Download Phase I and Pharmacokinetic Study of the Oral Fluoropyrimidine

P h a se I a n d P h a rm a c o ki n e t i c S t u dy o f t h e O r a l
F l uo ro p y rimid in e Cape c i t a bi n e i n C o m bi n a t i o n W i t h
P a c l i ta x e l in P a tie n ts W i t h Adv a n c e d S o l i d Ma l i g n a n c ies
By Miguel A. Villalona-Calero, Geoffrey R. Weiss, Howard A. Burris, Maura Kraynak, Gladys Rodrigues,
Ronald L. Drengler, S. Gail Eckhardt, Bruno Reigner, Judy Moczygemba, Hans Ulrich Burger, Tom Griffin,
Daniel D. Von Hoff, and Eric K. Rowinsky
Purpose: To evaluate the feasibility of administering
the oral fluoropyrimidine capecitabine in combination
with paclitaxel, to characterize the principal toxicities of
the combination, to recommend doses for subsequent
disease-directed studies, and to determine whether significant pharmacokinetic interactions occur between
these agents when combined.
Patients and Methods: Sixty-six courses of capecitabine and paclitaxel were administered to 17 patients in a two-stage dose-escalation study. Paclitaxel was
administered as a 3-hour intravenous (IV) infusion every 3
weeks, and capecitabine was administered continuously
as two divided daily doses. During stage I, capecitabine
was escalated to a target dose of 1,657 mg/m2/d, whereas
the paclitaxel dose was fixed at 135 mg/m2. In stage II,
paclitaxel was increased to a target dose of 175 mg/m2,
and the capecitabine dose was the maximum established
in stage I. Pharmacokinetics were characterized for each
drug when given alone and concurrently.
Results: Myelosuppression, predominately neutropenia, was the principal dose-limiting toxicity (DLT). Other
toxicities included hand-foot syndrome, diarrhea, hyperbilirubinemia, skin rash, myalgia, and arthralgia. Two
patients treated with capecitabine 1,657 mg/m2/d and
paclitaxel 175 mg/m2 developed DLTs, whereas none of
six patients treated with capecitabine 1,331 mg/m2/d
and paclitaxel 175 mg/m2 developed DLTs during course
1. Pharmacokinetic studies indicated that capecitabine
and paclitaxel did not affect the pharmacokinetic behavior of each other. No major antitumor responses were
noted.
Conclusion: Recommended combination doses of
continuous capecitabine and paclitaxel are capecitabine 1,331 mg/m2/d and paclitaxel 175 mg/m2/d
IV every 3 weeks. Favorable preclinical mechanistic
interactions between capecitabine and paclitaxel, as
well as an acceptable toxicity profile without clinically
relevant pharmacokinetic interactions, support the performance of disease-directed evaluations of this combination.
J Clin Oncol 17:1915-1925. r 1999 by American
Society of Clinical Oncology.
T
58-DFUR was shown to be much more than that of 5-FU,
and 58-DFUR was shown to possess potent anticachectic and
antimetastatic activity that was not noted with 5-FU.3,4,8,9 In
clinical studies, high and predictable oral bioavailability was
noted, and 58-DFUR showed prominent antitumor activity in
patients with breast, colorectal, and gastric cancers.10-12
However, diarrhea caused by the generation of 5-FU in
gastrointestinal tissues that also highly express pyrimidine
nucleoside phosphorylases limited the development of 58-
HE ORAL FLUOROPYRIMIDINE capecitabine (Xeloda; Hoffmann-LaRoche, Inc, Nutley, NJ; N4pentyloxycarbonyl-58-deoxy-5-fluorocytidine) was rationally synthesized to be efficiently absorbed from the
gastrointestinal tract as a prodrug and converted to fluorouracil (5-FU), preferentially in neoplastic tissues (Fig 1).1,2
After gastrointestinal absorption, capecitabine is metabolized to 58-deoxy-5-fluorocytidine (58-DFCR; Furtulon;
Roche, Basel, Switzerland) by hepatic carboxyl esterase,
and this intermediate is further metabolized to doxifluridine
(58-DFUR) by cytidine deaminase in both hepatic and
extrahepatic tissues, including malignant tumors.3,4 58DFUR is subsequently metabolized to 5-FU by the enzyme
thymidine phosphorylase (dThdPase), which is one of two
types of pyrimidine nucleoside phosphorylases.5 dThdPase
has recently been shown to be a potent tumor-associated
angiogenesis factor with higher expression in tumor cells than
normal cells, which may account for the intriguing results
reported by some investigators that malignant cells preferentially
convert 58-DFUR to 5-FU.6,7
The immediate precursor to capecitabine, 58-DFUR, underwent development as an oral fluoropyrimidine in the
1980s.3,4 In preclinical evaluations, the therapeutic index of
From the Institute for Drug Development, Cancer Therapy and
Research Center, and The University of Texas Health Science Center at
San Antonio, San Antonio; Brooke Army Medical Center, Fort Sam
Houston, TX; and Hoffmann-LaRoche, Inc, Nutley, NJ.
Submitted September 24, 1998; accepted February 11, 1999.
Supported in part by a grant from Hoffmann-LaRoche, Inc, and grant
no. MO1 RR01346 from the National Institutes of Health to the Frederic
C. Bartter Clinical Research Unit of The Audie Murphy Veterans
Administration Hospital.
Address reprint requests to Miguel A. Villalona-Calero, MD, Division
of Hematology Oncology, Ohio State University, B406 Starling-Louing
Hall, Columbus, OH 43210; email [email protected].
r 1999 by American Society of Clinical Oncology.
0732-183X/99/1706-1915
Journal of Clinical Oncology, Vol 17, No 6 (June), 1999: pp 1915-1925
1915
1916
VILLALONA-CALERO ET AL
Fig 1. The metabolism of capecitabine. After oral administration, capecitabine is metabolized to
5ⴕ-DFCR by acylamidase in the liver.
Cytidine deaminase, which is found
in both hepatic and extrahepatic
tissues, further metabolizes 5ⴕ-DFCR
to 5ⴕ-DFUR. dThdPase, which may
be preferentially expressed in neoplastic tissue and upregulated by
prior treatment with paclitaxel, metabolizes 5ⴕ-DFUR to 5-FU.
