Degradation of Paclitaxel and Related
`Compounds in Aqueous Solutions I Epimerization
`
`JIAHER TIAN12 VALENTINO J STELLA1
`
`1Department of Pharmaceutical Chemistry The University of Kansas Lawrence Kansas 66047
`
`2Wyeth Research Pearl River New York 10965
`
`Received 22 November 2006 revised 3 April 2007 accepted 11 May 2007
`
`Published online in Wiley InterScience wwwintersciencewileycom DOI 101002 ljps21112
`
`ABSTRACT
`have complex
`Paclitaxel and other
`structures including the
`taxanes
`presence of numerous hydrolytically sensitive ester groups and a chiral center
`that
`readily undergoes epimerization thus making their kinetics complex The present study
`attempts to understand the mechanism of epimerization at the 7 position of paclitaxel 7
`epitaxol 10deacetyltaxol 7epi10deacetyltaxol baccatin III and 10deacetylbaccatin
`and
`III Kinetics were studied as function of temperature pH and buffer concentration
`analyzed using a stability indicating assay and LCMS to identify degradation products
`Epimerization was base catalyzed with no evidence of acid catalysis noted The observed
`for epimerization K indicated a thermodynamically more favor
`equilibrium constant
`able Sepimer and a small free energy change between the two epimers For all of the
`compounds in this study removal of the C10 acetyl group increases the epimerization
`rate in basic aqueous solutions The observed base catalyzed epimerization in near
`to higher pH range suggests a possible rapid deprotonationprotonation of the C7
`neutral
`OH followed by a structural
`through a retroaldolaldol mechanism to
`rearrangement
`the rate limiting step of structure rearrangement most
`form the epimer Moreover
`likely occurs with the formation of an enolate intermediate © 2007 Wiley Liss Inc and the
`American Pharmacists Association J Pharm Sci 9712241235
`2008
`Keywords
`taxol 7epitaxol baccatin
`III epimerization degradation
`paclitaxel
`
`stability
`
`INTRODUCTION
`
`this work was to identify
`The overall goal of
`structural elements affecting the rate of epimer
`ization of paclitaxel and some of its analogues
`in aqueous solution As a novel anti tumor drug
`paclitaxel has been studied extensively since the
`late 1970s Despite many advances only limited
`information was available on the chemical stabi
`in aqueous solution Most early
`lity of paclitaxel
`studies were performed in mixed solvents or in
`the presence of cyclodextrin13 The degradation
`
`to Valentino J Stella Telephone 785864
`Correspondence
`3755 Fax 7858645736 Email stellakuedu
`Journal of Pharmaceutical Sciences Vol 97 12241235 2008
`0 2007 Wiley Liss Inc and the American Pharmacists Association
`
`1224
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MAR CH 2008
`
`kinetics were mistakenly interpreted as following
`pseudo first order in some instances while some
`results indicated that
`the loss of
`experimental
`paclitaxel did not follow first order kinetics in the
`basic pH range because of
`the epimerization
`competing with the hydrolysis reactions The
`C7 hydroxyl group was one of the most accessible
`functional groups on the taxane ring structures
`and was subject
`to epimerization Epimerization
`and various other
`baccatin
`III
`of paclitaxel
`taxanes during isolation and chemistry studies
`were noted411
`This paper focuses on the kinetics of the epim
`erization of paclitaxel and related compounds
`in aqueous solution The reaction kinetics was
`determined under various pH and temperature
`to clarify the mechanism
`conditions in an attempt
`e Imuy
`
`InterScience°
`
`Abraxis EX2028
`Actavis LLC v Abraxis Bioscience LLC
`1PR201701101 1PR201701103 1PR201701104
`
`

