Pharmacokinetics and Metabolism of Epidoxorubicin and Doxorubicin
`in Humans
`By K. Mross, P. Maessen, W.J.F. van der Vijgh, H. Gall, E. Boven, and H.M. Pinedo
`
`Pharmacokinetics of doxorubicin (DOX), epidoxorubi-
`cin (EPI), and their metabolites in plasma have been
`performed in eight patients receiving 40 to 56 mg/m2
`of both anthracyclines as a bolus injection in two
`sequential cycles. Terminal half-life and volume of
`distribution appeared to be smaller in case of EPI,
`whereas plasma clearance and cumulative urinary
`excretion was larger in comparison to DOX. The major
`metabolite of DOX was doxorubicinol (Aol) followed
`by 7-deoxy-doxorubicinol (7d-Aolon). Metabolism to
`glucuronides was found in case of EPI only. The area
`under the curves (AUC) of the metabolites of EPI de-
`creased in the order of the glucoronides E-glu > Eol-
`glu, 7d-Aolon > epirubicinol (Eol). The AUC of Eol was
`half of the value in its counterpart Aol. In the case of
`EPI, the AUC of 7d-Aolon was twice the level of that of
`the corresponding metabolite of DOX. The terminal
`half-lives of the cytostatic metabolites Aol and Eol
`were similar, but longer than the corresponding val-
`ues of their parent drugs. Half-lives of the glucuro-
`nides (E-glu, Eol-glu) were similar to the half-life of
`their parent drug. 7d-Aolon had a somewhat shorter
`
`E PIDOXORUBICIN (EPI; epirubicin Phar-
`marubicin [Farmitalia Carlo Erba, Milan,
`Italy]), the 4'-epimer of doxorubicin (DOX;
`Adriamycin [Farmitalia Carlo Erba]) (Fig 1), is
`one of the analogs presently being studied in
`phase II and III trials."3
`From preclinical studies with EPI it was con-
`cluded that differences in therapeutic and toxico-
`logic manifestations exist between EPI and
`DOX, reflecting apparent alterations in pharma-
`cologic properties and possible mode of action.4
`A comprehensive review of clinical activity and
`adverse effects of EPI was published recently in
`the Journal of Clinical Oncology.'
`Search for anthracycline analogs is necessary
`because clinical use of DOX is hampered by un-
`favorable side effects. Furthermore, human tu-
`mors of major importance, such as pancreatic
`cancer and lung cancer, are not generally respon-
`sive to DOX. One of the major drawbacks of
`DOX treatment is its cardiotoxic effect. The inci-
`dence of congestive heart failure (CHF) is 5% at
`a cumulative dose of 550 mg/m2 and 800 to 900
`mg/m2 of DOX and EPI, respectively.6' 7
`
`half-life in comparison to both DOX and EPI. Approxi-
`mately 6.2% of EPI and 5.9% of DOX were excreted
`by the kidney during the initial 48 hours. Aol was
`found in the urine of patients treated with DOX,
`whereas Eol, E-glu, and Eol-glu were detected in urine
`of patients treated with EPI. The cumulative urinary
`excretion appeared to be 10.5% for EPI and its metab-
`olites, and 6.9% for DOX and its metabolite. The plas-
`ma concentration v time curves of (7d)-aglycones
`showed a second peak between two and 12 hours
`after injection, suggesting an enterohepatic circula-
`tion for metabolites lacking the daunosamine sugar
`moiety. The plasma concentrations of the glucuron-
`ides were maximal at 1.2 hours for E-glu and 1.9
`hours for Eol-glu. All other compounds reached their
`maximum plasma concentration during the first min-
`utes after the administration of DOX and EPI. Deviat-
`ing plasma kinetics were observed in one patient,
`probably due to prior drug administration.
`J Clin Oncol 6:517-526.© 1988 by American Society
`of Clinical Oncology.
`
`The differences between DOX and EPI re-
`corded for preclinical and clinical pharmacologi-
`cal behavior are probably attributed to differ-
`ences in tissue distribution, metabolism, and
`pharmacokinetics. EPI and DOX do not differ
`substantially in their affinity to double-stranded
`DNA, most likely the main biological target of
`these two anthracyclines.'
`Human metabolism of DOX involves carbonyl
`reduction by aldo- keto reductase, the major en-
`zymatic conversion,' as well as reductive glyco-
`
`From the Department of Oncology, Free University Hospital,
`Amsterdam.
`Submitted March 19, 1987; accepted August 31, 1987.
`Supported by a grant from Farmitalia, Milano, Italy. K.
`Mross is a recipient of European Organization for the Research
`and Treatment of Cancer fellowship.
`Dr Mross' present address is University Clinic Eppendorf,
`Medical Clinic, Department of Oncology and Hematology,
`Hamburg, Federal Republic of Germany.
