Review
`
`Annals of Oncology 5: 495-505, 1994.
`© 1994 Kluwer Academic Publishers. Printed in the Netherlands.
`
`Paclitaxel (Taxol™) and docetaxel (Taxotere™): Not simply two of a kind
`
`J. Verweij,1 M. Clavelf2 & B. Chevalier3
`'Dept. of Medical Oncology, Rotterdam Cancer Institute, Rotterdam, The Netherlands; 2Dept. of Medical Oncology, Centre Leon Berard,
`Lyon; 3Dept. of Medical Oncology, Centre H. Becqucrel, Rouen, France
`
`Summary
`
`Paclitaxel and docetaxel are the two presently clinically avail-
`able representatives of the new class of taxane drugs. They
`share major parts of their structures and mechanisms of
`action, but differ in several other aspects. For instance, there
`is a difference in their tubulin polymer generation, and doce-
`taxel appears twice as active in depolymerization inhibition.
`In vitro docetaxel also tends to be more potent in different
`cell lines and investigational models. While in vitro and in
`
`vivo studies suggest that prolonged exposure to paclitaxel is
`better than a brief exposure, no such tendency is seen for
`docetaxel, indicating it to be a schedule-independent drug.
`Clinical studies have not confirmed an advantage for pro-
`longed exposure to paclitaxel; but do show differences in the
`toxicity profiles of the two drugs. These topics will be ad-
`dressed in detail.
`
`Key words: paclitaxel, docetaxel, taxoids
`
`Introduction
`
`Biochemistry and mechanism of action
`
`Recent clinical research on new drug development has
`been extremely interesting, mainly because of the intro-
`duction of two classes of drugs, the taxanes and the
`camptothecins, that already have longstanding histo-
`ries, rendering the designation 'new5 somewhat exag-
`gerated. This review will focus on the development of
`taxanes.
`Antimicrotubule agents have for many years been
`important anticancer agents, contributing significantly
`to the potentially curative treatment of diseases such as
`the malignant lymphomas, leukaemias and germ cell
`tumors. The search for new antimicrotubule agents
`during the past few decades has been less rewarding
`than had been hoped. Paclitaxel, extracted as early as
`the late 1960s from the bark of the Pacific Yew Taxus
`brevifolia, is the first compound with a taxene ring
`
`shown to possess antitumor activity [1]. Because of the
`scarcity of the drug, the difficulties in its isolation,
`extraction and formulation, the development of this
`antitumor agent was initially relatively slow. Once these
`problems were solved and after the discovery of its
`unique mechanism of action, development accelerated.
`The second drug in this new class of compounds is
`docetaxel, extracted in 1986 from the needles of the
`European Yew Taxus baccata [2]. This drug was more
`readily available because of the regenerating capacity
`of the source, and somewhat better soluble, and thus its
`development was more rapid than that of paclitaxel.
`This review will summarize preclinical and clinical data
`on both taxenes. Randomized studies comparing these
`two drugs have not yet been performed, so a true clini-
`cal comparison is not yet possible.
`
`Paclitaxel consists of an eight-member taxane ring with
`a four-member oxetane ring and a bulky ester side
`chain at C-13 that is necessary for antitumor activity [3]
`but which can be modified (Fig. 1). The chemical for-
`mula of paclitaxel is C47H51O14 and its molecular
`weight is 853.9. It is highly lipophilic and insoluble in
`water, but soluble in Cremophor EL, polyethylene
`glycols 300 and 400, chloroform, acetone, ethanol and
`methanol. For clinical use paclitaxel is formulated in
`50% Cremophor EL and 50% dehydrated alcohol.
`Docetaxel differs from paclitaxel in the 10-position
`on the baccatin ring and in the 3'-position of the lateral
`chain, and has a chemical formula of C43H53NO14 and
`a molecular weight of 807.9. It is insoluble in water, but
`soluble in 0.1 N hydrochloric acid, chloroform, dime-
`thylformamide, 95%-96% v/v ethanol, 0.1 N sodium
`hydroxide and methanol. The formulation used in the
`most recent clinical studies consists of 100% polysor-
`bate 80.
`Microtubules are among the most strategic subcel-
`lular targets of anticancer agents. Like DNA, micro-
`tubules are ubiquitous to all eukaryotic cells. They are
`composed of tubuline dimers consisting of an a- and a
`|3-subunit protein that polymerize and, with numerous
`microtubule-associated proteins (MAPs), decorate the
`exterior wall of the hollow micro tubule structure [4].