`

`DEGRADATION OF PACLITAXEL AND RELATED COMPOUNDS
`
`1225
`
`of the epimerization which was previously unclear
`The influence of the ring structure and substi
`tuents was also investigated
`Figure 1 shows the structures of paclitaxel 1
`and several
`related compounds also used in this
`study including 7epitaxol 2 10deacetyltaxol
`3 7 epi10deacetyltaxol 4 baccatin
`III 5
`III 6 Among these
`
`and 10deacetylbaccatin
`compounds baccatin III and 10deacetylbaccatin
`represented the diterpene ring structure of
`III
`paclitaxel and 10deacetyltaxol with or without
`to the C10 position
`the acetyl group attached
`respectively
`
`1 RI = CH3C0
`R2 = OH R3 = H
`2 RI = CH3C0
`R2 = H
`R3 = OH
`3 RI = H
`R2 = OH R3 = H
`4 RI = H
`R3 = OH
`R2 = H
`
`EXPERIMENTAL
`
`Chemicals and Materials
`
`7 epi10
`Paclitaxel 7ep itaxol 10deacetyltaxol
`baccatin
`deacetyltaxol
`III
`10deacetylbaccatin
`III were generous gifts from Tapestry Pharma
`ceuticals Boulder CO Methanol
`acetonitrile
`and other organic solvents were of HPLC grade
`and were purchased from Fisher Scientific Fair
`Lawn NJ Hydrochloric acid formic acid acetic
`acid benzoic acid sodium hydroxide sodium
`acetate sodium carbonate sodium bicarbonate
`salts monobasic
`and
`and sodium phosphate
`dibasic used in preparation of buffer solutions
`from Sigma Chemicals Co St
`were obtained
`Louis MO Water used in this study was purified
`through a Millipore MILLIQTM system and was
`glass distilled before use
`
`pH
`
`The pH of the solution was controlled throughout
`the reaction by using dilute hydrochloric acid
`appropriate buffers and sodium hydroxide solu
`tions Reactions at pH 1 2 and 3 were performed
`in dilute solutions of hydrochloric acid Buffer
`solutions of pH 4 and 5 were prepared from acetic
`acid and sodium acetate Buffer solutions of pH 6
`7 and 8 were prepared with sodium phosphate
`salts monobasic and dibasic Buffer solution
`of pH 9 was prepared with sodium borate and
`boric acid Buffer solution of pH 10 was prepared
`with sodium carbonate and sodium bicarbonate
`The kinetic measurements at pH 11 and 12 were
`performed in dilute sodium hydroxide solutions
`As listed in Table 1 the buffer concentration was
`10 mM except for experiments at pH 1 2 and 12
`The pH values were measured at
`the beginning
`and the end of
`the kinetic experiments No
`significant change of pH was observed through
`the reaction In addition no buffer catalysis
`out
`was observed The pH meter was standardized
`with appropriate standard buffer solutions of
`pH 1 4 7 and 10 at
`the
`the temperatures of
`kinetic experiments
`
`5 R1=CH3CO R2 = OH R3 = H
`6 RI = H
`R2 = OH R3 = H
`Figure 1 The structures of paclitaxel and related
`paclitaxel 1 7epitaxol 2 10deacetyl
`compounds
`taxol 3 7epi10deacetyltaxol 4 baccatin III 5 10
`III 6
`
`deacetylbaccatin
`
`HPLC Analytical Assays
`
`To simultaneously detect and quantify the pre
`sence of the starting compound its epimer and
`other degradation products a stability indicating
`HPLCUV assay was developed The HPLC system
`employed in this study consisted of a Shimadzu
`
`DOI 101002jps
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`

`

`1226
`
`TIAN AND STELLA
`
`Table 1 The Composition and Measured pH of
`the Buffer Used in Kinetic Experiments of
`III
`10Deacetylbaccatin
`
`Buffer Composition
`
`HC1
`HC1
`HC1
`CH3COOHCH3COONa
`CH3COOHCH3COONa
`NaH2PO4Na2HPO4
`NaH2PO4Na2HPO4
`NaH2PO4Na2HPO4
`H3B03Na2B407
`NaHCO3Na2CO3
`NaOH
`NaOH
`
`Concentration
`
`mM
`
`Measured pH
`
`1000
`100
`10
`10
`10
`10
`10
`10
`10
`10
`10
`100
`
`109
`199
`305
`453
`518
`674
`711
`761
`904
`981
`1079
`1182
`
`SCL 10A system controller a Sil10A auto injec
`tor two LC10AT pumps and a SPD10A UV
`spectrophotometric detector The detection wave
`length was set at 230 nm for all analytes The
`detection signal was directly recorded on to a PC
`and the chromatograms were reprocessed
`by
`the CLASS VP Chromatography Data System
`program of Shimadzu Isocratic elution at ambient
`temperature was conducted on a Hypersil ODS
`C18 column Alltech with particle size of 5 µm
`and dimensions of 250 mm x 46 mm The mobile
`phase acetonitrilewater mixture with 01 vv
`formic acid was delivered at a flow rate of 10 mL
`mm The acetonitrile and the aqueous portion
`were filtered separately and then mixed The final
`mixture was degassed by sonication for 30 min
`prior to using The percentage of acetonitrile was
`varied from 25 to 50 with different com
`pounds to provide optimum separation Formic
`acid was chosen over trifluoroacetic acid because
`its compatibility with mass spectrometry A
`of
`20 µL aliquot of the sample was injected each
`time through the autosampler For paclitaxel
`7 epi10deacetyl
`7 ep i taxol
`10deacetyltaxol
`III and 10deacetylbaccatin III
`taxol baccatin
`of individual compound
`the molar concentrations
`were then calculated based on the measured peak
`areas and the standard curves of HPLC assays
`While some pure compounds were unavailable
`they were assumed
`to have the same molar
`extinction coefficients for UV absorbance as their
`corresponding epimers ie 7 epi10deacetylbac
`due to the
`catin III vs 10deacetylbaccatin III
`similarity of their chemical structures
`
`Mass Spectrometry
`
`Identification and profiling of epimerization and
`degradation products was performed using a
`Waters Alliance 2690 HPLC system connected
`to a Micromass Quattro Micro Tandem Quadruple
`mass spectrometer The instrument was also
`equipped with a Waters 2487 Dual Absorbance
`UV detector The column used was an AllTech
`Hypersil C18 column with particle size of 5 µm
`and dimensions of 250 mm x 46 mm The mobile
`phase acetonitrilewater mixture with 01 vv
`formic acid was delivered at a flow rate of 10 mL
`mm The system was operated at an electrospray
`source block temperature of 120°C a desolvation
`temperature of 350°C a cone voltage of 18 kV
`For HPLCUVMS mode the flow rate was
`0833 mLmin versus 0167 mLmin for UV and
`MS respectively The molecules undergo electron
`spray ionization in the positive ion mode The
`sample injection volume was 40 µL
`With positive ion electrospray ES+ ionization
`formation was one of the common
`mode adduct
`complexities in LCMS For some molecules being
`ions like M+ Na M+ Ic+ or
`analyzed adduct
`EM + NH4+ were abundant in the spectra
`
`Kinetic Procedures
`
`Stock solutions of 10deacetylbaccatin III bacca
`tin III paclitaxel 7epitaxol 10deacetyltaxol
`and 7epi10deacetyltaxol were prepared by accu
`rately weighing sample powder and dissolving
`them in an appropriate amount of acetonitrile
`of 125 µgmL Stock
`to obtain a concentration
`solutions used for kinetic study were made fresh
`and were stored at 4°C in a refrigerator shortly
`before using From the stock solutions working
`standards were prepared by dilution with acet
`onitrile to various concentrations
`The reaction kinetics for interconversion between
`epimers was investigated in aqueous solutions
`at pH 112 For the kinetic experiments at 25°C
`242 mL of appropriate buffer solutions were
`equilibrated in a waterbath at 250 ± 01°C and
`the epimerization was initiated by adding 08 mL
`of the stock solution 125 µgmL in acetonitrile
`into the reaction buffer This resulted in an initial
`of 20 µgmL At various
`reaction concentration
`time intervals aliquots 08 mL of the reaction
`and analyzed
`solutions were withdrawn
`by
`HPLC The reactions at basic pH were quenched
`by adding dilute hydrochloric
`adjust the pH close to 5
`
`acid solution to
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`DOI 101002jps
`
`

