached much higher levels in small intestine than in tumor cells at the higher dosages. At a 3-mg/kg dose of methotrexate s.c., intracellular exchangeable drug levels persisted more than four times longer in L1210 cells than in small intestine, but differences in persistence (L1210 cell versus gut) diminished markedly with increasing dosage. At 96 mg/kg, the difference in persistence was less than 2-fold. In small intestine and L1210 cells, theduration of inhibition of DNA synthesis at different dosages correlated with the extent to which exchangeable drug was retained. Toxic deaths occurred when inhibition in small intestine lasted longer than 25 to 30 hr. Recovery of synthesis in small intestine and L1210 cells occurred synchronously and only below dosages of 400 mg/kg. Within 24 hr after dosages of greater than 24 mg/kg, the rate of tumor cell loss increased to a point characterized by a single exponential (t1/2=8.5 hr). The total cell loss, but not the rate of cell loss, was dose dependent.

Relative frequency and kinetic properties of transport-defective phenotypes among methotrexate-resistant L1210 clonal cell lines derived in vivo.

Information was sought on the relative extent to which transport-defective, methotrexate-resistant phenotypes emerge among the total subpopulation of resistant phenotypes during therapeutic challenge of leukemic cells in vivo. A number of monoclonal methotrexate-resistant sublines of the L1210 leukemia were derived during methotrexate therapy of leukemic mice and biochemically characterized. Of the total number of 14 sublines derived, five exhibited altered [3H]methotrexate transport alone, five exhibited increased dihydrofolate reductase content alone (2- to 18-fold), and four showed alterations in both of these properties. Methotrexate binding and substrate turnover rate for dihydrofolate reductase appeared to be unchanged in any of the resistant sublines. The relative resistance of each subline was accounted for by the biochemical alterations observed. Among the transport-defective sublines, one subcategory showed a 3- to 4-fold reduction in apparent influx Vmax for [3H]methotrexate, a second category showed both a 5-fold reduction in influx Vmax and a 3-fold increase in the apparent influx Km, and one subline showed only a 2-fold increase in Km. Otherwise, Michaelis-Menten saturation kinetics for influx was observed in each case and in the case of the parental line and the other resistant sublines. None of the resistant sublines exhibited altered efflux of [3H]methotrexate. Steady-state levels measured for intracellular exchangeable (osmotically active) fractions of drug accurately reflected the values for specific kinetic parameters determined for each sensitive and resistant cell line. These studies show that transport-defective phenotypes represent a major category of methotrexate-resistant cell types which emerge initially from leukemic cell populations under therapy in mice. Based on considerations discussed here, it is reasonable to assume that a similar relative occurrence of this phenotype would result during methotrexate therapy of leukemia patients.