`There is a continuous dynamic equilibrium between
`tubulin dimers and microtubules, i.e., a continuous bal-
`ance between polymerization and depolymerization.
`In addition to being an essential component of the
`mitotic spindle, and required for the maintenance of
`cell shape, microtubules are involved in a wide variety
`
`MYLAN - EXHIBIT 1019
`
`

`
`In vitro cytotoxkity
`
`Both taxanes have been found to be extremely potent
`against a wide variety of mouse and human cancer cell
`lines. Several in vitro studies have compared the activ-
`ity of both compounds [20-23]. Studies of paclitaxel
`have suggested a higher potency in prolonged-exposure
`experiments [14].
`Riou et al. compared the cytotoxicity of docetaxel
`and paclitaxel in several murine (P388, SVras) and
`human tumor cell lines (Calcl8, HCT116, T24, N417,
`KB). Docetaxel was found to be 1.3- to 12-fold more
`potent than paclitaxel after 96 hours of exposure, a
`result that could be explained by the higher affinity of
`docetaxel for microtubules [20]. There was partial
`cross-resistance of docetaxel to the P-glycoprotein-
`positive P388/DOX cell line resistant to doxorubicin,
`while for paclitaxel this cross-resistance was complete.
`Using the sulforhodamine B assay, Kelland and Abel
`[21] compared the cytotoxic properties of both com-
`pounds in nine ovarian carcinoma cell lines, three of
`which were rendered resistant to platin derivatives.
`Both taxanes were much more cytotoxic than cisplatin,
`etoposide and doxorubicin in these cell lines. In the cis-
`platin-sensitive cell lines docetaxel was 1.3- to 3.9-fold
`more potent than paclitaxel after continuous exposure
`and 1.2- to 10-fold more potent after a 2-hour expo-
`sure.
`No cross-resistance with either paclitaxel or doce-
`taxel was found in the cell lines with acquired resistance
`to platinum. In another model, using in vitro colony
`formation of freshly explanted human tumor cells in a
`capillary soft agar cloning system, docetaxel was stud-
`ied at concentrations of 0.025-10 jig/ml in short-term
`(1 hour) or continuous (14 days) incubations [22].
`There was a dose-dependent inhibition of colony for-
`mation in breast cancer, lung cancer, ovarian cancer,
`colorectal cancer and melanoma. In addition, a 1-hour
`exposure to paclitaxel at 10 ng/ml resulted in inhibition
`of colony formation, but in a head-to-head compari-
`son, 29 specimens were found to be more sensitive to
`docetaxel, while only 13 were more sensitive to pacli-
`taxel. Thus, these data suggested a lack of complete
`cross-resistance between the two drugs and a higher
`potency of docetaxel in the majority of specimens
`evaluated. The latter was confirmed in a similar study
`comparing docetaxel and paclitaxel at final concentra-
`tions of 0.04, 0.4 and 4.0 ^mol/l [23]. Again there was
`a concentration-dependent anti-tumor cell activity for
`both drugs with all of the types of tumor cells respond-
`ing, even those resistant to conventional antineoplastic
`agents. Docetaxel was significantly more potent by
`1-hour incubation than paclitaxel (p - 0.0002), which
`was also the case for the long-term (21-28 days) con-
`tinuous exposure.
`
`or 1H
`
`1 H
`I-
`><Oo
`-C-CH,
`II
`0
`
`1O
`
`496
`
`CH,- C-O.
`Me
`
`^
`
`0 - -- \
`
`Au
`
`:
`HO
`Ph-C-0
`II
`o
`
`OI
`
`Ic-
`
`C H-
`
`OH
`
`oI
`
`I
`P h - C - NH - C H-
`1Ph
`
`P „ OH
`
`/•fg. /. Chemical structure of paclitaxel (top) and docetaxel (bot-
`tom).
`
`of cellular activities such as cell motility and transport
`between organelles within the cell [5-8]. Furthermore,
`they may also have a role in modulating the interactions
`of growth factors with cell-surface receptors and the
`proliferative transmembrane signals produced by these
`interactions [5-9]. Many of the unique pharmacologic
`interactions of drugs with microtubules are caused by a
`dynamic equilibrium between microtubules and tubulin
`dimers [6,10]. Any disruption of the equilibrium within
`the microtubule system would be expected to disrupt
`cell division and normal cellular activities in which
`microtubules are involved.
