`The
`
` medicine
`of
`
`original article
`
`Docetaxel plus Prednisone or Mitoxantrone
`plus Prednisone for Advanced Prostate Cancer
`
`Ian F. Tannock, M.D., Ph.D., Ronald de Wit, M.D., William R. Berry, M.D.,
`Jozsef Horti, M.D., Anna Pluzanska, M.D., Kim N. Chi, M.D.,
`Stephane Oudard, M.D., Christine Théodore, M.D.,
`Nicholas D. James, M.D., Ph.D., Ingela Turesson, M.D., Ph.D.,
`Mark A. Rosenthal, M.D., Ph.D., and Mario A. Eisenberger, M.D.,
`for the TAX 327 Investigators
`
`abstract
`
`background
`Mitoxantrone plus prednisone reduces pain and improves the quality of life in men
`with advanced, hormone-refractory prostate cancer, but it does not improve survival.
`We compared such treatment with docetaxel plus prednisone in men with this disease.
`methods
`From March 2000 through June 2002, 1006 men with metastatic hormone-refractory
`prostate cancer received 5 mg of prednisone twice daily and were randomly assigned to
`receive 12 mg of mitoxantrone per square meter of body-surface area every three weeks,
`75 mg of docetaxel per square meter every three weeks, or 30 mg of docetaxel per square
`meter weekly for five of every six weeks. The primary end point was overall survival.
`Secondary end points were pain, prostate-specific antigen (PSA) levels, and the quality
`of life. All statistical comparisons were against mitoxantrone.
`results
`As compared with the men in the mitoxantrone group, men in the group given doce-
`taxel every three weeks had a hazard ratio for death of 0.76 (95 percent confidence in-
`terval, 0.62 to 0.94; P=0.009 by the stratified log-rank test) and those given weekly
`docetaxel had a hazard ratio for death of 0.91 (95 percent confidence interval, 0.75 to
`1.11; P=0.36). The median survival was 16.5 months in the mitoxantrone group, 18.9
`months in the group given docetaxel every 3 weeks, and 17.4 months in the group giv-
`en weekly docetaxel. Among these three groups, 32 percent, 45 percent, and 48 percent
`of men, respectively, had at least a 50 percent decrease in the serum PSA level (P<0.001
`for both comparisons with mitoxantrone); 22 percent, 35 percent (P=0.01), and 31
`percent (P=0.08) had predefined reductions in pain; and 13 percent, 22 percent
`(P=0.009), and 23 percent (P=0.005) had improvements in the quality of life. Adverse
`events were also more common in the groups that received docetaxel.
`conclusions
`When given with prednisone, treatment with docetaxel every three weeks led to superi-
`or survival and improved rates of response in terms of pain, serum PSA level, and qual-
`ity of life, as compared with mitoxantrone plus prednisone.
`
`From the Department of Medical Oncology
`and Hematology, Princess Margaret Hos-
`pital and University of Toronto, Toronto
`(I.F.T.); the Department of Medical Oncol-
`ogy, Erasmus University Medical Centre,
`Rotterdam, the Netherlands (R.W.); Raleigh
`Hematology Oncology Associates, Cary,
`N.C. (W.R.B.); the Department of Chemo-
`therapy and Clinical Pharmacology, Nation-
`al Institute of Oncology, Budapest, Hunga-
`ry (J.H.); the Department of Chemotherapy,
`Medical University, Lodz, Poland (A.P.);
`BC Cancer Agency, Vancouver, B.C., Cana-
`da (K.N.C.); Hôpital Européen Georges
`Pompidou, Paris (S.O.); Institut Gustav
`Roussy, Villejuif, France (C.T.); Cancer Re-
`search UK Institute for Cancer Studies, Bir-
`mingham, United Kingdom (N.D.J.); the
`Section of Oncology, Uppsala University
`Hospital, Uppsala, Sweden (I.T.); Cancer
`Trials Australia, Victoria, Australia (M.A.R.);
`and the Sydney Kimmel Comprehensive
`Cancer Center, Johns Hopkins University,
`Baltimore (M.A.E.). Address reprint re-
`quests to Dr. Tannock at the Department
`of Medical Oncology and Hematology,
`Princess Margaret Hospital, 610 Universi-
`ty Ave., Toronto, ON M5G 2M9, Canada,
`or at ian.tannock@uhn.on.ca.
