[CANCER RESEARCH 49. 5755-5760. October 15. ]989|

Pharmacokinetics of Diastereoisomers of (6l?,5)-Folinic Acid (Leucovorin) inHumans during Constant High-Dose Intravenous Infusion1

Edward M. Newman,2 James A. Straw, and James H. Doroshow

Division of Pediatrics [E. M. N.] and Department of Medical Oncology and Therapeutics Research [J. H. D.], City of Hope Cancer Research Center, Duarte, California9IOIO; and Department of Pharmacology [J. A. S.J, The George Washington University, Washington, DC 20037

ABSTRACT

Eleven patients treated with a 5.5-day continuous i.v. infusion of 500mg/nr/day of (6/f,S)-folinic acid in combination with daily bolus 5-fluorouracil had a median steady-state plasma concentration of 3.25 ¿/\i(65)-folinic acid (the bioactive diastereoisomer). The bioactive metabolite(65)-5-methyltetrahydrofolic acid, analyzed in six patients, reached amedian steady-state plasma concentration of 5.7 MM.The lowest plasmaconcentrations at steady-state were 1.86 MM(6i)-folinic acid and 3.12MM(6S>5-rnethyltetrahydrofolic acid. These concentrations are abovethe minimum concentrations shown by other investigators to producesynergism between (6A,S)-folinic acid and 5-fluorouracil in vitro. Themedian steady-state plasma concentration of ((>Wl-t'olinic acid was 38.2

MM,more than 10 times the concentration ot (6 S ) Inlinu acid. Along withother plasma pharmacokinetic parameters, terminal half-lives were estimated for (fi.V)'l'oliine acid (median, 45.4 min), (6/î)-folinicacid (median,

388 min), and (6S)-5-methyltetrahydrofolic acid (median, 446 min).Investigation of the renal pharmacokinetics confirmed the marked difference in the renal clearance of the two diastereoisomers of folinic acidwhich had been observed after low doses of (6A,5)-folinic acid (J. A.Straw, D. Szapary, and VV.T. \Vynn, Cancer Res., 44:3114-3119, 1984).However, the low renal clearance of (6/?)-folinic acid (median, 8.2 ml/min m ' i was attributable to the extensive binding of (6A)-folinic acid to

plasma proteins (median, 8.7% free), not to reabsorption in the kidney.

INTRODUCTION

The clinical use of high-dose (67v,5)-folinic acid and 5-fluorouracil to treat colorectal cancer and other neoplasms is increasing. The rationale for using this combination of drugs isbased on preclinical evidence demonstrating cytotoxic synergism between fluoropyrimidines and folates (1-4), the biochemical data that the 5,10-methylenetetrahydrofolate/fluorodeox-yuridine monophosphate/thymidylate synthase ternery complex is stabilized by high concentrations of the folate cofactor(5), and the correlation between the cell growth-inhibitorypotential of fluoropyrimidines and the stability of the ternarycomplex (2, 6).

Because of in vitro evidence that antineoplastic synergism isdependent on the concentration of (6/»,S)-folinicacid, the human pharmacokinetics of this compound are likely to have animportant impact on the potential for clinical success. Severalfactors complicate any study of the pharmacokinetics of(6/î,.S)-folinicacid. The natural (6S) and unnatural (6R) diastereoisomers of folinic acid display significantly different pharmacokinetic behavior (7). The plasma concentrations of theunnatural isomer cannot be dismissed as irrelevant because thetwo isomers may compete for membrane transport (8) and

Received 12/27/88; revised 6/28/89; accepted 7/6/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1Supported by a grant from American Cyanamid Co.. Lederle Laboratories,Pearl River. NY. by NIH Grant BRSG S07RR-0547I to the City of HopeNational Medical Center, and by NIH Grant CA33572, the City of Hope CancerCenter core grant.

- To whom requests for reprints should be addressed, at Division of Pediatrics.City of Hope National Medical Center. 1500 East Duarte Road. Duarte. CA91010.

metabolism (9). (65)-5-CH,-THF,' the major metabolite of(6S)-folinic acid, is also biologically active. However, the relative contributions of (65)-folinic acid and (6.S)-5-CH,-THF inplasma to the intracellular folate cofactor pools of the patients'

tumors are unknown. To examine only the total biologicallyactive folates would be both pharmacokinetically unsound andpotentially misleading clinically. The possibility that some human tumors may not metabolize (OSJ-S-CH^-THF efficiently(10) underscores the importance of analyzing (6S)-folinic acidand (65)-5-CH,-THF separately.

A series of Phase II and III clinical trials in breast, colon,and hepatocellular cancer have been conducted at the City ofHope Cancer Research Center with treatment regimens consisting of high-dose continuous i.v. infusions of (6/î,5)-folinicacid and daily bolus i.v. 5-fluorouracil (11-12). Because intracellular folates accumulate slowly even when cells are continuously exposed to (o/^SJ-folinic acid in tissue culture (6, 13) thefolinic acid infusion for these trials was started 24 h before thefirst 5-fluorouracil dose. Because folate derivatives may not beretained within tumors during periods of low folate concentrations in the plasma (14), the (o^SJ-folinic acid infusion wascontinued throughout the 5-day course of daily 5-fluorouraciltreatment to minimize the possibility that ternary complexwould dissociate to active thymidylate synthase between 5-fluorouracil doses. The dose of (oA^-folinic acid for thesetrials (500 mg/nr/day) was chosen in anticipation that it wouldproduce C„of (6S>folinic acid in excess of 1 /tM, a concentration which has been shown to be synergistic with 5-fluorouracilin vitro (1).