`Taxanes bind preferentially and reversibly to the p1-
`subunit of tubulin in the microtubules rather than to
`tubulin dimers [11-13]. The binding site to tubulin dif-
`fers from the one of vinca-alkaloids and podophyllo-
`toxins. While vincas inhibit polymerization and in-
`crease microtubule disassembly [14], the binding of
`taxanes enhances polymerization of the tubulin into
`stable microtubules and inhibits microtubule depoly-
`merization, thereby inducing the formation of stable
`microtubule bundles [14-17]. This disruption of the
`normal equilibrium ultimately leads to cell death.
`As an inhibitor of microtubule depolymerization,
`docetaxel is approximately twice as potent as paclitaxel
`[16, 17]. In addition, docetaxel generates tubulin poly-
`mers that differ structurally from those generated by
`paclitaxel [18] and does not alter the number of proto-
`filaments in the microtubules, while paclitaxel does
`[19].
`
`

`
`In vivo activity
`
`Paclitaxel and, even more, docetaxel have also been
`studied in many murine tumor models and human
`tumor xenografts.
`Paclitaxel was found to be active against intraperi-
`toneal B16 melanoma, MX-1 breast cancer, and, to a
`lesser extent, CX-1 colon carcinoma, LX-1 lung tumor
`and ip P388 and L1210 [14]. It was inactive against
`subcutaneous CD8FJ mammary and C38 colon can-
`cers and intravenous Lewis Lung carcinoma. Studies
`on schedule dependency in the P388 leukemia model
`suggested that an increased life-span was maximal
`when mice were treated every 3 hours, which approxi-
`mates a continuous administration schedule [14], but
`these studies did not use the maximum tolerable dose
`and therefore definitive conclusions can not be drawn.
`In s.c. B16 melanoma docetaxel was 2.7 times as
`potent as paclitaxel [24], and like paclitaxel, it was ac-
`tive against MX-1 mammary cancer [25]. In contrast to
`paclitaxel, docetaxel was active against C38 colon car-
`cinoma, and also highly active against CX-1 colon car-
`cinoma and LX-1 lung carcinoma [25,26].
`Docetaxel was active in murine PO3 pancreatic car-
`cinoma and s.c. C51 colon carcinoma as well, and in
`the xenografted KM 20L2 colon carcinoma, SK-
`MEL-2 melanoma and OVCAR-3, HOC8, HOC18
`and HOC22 ovarian carcinoma [25,27]. The activity of
`docetaxel was less pronounced in s.c. Lewis Lung car-
`cinoma, L1210 and P388 leukaemia and Glasgow
`osteogenic sarcoma [24].
`While conclusions as to the schedule dependency of
`paclitaxel would be preliminary, there is no doubt that
`the activity of docetaxel is schedule-independent [24].
`
`Animal toxicology
`
`Toxicology studies with paclitaxel in rodents were per-
`formed with intraperitoneal drug administration be-
`cause of the toxicity of the vehicle, and because of vol-
`ume constraints, in dogs the drug was given intra-
`venously [28].
`The single-dose LD10 in Sprague-Dawley rats was
`138 mg/m2, and the LD50 was 206 mg/m2. For the
`daily-times-five administration schedule these doses
`were 36 and 51 mg/mVday, respectively, in rats, and 67
`and 82 mg/m2/day in CD2F1 mice. The single-dose
`toxic dose low (TDL) in beagle dogs was 45 mg/m2.
`The major toxic effect in rats and dogs was reversible
`myelosuppression. In rodents oligospermia was noted.
`Gastrointestinal toxicity was most pronounced in the
`daily-times-five schedule in dogs and consisted of diar-
`rhea, mucosal ulcerations and emesis. In dogs hypoten-
`sion was also observed which was believed to be related
`to cremophor EL.
`For docetaxel the single-dose LD10 in mice was 345
`mg/m2, and the LD50 was 414 mg/m2. In dogs the TDL
`was 15 mg/m2 [29]. The main toxic effects of docetaxel
`
`497
`
`were reversible myelosuppression and epithelial necro-
`sis in the digestive tract. In mice, cumulative and rever-
`sible neurotoxicity was observed. Dogs experienced
`hypotension which was thought to be related to the
`vehicle Polysorbate 80.
`Although there are minor differences, animal toxicol-
`ogy in general appears similar for paclitaxel and doce-
`taxel, with similar target organs and more pronounced
`toxicity with repeated-administration schedules.
`The recommended starting doses for phase I studies
`were based on one-third of the TDL in dogs for both
`drugs: 15 mg/m2 for paclitaxel and 5 mg/m2 for doce-
`taxel.