`
`N Engl J Med 2004;351:1502-12.
`Copyright © 2004 Massachusetts Medical Society.
`
`1502
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`7, 2004
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`MYLAN - EXHIBIT 1013
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`p
`
`docetaxel versus mitoxantrone for prostate cancer
`
`rostate cancer is the most com
`-
`mon cancer among men, with approximate-
`ly 220,000 cases and 29,000 deaths annu-
`1
`ally in the United States.
` About 10 to 20 percent of
`men with prostate cancer present with metastatic
`disease, and in many others, metastases develop
`despite treatment with surgery or radiotherapy.
`Treatment of metastatic prostate cancer is pal-
`liative. In about 80 percent of men, primary andro-
`gen ablation leads to symptomatic improvement
`and a reduction in serum levels of prostate-specific
`antigen (PSA), but in all patients the disease even-
`tually becomes refractory to hormone treatment.
`The options then include symptomatic care with
`narcotic analgesics, radiotherapy to dominant sites
`of bone pain, treatment with bone-seeking isotopes
`such as strontium-89, and cytotoxic chemothera-
`py. Bisphosphonates may reduce skeletal compli-
`2-4
`cations,
` and low-dose prednisone or hydrocor-
`5,6
`tisone may be palliative in some patients.
`Chemotherapy can reduce serum PSA levels in
`patients with hormone-refractory prostate cancer
`and relieves pain in some patients, but tolerability
`is of concern, particularly since most patients are
`7
`elderly and many have other medical problems.
`A randomized trial showed that mitoxantrone plus
`low-dose prednisone relieved pain and improved
`the quality of life more frequently than did pred-
`8,9
`nisone alone.
` Consistent benefits of mitoxantro-
`ne plus a corticosteroid were observed in other ran-
`domized trials, but none found that this approach
`10-12
`improved survival.
` These trials established mi-
`toxantrone plus a corticosteroid as the treatment of
`reference for hormone-refractory prostate cancer.
`Phase 2 studies of the taxane docetaxel have re-
`ported PSA responses (defined as a reduction in
`serum PSA levels of at least 50 percent) in up to 50
`13-16
`percent of patients.
` Studies of docetaxel plus
`either estramustine or calcitriol have shown PSA
`17-19
`responses in up to 80 percent of patients.
` How-
`ever, outcomes of single-group studies are subject
`20
`to bias.
`We conducted a phase 3 study, the TAX 327
`Study, comparing docetaxel (given either every three
`weeks or weekly) plus daily prednisone with mito-
`xantrone plus prednisone. The docetaxel regimens
`were selected on the basis of their dose equivalence
`(a dose intensity of 25 mg per square meter of body-
`surface area per week and a maximal cumulative
`dose of 750 mg per square meter) and their activi-
`ty and tolerability in phase 2 studies. The primary
`hypothesis was that treatment with docetaxel plus
`
`prednisone would improve overall survival as com-
`pared with mitoxantrone plus prednisone.
`
`methods
`
`patients
`This randomized, nonblinded, phase 3 study in-
`volved centers in 24 countries. Eligible patients had
`histologically or cytologically confirmed adeno-
`carcinoma of the prostate with clinical or radiolog-
`ic evidence of metastatic disease, had had disease
`progression during hormonal therapy, and were re-
`ceiving primary androgen-ablation therapy as main-
`tenance therapy. At least four weeks had to have
`elapsed between the withdrawal of antiandrogens
`(six weeks in the case of bicalutamide) and enroll-
`ment, so as to avoid the possibility of confounding
`as a result of the response to antiandrogen with-
`21,22
`drawal.
` Another requirement was disease pro-
`gression, as indicated by increasing serum levels of
`PSA on three consecutive measurements obtained
`at least one week apart or findings on physical ex-
`amination or imaging studies.
