`Research
`
`Cellular but not plasma pharmacokinetics of lometrexol
`correlate with the occurrence of cumulative hematological
`toxicity.
`
`T W Synold, E M Newman, M Carroll, et al.
`
`Clin Cancer Res 1998;4:2349-2355. Published online October 1, 1998.
`
`Updated Version
`
`Access the most recent version of this article at:
`http://c|incancerres.aacrjourna|s.orgIcontent/4/10/2349
`
`Citing Articles
`
`This article has been cited by 1 Highwire-hosted articles. Access the articles at:
`http://c|incancerres.aacrjournals.orglcontent/4/10/2349#re|ated-urls
`
`Email alerts
`
`Sign up to receive free email-alerts related to this article or journal.
`
`Reprints and
`subscriptions
`
`To order reprints of this article or to subscribe to the joumal. contact the AACR Publications
`Department at pubs@aacr.org.
`
`Permissions
`
`To request permission to re-use all or part of this article, contact the AACR Publications
`Department at permissions@aacr.org.
`
`Downloaded from clincancerres.aacrjouma|s.org on February 1, 2011
`Copyright © 1998 American Association for Cancer Research
`
`Lilly Ex. 2047
`Sandoz v. Lilly IPR2016-00318
`
`
`
`Vol. 4. 2349-2355. October I998
`
`Clinical Cancer Research 349
`
`Cellular but not Plasma Pharmacokinetics of Lometrexol Correlate
`with the Occurrence of Cumulative Hematological Toxicity’
`
`Timothy W. Synold,2 Edward M. Newman,
`Ilnuuu r‘a\——a\" E‘:-s-ann I‘ “nn1wt':nu
`ivaaly \.rflIlUll, a-uun.u iva. ruuggna,
`
`Susan Groshen, Kay Johnson, and
`James H. Doroshow
`Departments of Medical Oncology and Therapeutics Research
`['l'. W. 5.. M. C.. J. H. D.] and Pediatrics (E. M. N.]. City of Hope
`National Medical Center. Duarte. California 9l0l0: and Division of
`Medical Oncology. USC-Norris Comprehensive Cancer Center. Los
`Angeles. California 90033 IF. M. M.. S. G.. K. I.1
`
`ABSTRACT
`
`Lometrexol inhibits the first folate-dependent enzyme
`in de nova purine biosynthesk and is avidly polyglutamated
`and retained in tissues exprewing folylpolyglutantate syn-
`thetase. Although clinical studim have been limited by cu-
`mulative toxicity. preclinical studies show that pretreatment
`with folic acid can protect normal tissue while maintaining
`t-.-m-.-r cytv.-I.-.-deity. '!‘!-.ere!v.-re. a Ply.-se ! study of !omet.re.~:e!
`every 21 days preceded by l.v. folc acid was initiated. Lorne-
`trexol was studied in six patient at 15 mg/in‘, in six patients
`at 20 mg/m‘, in three patients at 25 mum’, and in nine
`patients at 30 mg/nr’. Patients received either 5 mg of folic
`acid I h before or 25 mg/m’ 3 h before Iometrexol. Blood
`samples were obtained around the first course and weeltiy
`thereafter for deurmination of plasma and erythrocyte
`(RBC)
`lometrexol concentrations. Bioactive folates
`in
`plasma and RBCs were determined in a subset of patients.
`Lometrexol pharmacoltlnetics were best described by a
`three-compartment model. Mean clearance and volume of
`dktribution were 1.6 1: 0.6 Iiterslhlm’ and 8.9 t 4.1 litery
`m’. Mean half-lives were 0.23 3: 0.1, 2.9 :1: 1.4, and 25.0 1
`48.7 h. Pharmacokineti were independent of either lorrie-
`trexol or folic acid dose. In the weekly blood samples, RBC
`lometrexol levels rose, long after plasma lometrexol was
`“""“"'bie. '-{BC lornietrexoi levels
`iruiependent of
`folic acid or lometrexol dose. Bioactive folates measured in
`plasma and RBCs during this same time period did not
`accumulate. Rising RBC levels were correlated with a fall in
`hematocrit, hemoglobin, and platelet count. This study in-
`dicates that the cumulative toxicity of lometrexol is related
`
`Received 4/M198: revised 7/9/98: accepted 7ll0l98.
`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 in accordance with IS U.S.C. Section I734 solely to
`indicate this fact.
`' This work was supported by NIH Grants U01 CA 62505 and 5P30 CA
`33572.