`Address reprint requests to W.J.F. van der Vijgh, PhD, De-
`partment of Oncology, Free University Hospital, De Boelelaan
`1117, 1081 HV Amsterdam, The Netherlands.
`0 1988 by American Society of Clinical Oncology.
`0732-183X/88/0603-0008$3.00/0
`
`Journal of Clinical Oncology, Vol 6, No 3 (March), 1988: pp 517-526
`
`517
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0001
`
`

`
`518
`
`MROSS ET AL
`
`OH
`
`OH
`
`CH 3,
`
`Fig 1. Molecular structures of DOX (left) and EPI (right).
`
`NHJ
`
`CI-
`
`sidic and hydrolytic glycosidic cleavage. Molec-
`ular structures of the metabolites are shown in
`Fig 2. The main metabolite doxorubicinol (Aol),
`maintains anticancer activity, whereas the agly-
`cones are not active.' 0
`The pharmacokinetics of DOX using different
`schedules is well documented. 11, 12 The 4'-gluc-
`
`O OH
`
`o
`
`OH
`
`OH
`
`OH
`
`OH
`
`M.
`
`0
`
`HO
`
`H
`
`M
`
`0 HoH
`
`H
`
`DOX AGLYCONE
`
`(AON)
`
`AOL AGLYCONE (AOLON)
`
`O OH
`
`0
`
`OH
`
`o
`
`"OH
`
`0 On
`
`OH
`
`M.O
`
`0 HO
`
`7-DEOXY AON (7d-AON)
`
`7-DEOXY AOLON (7d-AOLON)
`
`0
`
`OH
`
`On
`
`0O
`
`MNO
`
`0 HO
`
`NH
`2
`
`OH
`
`M,
`
`HO NH
`2
`
`EPIRUBICINOL (EOL)
`
`DOXORUBICINOL (AOL)
`
`0
`
`OH
`
`0
`
`0
`
`OH
`
`OH
`
`(0
`
`0
`
`OHO
`
`COON
`
`MO
`
`O
`
`0W
`
`o
`
`M0
`
`0O
`
`o
`
`COOH
`
`On
`
`NM2
`
`OH
`
`2
`
`EPI-GLUCURONIDE (E-GLU)
`EOL-GLUCURONIDE (EOL-GLU)
`Fig 2. Molecular structure of all measured metab-
`olites derived from DOX and EPI.
`
`uronides, only present as metabolites in case of
`EPI, were first described by Weenen et al." By
`that time, pharmacokinetic studies of EPI and its
`metabolites were hampered by low recoveries of
`the very polar glucuronides due to existing liq-
`uid-liquid extraction procedures.14 This problem
`was solved by the introduction of a liquid-solid
`extraction procedure for anthracyclines.5 "' 16'17
`However, description of pharmacokinetic data
`for EPI has been incomplete with respect to all
`known metabolites up to now.
`A pharmacokinetic study of DOX and EPI in
`patients with advanced cancer, designed as a
`crossover study, was implemented to describe
`the pharmacokinetic behavior of both drugs and
`their metabolites. The results of this study and
`their relationship to suggested modes of action
`are discussed.
`
`PATIENTS, MATERIALS, AND METHODS
`Treatment Schedule
`EPI and DOX were obtained as a sterile lyophilized powder
`(50 mg/vial, Farmitalia, Carlo Erba SpA, Italy). The drug was
`reconstituted with 25 mL sterile water (United States Pharmaco-
`peia). The prescribed dose of DOX or EPI was administered
`subsequently within one to two minutes through the line of rapid
`saline infusion. The pharmacokinetic study was performed in
`patients receiving 40 to 56 mg/m 2 DOX or EPI in a crossover
`design with an interval of 3 weeks.
`Eight female patients, not previously treated with anthracy-
`clines, were included in the study after informed consent was
`obtained. All patients had advanced disease requiring treatment
`with anthracyclines either as single agent, in combination with
`fluorouracil and cyclophosphamide for breast cancer, or in com-
`bination with mitomycin for adenocarcinoma of unknown pri-
`mary origin. During the 48-hour sampling period no other cyto-
`static agents were administered. Thereafter, anticancer treat-
`ment was continued according to the treatment protocol.
`Patient characteristics and baseline laboratory values are sum-
`marized in Table 1. All patients had total bilirubin levels < 15
`jpmol/L and a normal serum creatinine. Three patients had meta-
`static lesions in the liver. Only one patient (J.O.) had been
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0002
`
`

`
`PHARMACOKINETICS OF DOXORUBICIN/EPIDOXORUBICIN
`
`519
`
`Table 1. Patient Characteristics
`Total
`Alkaline
`Total
`Bili-
`Phos-
`Crea-
`Pa- Age Primary
`tinine
`rubin
`phatase y-GT SGOT SGPT LDH Protein Albumin
`(U/L)
`(U/L)
`tient
`(yr)'
`(U/L)
`Tumor
`Metastatic Lesion
`(gmol/L) (p(cid:2)mol/L)
`(U/L)
`(U/L)
`(g/L)
`(g/L)
`Lungs, bones, pleura
`J.O. 62 Breast
`7
`315
`63
`154
`17
`9
`6
`64
`31
`K.U.