`
`Animal pharmacology
`
`Pharmacokinetic data on paclitaxel in animals are
`scarce because of the lack of a sensitive assay in the era
`when the drug was being investigated only in the la-
`boratory. None of the bioassays that were developed
`were suitable for detailed studies. With an assay with a
`detection limit of 0.1 u.M/1, however, it has been shown
`that in animals, paclitaxel was almost totally bound to
`proteins and had distribution and elimination half-lives
`of 2.7 and 42 minutes, respectively, in rabbits [30].
`In contrast a sensitive and selective high-perfor-
`mance liquid chromatographic (HPLC) assay for doce-
`taxel was available early in its development [31], and
`was used for studies in animals [29, 32]. In tumor-bear-
`ing mice t'/2a and tV2p were 7 min and 1.1 hour, respec-
`tively. The pharmacokinetics were linear. The plasma
`clearance was 2.2 1/h/kg and the apparent volume of
`distribution at steady state 2.2 1/kg. The AUC at doses
`of 13-62 mg/kg ranged from 4.5 to 29.6 [ig/ml/h.
`Administration of radiolabelled docetaxel led to a
`rapid diffusion into all tissues except those of the CNS,
`with the highest levels seen in liver, bile, intestines and
`gastric contents. The plasma protein binding in mice
`was 76%-89%, and the elimination was almost com-
`plete at 96 hours after administration. In the dog and
`mouse the primary route of elimination of radiola-
`belled docetaxel was by hepatic extraction and biliary
`excretion, while urinary excretion was < 10% [33].
`
`Phase I clinical trials
`
`Phase I trials on paclitaxel were begun in 1983, but a
`high incidence of severe hypersensitivity reactions led
`to the premature closure of many of them. The hyper-
`sensitivity reactions occurred more frequently when
`paclitaxel was infused over a short period. Therefore,
`the infusion durations were extended to 6 or 24 hours
`and empiric prophylactic anti-allergic medication con-
`sisting of corticosteroids, antihistamines and 5HT2
`antagonists were given. Only quite recently has sugges-
`tive evidence been obtained that extension of the infu-
`sion duration might not have been necessary with the
`
`

`
`Phase I studies combining paclitaxel with other cyto-
`toxic agents have already been performed and many
`are still in progress. In view of the marginal overlap in
`toxicities and the suggested additive effect observed in
`vitro, cisplatin was the first drug added to paclitaxel
`[45]. The MTD of this combination was 200 mg/m2 of
`paclitaxel and 75 mg/m2 of cisplatin, given every 3
`weeks. The dose-limiting toxicity was again neutro-
`penia with a sequence dependency and was more pro-
`nounced when paclitaxel was administered after cis-
`platin. On the basis of pharmacokinetic data it was
`suggested that this interaction was due to a decrease in
`paclitaxel clearance when cisplatin preceded paclitaxel,
`and it is therefore recommended that paclitaxel be
`administered before cisplatin. With the addition of
`G-CSF to this regimen a dose of 250 mg/m2 paclitaxel
`with 75 mg/m2 of cisplatin was found to be feasible for
`repeated administrations [46]. With the combination of
`these two drugs neuromuscular toxicity becomes an
`important side effect [45, 46]. Ah" further combination
`regimens including paclitaxel have routinely used
`G-CSF.
`Combining paclitaxel with doxorubicin creates ma-
`jor difficulties. Sequential [47] and concurrent [48]
`administrations have been studied. With sequential
`administration the MTD was 60 mg/m2 of doxorubicin
`preceded by 125 mg/m2 of paclitaxel. The most effi-
`cacious dose has not yet been defined. Dose-limiting
`toxicities were severe mucositis and neutropenia with
`infection [47]. With a concurrent 72-hour administra-
`tion the dose-limiting toxicity is diarrhea with abdomi-
`nal pain. The MTD was 60 mg/m2 of doxorubicin with
`180 mg/m2 of paclitaxel [48], but also at lower dose
`levels the number of hospital admissions for side ef-
`fects is impressive, and whether this schedule should be
`further explored is questionable.
`The data on the combination of paclitaxel with
`cyclophosphamide are still limited [49]. As for cisplatin
`there appears to be sequence-specific toxicity. The
`MTD in this study had not yet been reached by the time
`of the first publication, while toxicities mainly included
`neutropenia with fever. An ongoing study at the NCI is
`adding cisplatin to the above combination, and admin-
`istration of a high dose of all three drugs appears fea-
`sible (44].
`Finally, several phase I studies combining paclitaxel
`with carboplatin are ongoing. One study, using target
`AUC for dosing of carboplatin, has reached MTD [51].
`The dose-limiting toxicity was myelosuppression with a
`suggestion of cumulation. The MTD was defined as 135
`mg/m2 of paclitaxel with AUC - 7.5 for carboplatin.