`

`DEGRADATION OF PACLITAXEL AND RELATED COMPOUNDS
`
`1227
`
`For the stability studies at elevated tempera
`tures at 37 50 and 70°C vials containing sample
`cont
`solutions were placed in thermostatically
`rolled ovens The reaction solutions were main
`tained at
`the desired temperature throughout
`the kinetic study Portions were removed from
`the reaction solution at appropriate intervals The
`samples were quickly cooled in ice water to quench
`the reaction and followed immediately by HPLC
`analysis
`
`RESULTS AND DISCUSSION
`
`Epimerization of 10Deacetyl
`
`Baccatin
`
`III
`
`Order of Reaction and Observed Rate Constants
`A typical HPLC chromatogram of the epimeriza
`is shown
`baccatin
`in
`tion of 10deacetyl
`III
`Figure 2 indicating that
`the starting compound
`
`LCMSMS
`
`Sepimer and the epimerization and degradation
`can be separated The formation of one
`products
`initial product with a longer retention time than
`III was observed Using
`10deacetylbaccatin
`the initial product was identified as the
`7epimer of 10deacetyl baccatin III since it had
`the same molecular weight With more reaction
`time both epimers further degraded into frag
`mented products
`Figure 3 shows typical plots of the percent resi
`dual R and Sepimers of the initial concentration
`of 10deactyl baccatin III Sepimer throughout
`at pH 108 and 50°C The time
`the reaction
`indicated that
`course
`10deacetylbaccatin
`III
`and its epimer reached a relatively fast pseudo
`equilibrium and was followed by further degrada
`the loss of the starting
`tion Obviously neither
`compound nor the formation of products
`should
`follow simple first order kinetics
`
`1
`
`Channel
`An1
`159e5
`
`Scan ES+
`TIC
`678e8
`
`CC
`
`u
`
`CLC
`
`CD
`
`1965
`
`rn
`
`1956
`
`III
`
`10deacetylbaccatin
`
`899
`
`895
`
`070804003
`100
`
`197
`
`H
`
`166
`
`0 I
`070804003
`
`f
`
`1001
`
`1
`
`158220
`
`358
`N
`
`000
`
`250
`
`500
`
`750
`
`1000
`
`1250
`
`1500
`
`1750
`
`2000
`
`2250
`
`2500
`
`2750
`
`3000
`
`3250
`
`3500
`
`Figure 2 HPLC chromatogram and MS data for
`the degradation of 10deacetyl
`III and formation of its epimer as the initial primary product
`baccatin
`in aqueous
`solution at pH 108 25°C and 5 mm reaction time
`
`DOI 101002jps
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`