Since the initial investigation of the pharmacokinetics of (6R,5)-folinic acid in humans (7), other investigators have studiedthe plasma pharmacokinetics after high doses of (6/?,S)-folinicacid were administered by short infusions (15-17). This reportdeals specifically with the plasma and renal pharmacokineticsof high-dose (6/?,5)-folinic acid administered by continuousinfusion. Preliminary data from three patients have been reported previously (18). The Phase II clinical results are reportedelsewhere (11-12).

MATERIALS AND METHODS

Subjects and Samples. Eleven patients with hepatocellular (one),colorectal (eight), or breast (two) carcinoma were participating in phaseII or 111clinical trials with their informed consent. These patientsreceived a continuous infusion of 500 mg/nr/day of (6/?,5)-folinic acidfor 5.5 days and 5-fluorouracil (370 mg/nr) by i.v. bolus daily for 5days beginning 24 h after the (6/î,S)-folinicacid infusion was started.In Patients 1-6, plasma samples for determination of (65)-folinic acid,(6/?)-folinic acid, and (6S)-5-CH,-THF were obtained at the following

•¿�'The abbreviations used are: 5-CH,-THF, 5-methyltetrahydrofolic acid; AUC,

area under the plasma concentration versus time curve; AUMC, area under themoment (concentration x time) versus time curve; CL„„nonrenai clearance; C¿p,total plasma clearance; C£„renal clearance; C/t), plasma concentration as afunction of time; Csl, steady-state plasma concentration; FC, the fraction of (6S)-folinic acid converted to (65)-5-CH,-THF; HPLC. high-performance liquid chro-matography; K,0, elimination rate constant; MRT, mean residence time; r, correlation coefficient; t„„half-life; Vd,volume of distribution in a one-compartmentmodel; V„.steady-state volume of distribution.

5755

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

PHARMACOKINETICS OF HIGH DOSE <6R. SJ-FOLINIC ACID

times: prior to the infusion; 10 min, 20 min, 30 min. 2 h. 4 h, 6 h, andeach 24 h after the start of the infusion; prior to the end of the infusion;10 min, 20 min, 30 min, 2 h, 4 h, and 6 h after the end of the infusion.Urine was collected from 47 to 71 h and from 95 to 119 h after thestart of the (6/f,5)-folinic acid infusion. In Patients 7-11 fewer plasmasamples and no urine samples were obtained. Samples were obtainedfor 5-fluorouracil determination prior to, and 5, 10, 20, and 30 minafter the first 5-fluorouracil dose. Pertinent samples were also obtainedwith informed consent from four patients who received 5-fluorouracilin an identical manner, but without (6/?,5)-folinic acid. Ultrafiltratesof plasma were obtained from an additional five patients who received(6/î,5)-folinicacid at 500 mg/nr/day to determine the percentages of(6/?)-folinic acid, (65)-folinic acid, and (65)-5-CH,-THF bound toplasma proteins. Approximately 1 ml of fresh plasma was centrifugedat 1500 x g for 30 min through a Centrifree micropartition device(Amicon). The volumes of the ultrafiltrates were all less than 10% ofthe sample volumes.

Ascorbic acid (5 mg/ml) was added to plasma and plasma ultrafiltrates obtained for folate determinations; aliquots were stored at —¿�70°C.

(65)-Folinic acid, (6tf)-folinic acid, and (65)-5-CH,-THF concentrations, as determined by the following analytical procedures, remainedconstant for at least 2 years under these conditions. Plasma samples inwhich 5-fluorouracil was to be assayed were stored at —¿�70°Cwithout

ascorbic acid.Urine was collected into brown bottles which contained sufficient

ascorbic acid for the concentration at the end of the collection to be atleast 1 mg/ml. These containers were kept on ice during the collectionperiod, after which the total volume was measured and aliquots werestored at -70°C.

Reagents. (65)-Folinic acid was generously provided by Dr. R.Moran, Childrens Hospital of Los Angeles. Tri-n-octylamine and 1,1,2-trichlorotrifluoroethane were purchased from Aldrich Chemical Company; (6A,5)-folinic acid and (6/?,5)-5-CHi-THF, from Sigma Chemical Company; L-(+)-ascorbic acid, from J. T. Baker, Inc.; and HPLCgrade ammonium phosphate, monobasic, from Fisher Scientific. Redistilled trifluoroacetic acid was a gift from Dr. B. Clark, City of Hope.

Analytical Procedures. The procedures for deproteinization of theplasma samples and determination of biologically active folates by thegrowth of Lactobacillus casei and Pediococcus cerevisiae in folate-deficient growth medium have been previously described (19). (65)-Folinic acid was determined in the heat-deproteinized plasma directlyusing P. cerevisiae. Recovery of (65)-folinic acid from plasma whichhad been spiked in vitro was 99.5 ±4.2% (SE, n = 6).

