8
`
`
`
`The Taxoids
`
`
`
`Marie-Christine Bissery, PhD and Francois Lavelle, PhD
`
`CONTENTS
`
`INTRODUCTION
`DISCOVERY
`
`MECHANISM OF ACTION OF TAXOIDs
`ANTITUMOR PROPERTIES
`ANIMAL PHARMACOLOGY
`ANIMAL TOXICOLOGY
`CONCLUSION
`
`1 . INTRODUCTION
`
`Paclitaxel and docetaxel belong to the taxoid family, a new class of antineoplastic
`drugs. The name taxoids refers to compounds, natural or modified, having a taxane
`skeleton.
`
`Paclitaxel (Taxol®, NSC 125973) was extracted in the late 19605 from the bark of
`the Pacific Yew, Taxus brevifolia. Because of the scarcity of the drug and the difficul-
`ties of formulation, the development was initially slow. Once these problems were
`solved, development accelerated. Docetaxel (Taxotere®, PR 56976) was obtained by
`hemisynthesis, using the starting material, 10-deacetyl baccatin III extracted from the
`needles of the European Yew tree, Taxus baccata (Fig. 1). This drug was more readily
`available because of the renewability of the source, and somewhat more soluble, and
`thus its development was rapid. Paclitaxel consists of an eight-member taxane ring
`with a four-member oxetane ring and a side chain at the C-13 position (Fig. l). Doce-
`taxel differs from paclitaxel in the 10-position on the baccatin ring and in the 3 '-posi-
`tion of the lateral chain (Fig. 1).
`This chapter will summarize the key steps in the development process of these two
`new exciting antitumor agents.
`
`2. DISCOVERY
`
`2. 1. Paclitaxel
`
`In 1960, a vast screening program for antitumor agents derived from plants was
`initiated by the Cancer Chemotherapy National Service Center under J. L. Hartwell
`(1—3). In 1962, a US Department of Agriculture botanist, A. S. Barclay, collected 650
`
`From: Cancer Therapeutics: Experimental and Clinical Agents
`Edited by: B. Teicher Humana Press Inc., Totowa, NJ
`
`175
`
`Genentech 2070
`
`Hospira v. Genentech
`|PR2017-00737
`
`Genentech 2070
`Hospira v. Genentech
`IPR2017-00737
`
`

`

`1 76
`
`Part I / Cytotoxic Agents
`
`HO
`
`0
`
`HO In.
`
` OH
`(530006»:5
`
`10-Deacetyl Baccatin Ill
`
`
`
`Fl1 = ‘CeHs
`
`R2 = —COCH3
`
`paclitaxel
`
`F!1 = -OC(CH3)3
`
`R2 = -H
`
`docetaxel
`
`Fig. 1. Structures of paclitaxel (Taxol®), docetaxel (Taxotere®) and 10-deacetyl-baccatin 111.
`
`samples in the West Coast States of the US, including samples of T. brevifolia, the
`Pacific Yew tree. Initial screening of crude extracts showed cytotoxicity toward 9KB
`cells derived from a nasopharyngeal tumor. M. Wall at Research Triangle Institute
`was particularly interested in plants with 9KB activity because of his previous
`experience with Camptotheca extracts. This is why he received among other plants,
`T. brevifolia. The isolation procedure of the drug substance was laborious and involved
`numerous steps (ethanol extraction; partition of the ethanolic residue between water
`and chloroform; followed by Craig countercurrent distribution), each of them being
`monitored by an in vivo assay using the rat Walker 256 carcinosarcoma (2—4) or/and
`the P1534 leukemia (1 ). Approximately 0.5 g of paclitaxel could be isolated starting
`with 12 kg of dried stem and bark (yield 0.004%). The isolation of the pure com-
`pound was achieved in 1966(1). In 1971 the structure of paclitaxel was published, and
`antitumor efficacy was reported in L1210, P388, and P15341eukemias, and in Walker
`carcinosarcoma 256 (5). However, the activity levels seen against the L1210 and P388
`leukemia models were very modest compared to that of other compounds. The only
`tumor system showing good efficacy, the P1534 leukemia, was not thought to be of
`predictive value (I ). Finally, the compound was poorly soluble (i.e., the in vivo evalu-
`
`