`’To whom requests for reprints should be addressed. at City of Hope
`National Medical Center. 1500 East Duane Road. Duane. CA 9l0l0.
`Phone: (626) 359-81 l I: Fax: (626) 301-8233.
`
`to tissue concentration and not plasma pharmacoltinetics.
`RIIC lometrexol may be an indicator of cumulative dm
`exposure and effect.
`
`INTRODUCTION
`Lometrexol [(6R)-5.I0-dideaza-5.6.7.8-tetrahydrofolic acid]
`is an antlfolate antimetabolite with a broad spectrum of cytotoxic
`activity against preclinical murine tumor model systems and human
`tumor xenografts (I. 2). Unlike classical antifolates, iometrexol acts
`by inhibiting glycinamidc ribomiclcotidc formyllransfcmsc.
`the
`first folate-dependent enzyme in the de navo purine biosynthetic
`pathway (1). in addition to its unique mechanism of action. lume-
`Irexo! is an excellent substrate for FPC-S’ and is avidly po!yg!I_II.a-
`mated (l. 3). It has been demonstrated previously that the polyglu-
`tamates of lometrexol are more potent inhibitors of glycinarnide
`ribonucleotide formyluansferase titan lometrexol itself (4. 5). Fur-
`thermore. in addition to having a higher affinity for their target
`enzyme. polygiutamated anabolites are retained intraccilularly for
`prolonged periods of time (3).
`Because of promising preclinical activity. clinical trials of
`lometrexol began almost 10 years ago; however. the utility of
`lometrexol has been limited by serious toxicities (thrombocyto-
`penia. leultopenia, and mucositis). which appear to be cumula-
`tive and long-lasting (6. 7). it has been speculated that this
`cumulative toxicity may be related to the formation and reten-
`tion of polyglutamatcd anabolites. resulting in longer than ex-
`pected drug exposures in certain target tissues. in a previous
`Phase I study of lometrexol (7), unpublished data indicated that
`RBC folate levels appeared to increase over time and remain
`elevated for monms. Because the RBC fa “res in that study were
`measured using a commercially available assay that cross-reacts
`with lometrexoi. it was hypothesized that lometrexol was accu-
`mulating in RBCs. Tissue levels of other antifolates. such as
`methotrexate in erythrocytes. have been used as a measure of
`long-term drug exposure (8).
`Dietary folic acid supplementation has been shown to
`modulate lomettexol toxicity without diminishing antitumor ac-
`tivity in preclinical models (9). As a result, recent clinical trials
`of Iomctrcxol have all included the use of either prior or con-
`current folic acid treatment. One such trial using folic acid
`s-.-pplerreerttation beginning ! week prior and Co!!!ir!'.!i!!g for 1
`week following lomeuexol demonstrated that much higher
`lometrexol doses could be given without cumulative toxicity
`(l0). However. due to concerns tluit overly aggressive folic acid
`supplementation might rescue tumor as well as nomtal tissues.
`the current study evaluating the administration of lometrexol
`every 3 weeks preceded by a single i.v. dose of folic acid was
`undertaken. Because of its delayed toxicity profile. prolonged
`
`‘The abbreviations used are: FPGS. folylpolyglutamate synthetase:
`AUC. area under the concentration rer.tu.\‘ time curve.
`
`Downloaded from clincancerres.aacrjouma|s.org on February 1, 2011
`Copyright © 1998 American Association for Cancer Research
`
`Lilly Ex. 2047
`Sandoz v. Lilly IPR2016-00318
`
`
`
`2350 Plasma and RBC Pharmacokinetics of Lometrexol
`
`phannacokinetic and pharmacological monitoring of lometrexol
`were important correlates of this clinical trial.
`
`MATERIALS AND METHODS
`Patients. Patients with a histologically documented di-
`agnosis of a solid tumor were eligible for this study if they met
`the following inclusion criteria: (a) age of 221 years. (b) life
`expectancy of at least 4 months and Kamofsky performance
`status of 260%. (c) no chemotherapy or radiation therapy for at
`least 3 weeks prior to study entry. (d) evidence of adequate renal
`and hepatic function. (2) evidence of adequate bone manow
`function, and (I) ability to give voluntary infonned consent.
`Patients were ineligible if they met any of the following exclu-
`sion criteria: (a) concurrent use of cytotoxic therapy or multi-
`vitamins containing folic acid. (b) presence of partial bowel
`obstruction or inflammatory bowel disease, (c) concurrent treat-
`ment with allopurinol. (d) presence of active infection.