`

`rate constants of 10deacetylbaccatin III and its
`epimer due to hydrolysis of ester groups This
`scheme can be described by Eqs 1 and 2
`d S = k 1S + k2 R k3 S
`d R = k S k2 R k4 R
`
`1
`
`2
`
`Solving these differential equations leads to Eqs
`
`38
`
`where
`
`and
`
`S = c1k2 exit + c2k2 ex2t
`R = ci 4 + ki +k3 exit
`± C2 X2 ± kl
`
`k3k2 eX2t
`
`01 = So 2 +
`
`+ k3
`
`Rok2
`
`Soad +
`
`+k3 + Rok2
`
`C2 =
`
`III
`
`f10deacetylbaccatin
`
`Residualpercento
`
`100
`
`L 10
`
`1228
`
`TIAN AND STELLA
`
`100
`
`80
`
`60
`
`40
`
`20
`
`00
`
`200
`
`400
`
`600
`
`800
`
`1000
`
`1200
`
`Time minute
`
`Figure 3 Time courses for the loss of Sepimer 0
`formation and loss of Repimer V and total 10dea
`III disappearance M total residual of the
`S and Repimers at pH 108 and 25°C The total
`residual of 10deacetylbaccatin III
`is plotted semilogar
`time The solid lines represent
`the
`ithmically versus
`to Eqs 911
`best
`
`cetylbaccatin
`
`fit
`
`3 4 5 6
`
`7 8
`
`X=
`
`+ k2 + k3 + k4 + ki + k3
`
`k2
`
`k42 + 4kik2
`
`2
`
`ki + k3
`
`k2
`
`k42 + 4kik2
`
`2
`
`The semilogarithmic plot of the percent
`residual
`total 10deacetylbaccatin III sum of both epimers
`versus time was linear which indicated that
`the
`overall secondary degradation pathway followed
`The disappearance
`pseudo first order kinetics
`the Sepimer and the appearance of the R
`of
`epimer were also measured at various pH values
`112 and temperatures The epimerization as
`expected was found to be reversible in neutral
`and basic pH range while no significant epimer
`ization was observed under acidic pH conditions
`In addition the overall secondary degradation
`the relatively fast epimerization follows
`after
`pseudo first order kinetics at any given constant
`pH and temperature
`From these results the concentration time
`profiles at various pH values are thought
`to be
`a consequence of the reaction scheme illustrated
`in Figure 4a where k1 is the epimerization rate
`from the Sepimer to Repimer k2 is the
`constant
`reverse reaction and k3 and k4 the degradation
`
`Due to the similarity of the chemical structures of
`the R and Sepimers it was reasonable to assume
`that they have identical rate constants for further
`hydrolytic degradation A reasonable assumption
`to k4 Therefore the following
`is to set k3 equal
`can be written for the R and
`
`kinetic equations
`Sepimers
`
`S = S°k2 + R0k2
`k1 +k2
`
`expk3t
`
`S0k1
`
`R0k2
`
`k1
`
`k2
`
`exp ki + k2 + k3t 9
`
`R =
`
`S0k1 + R0k1
`ki +k2
`
`expk3t
`
`R0k2 expki +k2 + k3t 10
`50k1
`ki + k2
`
`R + 5 = Ro + No expk3t
`
`11
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`DOI 101002jps
`
`

`

`DEGRADATION OF PACLITAXEL AND RELATED COMPOUNDS
`
`1229
`
`0
`
`OH
`
`k1
`
`k2
`
`4
`
`CH3
`
`k3
`
`I
`
`products
`
`4
`
`k4
`
`I
`
`products
`
`F
`
`k k
`
`2
`
`4
`
`major
`
`Product
`
`
`
`ProductProduct
`
`k
`
`Product
`
`Product
`
`major
`
`I13C
`
`HsC
`
`4
`
`k2
`
`OH
`
`V
`
`More product
`
`More product
`
`Figure 4 The proposed pathways for the epimerization of 10deacetylbaccatin
`baccatin III b and secondary degradation under basic pH
`
`III a
`
`This assumption was supported by the linear plots
`of the total degradation profile on semilogarithmic
`scale shown in Figure 3 The semilogarithmic
`plots should curve at the early stage of the decom
`position if either the R or Sepimer decomposes
`faster than the other unless k1 and k2 are over
`whelmingly larger than k3
`
`might be expected to undergo hydrolysis simulta
`neously and the overall degradation rate constant
`is the sum of
`the individual hydrolytic
`rate
`three ester bonds A proposed
`constants
`of all
`reaction scheme is illustrated in Figure 4b
`
`Epimerization of Baccatin III
`
`The degradation of baccatin III
`in neutral
`to basic
`pH range seemed to be more complex than that of
`III Figure 5 shows typical
`10deacetylbaccatin
`residual R and Sepimers of
`plots of the percent
`III Sepimer at pH 108 and 250°C
`baccatin
`The time course indicates an initial epimerization
`reaction to generate the Repimer
`followed by
`further degradation leading to multiple products
`The ester groups at C2 C4 and C10 positions
`
`Effect of pH on Epimerization
`
`The first order rate constants of reversible epi
`merization k1 and k2 for 10deaceylbaccatin
`III
`and baccatin III under various temperature and
`pH conditions are given in Tables 2 and 3 The
`values were obtained
`individual
`rate constant
`from the best multivariance regression of experi
`
`mental data to Eqs 911 using SigmaPlot
`program v 7101 SPSS Inc Chicago IL The
`value of k3 was obtained from the pseudo first
`loss of the drug sum of
`order kinetics of the total
`R and Sepimers The data for the loss of
`the
`
`DOI 101002jps
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`

