Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`Cancer Therapy: Preclinical
`
`Clinical
`Cancer
`Research
`
`Preclinical Antitumor Activity of Cabazitaxel, a Semisynthetic
`Taxane Active in Taxane-Resistant Tumors
`
`Patricia Vrignaud, Dorothee Semiond, Pascale Lejeune, Herve Bouchard, Loreley Calvet, Cecile Combeau,
`Jean-Fran¸cois Riou, Alain Commer¸con, Fran¸cois Lavelle, and Marie-Christine Bissery
`
`Abstract
`
`Purpose: Taxanes are important chemotherapeutic agents with proven efficacy in human cancers, but
`their use is limited by resistance development. We report here the preclinical characteristics of cabazitaxel
`(XRP6258), a semisynthetic taxane developed to overcome taxane resistance.
`Experimental Design: Cabazitaxel effects on purified tubulin and on taxane-sensitive or chemotherapy-
`resistant tumor cells were evaluated in vitro. Antitumor activity and pharmacokinetics of intravenously
`administered cabazitaxel were assessed in tumor-bearing mice.
`Results: In vitro, cabazitaxel stabilized microtubules as effectively as docetaxel but was 10-fold more
`potent than docetaxel in chemotherapy-resistant tumor cells (IC50 ranges: cabazitaxel, 0.013–0.414 mmol/L;
`docetaxel, 0.17–4.01 mmol/L). The active concentrations of cabazitaxel in these cell lines were achieved
`easily and maintained for up to 96 hours in the tumors of mice bearing MA16/C tumors treated with
`cabazitaxel at 40 mg/kg. Cabazitaxel exhibited antitumor efficacy in a broad spectrum of murine and human
`tumors (melanoma B16, colon C51, C38, HCT 116, and HT-29, mammary MA17/A and MA16/C, pancreas
`P03 and MIA PaCa-2, prostate DU 145, lung A549 and NCI-H460, gastric N87, head and neck SR475, and
`kidney Caki-1). Of particular note, cabazitaxel was active in tumors poorly sensitive or innately resistant to
`docetaxel (Lewis lung, pancreas P02, colon HCT-8, gastric GXF-209, mammary UISO BCA-1) or with
`acquired docetaxel resistance (melanoma B16/TXT).
`Conclusions: Cabazitaxel is as active as docetaxel in docetaxel-sensitive tumor models but is more potent
`than docetaxel in tumor models with innate or acquired resistance to taxanes and other chemotherapies. These
`studies were the basis for subsequent clinical evaluation. Clin Cancer Res; 19(11); 2973–83. Ó2013 AACR.
`
`Introduction
`Microtubules are highly dynamic cytoskeletal fibers com-
`posed of 2 tubulin subunits (a and b). The polymerization
`and depolymerization of these molecules are crucial pro-
`cesses, not only to mitosis but also to intracellular traffick-
`ing. Microtubules are the main target of taxanes, which bind
`to a specific binding site on the tubulin b-subunit (1, 2). The
`taxanes paclitaxel and docetaxel suppress microtubule
`dynamics by promoting tubulin assembly and stabilizing
`microtubules (3), blocking mitosis at the metaphase/ana-
`phase transition, which results in cell death (ref. 4; Supple-
`mentary Fig. S1A and S1B). By stabilizing microtubules,
`taxanes also impact intracellular trafficking. This was recent-
`ly reported as one of the main mechanisms of taxane action
`
`Authors' Affiliation: Sanofi, Vitry-sur-Seine, France
`
`Note: Supplementary data for this article are available at Clinical Cancer
`Research Online (http://clincancerres.aacrjournals.org/).
`
`Corresponding Author: Patricia Vrignaud, Sanofi Oncology, 13 quai Jules
`Guesde, Vitry-sur-Seine 94403, France. Phone: 33-1-58-93-36-29; Fax:
`33-1-58-93-34-71; E-mail: patricia.vrignaud@sanofi.com
`
`doi: 10.1158/1078-0432.CCR-12-3146
`Ó2013 American Association for Cancer Research.
`
`in prostate cancer, where taxanes were shown to inhibit
`nuclear translocation of the androgen receptor, thereby
`preventing androgen receptor transcriptional activity and
`leading to prostate cancer cell death (5).
`Paclitaxel and docetaxel form the backbone of both first-
`line and salvage chemotherapy regimens for patients with a
`wide variety of tumor types. Paclitaxel is indicated for first-
`line treatment of ovarian, breast, and lung cancer and for
`second-line treatment of AIDS-related Kaposi’s sarcoma (6).
`Docetaxel is indicated for first-line treatment of breast, head
`and neck, gastric, lung, and prostate cancer and for second-
`line treatment of breast cancer (7). However, the use of both
`paclitaxel and docetaxel is limited by the development of
`tumor resistance (8–10). During the last 2 decades, con-
`siderable efforts have been made to understand, and devel-
`op new agents to overcome, taxane resistance.