`50 Breast
`8
`77
`87
`55
`33
`Liver, bones, pleura
`69
`54
`693
`63
`R.O. 50 Breast
`330
`203
`36
`72
`39
`Liver
`112
`10
`46
`-
`S.T.
`7
`7
`Bones
`41 Breast
`35
`71
`6
`140
`9
`211
`68
`L.A.
`53 Unknown Lungs, bones
`7
`11
`8
`65
`4
`94
`201
`71
`34
`B.O. 52 Breast
`8
`76
`33
`Lungs, pleura
`91
`6
`43
`13
`9
`182
`5
`10
`206
`59
`33
`B.A.
`54 Unknown Lungs, pleura
`61
`9
`111
`46
`V.R.
`51 Breast
`399
`72
`Lungs, liver
`35
`86
`6
`197
`93
`63
`122
`70-120
`9
`100
`5-15
`5-15
`Reference values
`175 60-80
`38-50
`34
`Abbreviations: y-GT, y-glutamyl transpeptidase. LDH, lactic dehydrogenase.
`
`analytical column. EPI was added as internal standard (2.5 x
`10-8 mol/L) to all plasma samples containing DOX and its
`metabolites, whereas DOX was added (2.5 x 10-8 mol/L) to
`all samples containing EPI and metabolites. Samples were pre-
`pared in duplicate. Six spiked plasma samples containing the
`two drugs and their metabolites (see Fig 2) (5 x 10 -0
`to 5 x
`10- mol/L) were included in each series to construct a within-
`run standard line. A typical chromatogram of a spiked blanc
`plasma sample is shown in Fig 3. The detection limit of the assay
`ranged from I x 10 - 0 mol/L for Aolon to 4 x 10 -0 mol/L
`for 7d-Aon and DOX. The other compounds could be detected
`within this range. Plasma samples were diluted with blank he-
`parinized plasma from a blood donor. For the first two samples
`
`C)
`
`a(
`02
`
`o
`
`treated with chemotherapy (cyclophosphamide, methotrexate,
`fluorouracil) and hormonal therapy (aminoglutethimide, hydro-
`cortisone) before.
`
`Blood and Urine Samples
`Blood samples of 10 mL were obtained at -5, 0, 5, 10, 15,
`30, 60 minutes, and 2, 4, 6, 9, 12, 24, 36, and 48 hours after
`bolus injection. Blood was collected in heparinized (150 IU Li-
`heparine) glass tubes (Terumo, Leuven, Belgium) and immedi-
`ately centrifuged at 4°C, 4,000 g for 15 minutes. The plasma
`was transferred to polystyrene tubes. Urine samples were ob-
`tained in portions of 6 hours over 48 hours. Plasma and urine
`samples were maintained at - 20°C until analysis. After thaw-
`ing, all samples were sonicated and centrifuged at 4'C for five
`minutes with 4,000 g to remove clotted material.
`
`High-Performance Liquid Chromatography
`Analysis
`The high-performance liquid chromatography (HPLC) assay
`used for detection and quantification of the two parent drugs and
`their metabolites was recently developed by Maessen et al. 17
`Briefly, the chromatographic system consisted of a WISP 710B
`injection system, a model 6000A solvent delivery system, and a
`data module with system controller 720 (Waters, Etten-Leur,
`The Netherlands). This system was provided with a stainless
`steel HPLC column (4.6 x 100 mm, 3 A.m CP MicroSpher,
`Chrompack, Middelburg, The Netherlands) including a guard
`column (4 x 4 mm, 5 prm LiChrosorb RP-18, Merck, Amster-
`dam) and a F-1000 fluorescence detector (excitation wave
`length, 480 nm; emission wavelength, 580 nm) from Merck-
`Hitachi, Amsterdam. An isocratic eluent was used consisting of
`0.02 mol/L NaH 2PO4 pH 4/acetonitrile (2.5/1 v/v) at a flow rate
`of 1 mL/min. DOX, EPI, and their metabolites were extracted
`from human plasma using C-18 Sep-Pak cartridges (Waters),
`pretreated with 5 mL methanol, 5 mL H20, and 5 mL buffer
`(0.02 mol/L NaH2PO4 pH 4/acetonitrile 9/1, v/v). One milliliter
`plasma was injected onto the cartridge, subsequently purged
`with 2 mL buffer, dried with a flow of air, and eluted with
`methanol/chloroform (3/1 v/v). The eluate was evaporated to
`dryness at 500 C under a stream of air. The residue was redis-
`solved in 50 ;LL buffer, of which 30 /,L was injected onto the
`
`8
`
`10
`
`12
`
`14
`
`0
`
`2
`
`4
`
`6
`(MIN.)