DFUR, although it is licensed in Japan.10 In contrast, the
carbamate structure of capecitabine precludes its conversion
to 5-FU in the intestines; therefore, capecitabine is much less
likely to induce diarrhea than 58-DFUR. Indeed, capecitabine administered daily for 3 weeks produced less gastrointestinal toxicity in animals and was more efficacious against the
human colon cancer xenografts HCT116 and CXF280
across a broader dosing range than 5-FU administered either
intravenously (IV) (three times weekly for 2 weeks) or
orally (daily for 3 weeks), or 58-DFUR given orally daily for
3 weeks.1,2 At equitoxic doses, treatment with capecitabine
also resulted in substantially greater intratumoral exposure
to 5-FU compared with 5-FU itself administered intraperitoneally.13
In early phase I and pharmacologic studies, capecitabine
showed efficient and predictable gastrointestinal absorption,
with 61% of the administered dose recovered in urine as
drug-related material, and the agent was rapidly converted to
58-DFCR and 58-DFUR (time to peak plasma concentration
[tmax], 30 minutes).14 In addition, 5-FU concentrations were
2.9-fold higher, on average, in malignant tissues compared
with nonmalignant tissues.14 Diarrhea and hand-foot syndrome have been the principal toxicities of capecitabine, and
severe myelosuppression has been uncommon on protracted
dosing schedules.2,15,16 Recommended phase II dose sched-
ules for capecitabine have included 1,331 mg/m2/d in two
divided doses daily for 12 weeks and 2,510 mg/m2/d in two
divided doses daily for 2 weeks every 3 weeks. Capecitabine
on the latter schedule has recently received regulatory
approval in the United States for treating patients with
paclitaxel-refractory metastatic breast cancer.
Because a principal determinant of the therapeutic index
of capecitabine-based treatment is the differential rate of
conversion of 58-DFUR to 5-FU in malignant and nonmalignant tissues, therapeutic strategies have focused on maximizing the ratio of dThdPase activity in malignant versus
nonmalignant tissues. Transfection of the dThdPase gene
into the DLD-1 human colon cancer cell line, which is
dThdPase-deficient and highly resistant to 58-DFUR, increases dThdPase activity by 1,068-fold and sensitivity to
58-DFUR by 1,070-fold.17 In addition, treatment of malignant tumors with various cytokines, such as tumor necrosis
factor-alpha, interleukin-1-alpha, and interferon-gamma
(IFN-␥), increases intratumoral dThdPase activity and enhances tumor sensitivity to 58-DFUR in vitro and in
vivo.17-19 For example, treatment of nude mice bearing the
human colon cancer xenograft COLO205 with human
IFN-␥ resulted in an 8.7-fold increase in dThdPase activity
in the tumors.19 Although IFN-␥ alone was ineffective
against this tumor, it substantially increased the activity of
1917
PHASE I STUDY OF PACLITAXEL AND CAPECITABINE
capecitabine, but did so only slightly for 5-FU. Furthermore,
treatment of the nude mice bearing a capecitabine-resistant
human colon cancer xenograft with paclitaxel, which has
been shown to upregulate both interleukin-1-alpha and
tumor necrosis factor-alpha, resulted in a 7.9-fold increase in
intratumoral dThdPase activity.19,20 In this model, dThdPase
activity increased 4 days after treatment with paclitaxel and
remained elevated for up to 10 days (Ishitsuka et al,
unpublished results). Paclitaxel, capecitabine, or 5-FU alone
did not appreciably inhibit tumor growth, whereas the
combination of 5-FU and paclitaxel was additive, and the
combination of capecitabine and paclitaxel produced synergy.
To date, these results with capecitabine provide both
mechanistic and clinical rationale to evaluate the feasibility
of administering it in combination with paclitaxel. In
addition, the overlapping antitumor activities of both singleagent capecitabine and paclitaxel in several tumor types,
particularly breast cancer, as well as the nonoverlapping
principal toxicities of these agents, support a rationale for
the evaluation of this combination regimen. The principal
objectives of this phase I and pharmacokinetic study were as
follows: (1) to determine the maximum-tolerated doses
(MTDs) of capecitabine administered twice daily in combination with paclitaxel as a 3-hour IV infusion every 3 weeks
to patients with advanced solid malignancies; (2) to describe
and quantitate the toxicities of the capecitabine-paclitaxel
regimen; (3) to characterize the individual pharmacokinetic
behavior of both paclitaxel and capecitabine and determine
whether significant pharmacokinetic interactions occur when
the agents are administered in combination; and (4) to seek
preliminary evidence of antitumor activity for this combination.
PATIENTS AND METHODS
Eligibility
Patients with histologically confirmed solid malignancies refractory
to conventional therapy or for whom no effective therapy existed were
candidates for this study. Eligibility criteria also included the following:
age ⱖ 18 years; Karnofsky performance status ⱖ 70% (ambulatory and
capable of self-care); a life expectancy ⱖ 3 months; no major surgery,
radiotherapy, or chemotherapy within 28 days (42 days for mitomycin
or nitrosourea); no history of major disorders or surgery involving the
stomach, small intestines, liver, or kidney that might affect the
gastrointestinal absorption or clearance of capecitabine; adequate
hematopoietic (WBC count ⱖ 3,000/µL, absolute neutrophil count
[ANC] ⱖ 1,500/µL, platelet count ⱖ 100,000/µL, and hemoglobin
level ⱖ 9.0 g/dL), hepatic (total bilirubin level ⬍ 1.5 times the upper
normal limits; AST, ALT, and alkaline phosphatase ⬍ 2.5 times the
upper normal limits), and renal (serum creatinine ⬍ 1.5 times the upper
normal limits) functions; no clinically significant cardiac disease (New
York Heart Association class III or IV); no history of seizures, CNS
disorders, or psychiatric disease that might affect study compliance; no
evidence of brain metastases; no history of hypersensitivity to paclitaxel
or other drugs formulated in Cremophor EL (polyoxyethylated castor
oil); no requirement for chronic corticosteroids except for inhalation
therapy; and no prior documentation of hepatitis B surface antigen,
hepatitis C antibodies, or human immunodeficiency virus type 1
antibodies, although such viral evaluations were not required for
eligibility. All patients gave informed written consent before treatment.
Dosage and Dose Escalation
The starting doses were paclitaxel 135 mg/m2 as a 3-hour IV infusion
and capecitabine 1,004 mg/m2/d orally in two divided doses every 12
hours. These dose schedules resulted in antitumor activity and minimal
toxicity when either drug was administered as a single agent in similar
patient populations.2,21 Paclitaxel was administered on day 1 and
repeated every 3 weeks. Capecitabine was begun on day 3 after the first
paclitaxel treatment and continued for the duration of the study. The
capecitabine dose levels evaluated were identical to those evaluated in
phase I single-agent studies to facilitate comparisons of results of
combination and single-agent studies. Doses of either capecitabine or
paclitaxel were escalated in each successive cohort of new patients in
two stages. In stage I, the goal was to progressively increase the dose of
capecitabine from 1,004 mg/m2/d to 1,331 mg/m2/d and then to the
target dose of 1,657 mg/m2/d, which was the MTD established on a
continuous daily dosing schedule,2 whereas the dose of paclitaxel was
fixed at 135 mg/m2. No dose escalation greater than 1,657 mg/m2 in the
capecitabine dose was to be undertaken. In stage II, the goal was to
increase the dose of paclitaxel from 135 mg/m2 to a clinically relevant
target dose of 175 mg/m2, with the initial dose of capecitabine being the
highest safest dose level established in stage I. A minimum of three new
patients were to be treated at all tolerable dose levels. Intrapatient dose
escalation was not permitted. Toxicities were graded according to the
National Cancer Institute Common Toxicity Criteria.22 Dose-limiting
toxicity (DLT) was defined as ANC less than 500/µL for longer than 5
days or less than 1,000/µL with fever; platelet count less than
25,000/µL; hemoglobin level less than 6.5 mg/dL; and nonhematologic
toxicity of at least grade 3 (including diarrhea, nausea, and vomiting)
that did not improve to at least grade 1 within 2 days after the institution
of appropriate therapy, or that led to the interruption of capecitabine
treatment or delayed paclitaxel treatment for more than 2 weeks. If
DLTs occurred, a maximum of six new patients was treated at that dose
level. The MTD level was defined as the highest capecitabine dose up to
1,657 mg/m2/d, which when combined with paclitaxel 175 mg/m2,
induced DLTs in less than two of six new patients during their first cycle
of therapy.