`Cabazitaxel (RPR 116258; XRP6258; TXD258; Jevtana)
`is a new semisynthetic taxane derived from 10-deacetyl-
`baccatin III, which is extracted from European yew needles
`(ref. 11; Supplementary Fig. S1C). Cabazitaxel was identi-
`fied using a 3-step screening process, assessing activity
`against microtubule stabilization, in vitro activity in resistant
`cell lines, and in vivo activity in a tumor model in which
`docetaxel resistance had been induced in vivo. This article
`
`www.aacrjournals.org
`
`2973
`
`Downloaded from
`
`
`
`on January 8, 2014. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org
`
`AVENTIS EXHIBIT 2033
`Mylan v. Aventis
`IPR2016-00627
`
`

`
`Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`Vrignaud et al.
`
`Translational Relevance
`Mechanisms of resistance to taxanes in patients have
`not been fully elucidated. In cell lines, overexpression of
`ATP-binding transporters, particularly P-glycoprotein,
`and alteration of microtubule dynamics are the most
`common mechanisms of taxane resistance. However,
`clinical data suggest that other mechanisms, including
`dysfunctional regulation of apoptotic and intracellular
`signaling, may operate in tumors escaping taxane ther-
`apy. To identify a docetaxel derivative with activity after
`taxane failure, we developed a clinically relevant doc-
`etaxel-resistant tumor model, mimicking tumor resis-
`tance development in patients who initially respond to
`docetaxel, but develop resistance over time. Cabazitaxel
`was selected from 450 derivatives based on activity in
`this model. Clinical proof-of-principle was achieved in a
`phase II study in patients with taxane-resistant metastatic
`breast cancer and a phase III study in metastatic hor-
`mone-refractory prostate cancer post-docetaxel therapy.
`The current study extends the characterization of caba-
`zitaxel, showing wide ranging in vitro and in vivo anti-
`tumor activity.
`
`describes the development and characterization of the in
`vivo–induced docetaxel-resistant tumor model, the mecha-
`nism of action of cabazitaxel on microtubules, and its
`preclinical evaluation in a wide range of taxane-sensitive
`and -resistant cell lines, both in vitro and in vivo.
`
`Materials and Methods
`Tubulin polymerization
`The effects of cabazitaxel on tubulin polymerization and
`cold-induced microtubule depolymerization were evaluat-
`ed using tubulin purified from porcine brain (12, 13).
`Tubulin was used at a concentration of 6 mmol/L for
`
`C) and 9 mmol/L for depo-
`polymerization assays (at 37
`
`lymerization assays (at 8
`C). Rates of polymerization/depo-
`lymerization were measured by optical density (OD) at
`350 nm and were expressed in DOD/min. Upper and lower
`limits for drug concentrations reducing polymerization lag
`time by 50% (LT50) and inhibiting cold-induced disassem-
`bly by 50% (dIC50) were determined.
`
`Microtubule and enzymatic parameters in tumors
`Microtubule parameters in B16 and B16/TXT tumors
`were characterized using real-time PCR (RT-PCR) analysis
`of at least 2 samples per tumor. PCR values in arbitrary units
`were obtained for the following genes: total a-tubulin
`(TUBA), total b-tubulin (TUBB), TUBB2, TUBB3, TUBB4A,
`TUBB4B, and TUBB7P.
`Glutathione S-transferase (GST) activity was assayed as
`previously described using 1-chloro-2,4-dinitrobenzene
`(CDNB) as the substrate (14). Formation of the CDNB
`glutathione (GSH) conjugate by cytosols was measured
`
`continuously in a spectrophotometer at 340 nm. The results
`were expressed as the quantity of CDNB conjugated per
`minute per milligram of cytosolic protein (nmol/min/mg).
`Total GSH concentration was determined as the sum of
`the reduced (GSH) and oxidized (GSSG) forms of GSH
`(15). In this assay, the sum of the reduced and oxidized
`forms of GSH is determined using a kinetic assay in which
`catalytic amounts of GSH or GSSG and GSH reductase bring
`0
`about the continuous reduction of 5,5
`-dithiobis(2-nitro-
`benzoic acid) by NADPH. The reaction rate is proportional
`to the concentration of GSH below 2 mmol/L. The forma-
`tion of 5-thio-2-nitrobenzoate was analyzed using a spec-
`trophotometer at 412 nm. The results were expressed as
`concentration per milligram of protein (nmol/mg).
`Cytochrome P450 3A (CYP3A) levels were determined
`using the Amersham ELISA system (code RPN 271; Amer-
`sham). This assay uses a rabbit primary antibody specific for
`rat CYP3A, a secondary conjugate of anti-rabbit immuno-
`globulin (Ig) and horseradish peroxidase antibody, and
`tetramethylbenzidine substrate. The horseradish peroxi-
`dase color that develops is proportional to CYP3A levels.
`This assay was validated against mouse CYP3A by the
`manufacturer. Protein concentrations of microsomes, cyto-
`sols, and homogenates were determined by the bicincho-
`ninic acid assay (16) using a commercial preparation
`(Pierce BCA Protein Assay Reagent).
`
`In vitro antiproliferative activity
`The HL60/TAX cell line (17) was a kind gift from Dr. K.