Samples of heat-deproteinized plasma were chromatographed on a4.6- x 250-mm Vydac 201HS 5-fim C,8 column (The SeparationsGroup) at a flow rate of 1.2 ml/min produced by a Spectra-Physics8700 solvent delivery system. After a 6-min isocratic elution with 7.5ITIMammonium phosphate buffer, adjusted to pH 2.5 using trifluoroacetic acid, a linear gradient was run over 20 min into 7.5 miviammoniumphosphate with 25% acetonitrile buffer, adjusted to pH 2.9 with trifluoroacetic acid, and elution was continued with the latter buffer for 3min. Between runs the column was requilibrated with the initial bufferfor 7 min. The absorbance of the eluate at 280 and 290 nm wasmeasured with a Beckman 165 detector. The following retention timeswere typical with this system: 5-CH,-THF, 17.3 min; 5,10-methenyI-tetrahydrofolic acid, 19.1 min; 10-formyltetrahydrofolicacid, 19.6 min;folinic acid, 20.9 min; dihydrofolic acid, 21.6 min; folie acid, 21.8 min.The chromatography ensured that 5-CH3-THF was well resolved fromother folates which would support the growth of L. casei. Resolutionof these potential contaminants from folinic acid was not a problembecause of the high concentrations of (6/?)-folinic acid in the plasmasamples. To adjust for the day-to-day variations in conditions, standardsof 5-CH.i-THF and folinic acid were run with each set of chromato-

grams.The total of (6.K)-folinic acid and (65)-folinic acid, which eluted as

a single peak, was determined by integration of the HPLC UV chro-matogram. The area obtained was converted to molarity by comparingit with the area of the external folinic acid standard. The concentrationof folinic acid in the standard was determined spectrophotometricallyusing ta«?= 37,200 at pH 7 (20). Recovery of folinic acid from plasma

which had been spiked in vitro was 93.6 ±0.95% (SE, n = 6). (6Ä)-Folinic acid was calculated as the difference between the total (byintegration of the UV peak) and the (65)-folinic acid (by P. cerevisiae).

Unknown components of plasma interfered with the quantitation of5-CH.vTHF by integration of the HPLC UV chromatogram. Therefore,(65)-5-CHi-THF was quantitated by L. casei bioassay of the appropri

ate HPLC fractions. This procedure also eliminated the possibility that(6/?)-5-CH,-THF could be mistaken for (65)-5-CH.,-THF. Recovery of(65)-5-CH,-THF from plasma which had been spiked in vitro was 93.6±3.4% (SE, n = 6) by bioassay.

In addition to (65)-folinic acid and (65)-5-CH,-THF, there was atleast one other compound in plasma which supported the growth of L.casei. This material, which was always less than 10% of (65)-5-CH.,-THF, eluted shortly before folinic acid. It could be distinguished fromthe latter compound by its inability to support the growth of P.cerevisiae. Although 10-formyltetrahydrofolic acid eluted in this position, it too should support the growth of P. cerevisiae (21). Folinic acidcan dehydrate to 5,10-methenyltetrahydrofolic acid at the acidic pH ofthe elution buffer (22), but this did not occur during chromatographyof pure (6/?,5)-folinic acid or during processing and chromatographyof (6/?,5)-folinic acid-spiked plasma. Thus, this relatively minor bioac-tive component of plasma was not identified.

For the determination of 5-fluorouracil, thawed aliquots of plasmawere deproteinized by the addition of an equal volume of ice-cold 0.6M trichloroacetic acid. Trichloroacetic acid was extracted from thesoluble fraction with tri-n-octylamine in 1,1.2-trichlorotrifluoroethane(23). 5-Fluorouracil was quantitated in the deproteinized plasma by thepeak area at 280 nm during isocratic HPLC using the system describedabove. The retention time of 5-fluorouracil was 4.8 min when elutedwith 10 HIM ammonium phosphate buffer, adjusted to pH 2.5 withtrifluoroacetic acid, at a flow rate of 1 ml/min. After each sample thecolumn was washed briefly with the same buffer containing 25% acetonitrile. Following initial determination of the 5-fluorouracil concentrations by comparison to an aqueous standard, a standard curve of aleast four drug concentrations was prepared in the patient's pretreat

ment plasma. These standards were analyzed in the same manner asthe unknown samples. Final determination of the 5-fluorouracil concentration in each unknown was calculated from the linear regressionof the standard curve.

Urine samples were sterilized by heat treatment identical to that usedto deproteinize plasma. The concentrations of (65)-folinic acid in thesterilized samples were then determined using the growth of P. cerevisiae (19). (65)-5-CH,-THF and the total of (6Ä)and (65)-folinic acidwere determined by HPLC as previously described (7). (6/?)-Folinicacid was calculated as the difference between the total (by integrationof the UV peak) and the (65)-folinic acid (by P. cerevisiae).

Pharmacokinetics. Pharmacokinetic analyses based on linear com-partmental models were performed using PCNONLIN (Statistical Consultants). All of the data points from both the approach to steady-stateduring the constant infusion and the decay after the end of the infusionwere used in the parameter estimations. The standard models providedwith the program were used to fit the plasma concentration versus timedata for (65)- and (6A)-folinic acid (24). This could not be done for(65)-5-CH,-THF because the dose could not be directly determined.The standard models were modified to treat FC as an additionalvariable. The parameter FC was constrained to be less than or equal tothe fraction of (65)-folinic acid not accounted for as (65)-folinic acidin the urine and more than or equal to the fraction of (65)-folinic acidwhich appeared as (65)-5-CH3-THF in the urine. These constraintslimited analysis to the six patients for whom urinary data were available.FC was assumed to be constant. Using Cr(t) and K,„determined by themodel for (65)-folinic acid, the input function for (65)-5-CH,-THF wasFC x [Kto of (65)-folinic acid] x [Cp(t) of (65)-folinic acid). It was alsoassumed that the central compartment for (65)-5-CH,-THF was thesame as the single compartment for (65)-folinic acid (see "Results")

and that all conversion took place within that compartment.The noncompartmental CLP, MRT, and Vs,values were derived from

AUC, AUMC, and the dose (25). These parameters were corrected forinput by continuous infusion (26, 27). CL, was calculated during a 24-h period at steady-state using the following equation: CL, = amount

5756

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

PHARMACOKINETICS OF HIGH-DOSE (6R. S1-FOL1NIC ACID

excreted/(plasma concentration x time). Clearances were normalizedbased on each patient's body surface area.