`Eligible patients had a Karnofsky performance-
`status score of at least 60 percent, no prior treat-
`ment with cytotoxic agents (except estramustine)
`or radioisotopes, no history of another cancer with-
`in the preceding five years (except basal or squa-
`mous-cell skin cancer), no brain or leptomeningeal
`metastases, no symptomatic peripheral neurop-
`athy of grade 2 or higher, and no other serious
`medical condition. At least four weeks had to have
`elapsed between prior surgery or radiotherapy (lim-
`ited to no more than 25 percent of the bone mar-
`row) and enrollment. Prior treatment with cortico-
`steroids was allowed. Normal cardiac function was
`required. Laboratory criteria for eligibility included
`a neutrophil count of at least 1500 per cubic milli-
`meter, a hemoglobin level of at least 10.0 g per deci-
`liter, a platelet count of at least 100,000 per cubic
`millimeter, a total bilirubin level below the upper
`limit of the normal range for each institution, and
`serum alanine aminotransferase, aspartate amino-
`transferase, and creatinine levels that were no more
`than 1.5 times the upper limit of the normal range.
`A clinical history was obtained, and a physical
`examination, with radiographic imaging, comput-
`ed tomography, and bone scanning, was performed
`within 14 days before randomization. Blood tests
`including measurement of serum PSA, electrocar-
`diography, and an evaluation of the left ventricu-
`lar ejection fraction by means of a multiple gated
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`7, 2004
`
`1503
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` new england journal
`The
`
` medicine
`of
`
`acquisition scan or echocardiography were per-
`formed. Pain, analgesic intake, and the quality of
`life were assessed at baseline. Pain was assessed by
`means of the Present Pain Intensity (PPI) scale from
`the McGill–Melzack questionnaire, which uses ver-
`bal descriptors; scores can range from 0 to 5, with
`23
`higher scores indicating greater pain.
` Patients
`recorded their daily PPI score and analgesic use in
`a diary. A daily analgesic score was calculated by
`assigning a score of 4 for a standard dose of a nar-
`cotic analgesic (e.g., 10 mg of morphine) and a score
`of 1 for a standard dose of a nonnarcotic analgesic.
`Patients were required to have stable levels of pain
`for at least seven days before randomization, de-
`fined by a daily variation of no more than 1 in the
`PPI score or of no more than 25 percent in the anal-
`gesic score. The quality of life was assessed with
`the Functional Assessment of Cancer Therapy–Pros-
`tate (FACT-P) questionnaire; scores on this self-
`administered questionnaire can range from 0 to
`156, with higher scores indicating a better quality
`24,25
`of life.
`All patients provided written informed consent,
`and the study was approved by all institutional re-
`view boards in accordance with the international
`standards of good clinical practice. An independent
`data and safety monitoring committee was estab-
`lished.
`
`randomization and treatment
`Randomization was centralized with the use of a
`stratified, permuted-block allocation scheme ac-
`cording to the baseline pain level (pain was classi-
`fied as present, as defined by a median PPI score of
`at least 2 or a mean analgesic score of at least 10,
`or as absent, as defined by a median PPI score of
`less than 2 and a mean analgesic score of less than
`10) and the baseline Karnofsky performance-status
`score (70 percent or less vs. 80 percent or more).
`Patients who were randomly assigned to the doc-
`etaxel groups received either 75 mg of docetaxel
`(Taxotere, Aventis) per square meter as a 1-hour in-
`travenous infusion on day 1 every 21 days or 30 mg
`of docetaxel per square meter as a 30-minute in-
`travenous infusion on days 1, 8, 15, 22, and 29 of a
`6-week cycle. Patients who were randomly assigned
`to the standard-therapy group received 12 mg of
`mitoxantrone (Novantrone, Immunex and Wyeth–
`Ayerst) per square meter as a 30-minute infusion
`on day 1 every 21 days. All patients received 5 mg
`of prednisone (or prednisolone, if prednisone was
`not available) orally twice daily starting on day 1. Pre-
`
`medication with dexamethasone was required in
`the docetaxel groups (8 mg given 12 hours, 3 hours,
`and 1 hour before the docetaxel infusion in the
`group treated every three weeks and 8 mg given
`1 hour before docetaxel in the group treated week-
`ly). Antiemetic medication was prescribed accord-
`ing to local practice.