`Bhalla (Medical University of South Carolina, Charleston,
`SC). Calc18/TXT and P388/TXT were developed internally
`from Calc18 or P388 parental cell lines. The P388/TXT cell
`line was selected by mutagenesis with ethyl methane sul-
`fonate and soft agar cloning in the presence of 0.06 mmol/L
`docetaxel. The Calc18/TXT cell line was established by 6-
`month exposure to increasing concentrations of docetaxel
`(up to 0.019 mmol/L). The cross-resistance pattern of these 2
`cell lines is shown in Supplementary Table S1. The other
`tumor cell lines were obtained from the National Cancer
`Institute (NCI; Bethesda, MD).
`Parental and resistant tumor cells were incubated with
`
`different drug concentrations for 96 hours at 37
`C; cell
`viability was measured in quadruplicate using neutral red
`uptake (18). The resistance factor for each drug was calcu-
`lated by dividing the IC50 in resistant cells by the mean IC50
`in sensitive/parental cells, using data from at least 3 inde-
`pendent experiments. Relative expression of ABCB1 mRNA
`was determined by Northern blotting using a human
`ABCB1 gene probe.
`
`Antitumor activity in tumor-bearing mice
`All experimental procedures were approved by Sanofi,
`Southern Research Institute (SRI; Birmingham, AL), and
`Molecular Imaging Research (MIR) Preclinical Services Lab-
`oratory Animal Care and Use committees. Protocol design,
`chemotherapy techniques, and methods of data analysis
`have been described previously (19–21). Briefly, tumors
`were implanted subcutaneously and bilaterally on day 0.
`
`2974
`
`Clin Cancer Res; 19(11) June 1, 2013
`
`Clinical Cancer Research
`
`
`
`Downloaded from on January 8, 2014. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org
`
`
`
`

`
`Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`Animals were randomly assigned to treatment (T) or control
`(C) groups. Tumors were measured using a caliper 2 to 5
`times weekly (according to tumor growth rate) until the
`tumor reached 2,000 mm3. Tumor volumes were estimated
`from 2-dimensional measurements using the formula:
`tumor volume (mm3) ¼ [length (mm)  width2
`(mm2)]/2. The day of death was recorded, and thoracic
`and abdominal cavities were examined macroscopically to
`assess probable cause of death.
`Mice. C57BL/6, B6D2F1 and Swiss nude mice were bred
`at Iffa Credo; C3H/HeN, BALB/c, BALB/c nude, and severe
`combined immunodeficient (SCID) mice were bred at
`Charles River; and ICR and NCR nude mice were bred at
`Taconic. All mice weighed more than 18 g at the start of
`treatment and had free access to food and water.
`Drugs. Cabazitaxel (RPR 116258; XRP6258; TXD258;
`Jevtana) and docetaxel (RP 56976) were prepared by mixing
`1 volume of ethanol stock solution, 1 volume of polysor-
`bate 80, and 18 volumes of 5% glucose in sterile water.
`Solutions were administered intravenously as a slow bolus
`(0.4 mL/mouse). Drug doses were adjusted on the basis of
`body weight at start of treatment. For cytotoxic compounds,
`such as docetaxel and cabazitaxel, a dose–response evalu-
`ation was conducted in each trial to determine the highest
`nontoxic dose (HNTD), defined as the highest drug dose
`inducing less than 20% body weight loss with no drug-
`related deaths. Animal body weights included the tumor
`weights.
`Tumor models. Murine tumors were obtained from Dr.
`Corbett (Wayne State University, Detroit, MI) and included
`colon C51 and C38 (19), pancreas P02 and P03 (22),
`mammary MA17/A and MA16/C (23), Lewis lung (24),
`and melanoma B16 (20). Tumors were maintained by serial
`passage in the mouse strain of origin. B16/TXT was isolated
`by treating C57BL/6 mice bearing docetaxel-sensitive B16
`melanoma at the HNTD of docetaxel (60 mg/kg) for 27
`passages, until the B16 tumor acquired full resistance to
`docetaxel. Human tumor cell lines were obtained from
`American Type Culture Collection and included prostate
`DU 145 (25), lung NCI-H460 (26) and A549 (27), pancreas
`MIA PaCa2 (28), and colon HT-29 (29), HCT 116 (30), and
`HCT-8 (31). Mammary UISO BCA-1 (32) and gastric GXF-
`209 (33) tumors were obtained from SRI and gastric N87
`(34) tumors from MIR Preclinical Services. Murine tumors
`were grafted into syngenic mice and human tumors were
`xenografted into immunocompromised mice.
`Plasma pharmacokinetics and tumor distribution. Caba-
`zitaxel concentrations in plasma and tumor tissue were
`evaluated in mice bearing advanced-stage (400 mm3)
`murine mammary adenocarcinoma MA16/C after admin-
`istration of the HNTD of cabazitaxel (40 mg/kg). Mice were
`treated on day 8 after subcutaneous tumor implantation
`with a single 45-second intravenous infusion of cabazitaxel
`in a polysorbate 80/ethanol/5% glucose solution, with a
`dosing volume of 25 mL/kg and a rate of infusion of
`1 mL/min. Blood and tumor samples were collected from
`3 animals per sampling time at 2, 5, and 15 minutes and 2,
`4, 8, 12, 24, 48, 96, and 168 hours after cabazitaxel treat-
`
`Preclinical Antitumor Activity of Cabazitaxel
`
`concentrations were analyzed by
`ment. Cabazitaxel
`liquid chromatography/tandem mass spectrometry, with
`limits of quantification of 2.5 ng/mL in plasma and 25
`ng/g in tumor tissue. Pharmacokinetic parameters were
`determined using WinNonLin software, Version 1.0 (Sci-
`entific Consulting Inc.), using a noncompartmental infu-
`sion model.