RESULTS

The plasma concentrations of (65)-folinic acid, (6/?)-folinicacid, and (oSJ-S-CH^-THF for all patients are plotted as afunction of time in Fig. 1. The combined data from the sixpatients with more extensive data sets were used to developlinear models for (65)- and (6/?)-folinic acid. During and 6 hfollowing the constant infusion, Cp(t) data for these two compounds each fit a single-compartment model with zero-orderinput and first-order decay (curves in Fig. 1). The data forindividual patients were then fit to this model. The pharmaco-kinetic parameters for (65")-folinic acid are listed in Table 1 and

the parameters for (6A)-folinic acid in Table 2. The results forone patient (Patient 8) differ substantially from the results ofthe remaining 10 patients. The high plasma concentrations of(65)-folinic acid for this patient led to a suspiciously low Vd.These plasma concentrations, obtained by direct bioassay ofplasma with P. cerevisiae, were confirmed by bioassay of theappropriate HPLC fractions with both P. cerevisiae and L.casei.

The noncompartmental CLP of (oS^-folinic acid and (6R)-folinic acid, given in Table 3, correlate well with the linearmodel-derived CLP (r = 0.963 and 0.997, P < 0.0001 in eachcase). The other noncompartmental parameters are dominated

0.01100.00

3! °-1(H

0.01100.00

0.10-

^i ! * i '. •¿�»»tí i î ' ì—L

(6S)-TOLINICACID

(SR)-FOLINICACID

-H 1 I- H 1 1 H

(6S)-5-CH3-THF

12 24 36 48 60 72 84TIME (HOURS)

96 108 120 132 144

Fig. 1. Plasma concentrations of (65)-folinic acid, (6/?)-folinic acid, and (651)-5-CH,-THF during and after a constant infusion of 500 mg/m2/day of (6R.S)-folinic acid for 5.5 days (132 h). Data points (•)represent individual plasmasamples. The curves represent the predicted values from the pharmacokineticmodels described in the text.

Table 1 (6S)-Folinic acid pharmacokinetic parameters

The plasma concentration versus time data for each patient were fit to a linearone-compartment model with constant i.v. input and first order output ("Model2," Ref. 24).

Pt°1234567891011MedianMeanSEC„(MM)1.863.252.203.462.401.883.457.953.562.913.633.253.320.51In(min)40.645.438.153.042.019.850.319.180.045.555.945.444.55.1CL,(ml/min/m2)1981131661061531951064610312510111312814CL,(ml/min/m2)45297127525748477CL„(ml/min/m2)1538495791011379810812r*(liter/m2)11.547.379.178.099.265.567.701.2611.888.248.158.158.020.87

' Patient number. Other abbreviations are the same as used in the text.

Table 2 (6R)-Folinic acid pharmacokinetic parameters

The plasma concentration versus time data for each patient were fit to a linearone-compartment model with constant i.v. input and first order output ("Model2," Ref. 24).

Pt°1234567891011MedianMeanSEC„23.052.130.177.647.137.538.223.853.930.145.838.241.74.8ty,(min)27052127759056426638816161319339638838550CL,(ml/min/m2)16.07.012.14.77.89.89.615.36.712.18.09.69.91.1CL,(ml/min/m2)12.95.810.93.87.78.88.28.31.4CL„(ml/min/m2)3.11.31.20.90.11.01.11.30.4Yé(liter/m2)6.215.284.854.026.323.745.373.555.963.384.564.854.840.32

1Patient number. Other abbreviations are the same as used in the text.

Table 3 Noncompartmental pharmacokinetic parametersThese parameters were derived from the AUC, the AUMC corrected for the

length of the infusion (26, 27), without reference to the number of pharmacokinetic compartments (25).

Pi"1234567891011MedianMeanSE(6S>FolinicacidCL,(ml/min/m2)17899158102108171102501021169510211611(6/?)-Folinic

acidCL,(ml/min/m2)15.97.612.25.27.910.310.315.97.412.18.610.310.31.0MRT(min)3946345537496704565057855127248950548657ya(liter/m2)6.274.816.733.885.264.705.201.234.083.304.204.704.510.45

" Patient number. Other abbreviations are the same as used in the text.

by the corrections for the prolonged infusion. For (6S)-folinicacid, which has an extremely short f./, value relative to theduration of the infusion, these corrections yielded net resultswith large standard errors and in three patients negative valuesfor MRT and Vss(data not shown). For (6/i)-folinic acid, whichhas a longer /./,, there were better correlations between thenoncompartmental MRT and the model-derived r./,(r = 0.842,

5757

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

PHARMACOKINETICS OF HIGH-DOSE (6Ä..V)-FOLINIC ACID

P = 0.001), and between the noncompartmental K, and themodel-derived Vu(r = 0.623, P = 0.04). The MRT and K„datafor (6A)-folinic acid are included in Table 3.