Drug Administration
Capecitabine was supplied by Hoffmann-LaRoche, Inc (Nutley, NJ)
as 100-, 150-, 500-, or 750-mg film-coated tablets. After the total dose
was calculated according to body surface area, it was then rounded off
to the nearest 50 mg. The precise dose of capecitabine, which was
packed individually in a mediplanner dispensing cassette (Health Care
Logistics, Circleville, OH), was administered with 200 mL of water
within 30 minutes of ingesting solid food. Capecitabine was to be taken
every 12 ⫾ 2 hours. Paclitaxel (Taxol; Bristol-Myers Squibb, Princeton,
NJ) was supplied in 30-mg (5-mL) vials, with each milliliter containing
6 mg of paclitaxel, 527 mg of polyoxyethylated castor oil (Cremophor
EL; BASF, Aktiengesellschaft, Germany), and 49.5% dehydrated
alcohol (United States Pharmacopeia). Paclitaxel was diluted with
either 0.9% sodium chloride solution or 5% dextrose solution to a final
concentration of 0.3 to 1.2 mg/mL administered as an IV infusion over 3
hours using a Harvard pump (Harvard Apparatus Inc, South Natick,
MA). The following medication was administered before paclitaxel:
1918
dexamethasone 20 mg orally 6 and 12 hours before treatment;
diphenhydramine 50 mg IV 30 to 60 minutes before treatment; and
either cimetidine 300 mg IV or ranitidine 50 mg IV 30 to 60 minutes
before treatment.
Dosage Modifications
Treatment with capecitabine was discontinued if patients developed
at least grade 2 nonhematologic (except isolated hyperbilirubinemia) or
grade 3 hematologic toxicity. Treatment was resumed when toxicity
resolved to grade 0 to 1 at either the original dose level (for grade 2
nonhematologic toxicity; grades 2 to 4 hematologic toxicity; grade 3
nausea and vomiting; and grade 3 palmar-plantar erythrodysesthesia) or
at the next lower dose level for all other grade 3 nonhematologic
toxicities. Patients were to be withdrawn from study if they developed
any grade 4 nonhematologic toxicity, unless an objective response was
documented, in which case treatment with capecitabine at the next
lower dose level was continued. If patients had persistent toxicity of at
least grade 2 in severity on the scheduled day of paclitaxel administration, treatment was delayed for 1 week. If the toxicity did not resolve to
ⱕ grade 1 after a 1-week treatment delay, the patient was withdrawn
from study unless clinical benefit was documented, in which case a
treatment delay of up to 2 additional weeks was permitted. For patients
who experienced DLTs during the previous course, the paclitaxel dose
was reduced to the next lower dose level or by 25% if DLTs occurred at
the 135-mg/m2 dose level.
Pretreatment Assessment and Follow-Up Studies
Histories, physical examinations, and routine laboratory studies were
performed before treatment and weekly after treatment. Routine
laboratory studies included serum electrolytes, chemistries, complete
blood cell count with differential WBC count, clotting times, and
urinalysis. If patients developed toxicity manifested by grade 3 or 4
abnormalities in hematologic or biochemical laboratory parameters, the
tests were repeated immediately and then daily until the toxicity
resolved. Tumors were measured after every two courses, and treatment
was continued in the absence of progressive disease or intolerable
toxicity. A complete response was defined as the disappearance of all
disease on two measurements separated by a minimum of 4 weeks. A
partial response required more than 50% reduction in the sum of the
products of the bidimensional measurements of all measurable lesions
documented by two measurements separated by at least 4 weeks.
Plasma Sampling and Assay
To assess the effects of capecitabine on the pharmacokinetics of
paclitaxel, blood was sampled for analysis of paclitaxel concentrations
after treatment on day 1 (before treatment with capecitabine) and day 22
(during treatment with capecitabine). Samples were collected from
indwelling venous catheters in the arm contralateral to the paclitaxel
infusion. For the analysis of paclitaxel, sampling times included
pretreatment, 2 hours into the infusion and immediately before the end
of the 3-hour infusion, and 3.5, 4, 5, 8, 11, 23, 30, and 48 hours after the
start of infusion. Paclitaxel plasma concentrations were measured by
liquid chromatography/mass spectroscopy as described previously.23
For analysis of capecitabine and capecitabine metabolites, blood
samples were collected before capecitabine treatment and at 0.5, 1, 2, 3,
4, 5, 6, 8, and 12 (immediately before the evening dose) hours after
treatment. Five-milliliter samples were collected into Vacutainer tubes
(Becton Dickinson, Rutherford, NJ) containing EDTA, except when
sampling times for paclitaxel and capecitabine coincided, in which case
10-mL samples were obtained. Blood samples were immediately
VILLALONA-CALERO ET AL
centrifuged to separate plasma, and at least 2 mL of plasma was
transferred into polypropylene tubes and frozen at ⫺20°C. Concentrations of capecitabine and its metabolites 58-DFCR, 58-DFUR, 5-FU,
58-68-dihydrofluorouracil (5-FUH2), and fluoro-beta-alanine (FBAL)
were measured by means of a validated liquid chromatography/mass
spectrometry technique, as previously described.24
Pharmacokinetic Analysis
Estimates of pharmacokinetic parameters for paclitaxel and capecitabine and its metabolites were derived from individual concentration-time
data sets using noncompartmental methods and the program SAS,
version 6.11 for Windows (SAS Institute, Inc, Cary, NC).25 The values
for area under the concentration-time curve (AUC) were calculated
using the linear trapezoidal method with extrapolation of the curve to
infinity (AUC0-⬁), and from time 0 to the last sampling time at which the
concentration could be measured (AUC0-t). The following pharmacokinetic parameters were also determined: maximum plasma concentration
(Cmax), tmax, and apparent half-life of elimination (t1/2). For paclitaxel,
total clearance (CLs) was estimated by dividing the dose (in milligrams)
by AUC0-⬁, and the volume of distribution (Vd) was calculated from
CLs/k, where k was the terminal rate constant of elimination.