`Several endpoints
`Assessments of antitumor activity.
`were used. Tumor growth delay (TC) was defined as the
`difference between tumors in the T and C groups in the
`median time (days) to reach a predetermined volume (750–
`1,000 mm3). Tumor doubling time (Td) in days was esti-
`mated from log linear tumor growth during the exponential
`phase (range, 100–1,000 mm3). Log cell kill was calculated
`using the formula (TC)/(3.32  Td), with antitumor
`activity defined as a log cell kill value  0.7 (21). SRI score
`was used to categorize antitumor activity based on log cell
`kill values as follows: <0.7¼ (inactive); 0.7–1.2¼þ; 1.3–
`1.9 ¼ þþ; 2.0–2.8 ¼ þþþ; > 2.8 ¼ þþþþ (highly active).
`Complete tumor regression (CR) was defined as tumor
`regression below the limit of palpation (62 mm3). Animals
`without palpable tumors at the end of the study were
`declared tumor-free survivors (TFS) and were excluded from
`the TC value calculation.
`Statistical analysis was conducted using either a pairwise
`Wilcoxon rank-sum test, with P value adjustment by the
`Holm method (N87 study), or by log-rank multiple com-
`parisons test versus control (with Bonferroni–Holm correc-
`tion for multiplicity) on individual values for time to reach a
`prespecified tumor size for treated and control groups
`(UISO BCA-1 study). A P value of less than 5% (P <
`0.05) was considered significant.
`
`Results
`Isolation and characterization of B16/TXT, a docetaxel-
`resistant melanoma
`To identify taxane derivatives with activity following
`taxane failure, a docetaxel-resistant tumor model (B16/
`TXT) was developed to mimic the gradual development of
`resistance to docetaxel observed in some patients following
`an initial tumor response to the agent. Mice bearing the
`sensitive murine B16 melanoma were treated with docetaxel
`at the HNTD (60 mg/kg per passage; log cell kill 1.7; Table 1).
`Resistance occurred very slowly, with 27 passages over 17
`months needed to obtain a fully docetaxel-resistant tumor
`(log cell kill < 0.7). B16/TXT was found to have similar Td
`(1.3–2 days) and histologic characteristics to the parental
`B16 tumor. Cross-resistance (no antitumor activity) was
`observed to the tubulin-binding drugs paclitaxel, vincris-
`tine, and vinblastine, but not to cyclophosphamide (log cell
`kill 2.9 in B16 vs. 3.0 in B16/TXT), CCNU (log cell kill 3.7 in
`B16 vs. 4.7 in B16/TXT), and etoposide (log cell kill 1.2 in
`both). B16/TXT was partially cross-resistant to doxorubicin
`(log cell kill 2.4 in B16 vs. 0.9 in B16/TXT). There was no
`difference between the docetaxel-sensitive and -resistant
`B16 tumors either in factors involved in drug resistance,
`such as GST activity (B16, 0.42  0.03 mmol/min/mg
`
`www.aacrjournals.org
`
`Clin Cancer Res; 19(11) June 1, 2013
`
`2975
`
`
`
`Downloaded from on January 8, 2014. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org
`
`
`
`

`
`Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`protein; B16/TXT, 0.39  0.04 mmol/min/mg protein) and
`GSH content (B16, 21.7  8.1 mmol/mg protein; B16/TXT,
`21.2  2.5 mmol/mg protein), or in activity of CYP3A,
`involved in TXT metabolism (B16, 2.4  0.8 mg/mg protein;
`B16/TXT, 2.9  0.06 mg/mg protein). Moreover, no over-
`expression of P-glycoprotein was found in B16/TXT, either
`by flow cytometry or Western blot analyses (data not
`shown). Analyses by RT-PCR of microtubule components
`revealed that B16/TXT expressed 3.13-fold higher levels of
`TUBB3 than the docetaxel-sensitive parental B16 tumor,
`whereas levels of other microtubule parameters were similar
`(Supplementary Table S2). As noted earlier, this model was
`pivotal in the selection of cabazitaxel, the characteristics of
`which are described hereafter.
`
`Microtubule stabilization
`Cabazitaxel had similar efficiency compared with doce-
`
`taxel for reducing the lag time for tubulin assembly (LT50 ¼
`0–0.1 mmol/L for both) and the rate of cold-induced micro-
`tubule depolymerization (dIC50 ¼ 0.1–0.25 mmol/L for
`both) in vitro (Table 2).