`

`Effect of Temperature on Epimerization
`
`The epimerization of 10deacetylbaccatin
`III was
`studied at pH 761 and 25 50 and 70°C The
`values of rate constants are given in Table 4 As
`expected both forward and reverse rate constants
`were increased rapidly with increasing tempera
`ture The Sepimer is more stable at all tempera
`tures while the forward rate constant kl S to
`Repimer increased slightly faster than k2 with
`increasing temperature
`When this data was fit
`to the Eyring equation12
`some
`the plot
`although linear did suggest
`curvature Estimated values of
`the activation
`
`enthalpy Alk and activation entropy
`listed in Table 5
`
`are
`
`pH Rate Profiles for the Epimerization of
`Ill and Baccatin Ill
`1 ODeacetylbaccatin
`
`The pH dependence of the first order epimeriza
`tion rate constants kl and k2 of 10deacetylbac
`catin III and baccatin III at 250°C are shown in
`Figures 6 and 7 respectively In both figures the
`observed rate constants
`the epimerization
`for
`increased rapidly and uniformly with increasing
`pH The slopes of these straightline portions of
`log k versus pH profiles are close to unity thus it
`is likely that the reversible epimerization is base
`catalyzed
`and basic pH range base
`In near neutral
`catalyzed epimerization seems predominant com
`pared to any water catalysis In this pH range no
`significant buffer catalysis was observed There
`fore these pH rate profiles can be described by
`Eq 12
`
`kpH=kB 101H
`
`12
`
`where Kw is the dissociation constant
`and 0111
`is the activity
`of hydrogen ions
`measured by a glass electrode
`
`for water
`
`as
`
`1230
`
`TIAN AND STELLA
`
`25000
`
`20000
`
`15000
`
`10000
`
`PeakArea
`
`5000
`
`200
`
`400
`
`600
`
`800
`
`1000
`
`1200
`
`1400
`
`1600
`
`Time minute
`
`Figure 5 Time course of the degradation of baccatin
`in aqueous solution at pH 1077 T= 25°C Symbol
`
`III
`
`10deacetylbaccatin
`
`key baccatin III 0 baccatin V 7epibaccatin III 0
`III V 7epi10deacetylbaccatin
`111v unidentified product 1 a unidentified product
`20 The solid lines represent the best fit
`to the Eqs 9
`and 10 according to the measured loss of baccatin III
`0 formation and loss of baccatin V 7epibaccatin III
`
`0 s
`
`tarting Sepimer and the formation of the R
`epimer are fitted into Eqs 9 and 10 to generate
`kl and k2 values Both forward and reverse rates of
`epimerization increased rapidly with increasing
`pH consistent with an apparent base catalyzed
`reaction
`for epimerization K
`The equilibrium constant
`from the ratio kk2 Since
`was also calculated
`both 10deacetylbaccatin III and baccatin III are
`neutral molecules K values should be indepen
`dent of pH This was seen at pH values higher
`than 9 while the values determined in the lower
`pH range might bear more experimental error due
`to the slow reactions The Sepimer is the thermo
`dynamically more stable isomer
`
`Table 2 FirstOrder Rate Constants k1 and k2 and Equilibrium Constant K of Epimerization of
`III at Various pH in Aqueous Solution at 25°C as Defined in Figure 4
`10Deacetylbaccatin
`
`711
`
`761
`
`pH
`
`904
`
`k1 h1
`k2 h1
`K kilk2
`
`64 ± 10 x 104
`54 + 66 x 103
`012
`
`25 ± 02x 103
`94 + 35x 103
`027
`
`46 ± 01 x 102
`0118+ 0005
`039
`
`1079
`
`1182
`
`229 ± 036
`644 + 109
`036
`
`335 ± 13
`834 + 31
`040
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`DOI 101002jps
`
`