`

`Chapter 8/ The Taxoids
`
`177
`
`ation was performed with drug suspension), and supplies were difficult to obtain.
`Because of this, paclitaxel was not selected for further preclinical development (1).
`Luckily in 1974—1975, extensive in vivo testing was conducted by the National Cancer
`Institute (NCI), and paclitaxel showed high activity against the murine B16 melanoma
`model, newly introduced to the NCI screen (1 ). This prompted the NCI to evaluate
`the compound further. Additional efficacy was noted in colon 26 carcinoma and in
`MX-l human breast carcinoma implanted under the renal capsule in nude mice (6).
`However, no efficacy was observed at that time in most of the other models tested,
`colon 38 adenocarcinoma, Lewis lung carcinoma, CD8F1 mammary, and human
`xenografts implanted subcutaneously (6). Preliminary studies of its mechanism of
`action indicated that it was a spindle poison that inhibited the cell proliferation at the
`GZ-M phase in the cell cycle and that it blocked mitosis (7). The turning point occurred
`with the demonstration that paclitaxel had a unique mechanism of action by Horwitz.
`It was established that paclitaxel stabilized microtubules and inhibited depolymeriza-
`tion back to tubulin. This differed from the mechanism of action of other spindle
`poisons, such as vinca-alkaloids, that bind to tubulin and inhibit its polymerization
`(8). On the basis of the in vivo B16 melanoma efficacy and the uniqueness of
`mechanism of action, the NCI initiated a very large effort to collect barks and wood
`to obtain enough material to initiate clinical trials (1—3).
`Formulation was also an issue, and in 1980, it was reported that toxicology studies
`would proceed if a satisfactory formulation was developed (6). These toxicology and
`clinical formulation development studies were completed by 1983. Clinical Phase I trials
`started in 1983. Progress of these trials was hampered by hypersensitivity reactions,
`which led to the premature closure of some Phase 1 studies (9, 10). Since these reactions
`were observed more commonly with infusions of shorter durations, a decision was
`made to pursue clinical trials using a 24-h continuous infusion, with premedication to
`lessen the reactions. The dose-limiting toxicity was neutropenia in seven out of the
`nine Phase I trial initiated (1,9,10).
`The final major step in paclitaxel’s development was the recognition of its activity
`against ovarian cancer with responses in approx 30% of patients, many of them hav-
`ing cisplatin or carboplatin refractory disease. These clinical results were reported in
`1989 (11) (i.e., 6 yr after first clinical entry). Even as evidence of paclitaxel’s activity
`increased, with the report of efficacy in breast and nonsmall—cell lung carcinoma
`(12,13), the clinical development was still prevented by the supply shortage. To address
`this issue, the NCI sought the assistance of the pharmaceutical industry. A coopera-
`tive research and development agreement (CRADA) was awarded to Bristol Myers
`Squibb in 1991 (3,9). In 1992, the company filed a New Drug Application. The Food
`and Drug Administration approved paclitaxel that same year for the treatment of
`patients whose ovarian carcinoma had progressed with other chemotherapy (3,9)
`and, in 1994, for metastatic breast cancer.
`
`2.2. Docetaxel
`
`The Institut de Chimie des Substances Naturelles (ICSN) of the Centre National de
`la Recherche Scientifique and thne-Poulenc were interested by the ongoing work on
`paclitaxel in the US and, in particular, of the newly described mechanism of action of
`paclitaxel (8). The ICSN had expertise in the chemistry and the biochemistry of anti-
`mitotic compounds, especially vinca-alkaloids, and it was using a test to measure the
`
`