Neither the dose of (6S>5-CH,-THF nor the function describing its input are known. Therefore, as described in the"Materials and Methods" section, the fraction of (oS')-folinic

acid converted to (6S)-5-CH.,-THF was treated as an additionalparameter to obtain a model for (65)-5-CH,-THF. A two-compartment model was required to obtain a good fit of the(6S>5-CH,-THF Cp(t) data (curve in Fig. 1). A one-compartment model seriously underestimated the plasma concentrations of (oS^-S-CHi-THF at early time points (data not shown).The pharmacokinetic parameters derived from the two-compartment model for each of the patients are shown in Table 4.Noncompartmental methods could not be applied to the datafor (6S)-5-CH3-THF because the dose rate was unknown.

For each patient, the C, of each compound reported in Tables1, 2, and 4 is a single model-derived estimate. Intrapatientvariation in the C,, was estimated in six of the patients bycomparing the daily samples from Days 2 to 5, time points atwhich the theoretical plasma concentrations were nearly constant. With one exception, the coefficients of variation (SD/mean) for (6S)-folinic acid, (6/f)-folinic acid, and (6S)-5-CH,-THF were all <15% in each patient. In one patient the coefficient of variation of (6/?)-folinic acid was 29%. This largevariation was caused by a single outlying value; the remainingthree data points were bunched closely around the model-derived estimate.

The poor renal clearance of (6/?)-folinic acid prompted us todetermine the protein binding of (6/?)-folinic acid, (6S>folinicacid, and (65)-5-CH,-THF in plasma. To minimize in vitroartifacts, ultrafiltration was done immediately on fresh plasma,prior to adding ascorbic acid or freezing the plasma. The resultsare presented in Table 5. The median binding percentages werethen used to calculate the CLr of the free compounds for thesix patients from whom urine had been collected. These clearances are presented in Table 6 relative to each patient's concur

rent creatinine clearance.The pharmacokinetic parameters for 5-fluorouracil derived

Table 4(6S)-5-CH,- THF pharmacokinetic parameters

The plasma concentration versus time data for each patient were fit to thefollowing system of equations:

</[(6S')-5-CH,-THF],/<ft = K x FC x [(6S)-folinic acid] - (K,0 + K,2)

x |(6S)-5-CH,-THFl, + K2, x [(6S)-5-CH3-THF]2 (A)

</«6S)-5-CH3-THF]2/</r= Kn

X [(6S)-5-CH,-THF], - AT21x [(6S)-5-CH3-THF]2 (B)

where: [(6S)-5-CH3-THF], is the concentration in compartment 1; ((6S)-5-CH3-THF]2 is the concentration in compartment 2; Kn. is the rate constant for the

transfer of (6S)-5-CH.,-THF from compartment x to compartment y, K is K,0 of(6£)-folinicacid; FC is the fraction of (óí'J-folinicacid converted to (65)-5-

CHj-THF (a variable); and [(65)-folinic acid] is the concentration of (6S)-folinicacid at time t as determined by the parameters given in Table 1 for each patient.

Table 5 Protein binding affolâtesin plasma

Pf123456MédianMeanSEca(jlM)3.127.453.987.408.784.005.705.780.96tv,(min)4753851107622417318446554118CL,(ml/min/m2)88.157.749.832.226.751.350.651.08.9CL,(ml/min/m2)41.319.739.520.225.843.032.731.64.4CL,,(ml/ya

min/m2)(liter/m2)46.838.010.312.00.98.311.219.47.529.916.253.219.615.017.818.725.36.0FC75755465645664.5654

Percentfree"Compound(65)-folinic

acid(6/?)-folinicacid(6.S>5-CH,-THFMedian72.48.749.2Min53.12.329.5Max96.434.882.9A'"14141

1Mean73.310.851.0SE3.32.14.5°Free solute was separated from protein-bound solute by ultrafiltration as

described in the "Materials and Methods" section.

One or more plasma samples were drawn from each of five patients. A/is thetotal number of samples analyzed for a given compound.

Table 6 Relative renal clearance of free fötalesThe C/.r-free for each compound at steady-state wras the amount excreted in

24 h/(plasma concentration x median percentage free x 24 h x body surfacearea). The results tabulated are the ratios of these CL,-free to each patient's

concurrent creatinine clearance.

CL,-Cree/CL„

CL„" (65)-Folinic (6Ä)-FolinicPt" (ml/min/m2) acid acid (6S)-5-CH.,-THF

123456MedianMeanSE571.10460.87462.1332.1543.6751.5446.3546

.4130.192.62

1.491.440.872.73

.751.36.282.06.221.98.712.02.382.031.390.23

0.14" Patient number.* Creatinine clearance normalized to the patient's body surface area. Other

abbreviations are the same as used in the text.

Table 7 5-Fluorouracil pharmacokinetic parameters

The plasma concentration versus time data for each patient were fit to a linearone-compartment model with bolus i.v. input and first order output ("Model I,"

Ref. 24).

With folinicacidPt°12345611MedianMeanSE(M(min)11.113.85.86.17.86.26.46.48.21.2CL,(ml/min/m2)145382115221010191775582010101185169Vt(liter/m2)23.216.412.89.721.66.77.512.814.02.5Pt"FIF2F3F4Withoutfolinicacidlu(min)9.814.07.28.59.19.91.5CLP(ml/min/nr)51765556354955657130Vt(liter/m2)7.313.25.96.77.08.31.7

' Patient number. Other abbreviations are the same as used in the text.

°Patient number; patients designated FI, F2, etc.. did not receive (6R,S)-

folinic acid. Other abbreviations are the same as used in the text.

from a linear one-compartment model are given in Table 7. Inaddition to data from seven patients receiving 5-fluorouracilcombined with (6/?,£)-folinic acid, data is presented for fourpatients receiving 5-fluorouracil alone.