`In the phase I studies on docetaxel various schedules
`of drug administration have been tested (Table 2) [52-
`57]. The major dose-limiting toxic effect throughout
`these studies was an early-onset, short-lasting, dose-
`dependent, schedule-independent and non-cumulative
`neutropenia. Other side effects were mild paresthesias,
`infrequent hypersensitivity reactions, general alopecia,
`nausea, vomiting, diarrhea and skin reactions (ery-
`
`498
`
`introduction of the premedication, which is one of the
`reasons that the optimal infusion duration has yet to be
`determined. The completed phase I studies, with dose-
`limiting toxicities, maximum tolerated doses and re-
`commended doses for phase II studies are listed in
`Table 1 [34-43].
`The major dose-limiting side effect of paclitaxel
`was neutropenia. Thrombocytopenia was uncommon.
`Other frequently occurring side effects were dose-de-
`pendent mucositis and peripheral neuropathy, arthral-
`gias/myalgias, hypersensitivity reactions, alopecia, nau-
`sea, vomiting and cardiac rhythm abnormalities. Fol-
`low-up studies in the United States have all been
`performed using 24-hour infusion schedules and in
`Europe shorter infusion times have also been studied.
`Patients with leukemias appear to tolerate a much
`higher dose [42], with mucositis being dose-limiting at
`the dose of 390 mg/m2. The recommended dose for
`phase II studies in leukemias is 315 mg/m2. Using a
`120-hour infusion every 3 weeks the MTD was 36 mg/
`m2/day (total dose/course 180 mg/m2), and the recom-
`mended dose for phase U studies 30 mg/m2/day (total
`dose per course 150 mg/m2) [43]. Again, the dose-lim-
`iting toxicity was neutropenia. With this schedule there
`were no neuro or cardiac toxic effects, which can partly,
`but certainly not fully, be explained by the dose given.
`Responses were reported in all phase I studies with
`paclitaxel.
`As neutropenia is the major dose-limiting side ef-
`fect, attempts have been made to increase dose inten-
`sity by adding granulocyte-colony stimulating factor
`(G-CSF). On a 3-weekly schedule the gain was limited,
`with the MTD of a 24-hour infusion being raised to
`300 mg/m2 [44]. However, with interval shortening, a
`worthwhile increase in dose intensity might be achiev-
`able and this question is the subject on ongoing studies.
`
`Table 1. Phase I studies of paclitaxel.
`
`Infusion
`duration
`(his)
`
`Rec. dose/
`MTD
`
`DLT
`
`Refer-
`ence
`
`NR
`NR/180
`210/255
`225/275
`280/275
`
`170/200
`250/275
`
`250/300
`315/390
`
`150/180
`
`20x5/40x5
`NR
`30x5/40x5
`
`HSR
`HSR
`Neutropenia
`Neutropenia
`Neutropenia/
`PNP
`Neutropenia
`PNP
`
`PNP
`Mucositis/
`PNP
`Neutropenia/
`mucositis
`Neutropenia
`HSR
`Neutropenia
`
`34
`41
`34
`35
`36
`
`40
`37
`
`44
`42
`
`43
`
`38
`39
`39
`
`1366 6
`
`24
`24
`
`24
`24
`
`120
`
`1 1 6
`
`Dose/
`schedule
`
`1 x q 3 wks
`1 x q 3 wks
`1 x q 3 wks
`1 x q 3 wks
`1 x q 3 wks
`
`1 x q 3 wks
`1 x q 3 wks
`1 x q 3 wks
`•G-CSF
`1 x q 3 wks
`(leukemia)
`1 x q 3 wks
`
`D x 5 q 3 wks
`D x 5 q 4 wks
`D x 5 q 4 wks
`
`HSR - hypersensitivity reaction, PNP - peripheral neuropathy,
`NR - not reported, DLT - dose-limiting toxicity.
`
`

`
`499
`
`phases [34-37, 39, 42]. It has been suggested that the
`pharmacokinetics of paclitaxel may be non-linear [61].
`In patients treated by 3-hour infusions, proportionally
`greater increases in maximum drug levels at the end of
`the infusion ( C ^) and AUC in relation to dose were
`observed. Since clearance decreased with increasing
`dose, a non-linear elimination process was indicated.
`The investigators fitted the pharmacokinetic data to a
`2-compartment open model with a saturable second-
`order process from the central to the peripheral com-
`partments and first-order processes for return from the
`peripheral compartment and central elimination. Thus,
`changes in schedule may result in different total drug
`exposures.