`Up to 10 cycles of treatment were planned for
`the group given docetaxel every three weeks and
`the mitoxantrone group and up to 5 cycles (of six
`weeks each) in the weekly-docetaxel group. Treat-
`ment delays of up to two weeks and up to two dose
`reductions were allowed. Dose reductions were
`specified for patients who had had grade 4 neutro-
`penia for at least seven days, an infection, or grade
`3 or 4 neutropenia with an oral temperature of at
`least 38.5°C. A dose reduction or treatment delay
`was also stipulated for patients who had an abso-
`lute neutrophil count of less than 1500 per cubic
`millimeter (for those on three-week schedules) or
`less than 1000 per cubic millimeter (for those re-
`ceiving weekly docetaxel) on a treatment day and
`for those with grade 3 or 4 thrombocytopenia. Treat-
`ment with granulocyte colony-stimulating factor
`was allowed for patients with febrile neutropenia.
`Systemic corticosteroids (other than dexametha-
`sone and prednisone) and bisphosphonates were
`not permitted.
`
`follow-up and outcomes
`Physical examinations and baseline blood tests
`were repeated at three-week intervals. Imaging
`studies to determine the extent of disease were per-
`formed at intervals of six to nine weeks and repeat-
`ed after four weeks to identify those with a response.
`The primary end point was overall survival. Sec-
`ondary end points were predefined reductions in
`pain, an improvement in the quality of life, a reduc-
`tion in serum PSA levels of at least 50 percent, and
`objective tumor responses.
`Patients with a PPI score of at least 2, an analgesic
`score of at least 10, or both (averaged over the pre-
`vious week) at baseline were assessed for the pain
`response at three-week intervals. A pain response
`was defined as a two-point reduction in the PPI
`score from baseline without an increase in the an-
`algesic score or as a reduction of at least 50 percent
`in the analgesic score without an increase in the
`PPI score, either of which was maintained for at
`least three weeks. Pain progression was defined as
`an increase in the PPI score of at least one point
`from the nadir, an increase from baseline of at least
`
`1504
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`,
`
`7
`
`2004
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`docetaxel versus mitoxantrone for prostate cancer
`
`25 percent in the analgesic score, or a requirement
`for palliative radiotherapy.
`Serum PSA was measured every three weeks,
`and a response (for patients with a baseline PSA
`level of at least 20 ng per milliliter) was defined as a
`reduction from baseline of at least 50 percent that
`was maintained for at least three weeks, whereas
`PSA progression was defined as an increase from
`the nadir of either at least 25 percent for men with
`no PSA response or at least 50 percent for all oth-
`ers. The duration of the PSA response and the pain
`response was defined as the time between the first
`and last evaluations at which the response criteria
`were met. For patients with at least one bidimen-
`sionally measurable lesion, tumor response was
`evaluated with the use of World Health Organiza-
`26
`tion criteria.
`The quality of life was assessed with the FACT-P
`questionnaire at baseline, every three weeks dur-
`ing therapy, and every month after the completion
`of therapy. All patients who answered the question-
`naire at baseline were included in the evaluation,
`and the FACT-P score was compared with the base-
`line value for each of these patients. Patients were
`defined as having a quality-of-life response if they
`had a 16-point improvement in their FACT-P score,
`as compared with baseline, on two measurements
`obtained at least three weeks apart.
`Adverse events were classified according to the
`Common Toxicity Criteria of the National Cancer
`Institute (version 2). Serious adverse events were
`fatal or life-threatening, required or prolonged hos-
`pitalization, resulted in persistent or substantial
`disability or incapacity, or were considered im-
`portant medical events. Treatment was stopped
`for any of the following reasons: completion of
`planned treatment, progression of disease, severe
`adverse events, or withdrawal of consent.
`
`statistical analysis
`There were three comparisons of interest between
`the docetaxel and mitoxantrone groups: docetaxel
`given every three weeks was compared with mito-
`xantrone, weekly docetaxel was compared with mi-
`toxantrone, and the combined docetaxel groups
`were compared with mitoxantrone. The study was
`designed to detect with 90 percent power a hazard
`ratio of 0.75 for death in the docetaxel groups as
`compared with the mitoxantrone group, with a
`two-sided type I error of 0.05 and with the data
`analyzed according to the intention to treat. The
`sample size was established as 1002 patients, and
`
`analysis was planned after 535 deaths had occurred.
`To allow for multiple comparisons, a P value of
`0.04 was considered to indicate statistical signif-
`icance for the comparison of the combined doce-
`taxel groups with the mitoxantrone group, and a
`P value of 0.0175 was considered to indicate sta-
`tistical significance for the comparison of each
`docetaxel group with the mitoxantrone group (all
`P values were two-sided), thus ensuring an overall
`significance level of 0.05.