`
`In vitro antiproliferative activity in chemotherapy-
`sensitive and -resistant cell lines
`Cabazitaxel showed similar antiproliferative activity
`compared with docetaxel in cell lines sensitive to chemo-
`therapy (murine leukemia P388, human tumor HL60 and
`KB, and breast Calc18), as shown by the similar IC50 ranges
`across different cell types (cabazitaxel, 0.004–0.041 mmol/L;
`docetaxel, 0.008–0.079 mmol/L; Table 3). In P-glycopro-
`tein–expressing cell lines with in vitro–acquired resistance to
`taxanes (P388/TXT, Calc18/TXT, and HL60/TAX) or to
`other chemotherapy agents (P388/DOX, P388/VCR, and
`KBV1), cabazitaxel was found to be more active than doc-
`etaxel (IC50 ranges: cabazitaxel, 0.013–0.414 mmol/L; doc-
`etaxel, 0.17–4.01 mmol/L). Resistance factors (an indication
`of the difference in drug concentrations needed to inhibit
`resistant vs. sensitive/parental cell lines) were 2 to 10 for
`cabazitaxel and 5 to 59 for docetaxel. Cell lines expressing
`moderate levels of P-glycoprotein (P388/TXT, P388/VCR,
`HL60/TAX, and Calc18/TXT), which may be more clinically
`representative, had minimal cross-resistance to cabazitaxel
`(resistance factors ¼ 2–4).
`
`Plasma pharmacokinetics and drug distribution in
`tumors
`The pharmacokinetic profile of cabazitaxel was evaluated
`in mice bearing docetaxel-sensitive murine mammary
`MA16/C adenocarcinoma tumors. Cabazitaxel was highly
`active in this tumor model, inducing CRs in 80% of mice and
`having a log cell kill of 3.7 at the HNTD of 40 mg/kg (Table
`1). This antitumor activity was consistent with drug uptake
`into the tumor, which was both rapid (maximum drug
`concentrations were reached 15 minutes after dosing) and
`sustained (at 48 hours post-dose, cabazitaxel concentrations
`were 40-fold higher in the tumor vs. plasma; Fig. 1). Ratios of
`cabazitaxel exposure in tumors versus plasma were 1.6 from
`0 to 48 hours and 2.9 over the entire experimental period.
`
`Vrignaud et al.
`
`bDefinitionofantitumoractivity:logcellkilltotal<0.7¼inactive;>2.8¼highlyactive.
`aMediantumorburdenatstartoftherapy:290,310,and400mm3forC38,P03,andMA16/Cstudies,respectively.
`Abbreviations:N/A,notavailableastreatmentconductedonearly-stagedisease;ND,notdeterminedinthesamestudy.
`(B16,B16/TXT,C38,P03,P02,and3LL).
`NOTE:MurinetumorsweregraftedinthesyngenicstrainofmiceoforiginofthetumorforMA17/A(C3H/HeN)andC51(BALB/c)andinB6D2F1micefortheC57BL/6syngenictumors
`
`ND
`ND
`ND
`ND
`ND
`0/5
`0/5
`0/5
`0/5
`
`TFS
`
`ND
`ND
`ND
`ND
`ND
`N/A
`0/5
`N/A
`N/A
`
`CR
`
`ND
`ND
`ND
`ND
`ND
`3.1
`3.1
`0.6
`1.7
`
`killb
`Logcell
`
`Docetaxel
`
`ND
`ND
`ND
`ND
`ND
`31.1
`29.1
`2.8
`6.6
`
`days
`TC,
`
`ND
`ND
`ND
`ND
`ND
`45
`60
`60
`60
`
`mg/kg
`HNTD,
`Total
`
`0/5
`0/5
`0/5
`0/5
`4/5
`0/5
`5/5
`0/5
`0/5
`
`TFS
`
`N/A
`N/A
`4/5
`N/A
`5/5
`N/A
`5/5
`N/A
`N/A
`
`CR
`
`1.2
`3.9
`3.7
`0.8
`
`—
`
`2.6
`
`—
`
`1.3
`2.1
`
`cellkillb
`log
`
`4.6
`15.7
`13.4
`6.6
`
`—
`
`25.8
`
`—
`
`5.5
`8.5
`
`days
`TC,
`
`Cabazitaxel
`
`58.5
`36
`40
`60
`60
`45
`60
`60
`60
`
`31.5,19.5,12.1,7.5(3,5,7)
`19.4,12,7.4,4.6(3,5,7)
`64.5,40,24.8,15.4(8a)
`32.2,20,12.4,7.7(3,5,7)
`32.2,20,12.4,7.7(17a,19,21)
`24.2,15,9.3,5.8(4,6,8)
`32.2,20,12.4,7.7(14a,17,20)
`32.2,20,12.4,7.7(3,5,7)
`32.2,20,12.4,7.7(3,5,7)
`
`mg/kg
`TotalHNTD,
`
`injection(scheduledays)
`Dose,mg/kgper
`
`Lung3LL(Td:1.2d)
`MammaryMA17/A(Td:1.2d)
`MammaryMA16/C(Td:1.1d)
`PancreasP02(Td:2.5d)
`PancreasP03(Td:3.5d)
`ColonC51(Td:3d)
`ColonC38(Td:2.8d)
`B16/TXTmelanoma(Td:1.3d)
`B16melanoma(Td:1.2d)
`Tumortype
`