`Although the drug is extensively protein-bound, it is
`rapidly eliminated from the plasma. The urinary excre-
`tion is only 1.4%-6.6%, thus making a minor contribu-
`tion to the systemic clearance which therefore should
`be related to metabolism, biliary excretion, or tissue
`binding [9].
`Unlike as is suggested for paclitaxel, docetaxel dis-
`position is not schedule-dependent. Pharmacokinetics
`are linear up to doses of 115 mg/m2 [52, 55, 56]. The
`disposition was triphasic [52] and interpatient variabil-
`ity appears limited whereas intrapatient variability is
`considerable [55]. As for paclitaxel, docetaxel is largely
`and rapidly protein-bound [62] and its urinary excre-
`tion is limited to 9%.
`For clinical use paclitaxel is formulated a.o in 50%
`cremophor EL. This vehicle has been shown to block
`the P-glycoprotein drug efflux pump in vitro. Interest-
`ingly, a pharmacokinetic study in patients treated with a
`3-hour infusion of paclitaxel at a dose of 135-175 mg/
`m2 has shown that the concentrations of cremophor EL
`found in plasma were sufficient to inhibit PGP in vitro,
`suggesting that this effect may contribute to the ob-
`served antitumor effect of paclitaxel [63].
`
`Phase II clinical trials
`
`In view of the responses observed with paclitaxel in
`phase I studies in chemotherapy-refractory patients
`with ovarian cancer, the initial focus in phase II studies
`was on this tumor type, although phase II studies in a
`wide variety of tumors have now been performed. One
`of the difficulties in summarizing these studies is the
`fact that different doses were applied in addition to dif-
`ferent infusion durations, although the majority of
`studies were performed with the 24-hour infusion. For
`docetaxel the phase II programme is less focused on a
`specific tumor type, and all studies to the present have
`used only one dose and one duration of infusion, facili-
`tating the interpretation of the results.
`It should be noted that many of the docetaxel data
`cited in the following are from abstracts. Data in the
`final papers to be published in the coming year might
`differ to some extent. Also, more mature data on pacli-
`taxel from randomized trials will soon be provided.
`
`Table 2. Phase I studies of docetaxel.
`
`Rec. dose/
`MTD
`
`DLT
`
`Refer-
`ence
`
`Neutropenia,
`skin
`Neutropenia,
`skin
`Neutropenia,
`skin,
`mucositis
`Neutropenia,
`mucositis
`Neutropenia,
`asthenia
`Neutropenia,
`mucositis
`
`52
`
`57
`
`54
`
`55
`
`53
`
`56
`
`100/115
`
`100/ND
`
`30/115
`
`Infusion
`duration
`(hrs)
`
`1-2
`
`1 2
`
`-5
`
`Dose/
`schedule
`
`1 x q 3 wks
`
`1 x q 3 wks
`
`1 x q 3 wks
`
`1 x q 3 wks
`
`24
`
`70/90
`
`110/110
`
`5x14/5x15
`
`1 1
`
`Dl-3 q 3 wks
`
`D x 5 q 3 wks
`
`DLT - dose-limiting toxicity, ND — not done.
`
`thema, desquamation and nail changes) and incidental
`fluid retention. Mucositis was mainly seen with longer
`durations of infusion, and was the main cause of com-
`plicated neutropenia. Cardiac toxicity was not report-
`ed.
`Responses were reported in different tumor types.
`Based on considerations such as dose intensity, toxicity
`profile and absence of preclinical data suggesting any
`schedule dependency, the recommended dose and
`schedule for phase II studies was 100 mg/m2 given as a
`1-hour infusion every 3 weeks, without prophylactic
`measures for hypersensitivity reactions and no prophy-
`lactic antiemetics.
`Phase I studies combining docetaxel with other cyto-
`toxic drugs are underway. It appears feasible to com-
`bine high doses of docetaxel and cisplatin [58]. In con-
`trast to paclitaxel [45], there seems to be no sequence-
`related difference in side effects, and docetaxel did not
`change the pharmacokinetics of cisplatin [59].
`
`Clinical pharmacology
`
`HPLC assays for both paclitaxel [60] and docetaxel [31]
`were used for pharmacokinetics in several phase I
`trials. The pharmacokinetic parameters of paclitaxel
`and docetaxel are summarized in Table 3.
`The disposition of paclitaxel in plasma has been
`characterized by a biphasic model with a rather wide
`variability between the distribution and elimination
`
`Table 3. Pharmacokinetic parameters of paclitaxel and docetaxel.