DISCUSSION

Pharmacokinetic modeling is useful in that it allows investigators to predict results at different doses and schedules. Alinear, one-compartment model fit the data for (6S)-folinic acidwell except for the last one or two data points (Fig. 1). A two-compartment model improved the fit after the infusion, but atthe expense of a good fit to the early data points (curve notshown). Another consideration in choosing between the twomodels was the specificity of the microbiological assay for (6S)-folinic acid. Although growth of P. cerevisiae is highly selectivefor (6S)-folinic acid, there may be other folates which wouldhave contributed insignificantly to the growth of P. cerevisiaeduring the infusion, but could have interfered with the microbiological assay when (6S)-folinic acid fell rapidly after the end

5758

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

PHARMACOKINETICS OF HIGH-DOSE (6Ä.S)-FOLINIC ACID

of the infusion. Forcing the model to fit the last data point,where the greatest potential for error existed, did not seemprudent. An a phase was not observed, but a compartment witha short t., would have been obscured by a lack of samples before10 min and the extremely long infusion time (28). Therefore,the one-compartment model was used. The mean CLr of (65)-folinic acid in this study (128 ±14 ml/min/m2, Table 1) differedat P < 0.01 from the mean CLr previously reported for low-dose bolus i.v. (6/?,5)-folinic acid (217 ±8 ml/min/m2, Ref. 7)

and this difference was accounted for entirely by a difference inCL,,r. On the face of it, this difference implied that the linearmodel was incorrect, i.e., that there was a dose-dependentchange in the metabolism of (65)-folinic acid. However, (65)-

folinic acid dropped below the limit of detection quite soon inthe low-dose study and there were insufficient early samples todetect an «phase (7). Incorrect extrapolation before and afterthe actual data points could have led to underestimates of theAUC and CL,,. With the 5.5-day infusion, the contribution ofthese extrapolated areas were insignificant. The pharmacoki-netic parameters derived from the current study simulated theCr observed in the earlier study quite well.4 Therefore, the

results reported here are more likely to be correct.The pharmacokinetic parameters for (6/?)-folinic acid (Table

2) did not differ significantly from the previously reportedvalues (7). Essentially all of the clearance was renal, but therewas sufficient scatter in the data that a minor degree of metabolism of the "biologically inactive" diastereoisomer could not

be ruled out. The results pertaining to the mechanism of urinaryexcretion (Table 6) were at odds with the previous conclusionthat (6/ï)-folinic acid was actively reabsorbed (7). The root ofthis discrepancy was the value used for the percentage of (6R)-folinic acid bound to plasma proteins. In the previous study(6/î)-folinicacid and (65)-folinic acid were found to have equivalent binding (mean ±SE, 54 ±6% free, Ref. 7), whereas, inthe current study (6/?)-folinic acid was bound to a significantlygreater extent (mean ±SE, 11 ±2% free, Table 5). We believethat the earlier finding is likely to be in error because ascorbicacid was added to the plasma before filtration. Although thedata in Table 6 indicated that (6/?)-folinic acid was activelyexcreted, this data was obtained using the median percentageof (6/?)-folinic acid bound in a separate group of samples.Because there was a wide range in the percentage of free drugin the samples analyzed (2.3 to 34.8%, Table 5), the excretionof (6/?)-folinic acid requires further investigation. Nonetheless,it can be concluded that active reabsorption was unlikely.

Analysis and interpretation of the (65)-5-CH,-THF data werefar less straight forward. The essential problems were thoseencountered with the analysis of any metabolite. The plasmaconcentrations and the amount of (6S)-5-CH,-THF excreted inthe urine were known, which allowed calculation of the CLr (32±4 ml/min/m2) directly. However, the amount of (65)-5-CH3-

THF formed was not known; the site of formation was notknown; and the time course of its formation was not known.Therefore, the pharmacokinetic model used for (65)-5-CH3-THF differed from standard models in that the fraction of (65)-folinic acid converted to (65)-5-CH.,-THF was included as a

variable. Several assumptions underlie the inclusion of thisvariable. Although a detailed discussion of the model is beyondthe scope of this report, the most important assumptions werethat all conversion took place within the central compartmentand that, for a given patient, the fraction converted remainedconstant during and after the infusion. The utility of a phar-

4J. A. Straw, unpublished data.

macokinetic model lies not in how closely it can fit data at onedose and schedule, but in its ability to predict the outcomeunder different conditions. The model parameters derived fromthe pooled data of the six patients were used to simulate theplasma concentrations following administration of an i.v. bolusof 28 mg/m2 of (6/?,5)-folinic acid. The simulated and observeddata were in good agreement.4 The mean terminal /.,, of (65)-5-CH,-THF (554 ±118 min) was skewed by a very long t,, in onepatient. Even the median t<,,(446 min) was twice as long as firstreported (7), but more in line with subsequent reports [mean,362 min (15); mean, 420 min (17)]. This long f., has importantclinical implications in that plasma levels of (65)-5-CH3-THFmay be better maintained between repeated intermittent dosesof (6/ï,5)-folinic acid than would have been predicted originally.