Descriptive statistics were used to summarize all pharmacokinetic
parameters. Mean, SD, coefficient of variation, minimum, maximum,
and median values were determined for all parameters. In addition,
geometric mean and geometric coefficient of variation were determined
for Cmax, AUC0-t, and AUC0-⬁. Clinically relevant pharmacokinetic
interactions between capecitabine and paclitaxel were sought. Plasma
samples obtained on day 1 were used to evaluate the pharmacokinetics
of paclitaxel alone, whereas day-15 samples were used to evaluate the
pharmacokinetics of capecitabine alone, and day-22 values were used to
evaluate the pharmacokinetics of both drugs given in combination. For
58-DFUR, 58-DFCR, and 5-FU, the AUC0-⬁ values estimated on days 15
and 22 were used for statistical comparisons. For both capecitabine and
5-FUH2, the AUC0-t was assessed because AUC0-⬁ could not be
estimated reliably in all patients. For FBAL, the AUC0-12 was evaluated.
Paired t tests were performed for statistical comparisons of AUC
between treatment days. Because several dosing groups of paclitaxel
and capecitabine were used in the analysis, a three-way analysis of
variance, with dose and day as fixed factors and subject nested in the
dose group as the random factor, was performed for the logarithmically
transformed AUC estimates of paclitaxel on day 1, of capecitabine and
metabolites on day 15, and of paclitaxel, capecitabine, and capecitabine
metabolites on day 22. Statistical analysis was performed using both
StatView version 5.0 statistical software program and the general linear
models program in PROC GLM of SAS version 6.11 (both from SAS
Institute).
RESULTS
General
The number of new and total patients, courses, and rates
of DLTs as a function of dose level are listed in Table 1.
Seventeen patients received 66 assessable courses of paclitaxel and capecitabine. Two other patients were not assessable for toxicity because treatment was discontinued early
during their first course as a result of rapidly progressive
disease. Patient characteristics are listed in Table 2. Sixteen
patients received prior chemotherapy and 10 subjects were
previously treated with both chemotherapy and radiation
therapy. During study stage I, the daily dose of capecitabine
1919
PHASE I STUDY OF PACLITAXEL AND CAPECITABINE
Table 1. Dose Escalation Scheme
Hematologic Toxicity
No. of Patients
Dose (mg/m2/d)
Paclitaxel
Capecitabine
New
135
135
135
175
175
Total
1,004
1,331
1,657
1,657
1,331
3
3
3
2
6
17
Reduced to
This Dose
Total
0
1
0
0
1
3
4
3
2
7
Total
Courses
12
18
10
2
24
66
New Patients
With DLT/
Total New
Patients*
0/3
0/3
0/3
2/2
0/6
*During course 1.
was increased from 1,004 to 1,657 mg/m2/d, and the dose of
paclitaxel was fixed at 135 mg/m2. Because there were no
dose-limiting events in patients treated with capecitabine
doses as high as 1,657 mg/m2/d, which was the dose
recommended for phase II studies of capecitabine as a single
agent on a continuous treatment schedule,2 the study was
amended, and further dose escalation of capecitabine was
not pursued. Instead, the dose of paclitaxel was increased to
175 mg/m2, and because no DLTs were observed with
capecitabine 1,657 mg/m2/d in study stage I, this dose was
administered to the first cohort of patients in stage II.
However, DLT, characterized by fever associated with
neutropenia, occurred during course 1 in the first two
patients treated at this dose level. Therefore, the daily dose
of capecitabine was decreased to 1,331 mg/m2/d in combination with paclitaxel 175 mg/m2 in the next cohort of patients.
None of the six new patients treated at this dose level
developed DLTs during course 1, and only two of 24 total
courses were associated with DLT, which consisted of grade
3 diarrhea that lasted longer than 2 days (course 2) and
severe neutropenia associated with fever (course 3). No
objective antitumor responses were observed.
Table 2. Characteristics of Patients
No. of
Patients
Characteristic
Male/female
Age, years
Median
Range
Karnofsky performance status, %
Median
Range
Previous treatment
Chemotherapy
Chemotherapy ⫹ radiation therapy
Disease site
Colorectal
Esophagus
Breast
Osteosarcoma
Unknown primary tumor
13/4
63
48-72
Myelosuppression, predominately neutropenia, was the
principal DLT in this study. Table 3 lists the number of
courses associated with grade 3 (ANC ⱖ 500/µL and ⬍
1,000/µL) and grade 4 (ANC ⬍ 500/µL) neutropenia, severe
(grade 3 to 4) neutropenia associated with fever, prolonged
(⬎ 5 days) grade 4 neutropenia, and severe thrombocytopenia (platelet count ⬍ 50,000/µL). Overall, 10 of 17 patients
(58%) developed either grades 3 or 4 neutropenia during 18
of 66 courses (27%). However, the duration of grade 4
neutropenia never exceeded 5 days, and the duration of
grade 4 neutropenia was short (median, 2 days). Fever
associated with grade 3 or 4 neutropenia was experienced by
three subjects (17%) during three courses (4.5%). Onset of
neutropenia occurred on days 7 to 15; the mean time to ANC
nadir was 15 days (range, 8 to 15 days); and the median time
to complete recovery was 7 days. Recovery of blood cell
counts to pretreatment levels was usually complete by day
22, and paclitaxel treatment delay because of incomplete
recovery of ANCs was required in only one of 66 courses.
Anemia and thrombocytopenia were typically mild and
infrequent, with only one patient experiencing grade 3
anemia and no patients experiencing grades 3 or 4 thrombocytopenia.
DLT, consisting of grades 3 to 4 neutropenia associated
with fever, was initially noted in two patients treated with
capecitabine 1,657 mg/m2/d and paclitaxel 175 mg/m2.
Another individual developed grade 3 neutropenia associated with fever during his third course of capecitabine/
paclitaxel at the next lower dose level (1,331 mg/m2/d and
175 mg/m2, respectively). Gram-negative bacteremia without sepsis was documented in two of these three patients.
Complete recovery of the ANC to pretreatment levels
occurred 5 days after the onset of neutropenia in all of these
patients. In one of the patients who was initially treated with
capecitabine 1,657 mg/m2/d and paclitaxel 175 mg/m2,
grades 2 and 3 neutropenia were noted on two occasions
soon after resumption of treatment with capecitabine 1,657
mg/m2/d.
Nonhematologic Toxicity
90
80-100
16
10
12
2
1
1
1
The principal nonhematologic toxicities of this regimen
of capecitabine and paclitaxel are listed in Table 4. Eleven
patients experienced diarrhea, which was typically mild to
moderate (grade ⱕ 2 in nine patients) and brief (median
duration, 2 days [range, 2 to 12 days]). Neither the incidence
nor severity of diarrhea seemed to be related to dose level.