`Fig 3. Typical chromatogram after extraction of a
`plasma sample spiked with DOX and EPI and their
`metabolites and separation of all compounds. The
`peaks in front of Eol-glu (hatched area) represent co-
`extracted plasma material with fluorescent proper-
`ties.
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0003
`
`

`
`520
`
`MROSS ET AL
`
`(t = 0 and 5 minutes) various dilutions were prepared in order to
`determine all metabolites as accurately as possible because of
`the large differences in drug and metabolite concentrations. The
`urine samples were sonicated after pH adjustment to 2.5 and
`subsequently centrifuged. After addition of an external standard
`(DOX, respectively, EPI), samples were directly injected onto
`the column.
`
`Pharmacokinetic Analysis
`Each set of concentration time [c(t)] values of DOX and EPI
`in plasma was fitted to the appropriate polyexponential equation
`using the program, NONLIN. 18 The final decision to describe
`the results according to a three compartment model was per-
`formed by AUTO AN. 19 The r2 of NONLIN least-squares fit
`was always better than .999, with one exception (Patient J.O., r2
`= .97). All fittings were performed to the three exponential
`equation:
`
`(1)
`
`cp = Ax e-t
`
`+ B x e-Pt + C x e-Yt.
`
`The pharmacokinetic parameters for the parent drugs were cal-
`culated using the following equations:
`
`(2) AUC = Ai, + Bi/ + C/i (nmol x L-' x min);
`(3) Clp = D/AUC (L x min-1) (normalized to 1.74 m2);
`A + B + C
`
`(4) Vd = D x
`
`02
`
`(5)
`
`2
`
`f2
`AUC2
`t½ = 0.693/k x 60 (k = a,
`
`(L) (normalized to 1 kg);
`
`, y).
`
`The abbreviations used are AUC, area under the curve; Clp,
`plasma clearance; Vd, volume of distribution; D, dose; t½, half-
`life; and k, rate constant. The terminal half-lives of DOX and
`EPI were also analyzed by the curve-stripping method using all
`time points from four hours onward. Additionally, the AUCs of
`the parent drugs and their metabolites were determined by means
`
`of the trapezoidal method because of irregular c(t) curves of the
`metabolites, not allowing calculations according to a pharmaco-
`kinetic model. The terminal half-lives of DOX, EPI, and their
`metabolites were determined from their concentrations in the
`final plasma samples by least square fitting.
`
`RESULTS
`The following results represent summarized
`data from seven of eight patients studied. All
`pharmacokinetic data from one patient (J.O.)
`and TAy of second patient (V.R.) were omitted
`from the calculation of mean values because
`these data differed <2 standard deviations from
`the mean.
`Pharmacokinetic parameters for the parent
`drugs, calculated according to the formulas 1
`through 5 described in the previous section, are
`shown in Tables 2 and 3. Besides the calculations
`with the NONLIN program, the terminal half-
`life was also calculated by curve-stripping. This
`procedure allowed a consequent calculation of
`all final half-lives over an interval of four to 48
`hours and may, therefore, provide additional in-
`formation about interpatient and interdrug vari-
`ation. The disappearance of both parent drugs
`was triphasic. The half-lives calculated for DOX
`were always longer than those calculated for
`EPI. The mean volume of distribution was larger
`in the case of DOX when compared with EPI,
`while the plasma clearance (normalized to 1.74
`m2 body surface area) was slower in the case of
`
`Table 2. Calculated Pharmacokinetic Parameters of DOX in Eight Patients
`
`PPC*
`(mol/L
`X 106)
`
`14.0§
`11.5
`16.0
`12.0
`14.9
`16.5
`10.0
`24.0
`
`15.0
`4.7
`
`Patient
`
`J.O.
`K.U.
`R.O.
`S.T.
`L.A.
`B.O.
`B.A.
`V.R.
`
`Mean
`+SD
`
`AUC
`(nmol
`min/L
`X
`10- 4)
`
`96.6§
`19.7
`16.5
`9.8
`13.0
`14.9
`8.3
`43.0
`
`17.9
`11.7
`
`Clpt
`(L/h)
`
`9.83§
`48.38
`53.99
`87.08
`65.46
`60.38
`87.46
`20.11
`
`60.40
`23.40
`
`Vd.
`(Ukg)
`
`6.59§
`31.14
`14.05
`20.37
`14.60
`12.29
`31.24
`44.63
`
`24.00
`12.00
`
`t,,
`(h)
`
`0.167§
`0.062
`0.050
`0.017
`0.046
`0.029
`0.034
`0.047
`
`0.041
`0.020
`
`t/23
`(h)
`
`1.49§
`3.02
`0.30
`0.07
`0.22
`0.09
`0.17
`1.68
`
`0.79
`1.13
`
`t,/y
`(h)
`
`48.2§
`48.4
`24.1
`17.6
`21.4
`18.7
`24.8
`157.8§
`
`25.8
`11.4
`
`t,/(cid:127)Y
`(h)
`
`15.8§
`19.9
`23.2
`24.1
`23.1
`24.2
`22.9
`55.0§
`
`22.9
`1.6
`
`NOTE. Explanation of the calculated parameters in Pharmacokinetic Analysis section.