`

`
`
`1 78 Part I / Cytotoxic Agents
`
`o ozcnzca,
`
`o OZCHZCCI,
`
`a ozcuzca3
`
`o ozcnzcm,
`
`
`
` 8
`
`2 R=H
`
`Fig. 2. First hemi-synthesis of docetaxel 2 and lO—deacetyl-paclitaxel 8.
`
`inhibition of polymerization or depolymerization of tubulin, based on the Shelanski’s
`method (14). Finally, the ICSN had a large supply of T. baccata, the European
`species of yew tree that is widely dispersed in France and Europe. Therefore, in 1979,
`the ICSN decided to undertake research in this area. In 1980, Rhone-Poulenc decided
`to stop a 20-year period of research on anthracyclines and signed a research agree-
`ment concerning taxoids with the ICSN with three main objectives: to explore the
`chemistry of taxoids, bearing in mind the issue of supply, to build structure—activity
`relationships, and to select new and patentable antitumor compounds in these series.
`From the beginning, P. Potier and colleagues were convinced that semisynthesis
`was the only realistic approach for preparing paclitaxel and analogs in sufficient
`quantities for pharmaceutical research and clinical trials. They decided to explore
`systematically and extensively the different chemical components present in the Euro-
`pean yew tree T. baccata, in particular in the needles, a renewable source of biological
`material. The purification of the components present in the needles was monitored by
`measuring their interactions with tubulin purified from calf brain (14). This purifica-
`tion led to the isolation of an abundant precursor of paclitaxel, lO-deacetyl-baccatin
`III, with a yield of 1 g/kg of fresh needles (15) (Fig. 1). This yield was important
`especially in light of the 150 mg of paclitaxel that could be extracted from 1 kg of
`dried bark (16). It was considered an interesting precursor for hemisynthesis of pacli-
`taxel and other taxoids (17). It was 50- to 100-fold less active than paclitaxel in
`inhibiting microtubules depolymerization (18). The closer precursor, baccatin III,
`was also detected but at much lower concentrations.
`
`2.2.1. FIRST ACCESS TO PACLITAXEL: DISCOVERY OF DOCETAXEL (FIG. 2)
`lO-Deacetylbaccatin III, protected at both the C-7 and 010 positions (compound
`6), was converted into the cinnamoyl ester at the C-13 position (compound 7) with a
`90% yield. The cinnamic double bond was then oxyminated (19), leading to docetaxel
`(compound 2) and to lO-deacetylpaclitaxel (compound 8) after cleavage of the Boc
`
`

`

`Chapter 8 / The Taxoids
`
`179
`
`
`
`0H
`
`(m
`
`Ocom
`
`10-désacélylbaccatine Ill 3 (R: H)
`
`Baccaline Ill 5 (R: COCH3)
`
`
`
`Fig. 3. Convergent synthesis of docetaxel 2 and paclitaxel 1.
`
`and reacylation (20). Paclitaxel was obtained using the same pathway, starting from
`baccatin 111. Similar to all new derivatives of paclitaxel, docetaxel was tested for its
`interaction with tubulin/microtubules and, surprisingly, was found to be twice more
`potent than paclitaxel in inhibiting the cold-induced reaction of microtubule depoly-
`merization. At the end of 1985, a small batch of docetaxel was available, and the first
`demonstration of its in vitro cytotoxicity and in vivo antitumor properties was obtained
`in the Oncology Department of Rhéne-Poulenc, using P388 leukemia. Further in
`vivo evaluation revealed efficacy against L1210, Lewis lung carcinoma and B16 mel-
`anoma (16,21 ).
`However, although very successful, this first semisynthetic approach was not appli-
`cable at an industrial scale owing to the use of toxic and very expensive reagents, such
`as osmium tetraoxide.
`
`2.2.2. CONVERGENT SYNTHESIS OF TAXOIDs (FIG. 3)
`
`The convergent synthesis of paclitaxel and docetaxel was performed by direct
`esterification of lO-deacetyl-baccatin III or baccatin III, with the acids corresponding
`to the lateral chains present in paclitaxel and docetaxel, respectively. The first asy-
`metric synthesis of the C-13 phenylisoserine chain of paclitaxel was done by Denis
`and collaborators (22). The lateral chain of docetaxel was prepared using benzalde-
`hyde and terbutyl chloroacetate (23). Finally, esterification of the acids by baccatin
`III and 10-deacetyl-baccatin III yielded paclitaxel (24) and docetaxel (23), respec—
`tively. Since these first experiments, the yield of the esterification methods and of the
`synthesis of the lateral chain have been improved by different teams (25—28).
`
`