However, diarrhea seemed to be cumulative, generally
occurring after a median of two courses. Except for the
diarrhea that occurred in one individual, all other episodes
were successfully managed with brief courses of either
1920
VILLALONA-CALERO ET AL
Table 3. Hematologic Toxicity of Capecitabine Plus Paclitaxel
No. of Courses With:
Dose Level
Neutropenia
Grade 4
Grade 4
⬎ 5 days
Grades 3-4
With Fever
Thrombocytopenia
Grades
3 and 4
0
1
0
1
6
0
0
0
0
0
0
0
0
2
1
0
0
0
0
0
ANC (/␮L) Nadir
Paclitaxel
(mg/m2)
Capecitabine
(mg/m2/d)
Total No.
of Patients/Courses
Median
Range
Grade 3
135
135
135
175
175
1,004
1,331
1,657
1,657
1,331
3/12
4/18*
3/10
2/2
7/24*
2150
1260
688
357
540
1,110-3,290
541-5,430
629-1,860
36-714
277-3,854
0
2
3
1
4
*Includes courses administered to two patients whose capecitabine/paclitaxel dose level was reduced after initial assignment to the 1,657/175-mg/m2 dose level.
loperamide or diphenoxylate hydrochloride. A 59-year-old
man with rectal cancer who was previously treated with
radiation to his pelvis and rectum developed grade 3
diarrhea 4 days after his second treatment with paclitaxel
during course 2 at the capecitabine/paclitaxel dose level of
1,331 mg/m2/d and 175 mg/m2, respectively, but these
complaints were temporally related to the initiation of
treatment with laxatives for constipation. His diarrhea was
still present, albeit grade 1 in severity, at the time of his death
from progressive disease 12 days after onset. A second
patient developed grade 3 diarrhea that lasted 1 day during
his third course of capecitabine/paclitaxel at 1,331 mg/m2/d
and 135 mg/m2, respectively.
Elevations in serum bilirubin concentrations occurred in
eight patients during 12 courses. The onset of bilirubin
elevations occurred between days 8 and 22 (median, day
15). Concomitant elevations in serum transaminases and
alkaline phosphatase were noted in one patient only. Fractionation of total bilirubin in two of the eight individuals showed
that unconjugated bilirubin accounted for most of the
elevation. The absolute elevations were generally modest
(median peak serum bilirubin concentration, 1.4 mg/dL
[range, 1.1 to 2.4 mg/dL]), and cumulative toxicity was not
observed. In addition, treatment with capecitabine was never
interrupted for hyperbilirubinemia, and hyperbilirubinemia
resolved after a median of 8 days in all individuals. For these
reasons, despite the fact that several of these elevations met
the criteria for grade 3 toxicity using the National Cancer
Institute Common Toxicity Criteria (1.5 to three times the
upper normal limit), these adverse effects were not considered dose-limiting.
One patient developed hand-foot syndrome. The patient, a
61-year-old man with metastatic colon cancer who was
treated with capecitabine 1,657 mg/m2/d and paclitaxel 135
mg/m2, developed a symmetrical erythematous macular rash
on both of his forearms and hands 15 days after beginning
treatment. Because of the concomitant development of grade
3 neutropenia, treatment with capecitabine was delayed for 5
days and then begun again at the same dose after complete
hematologic recovery and improvement of his rash. However, 7 days after treatment with his second dose of
paclitaxel on day 22, he developed abdominal cramping
(grade 2), nausea and vomiting (grade 2), diarrhea (grade 2),
and neutropenia (grade 3). His hands were tender, edematous, and erythematous, and a rash involving his arms,
shoulders, face, and trunk was also noted. These toxic
manifestations resolved after capecitabine and paclitaxel
were discontinued, and no further treatment was administered because of the development of progressive disease.
Two additional patients treated with capecitabine 1,657 or
1,004 mg/m2/d and paclitaxel 135 mg/m2 developed ery-
Table 4. Nonhematologic Toxicities of Capecitabine Plus Paclitaxel*
No. of Courses With:
Dose Level
Diarrhea
Hyperbilirubinemia
Skin Rash
Myalgia-Arthralgia
Paclitaxel (mg/m2)
Capecitabine (mg/m2/d)
No. of Courses
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
135
135
135
175
175
1,004
1,331
1,657
1,657
1,331
12
18
10
2
24
1
0
0
0
1
1
1
2
2
3
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
2
1
1
3
0
1
2
1
1
0
0
0
0
0
1
0
1
1
2
1
0
1
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
0
2
1
0
1
0
0
*Toxicity grading: Diarrhea: 1) increase by two to three stools/day over pretreatment; 2) increase by four to six stools/day or nocturnal stools; 3) increase by seven
to nine stools/day or incontinence; 4) increase by ⱖ 10 stools/day, grossly bloody diarrhea, or need for parenteral support. Hyperbilirubinemia: 1) not graded as; 2)
one to 1.5 upper normal limit; 3) 1.5 to three times upper normal limit; 4) ⬎ three times upper normal limit. Skin rash: 1) scattered macular or papular eruption, or
erythema that is asymptomatic; 2) scattered macular or papular eruption, or erythema with pruritus or other symptoms; 3) generalized symptomatic macular, papular,
or vesicular eruption; 4) exfoliative dermatitis or ulceration dermatitis. Myalgias–arthralgias: 1) mild; 2) moderate; 3) severe.
1921
PHASE I STUDY OF PACLITAXEL AND CAPECITABINE
thematous pruritic eruptions on their hands, which was not
typical of hand-foot syndrome, and therapy was not interrupted.
Seven patients experienced myalgia and/or arthralgia. The
onset of symptoms was typically 2 to 4 days after treatment
with paclitaxel 135 and 175 mg/m2. Myalgia and arthralgia
generally lasted 2 to 7 days and were managed successfully
with either nonsteroidal anti-inflammatory medications or
acetaminophen. Other mild to moderate drug-related toxicities, which seemed unrelated to dose level, included mucositis, nausea, vomiting, alopecia, conjunctivitis, facial flushing, lower extremity edema, paresthesia, and malaise. In
addition, a 72-year-old man with rectal carcinoma and lung
metastasis who was previously treated with radiation to his
primary rectal lesion, as well as interstitial irradiation for a
stage B prostate cancer, developed hemorrhagic cystitis 21
days after beginning his first course of capecitabine 1,331
mg/m2/d and paclitaxel 135 mg/m2. Cystoscopy showed a
large, friable necrotic mass involving the left trigone of his
bladder that was associated with extensive mucosal inflammation. Biopsy specimens of the mass and adjacent bladder
showed cystitis, which was believed to be caused by a
radiation recall phenomenon.
Pharmacokinetic Studies
Plasma samples for pharmacokinetic studies were obtained in all 17 patients on day 1 (treatment with paclitaxel
alone), 11 patients on day 15 (treatment with capecitabine
alone), and 14 patients on day 22 (concurrent treatment with
paclitaxel and capecitabine). Therefore, 11 paired samples
were available to assess the effects of paclitaxel on capecitabine pharmacokinetics, and 14 paired samples were available
to assess the effects of capecitabine on paclitaxel pharmacokinetics.