The rationale for the clinical protocol which formed the basisfor this pharmacokinetic study was derived from in vitro evidence that (6A,5)-folinic acid alters the pharmacodynamics of5-fluorouraciI. Specifically, (6/i,5)-folinic acid caused the sameconcentration of 5-fluorouracil to have a greater cytotoxic effect. In vivo, (6/?,5)-folinic acid might also alter the pharma-cokinetics of 5-fluorouracil, which would complicate interpretation of the clinical results. Although the pharmacokinetics of5-fluorouracil are not linear, plasma concentrations followingbolus i.v. administration are often well described by a linearone-compartment model (29), which was the case in our study.The mean value for the f.,,of 5-fluorouracil in patients receiving(6/?,5)-folinic acid (8.2 ±1.2 min, Table 7) was not significantlydifferent from the t,,. in patients receiving 5-fluorouracil alone(9.9 ±1.5 min). It was also comparable to the f./, reported forbolus 5-fluorouracil alone in several studies by other investigators (see Ref. 29). Although the CLP and the V,,appeared to besomewhat greater in the group receiving folinic acid than in thesmall 5-fluorouracil-alone group included in this study, thevalues for both parameters were within the range of valuesreported for 5-fluorouracil alone by other investigators (see Ref.29). Thus, concurrent administration of (6/i,5)-folinic acid didnot appear to alter the plasma pharmacokinetics of 5-fluorouracil.

From the viewpoint of determining whether or not the(6#,5)-folinic acid as administered in these protocols had thepotential to augment the antineoplastic activity of 5-fluorouracil, the most important results were the C„of (65)-folinic acidand (65)-5-CH,-THF. The C„of (65)-folinic acid exceeded 1.9UM in all patients (Table 1). The median was 3.25 ¿¿M.TheseC„values were greater than the concentrations required forsynergism with 5-fluorouracil in many in vitro studies (1-4).The Cssof the bioactive metabolite (65)-5-CH,-THF were even

higher, ranging from 3.1 to 8.8 ^M, with a median of 5.7 /¿M(Table 4). These concentrations were within the range shownby others to be synergistic with 5-fluorouracil in CCRF-CEMcells (30) and close to the concentration reported to producemaximum synergy in Sarcoma 180 and Hep-2 cells (10 pM,Ref. 2). If a patient's neoplastic cells were capable of using

either or both of these folates, then they might have exhibitedincreased sensitivity to 5-fluorouracil. Although no cause-and-effect relationship was established, our phase III colorectalcancer study has demonstrated a significantly higher responserate with the combination of (6/?,5)-folinic acid and 5-fluorouracil than with 5-fluorouracil alone (11). Likewise, we observed 10 responses to (6/î,5)-folinicacid and 5-fluorouracil inthe phase II breast cancer study in 60 patients whose diseasehad objectively progressed on previous chemotherapy with a 5-fluorouracil-containing regimen (12). Therefore, the objectiveof providing continuous exposure to concentrations of (65)-

5759

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

PHARMACOKINETICS OF HIGH-DOSE (6Ä..V)-FOLIMC ACID

folinic acid and (oS'J-S-CHj-THF known to be synergistic with

5-fluorouracil in vitro was clearly met. The clinical responsedata suggest that synergism also occurred in vivo.

REFERENCES

1. l 'liman. B., Lee. M., Martin, D. W., Jr., and Santi, D. V. Cytotoxicity of 5-fluoro-2'-deoxyuridine: requirement for reduced folate cofactors and antagonism by methotrexate. Proc. Nati. Acad. Sci. USA, 75: 980-983, 1978.

2. Evans, R. M., Laskin. J. D., and Hakala, M. T. Effect of excess folates anddeoxyinosine on the activity and site of action of 5-fluorouracil. Cancer Res.,41: 3288-3295, 1981.

3. Waxman, S., and Bruckner, H. The enhancement of 5-fluorouracil anti-metabolic activity by leucovorin, menadione and «-tocopherol. Eur. J. Clin.Oncol., IS: 685-692, 1982.

4. Keyomarsi, K., and Moran, R. G. Folinic acid augmentation of the effects offluoropyrimidines on murine and human leukemic cells. Cancer Res., 46:5229-5235, 1986.

5. Lockshin, A., and Danenberg, P. V. Biochemical factors affecting the tightness of 5-fluorodeoxyuridylate binding to human thymidylate synthetase.Biochem. Pharmacol., 30: 247-257, 1981.

6. Keyomarsi, K., and Moran, R. G. Mechanism of the cytotoxic synergism offluoropyrimidines and folinic acid in mouse leukemic cells. J. Biol. Chem.,263: 14402-14409, 1988.

7. Straw, J. A., Szapary. D., and Wynn, W. T. Pharmacokinetics of thediastereoisomers of leucovorin after intravenous and oral administration tonormal subjects. Cancer Res., 44: 3114-3119, 1984.

8. Sirotnak, F. M., Chello, P. L., Moccio, D. M., Kisliuk, R. L.. Combepine,G., Gaumont, Y., and Montgomery, J. A. Stereospecificity at carbon 6 offormyltetrahydrofolate as a competitive inhibitor of transport and cytotox-icity of methotrexate in vitro. Biochem. Pharmacol., 28: 2993-2997, 1979.

9. Sato, J. K., and Moran, R. G. Interaction of methotrexate (MTX) andcitovorum factor (CF) at folyl polyglutamate synthetase (FPGS). Proc. Am.Assoc. Cancer Res., 25: 312, 1984.

10. Houghton, J. A., Schmidt. C., and Houghton, P. J. The effect of derivativesof folie acid on the fluorodeoxyuridylate-thymidylate synthetase covalentcomplex in human colon xenografts. Eur. J. Cancer Clin. Oncol., IS: 347-354, 1982.