`

`180
`
`Part I / Cytotoxic Agents
`
`In 1989, the compound was obtained in sufficient amount to initiate extensive phar-
`macological and toxicological studies and a suitable formulation for iv evaluation was
`developed. Using the SC 316 melanoma murine model, it was found that iv docetaxel
`was more active than paclitaxel at an equitoxic dose (29).
`The compound was also found to be highly active against a large number of murine
`tumor models, most importantly when treated at an advanced stage (i.e., measurable
`disease), and schedule studies revealed that the compound was schedule-independent
`(29). Finally, a pharmacokinetic/distribution study in tumor-bearing mice showed
`that at optimal dosage, the area under the plasma and tumor concentration vs time
`curves (AUC) were much higher than the AUC required to kill human cancer cell
`lines in vitro. Toxicology studies were performed according to the NCI guidelines.
`Phase I clinical trials were initiated in 1990 in Europe and in the US (30). Five
`different schedules were investigated up front in record time. At the end of the Phase
`I trials, it was shown that neutropenia was the major dose-limiting toxicity, and
`responses were reported in different tumor types. Based on considerations, such as
`dose intensity, toxicity profile, and preclinical data, suggesting absence of schedule
`dependency, the recommended dose and schedule for Phase II studies were 100 mg/m2
`adminisetred as a 1-h infusion every 3 wk, without prophylactic measures. Broad
`Phase II testing was initiated in 1992 throughout Europe, North America, and Japan,
`and a CRADA was signed by the NCI and Rhone-Poulenc Rorer. A broad spectrum
`of efficacy was reported, including breast, nonsmall-cell lung, and ovarian cancers. A
`New Drug Application was filed in 1994, and docetaxel has now been approved in
`more than 30 countries.
`
`3. MECHANISM OF ACTION OF TAXOIDS
`
`3.1. The Cellular Target of Taxoids
`
`Together with actin microfilaments and intermediate filaments, microtubules form
`the cytoskeleton of eukaryotic cells. The microtubules are involved in a variety of
`cell functions, including chromosome movement and the regulation of cell shape and
`motility (31). These activities are modulated through associations with several bio-
`chemical components, such as guanosine triphosphate (GTP), and a wide range of
`proteins, the microtubule-associated proteins (MAPS). When a cell begins to divide,
`interphasic microtubules totally vanish and the mitotic spindle assembles. The
`depolymerization of mitotic spindle microtubules is essential for specific mitotic
`events, such as the movement of the chromosomes to the metaphase plate and their
`correct segregation during anaphase (32). Microtubules are long, hollow cylinders
`assembled from a heterodimeric (oz/B) globular protein called tubulin. They consist of
`13 aligned protofilaments within which the tubulin subunits interact through long-
`itudinal and lateral bonds (33). Not all the tubulin pool assembles into microtubules:
`a steady state is maintained between assembled tubulin and a concentration of free
`tubulin called the critical concentration.
`
`3.2. Taxoids Stabilize Microtubules
`
`The polymerization of tubulin purified from mammalian brain usually enhances
`the turbidity of the solution; thus, the degree of polymerization can be monitored
`simply by measuring turbidity (34). Figure 4 depicts, for example, the effects of doce-
`
`

`

`Chapter 8 / The Taxoids
`
`0.40
`
`181
`
`0.30 "
`
`0.20 “
`
`
`0.10
`
`CD.at350nm
`
`0.00
`
`.__-__L_—__L—_—_——L———J
`0
`1o
`20
`30
`40
`
`TIME (mln)
`
`Fig. 4. Effect of docetaxel on polymerization of tubulin and depolymerization of microtubules.
`Tubulin was polymerized by heating from 3 to 37 °C. Depolymerization of microtubules was obtained
`by cooling from 37 to 3 °C (arrow). Polymerization or depolymerization was monitored by following
`the turbidimetry at 350 nm. (A) 10 W porcine brain tubuline. (B) 10 “M tubulin and 3 uM docetaxel.
`
`taxel on the kinetics of tubulin assembly and disassembly. The lag time correspond-
`ing to the activation and nucleation of tubulin is notably reduced, and the rate of
`polymerization is increased (35). Finally, microtubules, stabilized by taxoids, do not
`depolymerize on cold treatment. In fact, paclitaxel and docetaxel analogs are usually
`evaluated on the basis of the drug concentration that inhibits half of the cold-induced
`depolymerization. Docetaxel is about twice as efficient as paclitaxel in this respect
`(36,3 7). The thermodynamic parameters of tubulin assembly are also modified by the
`taxoids, and the critical concentration is significantly reduced in the presence of pac-
`litaxel (8). Docetaxel is twice as efficient as paclitaxel in decreasing the critical con-
`centration of tubulin assembly (38).
`It should be pointed out that the mechanism of action of the taxoids is unique,
`since all other known mitotic spindle poisons, in particular, the vinca-alkaloids, shift
`the tubulin-microtubule equilibrium toward tubulin (39) (Fig. 5).
`
`3.3. Characterization of the Interaction Site
`
`Tritiated paclitaxel cosediments with microtubules and dissociates rapidly on the
`addition of paclitaxel. Thus, a rapid and reversible equilibrium exists between pac-
`litaxel and the microtubules. There is only one high-affinity binding site of paclitaxel
`per a/p tubulin subunit, indicating that the interaction between microtubules and
`paclitaxel is specific. The equilibrium dissociation constant was originally found to be
`870 nM (40). Docetaxel competes with paclitaxel for binding to the microtubules, but
`its equilibrium dissociation constant is two times less, i.e., it has better affinity (38).
`This difference could account for the higher efficiency of docetaxel to promote tubulin
`polymerization, to stabilize microtubules against cold-induced disassembly, and to
`
`