`
`Parameter
`
`Paclitaxel
`
`Docetaxel
`
`0.1
`0.6
`12.2
`350
`73
`<9
`
`0.29
`
`5-
`
`496
`110
`4.8
`
`V/i a (hrs)
`T'/2 B (hrs)
`TV2 Y (hrs)
`CL (ml/min/m2)
`VD (1/m2)
`UEXCR
`(%)
`
`

`
`500
`
`Ovarian cancer (Table 4)
`
`Table 4. Phase II studies in ovarian cancer.
`
`Drug
`
`Taxol
`
`Taxotere
`
`Dose
`(mg/m2)
`
`135
`175
`250
`100
`
`N
`
`862
`226
`91
`120
`
`RR(%)
`
`References
`
`20
`25
`36
`30
`
`67-69
`66,69
`64, 65, 70
`72-74
`
`gest a major role for paclitaxel in the first-line chemo-
`therapy of ovarian cancer.
`Phase HI studies with docetaxel have not yet been
`performed.
`
`Breast cancer (Table 5)
`
`Both taxanes have been extensively studied in meta-
`static breast cancer although phase HI studies have not
`yet been reported. The first reported phase II study on
`paclitaxel [75] indicated an overall 56% response rate
`in 25 evaluable patients, 14 of whom received only
`prior adjuvant chemotherapy while 11 had previously
`been treated with chemotherapy for metastatic disease.
`For interpretation purposes, in the following we will
`take into account only prior chemotherapy for meta-
`static disease; thus, first-line treatment in metastatic
`disease combines the data obtained in patients with or
`without prior adjuvant chemotherapy. In most studies
`the percentage of patients previously treated with adju-
`vant chemotherapy ranges between 25% and 40%.
`Using a dose of 250 mg/m2 as a 24-hour infusion the
`investigators observed 3 complete and 11 partial re-
`sponses, with a median response duration of 5+ months
`[75]. A confirmatory study using the same dose with
`the addition of G-CSF resulted in an overall response
`rate of 62% in 26 patients receiving the drug as first-
`line chemotherapy [76]. Both studies were performed
`as single-institution studies. In a large multicenter mul-
`tinational European study including 471 patients,
`lower doses of paclitaxel were administered and com-
`pared (135 and 175 mg/m2, respectively) using a
`
`Table 5. Phase II studies in breast cancer.
`
`Line
`
`Paclitaxel
`
`Docetaxel
`
`References
`
`77
`77
`75, 76, 79,
`81-83
`77
`77
`75,79,81-85
`78
`79
`
`RR
`(%)
`
`-
`71
`
`_ _
`
`62
`
`_ -
`
`N
`
`_ -
`
`95
`
`—
`31
`
`— -
`
`RR
`(%)
`
`29
`35
`56
`
`16
`24
`47
`
`25
`27
`
`N
`
`_ -
`
`40
`
`_ -
`
`36
`12
`51
`
`Dose
`(mg/m2)
`
`135
`175
`250
`
`135
`175
`250
`
`150
`250
`
`1"
`
`2Dd
`
`>3rd
`
`Many phase II trials with paclitaxel have been per-
`formed and reported [64-71]. As stated, the doses
`used in these studies were rather diverse and the strata
`not similar, thus preventing formulation of a straight
`foreward conclusion.
`In the first reported study [64] using paclitaxel doses
`of 110-250 mg/m2, 12 of 40 patients (30%) respond-
`ed, including 6 (24%) of 25 patients progressing during
`cisplatin treatment, and 6 (40%) of 15 patients progres-
`sing after cisplatin-free intervals of at least 6 months.
`Two other early studies [65, 66] yielded overall re-
`sponse rates of 20% and 29%. Because overall re-
`sponse throughout the studies is the easiest to compare,
`and detailed information on responses in the different
`strata is frequently not yet published, all further data
`and those in Table 4 are given as overall response data.
`As the table indicates, there appears to be some extent
`of a dose-response relationship, with response rates
`varying between 19% with a dose of 135 mg/m2, to
`36% with a dose of 250 mg/m2. Nevertheless, the data
`are still limited by patient numbers and we await ran-
`domized study results before drawing definitive con-
`clusions.
`The mean response duration at all dose levels is
`approximately 7 months. However, a large multination-
`al multicenter trial with a bifactorial design, comparing
`135 mg/m2 and 175 mg/m2, and 3-hour and 24-hour
`infusions in a total of 312 patients [69] showed no sig-
`nificant difference in the response rates of the two
`doses or infusion schedules in the first 286 patients
`evaluable for response. Even more important, this
`study showed that toxic effects in 298 evaluable pa-
`tients were more pronounced with the longer infusion
`duration, irrespective of dose, which suggests that the
`shorter infusion may be the most appropriate. Unfor-
`tunately, as already indicated, the majority of studies
`have been performed with the 24-hour infusion sched-
`ule.