`

`DEGRADATION OF PACLITAXEL AND RELATED COMPOUNDS
`
`1231
`
`Table 3 First Order Rate Constants k1 and k2 and Equilibrium Constant K of
`Epimerization of Baccatin III at Various pH in Aqueous Solution at 25°C as Defined
`in Figure 4
`
`pH
`
`707
`
`768
`981
`h1 48 ± 12 x 104 27 ± 40 x 103 59 ± 04 x 102 092 ± 007 877 ± 144
`k2 h1 10 ± 97 x 102 12 ± 50 x 102 87 ± 12 x 102 134 ± 021 128 ± 322
`K k1k2
`0048
`022
`069
`068
`065
`
`1077
`
`1193
`
`Table 4 First Order Rate Constants k1 and k2 and Equilibrium Constant K of
`in Aqueous Solution at pH 761 and Various
`Epimerization of 10Deacetylbaccatin
`Temperatures
`
`III
`
`Temperature °C
`
`25
`
`50
`
`70
`
`k1111
`k2 111
`K kilk2
`
`25 + 02>< 103
`94 + 35x 103
`027
`
`0159 ± 0035
`0358 ± 0099
`044
`
`0564 ± 0162
`0978 ± 0586
`058
`
`Table 5 Temperature Dependence of Epimerization of 10Deacetylbaccatin
`Measured at pH 761
`
`III
`
`AB kcal mol1
`
`k1 of epimerization
`k2 of epimerization
`
`244 ± 48
`208 ± 45
`
`AS eu
`45 ± 09
`137 ± 32
`
`AG kcal mol1
`
`258 ± 51
`249 ± 54
`
`values
`
`In Figures 6 and 7 the solid lines represent
`the theoretical curves calculated by Eq 12 The
`second order
`the
`rate constants
`for
`from the fit are reported in Table 6
`obtained
`The results show that
`the second order base
`catalyzed rate constants
`for the epimerization of
`10deacetylbacatin III both forward and reverse
`are faster compared to that of baccatin III
`
`Epimerization of Paclitaxel and Other Compounds
`
`Figures 811 show the epimerization of paclitaxel
`7 epi10deacetyl
`7 ep i taxol 10deacetyltaxol
`taxol at pH 772 and 70°C At
`this elevated
`temperature the solubility of these compounds
`The
`was sufficient
`for quantitative
`analysis
`the total
`the
`semilogarithmic
`loss of
`plot of
`starting compound sum of both epimers versus
`reasonably linear which indicated
`time was
`that
`the overall secondary degradation followed
`pseudo first order kinetics disregarding the epi
`merization The experimental data yielded good
`fits for Eqs 9 and 10 The obtained first order
`
`rate constants
`
`for reversible epimerization k1
`
`4
`
`10
`
`11
`
`12
`
`13
`
`14
`
`pH
`
`Figure 6 pH rate profile for the values k1111 and k2
`0 used to describe the epimerization of 10deacetyl
`baccatin III at various pH values 25°C The solid line
`to Eq 12 where km = 492 ±
`represents the best
`124 ± 013 x 104 M1 h1
`082 x 103 M1 h1 k2B
`
`fit
`
`DOI 101002jps
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`

`

`fpaclitaxel
`
`Residualpercento
`
`20
`
`40
`
`60
`
`80
`
`100
`
`120
`
`140
`
`160
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`0
`
`0
`
`Time minute
`
`Figure 8 Time courses for the loss of Sepimer 0
`formation and loss of Repimer V and total paclitaxel
`total of the S and Repimers at
`disappearance
`is plotted semi
`pH 772 and 70°C The total paclitaxel
`logarithmically versus time The solid lines represent
`to Eqs 911
`the best
`
`fit
`
`substituent
`
`rate
`on the epimerization
`effect
`can be seen As listed in Tables 2 3
`constants
`and 6 10deacetylbaccatin
`III exhibits faster base
`catalyzed epimerization rates than baccatin III at
`
`f7epitaxol
`
`Residualpercento
`
`100
`
`v
`
`100
`
`80
`
`60
`
`40
`
`20
`
`CD
`
`CI0
`
`0
`
`20
`
`40
`
`60
`
`80
`
`100
`
`120
`
`10
`
`Time minute
`
`Figure 9 Time courses for the loss of Sepimer V
`formation and loss ofRepimer 0 and total 7epitaxol
`disappearance M total of the S and Repimers at pH
`
`772 and 70°C The total 7epitaxol
`is plotted semilo
`garithmically versus time The solid lines represent the
`to Eqs 911
`best
`
`fit
`
`1232
`
`TIAN AND STELLA
`
`9
`
`10
`
`11
`
`pH
`
`12
`
`13
`
`fit
`
`Figure 7 pH rate profile for the values k1111 and k2
`0 used to describe the epimerization of baccatin III
`at various pH values 25°C The solid line represents
`to Eq 12 where
`km = 105 ± 034 x
`the best
`152 ± 041 x 103 1V11 h1
`103 1V11 h1 k2B
`S to Repimer and k2 R to Sepimer for
`paclitaxel and its analogues at pH 77 and 70°C
`are given in Table 7 The rate constants obtained
`from the epimerization of the S and Repimers
`starting from either epimer are virtually the
`same The Sepimer was the thermodynamically
`more stable specie for all of the compounds at this
`that paclitaxel and
`condition The results suggest
`its derivatives follow the similar reaction mechan
`ism for 10deacetylbaccatin III and baccatin III
`described previously in Figure 4
`
`Influence of the Substituents on Epimerization
`
`By comparing the epimerization data for pacli
`taxel and its analogues which have the same
`fundamental diterpene ring skeleton but different
`the C10 and C13 position the
`substituents at
`
`Table 6 The Second Order Rate Constants for
`III and Baccatin
`Epimerization of 10Deacetylbaccatin
`III at 25°C under Basic pH Conditions
`
`10Deacetylbaccatin
`
`III
`
`Baccatin
`
`III
`
`km 141h1
`
`k 2B M1 h1
`492 ± 082 x 103 124 ± 013 x 104
`
`105 ± 034 x 103 152 ± 041 x 103
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`DOI 101002jps
`
`