`

`
`182
`Part I / Cytotoxic Agents
`
`MICROTUBULES
`
`VI NCA
`ALCA LOI DS
`
`POLYMEHIZATION
`
`DEPOLYMERIZATION
`
`TAX O. D S
`
`TUBULIN
`
`Fig. 5. Mechanism of action of vinca alkaloids and taxoids.
`
`decrease the critical concentration of tubulin assembly. To acquire further insight into
`the taxoid—microtubule interaction at the molecular level, it is essential to locate the
`taxoid binding site. So far, it is known that the binding of taxoids is linked to the
`polymerization process. Furthermore, no binding of taxoids to dimeric tubulin has
`been detected, indicating that the site is located on assembled tubulin (38). This site
`does not overlap those of other known ligands, such as colchicine, podophyllotoxin,
`vinblastine, or GTP (41,42).
`
`3.4. Models of the Mechanism of Action
`
`Paclitaxel-bound microtubules, the structure of which has been resolved at 3 nm
`employing X-ray scattering, appear to be constituted of 12 protofilaments instead of
`the 13 protofilaments usually observed (43,44). The solution structure of microtu-
`bules induced by docetaxel has been also characterized using the same technique (45).
`The substructures of the microtubule walls are identical in paclitaxel- and docetaxel-
`induced microtubules; however, the population of docetaxel microtubules has an
`average of 13 protofilaments like the control microtubules. It is proposed that the
`chemical substitutions present in docetaxel side chain in its binding site increases
`slightly the contact angle between adjacent protofilaments. The simple working
`hypothesis is that taxoids bind between adjacent tubulin molecules, and such a hypoth-
`esis is fully compatible with the observed thermodynamic behavior of the taxoid-
`induced microtubule assembly system (43).
`
`4. ANTITUMOR PROPERTIES
`
`4.1. In Vitro Activities
`
`4. 1 . 1 . CELLULAR CYTOTOXICITY
`
`Both taxoids have been found extremely potent against a wide variety of murine
`and human cancer cell lines. Using the COMPARE computer program, it was con-
`
`

`

`Chapter 8/ The Taxoids
`
`183
`
`eluded that docetaxel response profile on 50 human tumor cell lines in the new NCI
`screening panel, correlated with the data pattern of test agents acting on the tubulin/
`microtubule system, the closest compound being paclitaxel (NCI, unpublished results).
`Several in vitro studies have been done comparing their activities under various exper-
`imental conditions (liquid medium, semisolid medium, short- and long-term expo-
`sures). The cytotoxicity of paclitaxel and docetaxel at submicromolar concentrations
`was compared in several murine, P388, SVras, and human tumor cell lines, breast
`Calc18, colon HCT 116, bladder T24, and nasopharyngeal KB (46). Docetaxel was
`found to be 1.3- to 12-fold more potent than paclitaxel. The cytotoxic properties of
`paclitaxel and docetaxel were also compared against nine cell lines established from
`human ovarian tumors and having intrinsic or acquired resistance to cisplatin. These
`cell lines were not crossresistant to the taxoids, and docetaxel was found to be active
`at a twofold lower concentrations than paclitaxel (47).
`In addition, the activities of docetaxel and paclitaxel, in a human tumor cloning
`stem cell assay (starting from fresh human tumor biopsies), were compared at con-
`centrations of taxoids similar to the plasma levels obtained after treatment of patients
`(48,49). Melanoma, breast, lung, ovarian, and colon tumors cells were significantly
`inhibited, regardless of the schedule of incubation (1-h exposure or continuous expo-
`sure for 14—28 (1). Interestingly, 29 samples were found to be more sensitive to doce-
`taxel, whereas 13 were more sensitive to paclitaxel, suggesting partial crossresistance
`between these two drugs (49). Finally, the in vitro cellular effects of docetaxel and
`paclitaxel have been recently assessed against a wide range of human normal and
`tumor samples, including tumor cell lines, primary cultures from tumor biopsies and
`normal bone marrow samples (50). ICSO (50°70 inhibitory concentrations) values of the
`two taxoids were in the nanomolar range and docetaxel appeared to be two- to four-
`fold more cytotoxic than paclitaxel (50).
`
`4.1.2. MECHANISM OF CYTOTOXICITY AND CELLULAR EFFECTS
`
`Uptake and efflux studies were performed on P388 leukemia cells in vitro with
`radiolabeled docetaxel and paclitaxel. Uptake experiments revealed that a threefold
`higher intracellular concentration of docetaxel was obtained as compared to pacli-
`taxel, for the same initial extracellular concentration (0.1 M4) (51). Efflux studies
`revealed that the half-time of efflux of docetaxel from P388 cells was at least three
`
`times slower than that of paclitaxel (150 vs 45 min, respectively).
`Thus, the higher potency of docetaxel observed in vitro may be explained by the
`combination of its higher affinity for microtubules, its higher achievable intracellular
`concentration, and the slower cellular efflux.
`Cell-cycle studies revealed that paclitaxel was mainly cytotoxic during mitosis (M
`phase), as demonstrated by experiments on CHO and A 2780 ovarian tumor cell lines
`(52). Inhibition of cytokinesis has been observed, but some cells can progress through
`new cell cycles, leading to the formation of polyploid cells (53,54). Using synchronized
`HCT116 cells,
`it was demonstrated that paclitaxel inhibits formation of mitotic
`spindles in cells without affecting function of preformed spindles and without arrest-
`ing cells in mitosis (54). Docetaxel has been found to be more active on proliferating
`than on nonproliferating KB cells (46) and to inhibit mitosis in several cell lines, such
`as J82 and KB (55). Surprisingly, using synchronized HeLa cells, it has been shown
`that docetaxel exerts cell killing specifically during the S phase of the cell cycle; no
`
`