`Docetaxel at the dose of 100 mg/m2 has been the
`topic of three studies [72-74]. The overall reported re-
`sponse rate in 120 patients was 30%, similar to the
`results with paclitaxel in this disease.
`In a phase HI study the Gynecology Oncology
`Group (GOG) [71] has compared the combination of
`paclitaxel 135 mg/m2 and cisplatin 75 mg/m2 (CT)
`every 3 weeks for 6 cycles with the combination of
`cyclophosphamide 750 mg/m2 and cisplatin 75 mg/m2
`(CP) in 388 eligible previously untreated patients with
`stage HI-TV debulked ovarian cancer with residual dis-
`ease of >1 cm. The response rate was significantly
`(p < 0.01) better for CT (79%) than for CP (63%), as
`was survival (17.9 months and 13.8 months respec-
`tively), but neutropenia was also significantly more pro-
`nounced with CT and cardiotoxicity was seen only with
`this regimen. Nevertheless, these data are important
`and if they are confirmed in similar studies and if sur-
`vival differences persist with longer follow-up they sug-
`
`

`
`501
`
`Refer-
`ences
`
`95,96
`88-92
`93
`100,101
`97-99
`102, 103
`104
`108
`105,107
`
`91
`94
`112
`
`Table 7. Phase II studies in other types of tumor.
`
`Tumor type
`
`Paclitaxel
`
`Docetaxel
`
`Dose
`(mg/m2)
`
`N
`
`N
`
`RR
`(%)
`
`RR
`(%)
`
`37
`28
`_
`26
`25
`21
`
`9_5
`
`33
`28
`21
`
`38
`76
`_
`34
`28
`24
`33
`-
`43
`
`18
`18
`28
`
`43
`22
`34
`5
`14
`13
`
`050 ---
`
`28
`49
`32
`22
`87
`23
`19
`22
`18
`
`---
`
`First line chemotherapy
`Head/neck cancer
`250
`NSCLC
`200-250
`250
`SCLC
`Gastric cancer
`250
`Melanoma
`200-250
`Pancreatic cancer
`250
`Colorectal cancer
`250
`Prostate cancer
`135-170
`Renal cancer
`250
`Second-line chemotherapy
`NSCLC
`SCLC
`Soft-tissue sarcoma
`
`---
`
`tients [91], further underscoring the importance of the
`latter observation. Although the median duration of re-
`sponse was relatively short, the lack of active drugs in
`the treatment of this disease suggests that studies com-
`bining taxanes with drugs such as cisplatin should be
`assigned priority.
`Data on the activity of taxanes in small-cell lung can-
`cer are rather limited. Paclitaxel was tested as first-line
`chemotherapy in 32 patients, yielding a response rate
`of 34% [93], while docetaxel was tested in second-line
`chemotherapy in 18 patients, resulting in a response
`rate of 28% [94].
`In head and neck cancer both paclitaxel and doce-
`taxel are very active, yielding response rates of 43%
`and 44%, respectively [95, 96], in a disease stage that
`usually responds poorly to chemotherapy and where
`methotrexate is still considered the standard drug.
`Recent studies with the latter drug indicated a response
`rate of only 15% in a similar patient population.
`In 2 malignant melanoma studies including a total of
`87 patients, paclitaxel yielded a response rate of 14%
`[97, 98], while docetaxel, in a not yet confirmed study,
`appears to be slightly more active, with a response rate
`of 25% [99]. Again, it should be noted that there are no
`randomized comparisons.
`There also appears to be a suggestion that docetaxel
`has higher potency in gastric cancer while paclitaxel at
`a dose of 250 mg/m2 yielded a disappointing response
`rate of 5% in non-pretreated patients [100], docetaxel
`resulted in 26% responses in this patient population
`[101].
`In pancreatic cancer the respective response rates of
`paclitaxel and docetaxel were 13% and 21% [102,103].
`In colorectal, renal and prostatic cancer no activity of
`any importance was observed [104-108].
`Finally, docetaxel was identified as one of the few
`active drugs in soft tissue sarcomas [109].
`
`3-hour infusion. Response rates have sometimes been
`reported without giving patient numbers [77], but they
`seem to indicate some dose-response relationship
`(Table 5). Secon