`(2)
`presence of active angina pectoris or other uncontrolled heart
`disease. or (f) presence of other uncontrolled medical or psy-
`chiatric condition.
`Treatment Protocol. The starting dose level of lome-
`trexol was 30 mglm’ by rapid i.v. injection repeated every 3
`weeks. Patients originally received 5 mg of folic acid by rapid
`i.v. iniectior.
`! it before escalating doses of lometrexol. D.-e to
`the occurrence of cumulative toxicity at the first dose level,
`lometrexol dose levels were subsequently lowered to 20 and
`then 15 mglm’. With the continued incidence of cumulative
`toxicity.
`the study was amended to increase the folic acid
`pretreatment dose to 25 mg/m’ given 3 h before lometrexol. The
`amended folic acid dose and schedule were chosen based on
`
`previously published experience with plasma S-methyltetrahy-
`drofolate levels (I I). With the revised folic acid pretreatment,
`the starting lometrexol dose was 15 mglm’. The tiose was
`subsequently escalated in cohorts of three patients to levels of
`20. 25. and 30 mg/m2.
`Pharmacokinetic Sampling. Peripheral blood samples
`were obtained from all patients around the first course at the
`following times: immediately prior to treatment and 5 and 30
`min and I. 2. 4. 6. I2, 24. and 48 h after lometrexol. Blood was
`drawn weekly thereafter for determination of both plasma and
`REC lometrexol and liioactive folate coin‘.-eiitrations for as long
`as a patient remained on study. Samples for plasma analysis
`were centrifuged, and 1-ml aliquots were transferred to tubes
`containing 1 mg of ascorbic acid. For the weekly blood samples,
`0.5 ml of whole blood was removed prior to separation of
`plasma. diluted l:l0 with 0.1% ascorbic acid, and incubated for
`30 min at 37°C to lyse the RBCs and cleave both the lometrexol
`and endogenous
`folate polyglutamates to monoglutamates
`through the action of endogenous conjugase (12). All samples
`were stored frozen at -70°C or colder until analysis. Ascorbic
`acid was added to all whole blood and plasma samples to
`prevent ex vivo oxidation of endogenous foletes.
`Detennlnatlon of Plasma and RBC Lornetrexol.
`Lometrexol in plasma and whole blood was determined using a
`previously published high-performance liquid chromatography
`assay (l3). Briefly. following a solid-phase extraction step using
`cation-exchange cartridges. chromatographic separation was
`achieved isocraticaliy across a phenyl column. Using UV and
`
`electrochemical detectors in series for low and high sensitivity,
`respectively. lometrexol was detennined over the entire range of
`ciinicaiiy aciiievabie concentrations with a single injection. This
`method has a percentage recovery of >85% and a limit of
`quantitation of 20 nglml and is accurate and precise to within
`5.5%.
`Determimltion of Plasma and RBC Bloactlve Folates.
`Bioactive folates in plasma and whole blood were r_leterm__i_nez_l
`using a previously published microbiological assay (l4). The
`assay is based on the growth of Lactobacillus rhamuosus
`(ATCC 7469. NCIB 10463). formerly lacrobacillus casei. Pre-
`vious investigators have shown that longer chain folate poly-
`glutamates are inactive in the L casei microbiological assay
`([5). Therefore. whole blood samples were first incubated in the
`presence of endogenous conjugase to convert folate polygluta-
`mates to monoglulamates (12). Moreover. because lometrexol
`appeared to inhibit the growth of L. rhamnosus, even when
`hypoxanthine or inosine was added to the assay medium. an
`extraction merited was devised to separate bioactive foletes
`from lometrexol. Blood was processed for both plasma and
`whole blood analysis. as described previously for the determi-
`nation of lometrexol. For plasma, 0.3 ml of sample was diluted
`with 0.6 ml of 0. l % ascorbic acid and 0.3 ml of 0.5 M acetic acid
`
`was added. For whole blood lysate. 0.l ml of 0.5 M acetic acid
`was added to 1.0 ml of the lysate. The resulting pH was in the
`range of 4-5 for both plasma and whole blood lysate. Baker-
`bond C“, columns (VWR Scientific, Cenitos. CA) were pre-
`conditioned with 3 ml of methanol. followed by 3 ml of distilled
`water and 3 ml of 0.1% ascorbic acid. without allowing the
`colonies to d.'_-,'. The aeidif'-..-.d whole blfi lymte and plasrra
`samples were loaded onto preconditioned columns. and folates
`were eluted with l.0 ml of 0.1 % ascorbic acid and 10 ml of 7.5%
`acetonittile in 0.l% ascorbic acid. The sample flow-through.