`

`184
`
`Part I / Cytotoxic Agents
`
`In Vivo Antitumor Activity of Paclitaxel Against Human Tumor Xenografts
`
`Table 1
`
`sc Human tumor
`
`A 2780 ovarian
`LX-l lung
`H 2981 lung
`L 2987 lung
`RCA colon
`HGT-116 colon
`A431 vulva
`
`Highest nontoxic
`iv dosage
`mg/kg/dose
`
`'
`
`18
`24
`24
`36
`36
`36
`36
`
`Schedule
`days
`
`7, 9, 11, 13, 15
`5, 7, 9,11,13
`5, 7, 9,11,13
`1
`, 16, 18, 20, 22
`4, 6, 8, 10, 12
`3, 5, 7, 9, 11
`3, 5, 7, 9, 11
`
`Activity“
`rating
`
`+ + +
`+ + + +
`+ +
`+ + +
`> + +
`> + + +
`+ + +
`
`‘1 Activity rating: + + + + = highly active (log cell kill > 2.8), + + + = highly active (log cell
`kill = 2.0 to 2.8), + + = active (log cell kill = 1.3—1.9; T/C 2 150% for L1210, + = active (log
`cell kill = 0.7-1.2 for s.c. tumors, T/C = 125—174% for P388), — = inactive.
`
`cytotoxicity was observed during mitosis, a different situation from what is observed
`with paclitaxel (56).
`Finally, it was found that paclitaxel greatly increases the pool of polymerized tubu-
`lin in cells, and new short microtubules free in the cytoplasm were observed (57). In
`addition, at high concentration, it induced the formation of microtubules bundles (58).
`Using J82 human bladder and KB 3-1 human carcinoma cells, it was shown that pac-
`litaxel and docetaxel lead to the formation of bundles and asters in a dose- and time-
`
`dependent manner (55). Asters were observed in mitotic cells, and bundles were seen
`in interphase cells. The effects of docetaxel as compared to paclitaxel appeared at a
`twofold lower concentration (55 ).
`
`Paclitaxel and even more docetaxel have been studied in many murine tumor
`models and human tumor xenografts.
`
`4.2. In Vivo Activity
`
`4.2. 1. PACLITAXEL
`
`The development of an adequate formulation for paclitaxel led to a re-evaluation
`of its in vivo antitumor efficacy, using better experimental conditions, i.e., avoiding
`the previous ip/ip evaluation and administering the drug iv at a site different from the
`tumor site. The formulation used was 10% Cremophor®, 10% ethanol, 80% NaCl
`0.9%. These studies have been recently reviewed, and most of them were performed
`after initial clinical trials (59,60). Indeed, these studies demonstrated that paclitaxel
`delivered iv was active against several tumors implanted in distal sites and treated at an
`early stage: sc Madison 109 murine lung carcinoma and A 431 vulva, A 2780 ovarian,
`H 2981 and LX-l lung, and RCA and HGT-116 colon human tumor xenografts
`implanted under the renal capsule of nude mice (Table 1) (60). When administered
`sc five times weekly for three consecutive weeks, paclitaxel caused the complete regres-
`sion of a human breast tumor xenograft, and significantly delayed the growth of
`endometrial, ovarian, brain, tongue, and lung human tumor xenografts (61). Pacli-
`taxel was also evaluated against ovarian carcinoma xenografts HOC8, HOC18, and
`HOC22, and was found to have similar efficacy to docetaxel (62).
`
`