At paclitaxel doses of 135 and 175 mg/m2, AUC0-⬁ values
on day 1 averaged 10,829 and 16,045 ng-h/mL, respectively.
Mean (⫾SD) paclitaxel pharmacokinetic parameters on day
1 included a Vd of 565.46 ⫾ 224.8 L, a t1/2 of 16.7 ⫾ 4.9
hours, and a CLs of 24.01 ⫾ 7.1 L/h. Paclitaxel pharmacokinetic parameters are listed in Table 5, and cumulative plasma
concentration-versus-time profiles for paclitaxel administered alone (day 1) and concurrent with capecitabine (day
22) are shown in Fig 2.
Scatterplots depicting AUC and Cmax values of capecitabine and capecitabine metabolites as a function of capecitabine dose are shown in Fig 3. Although the sample size was
small, the relationship between capecitabine dose and AUC
appeared linear in the capecitabine dosing range evaluated.
Figure 4 shows plasma concentration-versus-time curves for
capecitabine and capecitabine metabolites from a representative patient, and pharmacokinetic parameter values for
Table 5. Paclitaxel Pharmacokinetic Parameters Before Treatment With
Capecitabine (Day 1) and During Capecitabine (Day 22)*
Pharmacokinetic
Parameter
Paclitaxel 135
(n ⫽ 6)
Cmax, ng/mL
AUC0-⬁, ng-h/mL
t1/2, h
CLs, L/h
Vd, L
Paclitaxel 175 mg/m2 (n ⫽ 8)
Cmax, ng/mL
AUC0-⬁, ng-h/mL
t1/2, h
CLs, L/h
Vd, L
Before Capecitabine
Treatment (Day 1)
During Capecitabine
Treatment (Day 22)
Mean
%
Mean
%
2,614
10,829
15.5
25.8
550.3
20
27
35
34
61
2,390
10,191
15.0
26.7
540.9
26
28
32
33
57
3,980
16,045
16.9
21.2
503.3
35
21
26
19
30
4,434
18,664
14.0
18.2
343.2
21
19
38
24
30
mg/m2
*Values represent geometric mean and coefficient of variation (percent) for
Cmax, AUC0-t, AUC0-⬁, CLs, and Vd, and arithmetic means and coefficient of
variation (percent) for t1/2.
prodrug and metabolites in patients treated with two daily
doses of capecitabine 666.5 mg/m2 (1,331 mg/m2/d) on day
15 are listed in Table 6. Peak capecitabine plasma concentrations occurred at a median of 1.05 hours (range, 0.5 to 4.05
hours) after capecitabine administration, whereas maximum
plasma concentrations of 58-DFCR, 58-DFUR, and 5-FU
were noted at median of 2.02 hours (range, 0.50 to 4.05
hours) after treatment. Median tmax values for 5-FUH2 and
FBAL were 3 hours (range, 1.03 to 4.05 hours) and 3.03
hours (range, 1.98 to 5.13 hours), respectively. Despite
nearly identical tmax values for 58-DFUR and 5-FU, AUC
values for 5-FU were 26.5-fold lower, on average, than those
for 58-DFUR at the recommended dose of capecitabine
(1,331 mg/m2/d) in this study.
The potential influence of capecitabine on paclitaxel
pharmacokinetics was assessed. Paclitaxel AUC0-⬁ values
that resulted from the administration of paclitaxel alone (day
1) were compared with AUC0-⬁ values obtained when
paclitaxel and capecitabine were administered concurrently
(day 22) using paired data sets. Paclitaxel exposure was not
affected by capecitabine, as shown by paired comparisons of
AUC0-⬁ values obtained on days 1 and 22 (paired t test, P ⫽
.80). Similarly, the potential influence of paclitaxel on the
pharmacokinetics of capecitabine and capecitabine metabolites was assessed by comparing AUC and Cmax values
obtained when capecitabine was administered before treatment with paclitaxel (day 15), with parameter values
obtained when capecitabine and paclitaxel were administered concurrently (day 22). Paired analyses showed that
paclitaxel did not significantly affect the AUC of capecitabine or any of its metabolites (paired t test, P ⬎ .1 for all
comparisons). In addition, Cmax values for capecitabine,
1922
VILLALONA-CALERO ET AL
58-DFCR, 58-DFUR, 5-FU, 5-FUH2, and FBAL before and
during treatment with paclitaxel were similar (paired t test,
P ⬎ .10 for all comparisons). Similar analyses using
three-way analysis of variance for logarithmically transformed AUC estimates of paclitaxel on day 1, capecitabine
and capecitabine metabolites on day 15, and all moieties on
day 22 showed no significant effects of paclitaxel on the
AUC of capecitabine and capecitabine metabolites and visa
versa, except for the AUC of FBAL, which was significantly
lower when capecitabine and paclitaxel were administered
concurrently (P ⫽ .02).
DISCUSSION
The high and predictable oral bioavailability of capecitabine, as well as the preferential conversion of this prodrug to
5-FU in neoplastic tissues, as shown by several investigators,1,13,14 render capecitabine one of the most interesting
fluoropyrimidines undergoing development. dThdPase, which
Fig 3. Scatterplots depicting AUC and Cmax values of capecitabine (䊊),
5-DFCR (䉭), 5-DFUR (䊉) and 5-FU (ⴛ) as a function of capecitabine dose.
Fig 2. Mean plasma concentration-versus-time curves for paclitaxel
administered before treatment with capecitabine on day 1 and during
treatment with capecitabine on day 22.
is highly expressed in neoplastic tissue and may account for
the preferential metabolism of 58-DFUR to 5-FU in malignant neoplasms, has been shown to be upregulated in human
colon cancer xenografts after treatment with paclitaxel, and
synergy between paclitaxel and capecitabine has also been
noted.19 On the basis of this preclinical rationale and the
overlapping clinical antitumor spectra of these agents, this
phase I pharmacologic study was performed to evaluate the
feasibility of administering capecitabine and paclitaxel in
combination and to study the pharmacologic profiles and
potential for pharmacokinetic interactions between these
agents.
Myelosuppression, predominately neutropenia, was the
principal DLT. However, clinically relevant single-agent
doses of both paclitaxel and capecitabine were able to be
administered in combination. Because the incidence of
severe neutropenia associated with fever at the paclitaxel
175-mg/m2 and capecitabine 1,657-mg/m2/d dose level was
unacceptably high, the dose level of paclitaxel 175 mg/m2
and capecitabine 1,331 mg/m2/d, consisting of paclitaxel
1923
PHASE I STUDY OF PACLITAXEL AND CAPECITABINE
hyperbilirubinemia was not clinically significant.2,15,16 In
contrast to the results of early studies, four of the eight
individuals who developed hyperbilirubinemia in the present study did not have liver metastases, including two of the
three patients with grade 3 toxicity, and only one subject had
concomitant elevations of other liver function tests. In
addition, the elevation was largely caused by unconjugated
bilirubin.