`ascorbate wash. and 10 ml of 7.5% acetonitrile in 0.l% ascorbic
`acid were pooled together. The pooled eluate for each sample
`was analyzed for total bioactive folates. The solid-phase extrac-
`tion method developed for use in this study resulted in recovery
`of 8| 1' 11% of 5-methyltetrahydrofolate from spiked samples.
`Pharmacoltinetic Modeling. Plasma lometrexol data
`were fit with either a two- or three-compartmental pharmacoki-
`Fr:--nu. I I ‘\
`netic model using Al'.>AP'l‘ ll scum... run. The corr.pa.'tmental
`model that best described lometrexol disposition was identified
`and used to determine the primary (e.g.. V,_.. K,0, K.2. K2,, K”.
`and K3,) pharmacokinetic parameters. Secondary pharmacolo-
`netic parameters (Cl,,,,.,,,_, V,,. half-lives. and AUC) were de-
`rived from the model-generated primary parameters.
`Statistical Methods. The Altaike lnforrnation Criterion
`
`(17) was used to identify the compartmental pharmacokinetic
`model that best described the disposition of lometrexol. Distri-
`bution of the pharmacokinetic parameters were expressed as
`means 1' SD. For analysis of the plasma pharmacokinetic data.
`the natural !oga.ri'_b.vn transformation was used to correct for
`heteroscedasticity. After transfonnation. a general linear regres-
`sion model was used to evaluate the association between each
`pharmacoltinetic parameter and either the lometrexol or folic
`acid dose. Significance testing was done using an F test based
`on the regression model. with P < 0.05 being considered
`indicative of a significant difference.
`
`Downloaded from clincanoerres.aacrjouma|s.org on February 1. 2011
`Copyright © 1998 American Association for Cancer Research
`
`Lilly Ex. 2047
`Sandoz v. Lilly IPR2016-00318
`
`
`
`Clinical Cancer Research 235!
`
`E
`
`§3
`
`
`
`
`
`plumrilometrexol(ug/l)
`
`time (hours)
`
`24
`
`a
`lometrexel dose and AUC ca!-:-_-lat
`‘or-_<hip
`Fig. 2 R.-.
`dividing the dose by the model-derived systemic clearance. Li
`generated by linear regression analysis.
`
`lieu (Ida!)
`
`in
`Fig. I Representative plasma versus time curves for Iornetrexol
`patients receiving I5. 20. or 30 mg/rn’ lometrexol with i.v. folic acid
`supplementation. Curves were generated in ADAPT ll using a three-
`compartment phannacokinetic model.
`
`Table 2 Effect of folic acid pretreatment on lometrexol plasma
`phmneeclrjnetics
`
`Table I Plasma pharmacokinetics by Iometrexol dose
`Lometrexol
`
`dose
`(mglm‘)
`I5
`20
`25
`30
`
`rm
`....
`rm.
`Va
`Cl,_.’m_".__‘.
`(ha)
`(h '1)
`(h")
`n‘ (liters/h/nt‘)(literslm:)
`30.7
`2.9
`(M9
`6
`1.7
`l I.0
`7.9
`2.9
`0.25
`6
`L4
`6.8
`6.5
`2.9
`0.22
`3
`1.4
`l0.2
`38.7
`2.3
`0.24
`9
`I .7
`8.5
`24"
`l.6t0.6‘ 8.9t4.l"0.23:‘:0.l"2.9tl.4"fi.0t48.7"
`
`" Number of patients per dose level.
`" Total number of patients.
`’ Mean : SD.
`
`RESULTS
`Patient Characteristics. Between March and December
`of I995. 24 patients were entered into this study. There were 12
`men and 12 women. whose tumor types included colorectal (n =
`II). lung (n = 6). breast (n = 4), and miscellaneous (n = 3)
`caulcels. 3iit paiieuis each were ileaieti with i5 and 23 mg/In’
`lometrexol. whereas three patients received 25 mglm’. and nine
`patients were treated with 30 mglm’. Of the patients treated with
`15 and 20 mglm’ lornetrexol. three received 5 mg of i.v. folic
`acid I h before whereas the other three received 25 mg/m’ 3 h
`before the lometrexol. All of the patients treated at the 25 mglmz
`dose level received 25 mg/m‘ of folic acid 3 h before the
`lometrexol. At the 30 mg/ml lornetrexol dose level. three pa-
`tients received 5 mg of i.v. folic acid I h before whereas the
`other six received 25 mglmz 3 It before the lometrexol. As
`reported previously (I8). dose-limiting toxicities were encoun-
`tered on cycles 2 and 3 in
`patients. regardless of the
`Iometrexol dose level. Grade 3 and 4 toxicities included neu-
`
`thrombocytopenia. anemia. mucositis. and fatigue.