`*Peak plasma concentrations normalized to 50 mg/m 2.
`tPlasma clearance normalized to 1.74 m2 .
`fCalculations with curve stripping method from four hours onward.
`§Omitted for calculation of the mean.
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0004
`
`

`
`PHARMACOKINETICS OF DOXORUBICIN/EPIDOXORUBICIN
`
`521
`
`Table 3. Calculated Pharmacokinetic Parameters of EPI in Eight Patients
`AUC
`(nmol
`min/L
`x
`10 -)
`21.8§
`11.7
`12.9
`8.7
`6.7
`9.6
`8.1
`18.2
`10.4
`4.2
`
`PPC*
`(mol/L
`X 106)
`
`5.5§
`12.0
`19.5
`18.0
`8.5
`14.5
`21.1
`31.4
`
`17.9
`7.4
`
`Patient
`
`J.O.
`K.U.
`R.O.
`S.T.
`L.A.
`B.O.
`B.A.
`V.R.
`
`Mean
`±SD
`
`Clpt
`(L/h)
`46.00§
`81.57
`68.93
`99.01
`126.78
`95.62
`90.46
`47.44
`87.10
`24.95
`
`Vd,
`(L/kg)
`18.82§
`10.22
`11.95
`12.30
`23.25
`15.93
`13.49
`16.69
`14.8
`4.3
`
`t,
`(h)
`0.088§
`0.051
`0.020
`0.028
`0.035
`0.023
`0.023
`0.029
`0.030
`0.010
`
`tl/2,
`(h)
`2.96§
`1.07
`0.08
`0.37
`0.66
`0.12
`0.22
`0.91
`0.49
`0.39
`
`t,2
`(h)
`27.7§
`16.6
`12.9
`15.6
`17.3
`14.0
`15.5
`32.5§
`15.3
`1.6
`
`t,2Y
`(h)
`20.2§
`12.2
`17.3
`15.3
`13.3
`16.0
`16.7
`28.6§
`15.1
`2.0
`
`NOTE. Explanation of the calculated parameters in Pharmacokinetic Analysis section.
`*Peak plasma concentrations normalized to 50 mg/m2 .
`"tPlasma clearance normalized to 1.74 m2
`.
`:Calculations with curve stripping method from four hours onward.
`§Omitted for calculation of the mean.
`
`DOX than in the case of EPI. The AUC of DOX
`was larger than it was in the case of EPI.
`DOX and EPI were rapidly metabolized. Im-
`mediately after bolus injection all metabolites
`were detectable. As shown in Tables 2 and 3,
`peak plasma concentrations (PPC) (normalized
`to 50 mg/m2 ranged from 10 X 10-6 mol/L up to
`24 x 10-6 mol/L for DOX, and 5.5 x 10-6
`mol/L to 32 x 10-6 mol/L for EPI-treated pa-
`tients. The mean PPC (normalized to 50 mg/m2)
`for the glucuronides was 4.7 ± 3.2 x 10-7
`mol/L E-glu occurring after 1.2 ± 0.6 hours,
`and 1.8 ± 2.3 x 10-7 mol/L Eol-glu occurring
`after 1.9 ± 1.1 hours (Table 4).
`Because of the irregular plasma concentration
`v time curves of the metabolites, it was not possi-
`ble to fit these curves with exponential terms.
`
`Table 4. Peak Plasma Levels of the Glucuronides
`E-glu and Eol-glu and Time Points of Peak
`
`E-glu
`(x10-7
`mol/L)
`
`Time
`(h)
`
`Eol-glu
`(x10-7
`mol/L)
`
`Time
`(h)
`
`Patient
`
`J.O.
`5.63
`2
`4.10
`6
`K.U.
`11.00
`1
`1.35
`2
`R.O.
`5.00
`2
`0.70
`4
`S.T.
`4.40
`1
`0.75
`1
`L.A.
`2.40
`2
`0.95
`2
`B.O.
`7.00
`1
`1.10
`1
`B.A.
`3.38
`0.5
`1.16
`1
`1
`V.R.
`5.35
`0.63
`2
`Mean±SD 4.7-3.2 1.2-0.6 1.8-2.3 1.9t1.1
`
`However, it was feasible to calculate the terminal
`half-lives of the metabolites from 12 hours on-
`ward (Table 5). The longest half-life was found
`in case of Aol and epirubicinol (Eol). The short-
`est half-life was calculated for 7d-Aon. The half-
`lives of the two glucuronides were similar to that
`of their parent drug. Half-lives for Aon and Ao-
`Ion could not be determined accurately.