`Table1.Dose–responseantitumoractivityofcabazitaxelanddocetaxelinmicebearingmurinetumors
`
`2976
`
`Clin Cancer Res; 19(11) June 1, 2013
`
`Clinical Cancer Research
`
`
`
`Downloaded from on January 8, 2014. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org
`
`
`
`

`
`Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`Preclinical Antitumor Activity of Cabazitaxel
`
`Table 2. Effects of taxoids on the assembly–disassembly process of pure tubulin
`
`Rate of cold-induced microtubule
`disassembly, DOD/min
`
`Lag time, min
`
`Docetaxel
`Cabazitaxel
`2 (n ¼ 5)
`7.42  10
`5.36  10
`2 (n ¼ 4)
`
`6  102 (n ¼ 3)
`0.89
`2 (n ¼ 5)
`
`3.05  102 (n ¼ 5)
`2.82  10
`0.92
`1.8  10
`2 (n ¼ 6)
`
`1.95  102 (n ¼ 5)
`0.92
`1.2  10
`2 (n ¼ 5)
`
`1.25  102 (n ¼ 4)
`0.96
`2 (n ¼ 4)
`
`0.5  102 (n ¼ 6)
`0.61  10
`1.22
`0.56  10
`2 (n ¼ 3)
`
`0.38  102 (n ¼ 3)
`1.47
`LT50: 0–0.1 mmol/L
`LT50: 0–0.1 mmol/L
`dIC50: 0.1–0.25 mmol/L
`dIC50: 0.1–0.25 mmol/L Mean: 1.06
`NOTE: Tubulin was used at a concentration of 6 mmol/L for polymerization (at 37C) and 9 mmol/L for depolymerization (at 8C). OD was
`measured at 350 nm. Rates of depolymerization were expressed in DOD/min. Ratios between depolymerization rates were calculated
`for each drug concentration. Boundaries of drug concentrations for dIC50 and LT50 are given.
`
`Drug concentration,
`mmol/L
`Control
`0.1
`0.25
`0.5
`1
`2.5
`5
`
`Ratio
`
`Cabazitaxel
`
`Docetaxel
`
`40
`
`4.18
`1.57
`1.0
`0.5
`
`9.16
`2.61
`1.04
`0.65
`
`Cabazitaxel concentrations were maintained above the range
`of cellular antiproliferative IC50 values [0.004–0.041 mmol/L
`(see Table 3), corresponding to 3–29 ng/mL, 4-day exposure]
`for 24 hours in plasma and 96 hours in the tumor.
`
`Schedule of administration
`The optimal schedule of cabazitaxel administration in
`vivo was initially determined by assessing the total dose that
`could be injected without undue toxicity for different sche-
`
`dules in nontumor-bearing B6D2F1 female mice (Supple-
`mentary Table S3). Three schedules of intravenous cabazi-
`taxel were administered: intermittent [days 1 and 5 (A1)],
`daily [days 1–5 (A2)] and split-dose [days 1–5, 3 times daily
`(A3)]. HNTDs were 58 mg/kg (A1), 29 mg/kg (A2), and 12
`mg/kg (A3), suggesting a trend for schedule dependency.
`These results indicate that, compared with intermittent
`treatment (A1) of the same duration, the daily (A2) and
`split-dose (A3) schedules require 2-fold and 4.8-fold dose
`
`Table 3. In vitro antiproliferative effects of cabazitaxel and docetaxel against sensitive and P-glycoprotein–
`expressing resistant cell lines
`
`Mean IC50, mmol/L,  SD
`
`Resistance factora
`
`ABCB1 mRNA levelb
`Cabazitaxel
`Docetaxel
`Docetaxel
`Cell line
`Cabazitaxel
`
`0.041  0.017
`0.079  0.004
`
`P388 murine leukemia
`þþþ
`0.414  0.036
`4.01  0.28
`51
`P388/DOX
`10
`
`0.013  0.005
`0.039  0.012
`
`P388 murine leukemia
`þþ
`0.024  0.015
`0.188  0.022
`5
`P388/TXT
`2
`
`0.013  0.005
`0.039  0.012
`
`P388 murine leukemia
`þþ
`0.024  0.003
`0.227  0.038
`6
`P388/VCR
`2
`
`0.022  0.010
`0.031  0.004
`
`HL60 human leukemia
`þþ
`0.060  0.029
`0.25  0.11
`8
`HL60/TAX
`3
`
`0.004  0.002
`0.008  0.002
`
`Calc18 human breast adenocarcinoma
`þþ
`0.016  0.004
`0.17  0.04
`21
`Calc18/TXT
`4
`
`0.035  0.026
`0.042  0.0212
`
`KB human epidermoid carcinoma
`þþþþ
`0.27  0.013
`2.48  0.12
`59
`KB V1
`8
`NOTE: Cells were incubated for 96 hours at 37C in liquid medium with drugs at different concentrations. Viability was assessed by
`neutral red, with the mean of at least 3 results obtained.