`tropenia.
`Anemia was the most common toxicity encountered. with S of
`24 patients requiring transfusions.
`lometrexol
`Plasma
`Pharmacoltinetits. First-dose
`plasma pharmacokinetics were determined in all 24 patients
`
`5 mg i.v. folic acid
`I h before
`Iometrexol‘
`(n = 9)
`l.7 : 0.6
`
`Pharmacolrinetic
`parameter
`lI2A-.aJLl._-u
`uuctauum I
`Clsyntetnic
`‘
`V, (liters/m’)
`rm“ tn“)
`r,,,.(tr‘)
`lie ('14)
`“ Mean 1' SD.
`" F test based on regression model.
`
`s.s : L6
`0.2 1 0.2
`2.: : l.l
`32.l 1' 73.7
`
`25 mg/m= i.v. folic
`acid 3 h before
`lomerrexol"
`(n = I5)
`L5 2 0.6
`
`9.0 : 5.1
`0.2 : 0.4
`3.3 : 1.4
`20.7 I 27.2
`
`P“
`0.30
`
`0.54
`0.86
`0.05
`0.43
`
`I shows representative plasma concen-
`entered on study. Fig.
`tration versus time curves for each of three lometrexol doses.
`
`Lometrexol elimination showed a triexponential pattern of de-
`cay, and the pharmacoltinetics were best described by a three-
`compartment model
`in I6 of 24 patients. according to the
`Akailte Information Criterion. As shown in Table I, the mean
`
`lometnexol Cl,,,,m,,_ and V0 were [.6 i 0.6 liters/hlrnz and 8.9 :-
`4.l
`liters/fit’. resi:-actively.
`lviearr a, 3. and 'y half-lives were
`0.23 : 0.I. 2.9 : L4. and 25.0 : 48.7 h. respectively. Cl,,.,,,,,,,,
`was independent of dose. as evidenced by a linear increase in
`AUC with increasing dose (Fig. 2).
`The effect of folic acid pretreatment on lometrexol plasma
`pharmacokinetics was explored. and the results are shown in
`Table 2. A total of 9 patients received 5 mg of i.v. folic acid I h
`before lometrexol. and I5 patients received 25 mglm’ 3 h before
`lometrexol. As shown in Table 2, with the exception of rum.
`there were no significant differences in any of the lometrexol
`pharmacokinetic parameters in patients receiving either of the
`pa reatment regimens. Fig. 2 illitstrates the sigrzifica.-it intc.-pa=
`tient variability observed in plasma AUC. with a 4-fold range in
`patients treated at the 30 mg/m2 dose level. Despite this wide
`variation in plasma lometrexol exposure, pharmacodynamic
`analyses were unable to identify any relationships between
`plasma pharmacokinetics on the first course and the occurrence
`of dose-limiting toxicity.
`
`Downloaded from clincanoerres.aacrjouma|s.org on February 1. 2011
`Copyright © 1998 American Association for Cancer Research
`
`Lilly Ex. 2047
`Sandoz v. Lilly IPR2016-00318
`
`
`
`2352 Plasma and RBC Pharmacokinetics of Lometrexol
`
`eon. coeff.-0.7
`p<‘0.00l
`
`Fig. 3 Representative RBC lometrexol versus time plots for patients
`receiving either 5 mg i.v. folic acid I h before (A) or 25 mg/m‘ 3 in
`before (B) lometrexol. RBC lometrexol was expressed as ng/ml of
`packed RBCs.
`
`RBC Pltarlnacoltlnetlts. Weekly RBC lometrexol data
`were available in all 24 patients enrolled on study, with a
`median follow-up of 39 days (range. l4-l I I). Fig. 3 shows the
`RBC lometrexol levels versus time for each patient receiving
`either 5 mg of i.v. folic acid 1 h before (Fig. 3A) or 25 mglm’
`3 h before (Fig. 3B) lometrexol. Although plasma lometrexol
`was undetectable in all patients within 72 h after a dose and
`remained undetectable throughout the follow-up period. RBC
`iornetrexoi icvcls measured in
`blood flplcs mm be-
`tween courses. RBC lometrexol levels were detectable starting
`at around 4 days following the first dose and continued to rise
`weekly in every patient. Furthermore. the rate and extent of
`apparent lometrexol accumulation in RBCS are independent of
`plasma pharmacokinetics and either the lometrexol or folic acid
`dose.