`In the primary analysis, overall survival was an-
`alyzed by means of the Kaplan–Meier method,
`with log-rank comparisons stratified according to
`the level of pain and the Karnofsky performance-
`status score. Pain, PSA, tumor, and quality-of-life
`responses were compared by means of the Coch-
`ran–Mantel–Haenszel test. All randomized patients
`were included in the analysis of survival, and all
`treated patients were included in the evaluation of
`adverse effects.
`Hazard ratios for death were calculated after
`adjustment for any chance imbalance in potential
`prognostic factors between the groups. The follow-
`ing factors were entered into a full stratified Cox
`proportional-hazards model and a backward selec-
`tion model in which nonsignificant factors were
`eliminated sequentially at a P level of 0.10: age (less
`than 65 years vs. 65 years or older); visceral involve-
`ment (yes vs. no); liver involvement (yes vs. no);
`number of prior hormonal therapies (two or fewer
`vs. more than two); prior estramustine (yes vs. no);
`presence of rising serum PSA levels alone, as com-
`pared with the presence of other indications of pro-
`gression; baseline hemoglobin level; and baseline
`serum level of alkaline phosphatase. One planned
`interim analysis of safety was conducted after the
`recruitment of 120 patients. No interim analysis
`for efficacy was performed.
`The study was designed by Dr. Tannock in col-
`laboration with Aventis personnel, and the proto-
`col was finalized after being reviewed by the other
`study cochairs, Drs. de Wit and Eisenberger. The
`data were collected and maintained by Aventis, but
`the cochairs handled all questions regarding the
`management of the study. Only the data and safety
`monitoring committee saw the results of the inter-
`im safety analysis; no analysis was undertaken nor
`were the results seen by Aventis, the study cochairs,
`or any other investigator until the predefined num-
`ber of events had occurred. The protocol contained
`a plan for analysis and publication at that time. All
`data were provided to the cochairs at the comple-
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`7, 2004
`
`1505
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` new england journal
`The
`
` medicine
`of
`
`tion of the study. Aventis personnel undertook the
`statistical analysis. The article was drafted by Dr.
`Tannock and modified after being reviewed by the
`cochairs and other coauthors. Aventis reviewed the
`manuscript, but its final content was entirely deter-
`mined by the investigators.
`
`results
`
`characteristics of the patients
`and treatment
`A total of 1006 patients underwent randomiza-
`tion from March 2000 through June 2002. The da-
`
`Table 1. Baseline Characteristics of the Patients.*
`
`Characteristic
`
`Docetaxel Every 3 Wk Weekly Docetaxel Mitoxantrone Every 3 Wk
`
`No. randomized
`Ineligible (%)
`Age
`Median (yr)
`Range (yr)
`≥75 Yr (%)
`Gleason score (%)
`≤7
`8–10
`Not available
`Prior treatment (%)
`Prostatectomy
`Radiotherapy
`Estramustine
`Hormonal manipulations (%)†
`1
`2
`>2
`Karnofsky performance-status score ≤70% (%)
`Pain (%)‡
` Serum PSA
`Median (ng/ml)
`≥20 ng/ml (%)
`Extent of disease (%)
`Bone metastases
`Visceral disease
`Measurable lesions
`Evidence of progression at entry (%)§
`Bone scan
`Increase in measurable lesions
`Increase in nonmeasurable lesions
`Increased PSA
`
`
`
`335
`12
`
`68
`42–92
`20
`
`42
`31
`26
`
`19
`52
`19
`
`9
`68
`23
`13
`45
`
`114
`87
`
`90
`22
`40
`
`71
`28
`13
`72
`
`334
`12
`
`69
`36–92
`21
`
`40
`31
`29
`
`24
`44
`18
`
`8
`72
`21
`12
`45
`
`108
`84
`
`91
`24
`39
`
`69
`30
`16
`66
`
`337
`12
`
`68
`43–86
`20
`
`42
`28
`30
`
`21
`51
`20
`
`6
`69
`25
`14
`46
`
`123
`89
`
`92
`22
`40
`
`69
`28
`15
`68
`
`* All patients were included in the intention-to-treat analysis. Because of rounding, not all percentages total 100.