`

`DEGRADATION OF PACLITAXEL AND RELATED COMPOUNDS
`
`1233
`
`Table 7 First Order Rate Constants k1 and k2 and
`Equilibrium Constant K of Epimerization of Paclitaxel
`and Related Compounds at pH 772 70°C According to
`Figure 4
`
`Starting Compound
`
`Paclitaxel
`
`7epitaxol
`
`10Deacetyltaxol
`7epi10deacetyltaxol
`
`k2 h1 K
`k1 h1
`281±101 342 ± 282
`203 ± 138 248 ± 056
`331 ± 066 593 ± 151
`296 ± 144 578 ± 192
`
`k2
`
`0822
`0821
`0558
`0512
`
`the
`
`removal of the C10 acetyl group increases
`epimerization rate in aqueous solution
`Both 10deacetylbaccatin
`III and 10deacetyl
`taxol have the identical taxane ring structure the
`latter has an additional side chain attached to the
`C13 position As seen by comparing the rate
`in Tables 4 and 7 10deacetyltaxol
`constants
`shows faster epimerization rates than 10deace
`tylbaccatin III at 70°C after the influence from the
`small pH variance is corrected This difference
`could possibly be due to the
`conformational
`the taxane ring caused by the side
`change of
`chain through hydrophobic interaction between
`the C4 acetyl group and the two phenyl groups
`This additional
`the
`interaction might distort
`taxane ring and cause a slightly higher strain
`the
`therefore increase the free energy level of
`ground states of both epimers which then leads to
`faster epimerization in both directions
`Finally the Sepimer is found to be more favor
`able than the Repimer for all of the compounds
`the 10 acetyl and
`in this study with or without
`side chain groups This observation was not
`in
`total agreement with some previously published
`information13 The 7epitaxol and other a
`epimers Repimer were determined as the more
`stable form in alcohol and other mixed solvents
`to be due to the
`This observation was thought
`strong hydrogen bonding formed between the 7a
`to the acyl oxygen of the Liu acetate first
`hydroxyl
`crystal structure
`observed
`in baccatin
`III
`However NMR studies of paclitaxel and other
`taxanes in solution indicated that some original
`intramolecular
`hydrogen bonding could be lost
`the new
`with the concomitant
`of
`appearance
`hydrophobic interactions1115 For paclitaxel and
`its derivatives investigated in this study their
`for epimerization K does
`equilibrium constant
`not show big variance from a range of 02 to 1
`under most experimental conditions That obser
`vation indicates a small
`free energy change
`between the two epimers For example a typical
`
`f10deacetyltaxol
`
`Residualpercento
`
`100
`
`10
`
`C
`
`20
`
`40
`
`80
`
`100
`
`120
`
`140
`
`Time minute
`
`Figure 10 Time courses for
`
`the loss of Sepimer
`
`0 formation and loss of Repimer 7 and total
`10deacetyltaxol disappearance M total of the S and
`
`Repimers at pH 772 and 70°C The total 10deacetyl
`versus time The
`taxol
`is plotted semilogarithmically
`to Eqs 911
`solid lines represent the best
`
`fit
`
`25°C In addition 10deacetyltaxol and its epimer
`show faster epimerization rates than paclitaxel
`and epitaxol at 70°C This is compatible with the
`results for 10deacetylbaccatin III and baccatin
`although to a smaller extent Therefore
`
`III
`
`f7epi10deacetyltaxol
`
`Residualpercento
`
`100
`
`10
`
`20
`
`40
`
`60
`
`80
`
`100
`
`120
`
`100
`
`80
`
`60
`
`40
`
`20
`
`a C
`
`Time minute
`
`Figure 11 Time courses for the loss of Sepimer 7
`formation and loss of Repimer 0 and total 7epi10
`the S and
`total of
`deacetyltaxol disappearance
`Repimers at pH 772 and 70°C The total 7epi10
`is plotted semilogarithmically versus time
`deacetyltaxol
`to Eqs 911
`The solid lines represent the best
`
`fit
`
`DOI 101002jps
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`