`Although no relationship between plasrm pharmacokinet-
`ics and toxicity could be identified. a significant linear relation-
`ship existed between the accumulation of lometrexol in RBCs
`and the percentage fall in hemoglobin. hematocrit. and platelet
`count. Fig. 4 shows the relationship between
`RBC
`lometrexol level and the percentage fall in hemoglobin (Fig.
`4A). hematocrit (Fig. 4B). and platelet count (Fig. 4C). The
`correlation coefficients between the change in RBC lomeuexol
`and the fall in these hematological parameters were -0.68 (P <
`0.001). -0.66 (P < 0.001). and -0.33 (P = 0.002). respec-
`tively. No clear trends or differences in these relationships couid
`
`'/odecrease
`
`RBC lometrexol (ng/ml R303)
`0 15mg/m2 o20rngIm2 A25mflm2 -30mg/m2
`
`Fig. 4 Relationship between RBC lometrexol and percentage decrease
`in hemoglobin (A). hematocrit (B). and platelets (C). Different symbols
`represent different lometrexol dosages. Percentage decrease was calcu-
`lated by subtracting the measured value on the given day from the
`pretw.-.tment baseline reieas-_-seine:-.t. Liter: dividing by the I.-Leline value
`and multiplying by I00.
`
`be identified when the different lometrexol dosage levels were
`compared.
`Plasma and RBC Bioaetive Folates. Bioactive folates
`in plasma and RBC: were determined in a subset of six patients
`over the same time period as that for lometrexol. Although the
`data confinn that significantly higher plasma folate levels were
`achieved with the higher folic acid pretreatment dose (data not
`shown). -iuiaiysis of weekiy biood saiiipies faiied to show accu-
`mulation of folates in the RBCS. Fig. 5 demonstrates the failure
`of RBCs to accumulate bioactive folates in patients receiving
`either 5 mg or 25 mg/m2 i.v. folic acid prior to lometrexol.
`
`DISCUSSION
`Initial Phase I studies of lometrexol demonstrated that
`drug-induced toxicity was related to the cumulative dose rather
`than the result of a single dose. These toxicities were severe and
`long-lasting. usually occurring following the second or third
`course of therapy (6. 7). Laboratory investigations pcrfomied
`after the initial Phase i studies found that iomeirexoi-induced
`
`Downloaded from clincanoerres.aacrjouma|s.org on February 1. 2011
`Copyright © 1998 American Association for Cancer Research
`
`Lilly Ex. 2047
`Sandoz v. Lilly IPR2016-00318
`
`
`
`Fig. 5 RBC bioactive folate levels versus time
`in a subset of patients receiving either 5 mg i.v.
`ioiic acid i h before (0) or 25 mgimz 3 h before
`(0) lometrexol. El. second course of lometrexol
`was given (between days 20 and 22).
`
`
`
`
`
`liioactiveFolntea(tall)
`
`Clinical Cancer Research 2353
`
`81
`
`Days
`
`cumulative toxicity could be modulated by folic acid supple-
`mentation without affecting tumor cytotoxicity (9). As a result,
`subsequent clinical development of lometrexol has included
`various schedules of folic acid supplementation. One such
`study. consisting of daily doses of oral folic acid beginning I
`week prior to lometrexol and continttihg for an additional week
`after. found that repeated courses of much higher lometrexol
`doses could be given safely (I0). However. due to fears that
`folic acid may rescue tumor as well as normal tissue if the
`pretreatment regimens were too aggressive, we initiated a trial
`of lometrexol every 3 weeks. preceded by a single i.v. dose of
`ioiic acid. As reported previously (l8). this trial was iirnited by
`the occurrence of cumulative and long-lasting hematological
`toxicities. Recognizing that lometrexol-induced cumulative tox-
`icity might be due to prolonged tissue retention and delayed
`elimination. plasma and whole blood dnrg levels were measured
`in each patient throughout his/her clinical course in an effort to
`correlate the pharmacokinetics of lometrexol with toxicity.