11. Doroshow, J. H., Bertrand, M., Newman, E., Multhauf, P., Leong, L.,Blayney, D., Goldberg, D., Margolin, K., Carr, B., Akman, S.. and Metter,G. Preliminary analysis of a randomized comparison of 5-fluorouracil versus5-fluorouracil and high-dose continuous-infusion folinic acid in disseminatedcolorectal cancer. Nati. Cancer Inst. Monogr., (Issue 5): 171-174, 1987.

12. Doroshow, J. H., Leong, L., Margolin, K., Flanagan. B., Goldberg, D.,Bertrand, M., Akman, S., Carr, B., Odujinrin, O., Newman, E., and Litch-field, T. Refractory metastatic breast cancer: salvage therapy with 5-fluorouracil and high-dose, continuous infusion folinic acid. J. Clin. Oncol., 7:439-444, 1989.

13. Yin, M. B., Zakrzewski, S. F., and Hakala, M. T. Relationship of cellular

folate cofactor pools to the activity of 5-fluorouracil. Mol. Pharmacol., 23:190-197, 1982.

14. Houghton, J. A., Williams, L. G., Raddparvar, S., de Graff, S., and Hough-ton, P. J. Effect of leucovorin (LV) on pools of methylenetetrahydrofolate(CH2-H4PteGlu) and tetrahydrofolate (H4PteGlu) in xenografts of humancolon adenocarcinoma (HCA). Proc. Am. Assoc. Cancer Res., 29:473, 1988.

15. Machover, D., Goldschmidt, E., Chollet, P., Metzger, G., Zittoun, J., Mar-quet, J., Vandenbulcke, J. M., Misset, J. L., Schwarzenberg. L., Fourtillan,J. B., Gaget, H., and Mathe, G. Treatment of advanced colorectal and gastricadenocarcinomas with 5-fluorouracil and high-dose folinic acid. J. Clin.Oncol., 4: 685-696, 1986.

16. Trave, F., Frank, C., Petrelli, N., Herrera, L., Mittelman, A., and Rustum,Y. Pharmacokinetics of folates in patients with colorectal carcinoma. Proc.Am. Assoc. Cancer Res. 27: 171, 1986.

17. Arbuck, S. G., Douglass, H. O., Jr., Trave, F., Milliron, S., Baroni, M.,Nava, H., Emrich, L. J., and Rustum, Y. M. A phase II trial of 5-fluorouraciland high-dose intravenous leucovorin in gastric carcinoma. J. Clin. Oncol.,5: 1150-1156, 1987.

18. Straw, J. A., Doroshow, J. H., and Newman, E. M. Pharmacokinetics ofleucovorin (D,L-5-formyltetrahydrofolate) after intravenous injection andconstant intravenous infusion. Nati. Cancer Inst. Monogr., (Issue 5): 41-45,1987.

19. Newman, E. M.. and Tsai, J. F. Microbiological analysis of 5-formyltetra-hydrofolic acid and other folates using an automatic 96-well plate reader.Anal. Biochem., 154: 509-515, 1986.

20. Blakley, R. L. The Biochemistry of Folie Acid and Related Pteridincs. p. 94,New York: Elsevier. 1969.

21. Baugh. C. M., and Krumdieck. C. L. Naturally occurring folates. Ann. NYAcad. Sci., 186:1-2». 1971.

22. Rabinowitz, J. C. Folie acid. In: P. D. Boyer. H. Lardy, and K. Myrbäck(eds.). The Enzymes, Vol. 2. Ed. 2. pp. 185-252. New York: Academic Press,1960.

23. Khym, J. X. An analytical system for rapid separation of tissue nucleotidesat low pressures on conventional aniónexchangers. Clin. Chem., 21: 1245-1252, 1975.

24. Metzler, C. M., and Weiner, D. L. PCNONLIN user's guide. Lexington,

KY: Statistical Consultants, Inc., 1985.25. Benêt.L. Z., and Galeazzi, R. L. Noncompartmcntal determination of the

steady-state volume of distribution. J. Pharm. Sci.. 68: 1071-1074, 1979.26. Lee, C. S., Brater, D. C., Gambertoglio, J. G., and Benet, L. Z. Disposition

kinetics of ethambutol in man. J. Pharmacokinet. Biopharm., 8: 335-346.1980.

27. Straughn, A. B. Model-independent steady-state volume of distribution. J.Pharm. Sci., 71: 597-598, 1982.

28. Loo, J. C. K., and Ricgelman, S. Assessment of pharmacokinetic constantsfrom postinfusion blood curves obtained after i.v. infusion. J. Pharm. Sci.,59:53-55, 1970.

29. Collins. J. M., Dedrick. R. L., King, F. G., Speyer, J. L., and Myers, C. E.Nonlinear pharmacokinetic models for 5-fluorouracil in man: intravenousand intraperitoneal routes. Clin. Pharmacol. Ther., 28: 235-246, 1980.

30. Mini, E., Mazzei, T., Coronnello, M., Criscuoli, L., Gualtieri, M., Periti, P.,and Berlino, J. R. Effects of 5-methyltetrahydrofolate on the activity offluoropyrimidines against human leukemia (CCRF-CEM) cells. Biochem.Pharmacol., 36: 2905-2911, 1987.

5760

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

1989;49:5755-5760. Cancer Res   Edward M. Newman, James A. Straw and James H. Doroshow  Intravenous Infusion(Leucovorin) in Humans during Constant High-Dose

)-Folinic AcidR,SPharmacokinetics of Diastereoisomers of (6

  Updated version

  http://cancerres.aacrjournals.org/content/49/20/5755

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/49/20/5755To request permission to re-use all or part of this article, use this link

Research. on November 21, 2018. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from