`Representative plasma decay curves for DOX,
`EPI, and their metabolites are illustrated in Figs 4
`and 5, respectively. AUCs (0 to 48 hours) for the
`parent drugs and their metabolites, normalized to
`50 mg/m 2, calculated with the trapezoidal rule
`are listed in Table 6. It can be deduced that Aol
`and 7d-Aolon represent the major metabolites in
`case of DOX. In case of EPI, E-Glu, the glucuro-
`nidated parent drug,
`is more prominent. The
`AUC of this metabolite reached nearly twice the
`value of its parent drug (EPI) and gave rise to a
`doubling of the total AUC due to EPI compared
`with DOX. The AUC for Eol was only half of
`
`Table 5. Terminal Half-Lives for the Metabolites of
`DOX and EPI
`
`Compounds
`DOX/EPI
`Aol/Eol
`7d-Aolon
`7d-Aon
`Eol-glu
`E-glu
`
`Dox (h)
`28.3 2.8
`32.8 ± 1.7
`16.8-6.3
`-
`
`EPI (h)
`19.0 t 2.4
`31.5 +6.0
`17.5±4.9
`13.9±5.2
`18.3±4.0
`18.6±2.1
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0005
`
`

`
`522
`
`10000
`
`1000
`
`100
`
`10
`
`* DOX
`* AOL
`A 7d AOLON
`3 7d AON
`o AOLON
`A AON
`
`0
`
`1
`
`2
`
`46
`
`9 12
`
`24
`
`36
`
`48
`
`hrsl
`
`Fig 4. Typical concentration v time curves of DOX
`and its metabolites in patient S.T. receiving 50 mg/m2
`as an IV bolus injection.
`
`that of the corresponding metabolite Aol. The
`AUCs for the 7-deoxy-aglycones were always
`higher in case of EPI and most pronounced in
`case of 7d-Aolon. The aglycones, especially 7d-
`Aolon, behaved irregularly in terms of their plas-
`ma concentrations v time. Highest concentra-
`tions were found immediately after injection fol-
`lowed by a fast decrease immediately thereafter.
`Invariably, a second peak could be detected at
`two to 12 hours after drug administration.
`
`* EPI
`* EOL
`A 7d AOLON
`0 7d AON
`O AOLON
`* AON
`* E GLU
`0 EOL GLU
`
`[xlog M
`
`10000
`
`1000
`
`100
`
`10
`
`1
`
`O,1
`
`0
`
`1
`
`2
`
`46
`
`9 12
`
`24
`
`36
`
`48
`
`lhrsl
`
`Fig 5. Typical concentration v time curves of EPI
`and its metabolites in patient B.A. receiving 40
`mg/m 2 as an IV bolus injection.
`
`MROSS ET AL
`
`Table 6. Areas Under the Plasma C(t) Curve (0 to
`48 Hours) of Dox, EPI, and Metabolites [(nmol-
`minl10- 4 )/L] Normalized to a Dose
`of 50 mg/m 2
`Dox
`
`Metabolites
`
`EPI
`
`DOX/EPI
`Aol/Eol
`7d-Aolon
`Aolon
`7d-Aon
`Aon
`Eol-glu
`E-glu
`
`13.25 2.60
`4.56± 1.56
`1.84 ±0.88
`0.18±0.19
`0.59±0.31
`0.28 0.26
`-
`-
`
`Total AUC
`
`20.70
`
`11.34 3.30
`2.31 ± 1.05
`4.35± 2.44
`0.23+0.19
`0.93±0.61
`0.03 0.04
`4.36 + 1.54
`20.04± 10.9
`
`43.59
`
`The cumulative urinary excretion (0 to 48
`hours) was similar for both unchanged drugs.
`Aol and Eol were detectable, but not in signifi-
`cant amounts. Eol-glu and E-glu were excreted
`into the urine. Eol-glu did not contribute much to
`the excretion of EPI as a metabolite, whereas the
`urinary excretion of E-glu was remarkable (see
`Table 7).
`One patient (J.O.) showed a deviating phar-
`macokinetic pattern (Tables 2 and 3). For this
`patient, the plasma v time curves for DOX and
`EPI including their metabolites are shown in Figs
`6 and 7. Comparison of these figures with Fig 4
`and 5, as well as the calculated pharmacokinetic
`parameters from Tables 2 and 3, show that most
`parameters were severely affected. All half-lives
`were longer, the AUCs of DOX and EPI were
`higher, and the plasma clearance was much low-
`er. The AUCs of nearly all metabolites were
`higher, but not uniformly with a constant factor.