`The plasma pharmacoltinetics of lomeuexol in patients
`treated with and without folic acid supplementation have been
`described previously. Similar to our study. Wedge er al. (10)
`reported that
`lometrexol plasma phannacokinetics were best
`4.nnna:Ln
`u-...n...u..d by a °.hrcc-comparrrrtcrrt phmrnaokiructic model 3
`that the pharmacokinetics were independent of lometrexol dose
`and folic acid supplementation. Moreover, lometrexol plasma
`clearance. volume of distribution. and elimination half-lives
`reported here are in good agreement with the results of Wedge
`er al. ( I0). Although this study confimts previous plasma phar-
`macokinctic findings. the determination that lometrexol accu-
`mulates and is retained in RBCs indicates a complex clinical
`pharmacology that may help to explain the unique toxicity
`profile of this agent.
`Accumulation of lometrexol in RBCs was previously pro-
`posed because of unpublished reports of high RBC folate levels
`in patients receiving lomeuexol. During an earlier Phase I trial
`of lometrexol. Ray and colleagues (7) found that weekly RBC
`folate levels increased over time and remained elevated for
`months. These investigators used a commercially available fo-
`late immunoassay that potentially cross-reacts with lometrexol.
`it was hypothesized that it was actually lometrexol that was
`
`accumulating in RBCs. Our study confinns this hypothesis by
`showing definitively that lometrexol levels in RBCs increase
`continuously over time and that this apparent accumulation is
`independent of either lometrexol or folic acid dose. Moreover.
`bioactive folates. as measured by a functional microbiological
`a_ssay rather than by are itrI_rnunoassay, did not act:-_Im.ulate over
`the same time period.
`Once natural folate cofactors and antifolates enter mam-
`malian cells. several glutamyl residues are added by FPGS.
`increasing in virro and clinical evidence indicates that polyglu-
`tamation by FPGS is central to the therapeutic activity of many
`cytotoxic antifoiates. Although circulating RBCs have undetect-
`able FPGS activity, purified norrnoblasts express very high
`levels of activity (I9. 20). Therefore. it has been proposed that
`folates accumulate in nonnoblasts in the bone marrow as poly-
`glutamates and are retained for the life of the cell. As a result.
`folate accumulation in erythrocytes. unlike plasma folates. is not
`influenced by recent dietary intake and better reflects the long-
`tenn folate (or antifolate) status of the host. Likewise, studies
`using age fractionated erythrocytes have concluded that meth-
`otrexate is incorporated in RBC precursors of the bone marrow
`(2l).
`
`The RBC lometrexol levels in this sandy were t.'iéiir—n‘rir‘rEt'.i
`by taking the difference between the measured whole blood and
`plasma levels. Although lometrexol was undetectable in patients
`plasma starting 72 h after the dose. the whole blood concentra-
`tions continued to rise during the weeks between doses.
`lt
`reasonable to assume. therefore, that the measured lometrexol
`was present primarily in the cellular fraction and that the in-
`crease over time probably reflects the maturation and release
`from the bone marrow of hematopoietic cells. which were
`loaded with long chain lometrexol polyglutamates. Moreover. it
`is likely that the intracellular lometrexol was present in the
`RBCs in its polyglutarttttted form. and through the action of
`endogenous plasma conjugase. the polyglutamates were con-
`vened to monoglutamate prior to high-perfonnance liquid chro-
`matography analysis. Although this hypothesis conld have been
`tested. we were unable to perfotrn the necessary experiments
`because all of the whole blood samples were incubated with
`endogenous conjugase at the time of collection.
`
`Downloaded from clincanoerres.aacrjouma|s.org on February 1. 2011
`Copyright © 1998 American Association for Cancer Research
`
`Lilly Ex. 2047
`Sandoz v. Lilly IPR2016-00318
`
`
`
`2354 Plasma and RBC Pharmacokinetics of Lometrexol
`
`Although lometrexol plasma pharmacokinetics were not
`related to the occurrence of any toxicity. RBC lometrexol levels
`were coneiated with the percentage faii in hemogiobin. hema-
`tot.-n't, and platelet count. Because anemia requiring transfusions
`was the most commonly encountered toxicity on this Phase I
`study. usually occurring following course two or later, it is
`possible that the occunence of delayed—onset anemia is a result
`of the accumulation of lometrexol in the RBCs. Although it is
`not clear why RBC lometrexol levels are related to delayed
`toxicity. one possibility is that the RBCs are acting as a drug
`reservoir. As drug-loaded RBCs lose membrane integrity over
`time. lometrexol may be slowly released back into the circula-
`tion. However. we were unable to detect lometrexol in plasma
`more than 72 h after i.v. adntinistration during this study. over.
`at the times corresponding to the highest RBC concentrations.