The relative early tmax values for all metabolites, particularly 58-DFUR, and higher Cmax and AUC values for
58-DFUR relative to both capecitabine and 58-DFCR, indicate that capecitabine is rapidly and extensively metabolized
to 58-DFUR. Interestingly, systemic exposure to 58-DFUR,
as estimated from AUC values in plasma, was nearly 30-fold
greater, on average, than that of 5-FU. However, because
conversion of 58-DFUR to 5-FU is largely an intracellular
process that primarily occurs in peripheral tissues,5 the
relative exposures of capecitabine and its metabolites based
on AUC value estimates from plasma concentrations in this
study may not accurately reflect the relative exposure of
peripheral tissues and malignant tumors to capecitabine and
capecitabine metabolites. Nevertheless, significant pharmacokinetic interactions between paclitaxel, capecitabine, and
capecitabine metabolites were not evident. Although hepatic
metabolism and/or biliary excretion play a principal role in
the metabolism and clearance of both capecitabine and
paclitaxel, significant pharmacokinetic interactions would
not have been anticipated because the principal mechanisms
of hepatic involvement in the metabolism and disposition of
capecitabine and paclitaxel are different. Paclitaxel is primarily metabolized by hepatic cytochrome P450 isoform CYP2C
to 6-alpha-hydroxypaclitaxel, with hepatic cytochrome P450
isoform CYP3A accounting for a minor degree of drug
disposition,27 whereas capecitabine is not metabolized by
hepatic cytochrome P450–dependent processes (Hirose et al,
unpublished results). In fact, capecitabine neither inhibits
nor induces hepatic cytochrome P450 mixed function oxidases.
Fig 4. Plasma concentration-versus-time curves for capecitabine and
metabolites on day 15 in a patient treated with capecitabine 1,331
mg/m2/d.
175 mg/m2 every 3 weeks and capecitabine in two divided
daily doses of 666 mg/m2, was the MTD and recommended
dose level for subsequent evaluations. Overall, both the
qualitative and quantitative aspects of the hematologic
effects at the paclitaxel 175-mg/m2 and capecitabine 1,331mg/m2/d dose level were similar to those observed with
paclitaxel 175 mg/m2 administered as a single agent over 3
hours.21,26 However, several observations, such as the recurrence of neutropenia when capecitabine was restarted after a
treatment delay in a patient who had recovered from severe
neutropenia, presumably caused by paclitaxel, indicates that
capecitabine may also contribute to the neutropenia that
occurs with the paclitaxel/capecitabine regimen.
Qualitatively, the nonhematologic effects of the capecitabine/paclitaxel combination were similar to the toxicities of
either capecitabine (diarrhea, rash, flushing, hand-foot syndrome, nausea, vomiting, and mucositis) or paclitaxel (myalgia, arthralgia, asthenia, peripheral neurotoxicity, malaise,
and alopecia) administered as single agents.2,15,16,22 The
development of hyperbilirubinemia in eight patients during
12 courses is of interest, but increases in serum bilirubin
levels were also noted in early studies of capecitabine
administered as a single agent, and, as in the present study,
Table 6. Pharmacokinetic Parameters for Capecitabine and Metabolites Before Treatment With Paclitaxel (Day 15) at the Capecitabine/Paclitaxel Dose Level of
175/1,331 mg/m2*
Pharmacokinetic
Parameter
Cmax, ␮g/mL
tmax, hours
Median
Range
AUC0-t, ␮g-h/mL
AUC0-⬁, ␮g-h/mL
t1/2, hours
Capecitabine
58-DFCR
58-DFUR
5-FU
5-FUH2
FBAL
Mean
%
Mean
%
Mean
%
Mean
%
Mean
%
2.93
74
1.78
105
5.10
18
0.196
42
0.736
25
1.01
0.50-3.00
3.59
44
3.83
56
0.51
18
1.99
0.52-4.02
3.32
87
3.44
84
0.79
22
1.99
0.52-4.02
8.92
13
9.04
14
0.69
16
1.99
0.52-4.02
0.334
31
0.341
30
0.87
49
*Patients treated with capecitabine 1,331 mg/m2/d in two divided daily doses of 665.5 mg/m2.
2.01
1.03-4.02
2.09
8
2.22
9
1.23
40
Mean
3.60
%
30
2.99
1.98-5.13
20.2
33
20.9
35
3.57
24
1924
VILLALONA-CALERO ET AL
The fact that there were no major antitumor responses
noted in the present study must be tempered by the
knowledge that 15 of the 17 subjects were previously shown
to be clearly refractory to 5-FU, with progressive tumor
growth occurring during prior treatment with 5-FU–based
regimens, and that at least 13 of the 17 patients had
malignant neoplasms that are considered innately resistant
to the taxanes. Still, the experimental evidence that paclitaxel upregulates dThdPase in malignant cells in vitro is
intriguing and provides a rationale for further evaluations of
combination regimens consisting of paclitaxel and capecitabine.19 In addition to this specific molecular interaction,
perhaps the most impressive cytotoxic interactions between
the taxanes and other cytotoxic agents have been between
the taxanes and 5-FU, which further supports the development of this drug combination.28,29
Although the specific capecitabine/paclitaxel regimen that
was evaluated in the present trial was selected, in part, to
achieve protracted exposure to 5-FU and maximal convenience for patients, other capecitabine/paclitaxel treatment
schedules may increase the likelihood that favorable pharmacologic and molecular drug interactions between these
agents will occur. For example, administering paclitaxel on a
more frequent schedule, such as on a weekly low-dose
schedule, may result in more protracted periods of dThdPase
upregulation, as well as coexposure of tumors to both
capecitabine and paclitaxel. Alternatively, the administration
of higher doses of capecitabine has been shown to be more
feasible on less protracted capecitabine daily dosing (eg,
14-day) schedules compared with continuous daily treatment schedules.15 Also, the use of shorter, albeit more
intensive, capecitabine regimens may hypothetically provide maximal concentrations of the 5-DFUR substrate to
dThdPase during periods when the enzyme is maximally
upregulated after treatment with paclitaxel.
The overall results of this study indicate that the administration of clinically relevant single-agent doses of both
capecitabine and paclitaxel is feasible in patients with solid
malignancies, and that there are no relevant pharmacologic
interactions between capecitabine and paclitaxel that might
complicate the further development and broad usage of the
regimen. On the basis of the toxicologic profile of the
capecitabine/paclitaxel regimen shown in the present study,
paclitaxel 175 mg/m2 as a 3-hour IV infusion every 3 weeks
administered in combination with capecitabine 1,331 mg/
m2/d in two divided doses is recommended for diseasedirected evaluations. From a mechanistic standpoint, further
evaluations of capecitabine/paclitaxel regimens that are
rationally designed to optimize favorable subcellular interactions are warranted.