`Factors ranged from 3.5 (E-glu) up to 15 times
`(Eol-glu) higher than normal. The differences
`between the AUCs of the (7d-)aglycones were of
`minor significance. The cumulative urinary ex-
`cretion was significantly lower than normal. This
`patient featured not only deviations in the phar-
`
`Table 7. Cumulative Urinary Excretion of EPI/DOX
`and Their Metabolites in Percent of the Dose (0 to 48
`Hours)
`
`Compound
`
`DOX/EPI
`Aol/Eol
`Eol-Glu
`E-glu
`
`Total
`
`Dox
`
`5.9 + 1.6
`1.0±0.3
`
`-
`
`6.9
`
`EPI
`
`6.2 - 1.5
`0.6±0.3
`0.4±0.2
`3.3 ± 1.2
`
`10.5
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0006
`
`

`
`PHARMACOKINETICS OF DOXORUBICIN/EPIDOXORUBICIN
`
`523
`
`nic acid. However, the analytical procedure was
`not totally satisfactory at that time with regard to
`a relatively low extraction efficiency and the as-
`sumption that fluorescence properties of parent
`drug and all metabolites were identical. Our
`present technique has a high recovery for all me-
`tabolites, while each compound has been quanti-
`tated with a highly sensitive fluorescence detec-
`tor using individual calibration lines for each of
`the compounds.'" In a study investigating the
`property of EPI to undergo glucuronidation, our
`group proposed a causal relationship between
`this characteristic of the analog and the differ-
`ences with the parent drug with respect to phar-
`macokinetics and toxicity.
`The half-life of distribution was shown to be
`extremely short for both EPI and DOX, while in
`all patients half-lives for DOX were longer than
`for EPI, being consistent with the results report-
`ed by Camaggi et al. 24
`EPI and DOX appeared to be distributed into a
`deep tissue compartment from which they were
`slowly released. Elimination of DOX was slower
`than that of EPI, which may be explained by the
`unique glucuronidation of EPI.
`The terminal half-lives of Aol and Eol are
`longer, whereas half-lives of the other metabo-
`lites were similar or shorter than those of the
`respective parent drugs. It is known that Aol and
`probably Eol are cytotoxic and may greatly con-
`
`1x109MI
`10000
`
`1000
`
`100
`
`10
`
`0.1
`
`0
`
`1
`
`2
`
`4 6 9 12
`
`24
`
`36
`
`48
`
`hrs
`
`Fig 7. Concentration v time curves of EPI and its
`metabolites of patient J.O. with appreciably deviat-
`ing kinetics and metabolism after administration of
`56 mg/m 2 EPI as an IV bolus injection.
`
`0
`
`1
`
`2
`
`4 6 9 12
`
`24
`
`[hrsi
`Fig 6. Concentration v time curves of DOX and its
`metabolites of patient J.O. with appreciably deviat-
`ing kinetics and metabolism after administration of
`56 mg/m 2 DOX as an IV bolus injection (Aon was not
`detectable).
`
`macokinetics of the parent drugs, but also sig-
`nificant changes in metabolism.
`
`DISCUSSION
`Although EPI and DOX differ only in the ori-
`entation of the 4'-hydroxy group in the sugar
`moiety, this structural modification appears to be
`critical as pharmacokinetics and metabolism of
`the two drugs differ greatly. Our present study
`concerned a comparison of the pharmacokinetics
`of DOX and EPI in eight patients with advanced
`cancer.
`The plasma elimination of DOX and EPI ap-
`pears to be triphasic after intravenous (IV) bolus
`injection, which is in agreement with previously
`reported data.20 23 However, contrary to earlier
`reports, 23,24 we failed to detect exceptionally
`large interindividual variations.
`Our group was the first to report on pharmaco-
`kinetics and metabolism of EPI,"3 showing the
`ability of EPI and Eol to conjugate with glucuro-
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0007
`
`

`
`524
`
`MROSS ET AL
`
`tribute to myelotoxicity and cardiotoxicity.12
`While the degree of cytotoxicity of the glucuro-
`nides is unknown, the aglycones were reported to
`lack antitumor activity. 0 Robert et al25 reported
`terminal half-lives for DOX, EPI, Aol, Eol, Eol-
`glu, and E-glu which generally corresponded
`with our results, although slightly shorter, which
`may be related to the delayed administration of 5-
`fluorouracil and cyclophosphamide at two hours
`after the anthracyclines.
`In contrast to previous reports, the major me-
`tabolite E-glu appeared to have an AUC almost
`twice as high as that of EPI.26' 2
`In addition, two
`prominent metabolites emerging in the case of
`EPI were Eol-glu and 7d-Aolon. In the case of
`DOX, Aol was most prominent, whereas all oth-
`er metabolites were of minor importance. A re-
`markable difference between the metabolism of
`the two drugs was observed by the formation of
`7d-Aolon. In the case of EPI, the AUC of the
`latter metabolite was twice as high as that of the
`corresponding metabolite of DOX.
`Except for the glucuronides, all metabolites
`achieved their maximum plasma concentration
`within the first minutes after bolus injection, in-
`dicating a rapid metabolism of DOX and EPI.