`† Hormonal manipulation was defined as bilateral orchiectomy or hormone therapy.
`‡ Pain was defined by a score of 2 or more on the Present Pain Intensity scale or an analgesic score of at least 10.
`§ Patients may have more than one indication for progression of disease.
`
`1506
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`,
`
`7
`
`2004
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`docetaxel versus mitoxantrone for prostate cancer
`
`Table 2. Treatment.*
`
`Variable
`
`No. randomized
`
`No. treated with chemotherapy
`
`No. treated with prednisone
`
`No. of cycles
`
`Median
`
`Range
`
`≥1 Infusion delayed (%)
`
`Dose reduction (%)
`
`Major protocol violation (%)
`
`Reasons for stopping treatment (%)
`
`Completed treatment
`
`Progression of disease
`
`Docetaxel
`Every 3 Wk
`
`Weekly
`Docetaxel
`
`Mitoxantrone
`Every 3 Wk
`
`335
`
`332
`
`332
`
`9.5
`
`1–11
`
`24
`
`12
`
`7
`
`46
`
`38
`
`334
`
`330
`
`330
`
`4
`
`1–6
`
`34
`
`9
`
`8
`
`35
`
`35
`
`337
`
`335
`
`335
`
`5
`
`1–11
`
`21
`
`8
`
`7
`
`25
`
`56
`
`Adverse event
`
`Withdrawal of consent
`
`Death
`
`Other
`
`Crossover to other drug (%)
`
`11
`
`1
`
`1
`
`4
`
`27
`
`16
`
`6
`
`2
`
`6
`
`24
`
`10
`
`3
`
`2
`
`5
`
`20
`
`*
`Percentages relate to the number of patients treated in each group. Because of
`rounding, not all percentages total 100.
`
`Weekly
`docetaxel
`
`Docetaxel
`every 3 wk
`
`0
`
`3
`
`6
`
`9
`
`12
`
`15
`18
`Months
`
`Mitoxantrone
`
`21
`
`24
`
`27
`
`30
`
`33
`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`
`Probability of Overall Survival (%)
`
`No. at Risk
`Docetaxel every
`3 wk
`Weekly docetaxel
`Mitoxantrone
`
`335
`
`334
`337
`
`296
`
`297
`297
`
`217
`
`200
`192
`
`104
`
`105
`95
`
`37
`
`29
`29
`
`5
`
`4
`3
`
`Figure 1. Kaplan–Meier Estimates of the Probability of Overall Survival
`in the Three Groups.
`
`tabase was locked on August 6, 2003, after the req-
`uisite number of deaths, specified in the statistical
`plan, had occurred.
`The baseline characteristics of the patients were
`well balanced among the three treatment groups
`(Table 1). The median age was 68 years; about 20
`percent of the patients were at least 75 years old.
`About 45 percent had pain, and about 40 percent
`had measurable soft-tissue lesions. The most com-
`mon indicators of disease progression before study
`entry were an increasing serum PSA level and evi-
`dence of an increase in bone metastases on bone
`scanning.
`Only nine patients (1 percent) did not receive
`chemotherapy and prednisone (Table 2). Patients
`tended to receive more cycles of the regimen in
`which docetaxel was given every three weeks than
`of the regimen in which mitoxantrone was given
`every three weeks. Most patients received the pre-
`scribed doses on schedule, with 8 to 12 percent
`requiring a dose reduction and 21 to 34 percent
`requiring at least one chemotherapy infusion to be
`delayed. Twenty percent of the patients who were
`randomly assigned to receive mitoxantrone sub-
`sequently received docetaxel, and 27 percent of
`those in the group given docetaxel every three
`weeks received subsequent mitoxantrone, as did
`24 percent of those in the weekly-docetaxel group.