ACKNOWLEDGMENT
This article is dedicated to the memory of Maura Kraynak.
REFERENCES
1. Ishitsuka H, Miwa M, Ishikawa T, et al: Capecitabine: An orally
available fluoropyrimidine with tumor selective activity. Proc Am Assoc
Cancer Res 36:407, 1995 (abstr 2426)
2. Budman D, Meropol N, Reigner B, et al: Preliminary studies of a
novel oral fluoropyrimidinecarbamate: Capecitabine. J Clin Oncol
16:1795-1802, 1998
3. Bollag W, Hartmann H: Tumor inhibitory effects of a new fluorouracil
derivative: 5-Deoxy-5-fluorouridine. Eur J Cancer 16:427-432, 1980
4. Hartmann H, Bollag W: 58-Deoxy-5-fluorouridine: A new antineoplastic pyrimidine antimetabolite. J Suisse Med 110:1078-1080, 1980
5. Kono A, Hara Y, Sugata S, et al: Activation of 58-deoxy-5fluorouridine by thymidine phosphorylase in human tumors. Chem
Pharm Bull 31:175-178, 1983
6. Takebayashi Y, Yamada K, Miyadera K, et al: The activity and
expression of thymidine phosphorylase in human solid tumors. Eur J
Cancer 32A:1227-1232, 1996
7. Takebayashi Y, Akiyama S, Akiba S, et al: Clinicopathologic and
prognostic significance of an angiogenic factor, thymidine phosphorylase, in
human colorectal carcinoma. J Natl Cancer Inst 88:1110-1117, 1996
8. Tanaka Y, Eda H, Fujimoto K, et al: Anticachectic activity of
58-DFUR in a murine tumour cachexia model, colon 26 adenocarcinoma. Cancer Res 50:4528-4532, 1990
9. Ishikawa T, Ura M, Yamamoto T, et al: Selective inhibition of
spontaneous pulmonary metastasis of Lewis lung carcinoma by 58DFUR. Int J Cancer 61:516-521, 1995
10. Bajetta E, Colleoni M, Di Bartolomeo M, et al: Doxifluridine and
leucovorin: An oral treatment combination in advanced colorectal
cancer. J Clin Oncol 13:2613-2619, 1995
11. Di Bartolomeo M, Bajetta E, Somma L, et al: Doxifluridine as
palliative treatment in advanced gastric and pancreatic cancer patients.
Oncology 53:54-57, 1996
12. Alberto P: Phase II study of an oral combination of doxifluridine,
prednimustine and idarubicin (FUPRIDA) for first line treatment of
advanced breast cancer. Ann Oncol 4:423-425, 1993
13. Ishikawa T, Utoh M, Sawada N, et al: Xeloda (capecitabine): An
orally available tumor-selective fluoropyrimidine carbamate. Proc Am
Soc Clin Oncol 16:208a, 1997 (abstr 727)
14. Schuller J, Cassidy J, Reigner B, et al: Tumor selectivity of
Xeloda in colorectal cancer patients. Proc Am Soc Clin Oncol 16:227a,
1997 (abstr 797)
15. Mackean M, Planting A, Twelves C, et al: A phase I and
pharmacologic study of intermittent twice-daily oral therapy with
capecitabine in patients with advanced and/or metastatic cancer. J Clin
Oncol 16:2977-2985, 1998
16. Taguchi T, Ishitani, K, Saitoh K, et al: A Japanese phase I study
of continuous twice daily treatment with capecitabine in patients with
advanced and/or metastatic solid tumors. Ann Oncol 7:87, 1996 (suppl
1) (abstr 299)
17. Eda H, Fujimoto K, Watanabe S, et al: Cytokines induce
thymidine phosphorylase expression in tumor cells and make them
PHASE I STUDY OF PACLITAXEL AND CAPECITABINE
more susceptible to 5’-deoxy-5-fluorouridine. Cancer Chemother Pharmacol 32:333-338, 1993
18. Eda H, Fujimoto K, Watanabe S, et al: Cytokines induce uridine
phosphorylase in mouse colon 26 carcinoma cells and make the cells
more susceptible to 58-deoxy-5-fluorouridine. Jpn J Cancer Res 84:341347, 1993
19. Sawada N, Ishikawa T, Fukase Y, et al: Induction of thymidine
phosphorylase activity and enhancement of capecitabine efficacy by
Taxol/Taxotere in human cancer xenografts. Clin Cancer Res 4:10131019, 1998
20. Bogdan C, Ding A: Taxol, a microtubule stabilizing antineoplastic agent, induces expression of tumor necrosis factor alpha and
interleukin-1 in macrophages. J Leuk Biol 52:119-121, 1992
21. Rowinsky E, Donehower R: Paclitaxel (Taxol). N Engl J Med
332:1004-1114, 1995
22. National Cancer Institute: Guidelines for Reporting of Adverse
Drug Reactions. Bethesda, MD, Division of Cancer Treatment, National
Cancer Institute, 1988
23. Royer I, Alvinerie P, Armand J, et al: Paclitaxel metabolites in
human plasma and urine: Identification of 6 alpha-hydroxytaxol,
7-epitaxol and taxol hydrolysis products using liquid chromatography/
atmospheric-pressure chemical ionization mass spectrometry. Rapid
Commun Mass Spectrom 9:495-502, 1995
1925
24. Reigner B, Verweij J, Dirix L, et al: Effect of food on the
pharmacokinetics of capecitabine and its metabolites following oral
administration in cancer patients. Clin Cancer Res 4:941-948, 1998
25. Gibaldi M, Perrier D: Noncompartmental analysis based on
statistical moment theory, in Gibaldi M, Perrier D (eds): Pharmacokinetics (ed 2). New York, Marcel Dekker, 1992, pp 409-417
26. Eisenhauer E, ten Bokkel Huinink W, Swenerton K, et al:
European-Canadian randomized trial of paclitaxel in relapsed ovarian
cancer: High-dose versus low-dose and long versus short infusion. J
Clin Oncol 12:2654-2666, 1994
27. Cresteil T, Monsarrat B, Alvinerie P, et al: Taxol metabolism by
human liver microsomes: Identification of cytochrome P450 isoenzymes involved in its biotransformation. Cancer Res 54:386-392,
1994
28. Klaassen U, Wilke H, Harstrick A, et al: Paclitaxel in combination with weekly 24-hour infusional 5-fluorouracil plus leucovorin in
the second-line treatment of metastatic breast cancer: Results of a phase
II study. Ann Oncol 9:45-50, 1998
29. Bokemeyer C, Lampe CS, Clemens M, et al: A phase II trial of
paclitaxel and weekly 24 h infusion of 5-fluorouracil/folinic acid in
patients with advanced gastric cancer. Anticancer Drugs 8:396-399,
1997