`Abbreviations: ABCB1, ATP-binding cassette, sub-family B, member 1; Calc18/TXT, Calc18 human breast adenocarcinoma resistant
`to docetaxel; HL60/TAX, HL60 human leukemia resistant to paclitaxel; KB V1, KB human epidermoid carcinoma resistant to vinblastine;
`P388/DOX, P388 murine leukemia resistant to doxorubicin; P388/TXT, P388 murine leukemia resistant to docetaxel; P388/VCR, P388
`murine leukemia resistant to vincristine.
`aResistance factor ¼ IC50 (resistant)/IC50 (parental) from the same experiment.
`bRelative expression obtained from Northern blot experiments using the human ABCB1 gene as probe.
`
`www.aacrjournals.org
`
`Clin Cancer Res; 19(11) June 1, 2013
`
`2977
`
`
`
`Downloaded from on January 8, 2014. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org
`
`
`
`

`
`Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`Plasma
`Tumor
`
`Figure 1. Pharmacokinetics of
`cabazitaxel in plasma and tumor
`tissue in mice. Cabazitaxel
`concentrations in plasma and
`tumor tissue were measured after a
`single intravenous dose of
`cabazitaxel at its HNTD of 40
`mg/kg in C3H/HeN female mice
`bearing advanced-stage
`(400 mm3) mammary
`adenocarcinoma MA16/C. Mean
`concentration  SD from 3 animals
`was determined at each sampling
`time.
`
`Vrignaud et al.
`
`105
`
`104
`
`103
`
`102
`
`101
`
`100
`
`Cabazitaxel concentration (ng/mL or ng/g)
`
`0
`
`10
`
`20
`
`30
`
`40
`
`50
`Time (h)
`
`60
`
`70
`
`80
`
`90
`
`100
`
`reductions, respectively. In addition, host recovery time
`(time from last treatment to recovery of initial body weight)
`was shorter with the intermittent schedule (A1; 8 days)
`compared with the daily (A2; 12 days) and split-dose (A3;
`15 days) schedules. Results of these dose-scheduling studies
`were consistent with further evaluations conducted in C3H/
`HeN mice bearing mammary adenocarcinoma MA17/A
`tumors, which showed a 4.5-fold lower HNTD for a
`split-dose schedule (days 3–7, 3 times daily; HNTD 42
`mg/kg) compared with an intermittent schedule (days 3
`and 7; HNTD 9.3 mg/kg; Supplementary Table S4). Caba-
`zitaxel antitumor activity against MA17/A tumors was also
`lower with the split-dose schedule compared with the inter-
`mittent schedule (log cell kill 1.3 vs. 4.6, respectively). These
`differences between split-dose and intermittent schedules
`were confirmed in a second tumor model, murine mam-
`mary adenocarcinoma MA16/C, which showed a 4.2-fold
`reduction in HNTD (11 vs. 46 mg/kg, respectively) and
`decreased antitumor activity (log cell kill 1.6 vs. 5.4, respec-
`tively). As the intermittent schedule allowed the highest drug
`dose to be administered, with the best host recovery, and had
`the greatest antitumor activity of the schedules tested, inter-
`mittent dosing was selected for further evaluation.
`
`Antitumor activity in docetaxel-sensitive tumors
`The antitumor properties of cabazitaxel in vivo were eval-
`uated using murine tumors grafted in syngenic mice and
`human tumors xenografted in immunocompromised mice
`(Tables 1 and 4). Cabazitaxel was found to be very active
`against murine B16 melanoma (log cell kill 2.1 at HNTD of
`20 mg/kg per injection on days 3, 5, and 7), murine colon
`adenocarcinoma C51 (log cell kill 2.6 at HNTD of 9.3 mg/kg
`per injection on days 4, 6, and 8), and mammary adenocarci-
`nomas MA16/C and MA17/A (reported earlier). Cabazitaxel
`
`was highly active, with CRs observed against the advanced-
`stage murine tumors colon C38 (5 of 5 CR; 5 of 5 TFS) and
`pancreas P03 (5 of 5 CR; 4 of 5 TFS) as well as human tumor
`xenografts including: prostate DU 145 (6 of 6 CR; 5 of 6 TFS);
`colon HCT 116 (7 of 7 CR; 2 of 7 TFS) and HT-29 (6 of 6 CR);
`pancreas MIA PaCa2 (6 of 6 CR; 6 of 6 TFS); breast Calc18
`(5 of 8 TFS); lung NCI-H460 (2 of 6 CR) and A549 (2 of 6 CR);
`head and neck SR475 (6 of 6 CR; 6 of 6 TFS); and kidney Caki-
`1 (5 of 6 CRs). In most of the above models, cabazitaxel and
`docetaxel exhibited similar antitumor activity.
`Dose–response effects were examined in the advanced
`human gastric carcinoma N87 model, a tumor expressing
`HER2 (ref. 34; Fig. 2A). At the HNTD (24.4 mg/kg on days
`27, 31, and 35), cabazitaxel was highly active and delayed
`tumor growth by 101 days (log cell kill > 6, 1 of 8 TFS; P <
`0.0001). The 2 dose levels below the HNTD (15 and 9.3 mg/
`kg per injection) also had a high level of antitumor activity
`(4.5 and 2.5 log cell kill; P < 0.0001 and 0.091, respectively).