`Glucuronides reached their peak concentrations
`on average at 1.2 hours in the case of E-glu, and
`at 1.9 hours in the case of Eol-glu. Deviant distri-
`bution and disposition may be expected of the
`polar glucuronides compared with the more apo-
`lar compounds, as glucuronides tend to be con-
`fined to the extracellular space. Assuming that
`biliary excretion is the major elimination path-
`way of such polar metabolites, EPI and Eol can
`be reabsorbed from the intestines due to the rela-
`tively high activity of 6/-glucuronidase in the in-
`testines. Furthermore, DOX and EPI can under-
`go deglycosidation
`and
`reduction
`in
`the
`intestines, thereby promoting reabsorption and
`enterohepatic cycling of aglycones. This may ex-
`plain the fact that the 7d-aglycones showed a
`second peak in the plasma c(t) curve after two to
`12 hours, a phenomenon which has not been
`previously described in pharmacokinetics studies
`of EPI or DOX. However, it was found in rats
`and rabbits that DOX and its metabolites are not
`extensively reabsorbed from in the small intes-
`tines.29' 30
`Total excretion of drug via urine, including
`metabolites, was higher for EPI than for DOX.
`
`These results are in agreement with our findings
`reported earlier.26
`In the case of liver function impairment with
`an elevated serum bilirubin level (a late sign), the
`dose of DOX should be reduced. One patient (not
`included in this series) with severe liver function
`impairment, receiving 15 mg/m2 DOX, appeared
`unable to metabolize the drug. On the other
`hand, as shown in Tables 2 and 3 and in Figs 6
`and 7, one heavily pretreated patient (J.O.) and
`two patients with liver metastases (K.U., V.R.)
`had pharmacokinetic parameters with
`longer
`half-lives and larger AUCs than the remaining
`patients. Consequently, no clear relationship be-
`tween the degree of liver damage and impairment
`of elimination of EPI and DOX could be estab-
`lished.
`In the case of J.O., the plasma elimination of
`DOX was delayed in each phase, while the con-
`centration of Aol was even higher than that of
`DOX at four hours after drug administration. A
`similar observation was made for Eol at six hours
`after EPI administration. Glucuronide concentra-
`tions (Eol-glu and E-glu) were much higher than
`in the other patients, a finding that might be
`related to glucuronyl transferase induction. In-
`duction of aldoketo-reductase may explain the
`higher concentrations of Eol and Aol. A third
`enzyme system, the cytochrome P-450 reduc-
`tase, involved in the formation of (7d-)agly-
`cones, seems to be reduced in this patient, be-
`cause pharmacokinetic parameters
`indicate a
`delayed excretion of the drugs. This patient had
`been pretreated with tamoxifen, aminoglutethi-
`mide (AG) plus hydrocortisone, and later treated
`with chemotherapy consisting of cyclophospha-
`mide, methotrexate, and 5-fluorouracil. Among
`these drugs AG has been shown to be a potential
`inducer of liver enzymes.31 Cyclophosphamide
`which is activated by liver microsomal mixed
`function oxydases,32 is known to cause a de-
`crease of the activity of several liver enzymes, ie,
`the cytochrome P-450 content. In this respect it is
`of interest that an interaction between the metab-
`olism of cyclophosphamide and DOX has recent-
`ly been described.33 Therefore, drug-DOX/EPI
`interactions may have been responsible for the
`observed alterations in metabolism and pharma-
`cokinetics in this patient. Considering the fact
`that biliary excretion of anthracyclines is an ac-
`tive, capacity-limited process, which may be
`
`Downloaded from ascopubs.org by 66.28.38.188 on November 2, 2016 from 066.028.038.188
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Ex. 1051-0008
`
`

`
`PHARMACOKINETICS OF DOXORUBICIN/EPIDOXORUBICIN
`
`525
`
`subjected to competitive inhibition, enhanced ir-
`regular pharmacokinetics may be observed in pa-
`tients with apparent normal liver functions re-
`ceiving concomitantly administered drugs.
`There has been a growing tendency to associ-
`ate anthracycline redox cycling and radical pro-
`duction with chronic cardiotoxicity.37 Redox
`chemistry includes the reductive cleavage of the
`glycoside bond from the hydroquinone which
`leads to the quinone-methide. Electrophilic and
`nucleophilic substitution leads to the aglycones
`and 7d-aglycones, respectively. Pharmacokinet-
`ics of the (7d)-aglycones seems to be highly in-
`teresting in this respect.38 In contrast to Robert"9
`who stated that (7d-)aglycones are artifactual and
`if present only in negligible amounts, it is evident
`from our results and a recently published study,'4
`that in vivo formation of this metabolite can oc-
`cur. Interestingly, the absolute concentration of
`7d-Aolon is at least twice as high after EPI than
`
`after DOX administration, suggesting an inverse
`relationship between the generation of free radi-
`cals and the formation of 7d-aglycones.
`The unique glucuronidation pathway and rapid
`elimination may partly expl