`

`1234
`
`TIAN AND STELLA
`
`K value of 03 or 3 is translated to a free energy
`change of approximately 05 kcal mol1 More
`that solvent polarity exhibits
`over others suggest
`a strong effect on the conformation of taxanes in
`solution16 Such a conformation change ie from
`methanol
`lead to
`to aqueous solution should
`an energy change enough to cause the shift of
`is quite reasonable to
`equilibrium Therefore it
`and its analogues
`to show
`expect paclitaxel
`various ratios of epimers in aqueous solution
`organic solvent and mixed solvents
`
`paclitaxel
`
`Mechanism of Base Catalyzed Epimerization
`A Retroaldolaldol
`intramolecular mechanism
`was proposed by McLaughlin
`et al6 as the
`mechanism of
`epimerization
`for
`Scheme 1 In this proposed mechanism the
`hydroxyl hydrogen is transferred from C7 to the
`C9 carbonyl with concomitant
`aldehyde formation
`at C7 ring cleavage
`between C7 and C8
`doublebond formation between C8 and C9 and
`at C9 in a concerted
`hence
`enol
`formation
`mechanism Free rotation of
`the single bond
`between C6 and C7 allows the electron density
`in the C8C9 alkene
`to attack
`the aldehyde
`the C7 carbon
`from either
`carbonyl
`forming either paclitaxel or 7epipaclitaxel
`This mechanism should show minimal pH
`is an intramolecular
`reaction
`dependence since it
`in a relatively closed system However a strong
`pH dependence has been observed for both 10
`III and baccatin III epimeriza
`deacetylbaccatin
`
`face of
`
`the
`tion in this study This indicated that
`epimerization more likely occurs with forming
`an enolate intermediate shown in Scheme 2 This
`mechanism can be written in terms of
`the fast
`abstraction of the proton of the 7 hydroxyl group
`by an external base The deprotonated specie then
`in the rate
`undergoes structural rearrangement
`step in which the
`single bond
`determining
`between C7 and C8 is cleaved and the enolate
`on C9 is formed simultaneously as described in
`Scheme 2 The reaction is reversible via the same
`intermediate This proposed mechanism is com
`patible to the observed base catalyzed epimeriza
`tion in high pH range and the lack of buffer
`It would still
`dependency
`concentration
`be
`considered a retroaldolaldol reaction
`This proposed mechanism implies that a sub
`stituent with stronger electron withdrawing effect
`might stabilize the deprotonated
`specie making
`the C7C8 bond easier to cleave thus increasing
`the epimerization rate However under basic pH
`III exhibits for
`condition 10deacetyl baccatin
`ward and reverse epimerization rate constants
`faster than those of baccatin III shown in Tabs 2
`and 3 even though the latter has more electron
`withdrawing influence from the C10 acetyl group
`A possible explanation is that
`the 10 hydroxyl
`group effectively stabilized the enolate intermedi
`ate by forming a hydrogen bond during the
`reaction which is impossible for baccatin
`III
`see Scheme 3 A similar comparison can be made
`between paclitaxel and 10deacetyltaxol shown in
`Tab 7 although to a smaller extent
`
`slowc II
`
`0
`
`H3
`
`OH
`
`fast
`
`OH
`
`Pi
`
`CH3
`
`fast
`
`slow
`
`Scheme 1 Mechanism of epimerization proposed by McLaughlin et al6
`
`H
`
`fast
`
`fast
`
`0
`
`cH 3
`
`slow
`
`fast
`
`c
`
`0
`CH 11
`HC
`
`alatru
`
`tast
`
`411
`
`slow
`
`CP
`
`gr
`113 7
`
`fast
`
`last
`
`Scheme 2 Proposed mechanism for base catalyzed epimerization
`
`JOURNAL OF PHARMACEUTICAL SCIENCES VOL 97 NO 3 MARCH 2008
`
`DOI 101002jps
`
`

`

`DEGRADATION OF PACLITAXEL AND RELATED COMPOUNDS
`
`1235
`
`HQ
`
`r
`
`0
`
`CH3
`
`slow
`
`fast
`
`uN
`
`cH3
`
`HC
`
`SI OW
`
`HO
`
`0
`
`cH3
`
`nnr
`
`Scheme 3
`
`Proposed mechanism for epimerization of 10deacetylbaccatin
`
`III
`
`CONCLUSIONS
`
`Paclitaxel and some related taxanes including
`7epi taxol 7epitaxol 10deacetyltaxol 7epi10
`deacetyltaxol baccatin III and 10deacetylbacca
`epimerize reversibly under basic and
`tin III
`neutral pH conditions in aqueous solution Both
`S and Repimers further hydrolyze while no
`other reaction such as dehydration of hydroxyl
`group or hydrolytic opening of the oxetane ring
`was observed under base conditions Removal of
`the C10 acetyl group increases the epimerization
`rate in basic aqueous solutions The observe