`In comparison, docetaxel was also highly active at the
`HNTD, but to a lesser extent (TC ¼ 67 days; log cell kill
`4.5; 1 of 8 TFS; P < 0.0001), and the activity was observed
`only at one dose below the HNTD. Thus, cabazitaxel had a
`greater therapeutic index (3 active dose levels) than doc-
`etaxel (2 active dose levels) in this gastric tumor model.
`
`Antitumor activity in tumor models poorly or not
`sensitive to docetaxel
`Cabazitaxel showed antitumor activity against the fully
`docetaxel-resistant B16/TXT tumor model (log cell kill 2.1
`in B16 vs. 1.3 in B16/TXT), but no antitumor activity was
`obtained against the P-glycoprotein–overexpressing tumor
`Calc18/TXT in which docetaxel resistance was induced
`in vitro (log cell kill 3.4 in Calc18 vs. 0.5 in Calc18/TXT).
`Cabazitaxel was also found to be active against 2 aggressive
`
`2978
`
`Clin Cancer Res; 19(11) June 1, 2013
`
`Clinical Cancer Research
`
`
`
`Downloaded from on January 8, 2014. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org
`
`
`
`

`
`Published OnlineFirst April 15, 2013; DOI: 10.1158/1078-0432.CCR-12-3146
`
`Preclinical Antitumor Activity of Cabazitaxel
`
`2/50/5
`1/60/6
`0/80/8
`N/A1/8
`0/50/5
`NDND
`N/A0/5
`NDND
`NDND
`6/66/6
`8.8
`0.8
`0/50/5
`45.8
`3.4
`6/60/6
`ND
`ND
`NDND
`ND
`ND
`NDND
`days
`cellkillaCRTFS
`Log
`TC,
`Docetaxel
`
`21.3
`40.5
`6.8
`66.8
`40
`ND
`4.1
`ND
`ND
`
`—
`
`1.4
`2.5
`0.5
`4.5
`1.9
`ND
`0.6
`ND
`ND
`
`—
`
`cDocetaxelgroupswerenottreatedonday27.
`HCT-8,MIAPaCa-2andSR475studies:41.7,25and15mg/kgperinjection.
`bThedose–responsepatternfordocetaxelwasdifferentfromthatofcabazitaxelinthefollowingstudies:HT-29,A549andCaki-1studies:51.9,32.2,20and12.4mg/kgperinjection;
`aDefinitionofantitumoractivity:logcellkilltotal<0.7¼inactive;>2.8¼highlyactive.
`Abbreviations:N/A,notavailableastreatmentconductedonearly-stagedisease;ND,notdeterminedinthesamestudy.
`
`5/60/6
`6/66/6
`0/80/8
`N/A1/8
`2/60/6
`2/60/6ND
`N/A0/5
`45
`N/A0/8ND
`N/A5/8ND
`75b
`6/66/6
`50b
`0/50/5
`96.6b
`6/60/6
`7/72/7ND
`6/65/6ND
`
`64.4b
`45b
`23.1
`73.2
`96.6b
`
`1.7
`
`—
`
`25.8
`
`—
`
`18.0
`1.4
`100.9>6
`2.2
`46.0
`2.7
`17.8
`>6
`75.0
`0.5
`7.3
`3.4
`50.2
`
`—
`
`1.9
`2.0
`3.4
`
`—
`
`—
`
`21.5
`27.2
`48.2
`
`—
`
`mg/kg
`TotalHNTD,
`
`cellkillaCRTFS
`log
`
`days
`TC,
`Cabazitaxel
`
`24.0
`42.0
`37.2
`73.2
`36.0
`24.0
`45.0
`38.1
`61.5
`48.0
`28.0
`22.2
`36.0
`48.0
`mg/kg
`TotalHNTD,
`
`KidneyCaki-1(Td:4.6d)
`140
`HeadandneckSR475(Td:4.9d)22.5,14,8.6,5.4(16,20,24)
`250
`GastricGXF-209(Td:4d)
`32.3,20,12.4,7.7(14,17,20)
`130
`GastricN87(Td:4.5d)
`39.4,24.4,15,9.3(27,31,35)
`140
`LungA549(Td:6.4d)
`19.4,12,7.4,4.6(21,27,33)
`130
`LungNCI-H460(Td:2d)
`19.4,12,7.4,4.6(10,13)
`130
`BreastUISOBCA-1(Td:2.1d)
`24.2,15,9.3,5.8(13,16,19)
`70
`BreastCalc18/TXT(Td:4d)
`33,20.5,12.7,7.9(5,7,9)
`N/A
`BreastCalc18(Td:4.5d)
`33,20.5,12.7,7.9(5,7,9)
`N/A
`PancreasMIAPaCa-2(Td:3d)
`19.4,12,7.5,4.6(15,19,23,27c)310
`ColonHCT-8(Td:3.5d)
`250
`22.4,14,8.6,5.4(13,17)
`ColonHT-29(Td:4d)
`140
`19.4,12,7.5,4.6(8,12,16)
`ColonHCT116(Td:4.3d)
`220
`19.4,12,7.5,4.6(16,19,22)
`ProstateDU145(Td:4.5d)
`210
`19.4,12,7.4,4.6(24,30,36,42)
`(scheduleindays)
`Tum