`

`Chapter 8 / The Taxoids
`
`185
`
`In Vivo Antitumor Activity of Docetaxel Against Human Tumor Xenografts
`
`Table 2
`
`sc Human tumor
`
`Calc 18 mammary
`MX-l mammary
`LX-l lung
`SKMEL-Z melanoma
`CX-l ovarian
`KM20L2 ovarian
`OVCAR-3
`
`Highest nontoxic
`iv dosage
`mg/kg/dose
`
`32.2
`22
`22
`33
`15
`33
`33
`
`Schedule
`days
`
`11, 15, 19
`11, 15, 19
`9, 13, 17
`27, 31, 35
`12,16, 20
`14,18,22
`3, 7, 11
`
`Activitya
`rating
`
`+ + +
`+ + + +
`+ +
`+ + + +
`+ + +
`+ +
`+ + + +
`
`“For activity rating, see Table 1.
`
`Further schedule-dependency studies were performed and showed that daily injec-
`tion for 7 d was the best schedule as opposed to longer spaced administration (60).
`
`4.2.2. DOCETAXEL
`
`The experimental antitumor activity of docetaxel has been evaluated against a
`panel of 30 tumors of mice and human tumors xenografted in nude mice, represent-
`ing a variety of tissue types and chemosensitivity patterns. The tumors were grafted in
`distal sites, and several tumors were treated at advanced and metastatic stages. Dose
`response was evaluated in all trials to determine accurately the maximum tolerated
`dose. The formulation used was a 1:1 ethanol polysorbate 80 solution, administered
`after a 1:10 dilution in glucose 5% in water (29). Docetaxel had a broad spectrum of
`antitumor activity, since 28/30 models responded to this agent (29, 63—65 ) (Tables 2
`and 3). The experimental antitumor activities of paclitaxel and docetaxel were com-
`pared by testing these drugs against B16 melanoma, a tumor sensitive to taxoids using
`an intermittent schedule, every 2 d x 3. Antitumor activity was expressed by the
`tumor growth delay (T—C) and by the log cell kill (LCK) of tumor cells obtained at the
`maximal tolerated dose (MTD) of each drug. According to these criteria, docetaxel
`was approximately two times more active and potent than paclitaxel (docetaxel:
`T-C = 12.2 d, LCK = 2.9, MTD = 11.3 mg/kg/d; paclitaxel: T-C = 4.7 d, LCK =
`1.1, MTD = 21.7 mg/kg/d) (29) (Fig. 6). Among the murine models tested, good
`activity was observed, with in some cases cures of early stage tumors. However, the
`most meaningful data were those obtained against advanced stage tumors (i.e.,
`tumors at least 200 mg at start of therapy) where complete tumor regressions could be
`observed. This occurred with the murine mammary adenocarcinomas 16/C and
`13/C, pancreatic ductal carinoma O3, colon 38 adenocarcinoma, and the human
`xenografts MX—l mammary and SK-MEL-Z melanoma (63,64). Prolonged tumor
`growth delays were also observed with Calc-18 breast, LX-l lung, CX-l colon, head
`and neck HNX-14C, and HNX-22B xenografts (64,66). Since the clinical activity of
`paclitaxel against ovarian tumors was impressive, five human ovarian xenografts hav-
`ing different sensitivities to the reference drug cisplatin were included in this study.
`Docetaxel was active against the three tumors sensitive to cisplatin; interestingly, it
`was also active against OV-Pe, which is resistant to cisplatin (65).
`
`