`Because the assay used to examine lometrexol levels in this
`study has a lower limit of detection of 10 ng/ml,
`it seems
`unlikely that delayed toxicity is due to prolonged exposure to
`circulating drug in plasma.
`Yet another possible explanation for the conelation be-
`tween RBC lometrexol levels and delayed toxicity is that the
`RBCs could be acting as a surrogate tissue for the true site of
`action. Lometrexol may be accumulating in other tissues. par-
`ticularly those in which toxicity is occurring, in parallel to the
`increases seen in RBCs. Finally. it is possible that the increase
`in RBC lometrexol levels and the observed decreases in hemo-
`
`globin, hematocrit. and platelets. although they occur in parallel,
`are unrelated. The cumulative anemia could be the result of
`folate deficiency caused by depletion of total body folate stores
`by lometrexol. This would explain why trials of lometrexol that
`included daily folic or folinic acid supplementation were able to
`give much higher doses safely (10. 22). However. lometrexol-
`induced folate deficiency as the cause of the chronic anemia
`appears unlikely in this case because we did not see a decrease
`in either plasma or whole blood folates in any of the patients
`studied. Therefore. the most probable explanation for the cumu-
`lative anemia seen on this study is the direct effect of intracel-
`lular lometrexol polyglutamates on RBC production.
`Persistent intracellular levels of methotrexate have long
`been reported in the erythrocytes and livers of patients receiving
`chronic therapy (23. 24). Moreover, the cellular accumulation of
`methotrexate has been associated
`a concomitant ioss of
`
`folates (24). As a result, it has been suggested that persistent
`methotrexate concentrations and/or the resulting folate defi-
`ciency is responsible for methotrexate-associated chronic toxic-
`ity. In children with acute lymphocytic leukemia. accumulation
`of methotrexate in erythrocytes as polyglutamates also serves as
`an indicator of overall methotrexate exposure during the main-
`tenance phase of therapy and has been reported to correlate with
`survival (25). Several prospective studies have used erythrocyte
`methotrexate levels as an index of chronic drug exposure and as
`a method of assessing patient compliance (8. 26, 27).
`This study is the first to definitively show that lometrexol
`accumulates in RBCs. It is likely that lometrexol. similar to
`methotrexate. accumulates as the polyglutamates in RBC pne-
`cursors in the bone marrow. The apparent rise in peripheral RBC
`lometrexol levels probably reflects maturation and release of
`drug-loaded RBCs into the circulation. The demonstration that
`this apparent accumuiation in RBCs is independent of dose
`
`indicates that lometrexol tissue uptake is limited by either trans-
`port or FPGS. Unlike plasma levels of lometrexol, RBC lome-
`trexol concentrations conelate with the occunence of cumula-
`
`toxicity. Although it remains to be seen
`tive hematological
`whether there is actually a causal relationship between RBC
`lometrexol and toxicity.
`the RBCs levels may serve as an
`indicator of general tissue uptake and retention and. therefore.
`may be useful as a predictor of delayed toxicity. In addition to
`methotrexate and lometrexol. several new agents that are sub-
`strates for FPGS are in various stages of development. It is quite
`possible that. as with lometrexol. plasma pharmacokinetics will
`not be the best measure of drug exposure for these agents that
`exhibit prolonged tissue retention.
`
`REFERENCES
`l. Beardsley. G. P.. Moroson. B. A.. Taylor. E. C.. and Moran. R. G. A
`new antifolate antimetabolite. 5.10-dideazatetrahytlrofolate. is a potent
`inhibitor of de novo purine synthesis. 1. Biol. Chem. 264: 328-333.
`I363.
`
`2. Taylor. E. C.. Hamby. J. M.. Shih. C.. Grindey. G. B.. Rinzel. S. M..
`Beardsley. G. P.. and Moran, R. G. Synthesis and antitumor activity of
`5-deaza-5.6.7.8-tetrahydrofolic acid and its N10-substituted analogues.
`J. Med. Chem.. 32: l5l7-I522. I989.
`3. Pizzorno. G.. Soltolosky. J. A.. Cashmore, A. R.. Moroson. B. A.,
`Cross. A. D.. and Beardsley. C-. P.
`lnttecell-.:!.=..' tr-.etabo!is.'n of 5.!!!-
`dideazatetrahydrofolic acid in human leukemia cell lines. Mol. Phanna-
`col.. 39: 85-89. I991.
`4. Sanghani, S. P.. and Moran. R. 6. Tight binding of folate substrates
`and inhibitors to recombinant mouse glycinarnide ribonucle