`

`186
`
`Part I / Cytotoxic Agents
`
`In Vivo Antitumor Activity of Docetaxel Against Murine Tumors
`
`Table 3
`
`Tumor
`
`Solid tumors sc
`
`Melanoma B16 early
`Pancreas
`
`P02
`P03 early
`P03 advanced
`
`Mammary
`MA16/C early
`MA16/C advanced
`MA13/C early
`MA13/C advanced
`MA44 early
`Colon
`
`C26 early
`C38 early
`C38 advanced
`C51 early
`C51 advanced
`
`Lewis lung early
`
`Osteosarcoma GOS early
`
`Hystiocytosarcoma M5076 early
`
`Leukemias ip
`P388 106 cells
`L1210 10‘ cells
`
`a For activity rating, see Table 1.
`b CR = complete regressions.
`
`Highest nontoxic
`iv dosage
`mg/kg/dose
`
`Schedule
`days
`
`Activity”
`rating
`
`24
`
`32.2
`20.5
`18.3
`
`15
`10.8
`14.2
`15
`22
`
`5
`23.5
`26.8
`12.7
`15.2
`
`23.2
`
`18.6
`
`8.6
`
`23.2
`21.7
`
`3, 5, 7, 9
`
`+ + + +
`
`3, 5, 7
`3, 5, 7, 9
`22, 24, 26, 28
`
`i
`+ + + +
`5/6 CR
`
`3, 5, 7
`7, 9, 11
`3, 5, 7
`24, 27, 30
`3, 5, 7
`
`1—4
`3, 5, 7
`14, 16, 18
`3, 5, 7
`10, 12, 14
`
`3—7
`
`3—7
`
`3—7
`
`1-4
`1—4
`
`+ + +
`5/5 CR
`+ + + +
`3/5 CR”
`i
`
`+
`+ + + +
`5/5 CR
`+ + +
`+ +
`
`+
`
`+
`
`—
`
`+
`+ +
`
`Scheduling studies were performed against advanced colon 38 adenocarcinoma.
`Docetaxel was tested using three different schedules comparing the effect of 2, 3, and
`10 administrations over the same duration of treatment. Overall, the administration
`schedule did not influence markedly the total dosage that can be administered and,
`thus, the compound was considered schedule-independent for the MTD (29).
`
`4.3. Combination Chemotherapy
`
`Since taxoids have clinical activity in ovarian, breast, and lung tumors, most of the
`experimental studies have been done with drugs active in these diseases: doxorubicin,
`5-fluorouracil, cyclophosphamide, cisplatin, and etoposide.
`
`4.3. 1. PACLITAXEL
`
`Both in vitro and in vivo studies were performed. The efficacy of combination
`therapy consisting of paclitaxel plus a topoisomerase II inhibitor, doxorubicin or
`
`

`

`
`
`Chapter 8 / The Taxoids 187
`10000
`
`CONTROL
`
`/
`PAQLITAXEL
`
`DOCETAXEL
`
`1000
`
`100
`
`
`
`
`
`TUMORWEIGHT(mg)
`
`:
`— — — L.
`,' 12 2 dayss‘
`Limit of palpation
`
`21.7
`
`L.C.K.
`
`COMPOUND
`
`DOCETAXEL
`PACLITAXEL
`
`M.T.D.
`mglkg/day
`13.4
`
`0
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`DAYS POST TUMOR IMPLANTATION
`
`Fig. 6. Comparative in vivo activities of paclitaxel and docetaxel against 316 melanoma. Detailed
`experimental conditions were described in ref. (29). Briefly, BGDZFl mice (7 mice/group) were
`grafted sc on day 0 with 30 mg 816 tumors fragments. Drugs were injected iv on days 4, 6, 8, and 10
`at the MTD (21.7 mg/kg/d for paclitaxel; 13.4 mg/kg/d for docetaxel). Tumor growths were mea-
`sured biweekly. Activity is expressed by the TC (where T and C are the median time in days neces-
`sary for the tumors of the treated group '1‘ and the control group C to reach a size of 1 g). Activity is
`also expressed by the LCK, which quantifies the number of tumor cells killed by the chemotherapy.
`
`etoposide, against various cell lines has been studied in vitro; better results were ob-
`tained when cells were first incubated with paclitaxel (67). Cisplatin—paclitaxel com-
`bination was evaluated using L1210 leukemic cells: maximal effects were observed
`when cells were incubated for 24 h with paclitaxel, and then treated for 30 min with
`cisplatin (68). Combinations of taxoids and tubulin-interactive agents are of interest
`because of their complementary mechanism of action. Paclitaxel—estramustine was
`found to give supra-additive cytotoxic effects on several lines of human prostatic
`adenocarcinoma. No additive properties were noted when taxoids were combined
`with vinblastine (69).
`In vivo combination chemotherapy studies have also been performed with pac-
`litaxel using the M109 tumor model (60). The combined agents included cisplatin,
`etoposide, doxorubicin, cyclophosphamide, methotrexate, pentamethylmelamine,
`and bleomycin. Taxol-cisplatin and Taxol-bleomycin were the two