`
`efficacy
`The median duration of follow-up was similar
`among the three groups: 20.8 months in the group
`given docetaxel every 3 weeks and 20.7 months in
`the other two groups. There were 166 deaths (50
`percent; hazard ratio for death, 0.76; 95 percent
`confidence interval, 0.62 to 0.94) in the group giv-
`en docetaxel every three weeks and 190 deaths (57
`percent; hazard ratio, 0.91; 95 percent confidence
`interval, 0.75 to 1.11) in the group given weekly
`docetaxel, as compared with 201 deaths (60 per-
`cent) in the mitoxantrone group. When the two
`docetaxel groups were combined and compared
`with the mitoxantrone group, the hazard ratio for
`death was 0.83 (95 percent confidence interval, 0.70
`to 0.99; P=0.04). As compared with the survival
`rate in the mitoxantrone group, the survival rate
`was significantly higher (P=0.009) in the group
`given docetaxel every three weeks but not in the
`group given weekly docetaxel (P=0.36). The medi-
`an duration of survival was 18.9 months (95 per-
`cent confidence interval, 17.0 to 21.2) in the group
`given docetaxel every 3 weeks, 17.4 months (95 per-
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`7, 2004
`
`1507
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` new england journal
`The
`
` medicine
`of
`
`Table 3. Response to Treatment, as Measured by Decreases in Pain, PSA Level,
`and Tumor Burden and Improvements in the Quality of Life.*
`
`Variable
`
`Pain†
`
`Docetaxel
`Every 3 Wk
`
`Weekly
`Docetaxel
`
`Mitoxantrone
`Every 3 Wk
`
`No. who could be evaluated
`
`153
`
`154
`
`157
`
`Response (%)
`
`Rate
`
`95% CI
`
`P value
`
`Duration (mo)‡
`
`Median
`
`95% CI
`
`35
`
`27–43
`
`0.01
`
`31
`
`24–39
`
`0.08
`
`22
`
`16–29
`
`3.5
`
`5.6
`
`4.8
`
`2.4–8.1
`
`2.8–6.8
`
`4.4–indeter-
`minate
`
`≥50% Reduction in serum PSA
`
`No. who could be evaluated
`
`Response (%)
`
`Rate
`
`291
`
`45
`
`282
`
`48
`
`300
`
`32
`
`cent confidence interval, 15.7 to 19.0) in the group
`given weekly docetaxel, and 16.5 months (95 per-
`cent confidence interval, 14.4 to 18.6) in the mito-
`xantrone group. Kaplan–Meier survival curves for
`the three groups are shown in Figure 1.
`The result of the sensitivity analysis, in which
`survival was adjusted for possible imbalances in
`potential prognostic factors, was consistent with
`the primary result. The hazard ratio for death in the
`group given docetaxel every three weeks, as com-
`pared with the mitoxantrone group, was 0.76 with-
`out adjustment and 0.74 and 0.75 after adjustment
`in the full stratified and backward Cox propor-
`tional-hazards models, respectively. As expected,
`visceral involvement, a high baseline alkaline phos-
`phatase level, and a low hemoglobin level were
`negative prognostic factors in the multivariate mod-
`els, whereas a rising serum PSA as the sole indica-
`tor of progression was a favorable factor. Post hoc
`analysis indicated that a high Gleason score (8, 9, or
`10) was an adverse prognostic factor for survival.
`The survival benefit of docetaxel given every three
`weeks was consistent across subgroups defined ac-
`cording to the presence or absence of pain at base-
`line, the Karnofsky performance-status score (70
`percent or less vs. 80 percent or more), and age
`(younger than 65 years vs. 65 years or older) (data
`not shown).
`A reduction in pain was more frequent among
`patients receiving docetaxel every three weeks than
`among those treated with mitoxantrone (35 per-
`cent vs. 22 percent, P=0.01) (Table 3), but the per-
`centage of patients with reduced pain in the weekly
`docetaxel group (31 percent) did not differ signifi-
`cantly from that of the mitoxantrone group. The
`median duration of reduced pain was 3.5 to 5.6
`months and did not differ significantly among the
`groups.
`Rates of PSA response were significantly higher
`in the docetaxel groups (45 percent in the group
`given docetaxel every three weeks and 48 percent in
`the group given weekly docetaxel, P<0.001 for both
`comparisons) than in the mitoxantrone group (32
`percent) (Table 3). The median duration of the PSA
`response ranged from 7.7 to 8.2 months and did
`not differ significantly among the three groups.
`Although patients with measurable soft-tissue
`lesions who received docetaxel every three weeks
`had a somewhat higher rate of tumor response than
`such patients who received mitoxantrone every
`three weeks (12 percent vs. 7 percent, P=0.11), this
`difference was not significant (Table 3).
`
`95% CI
`
`P value
`
`Duration (mo)‡
`
`Median
`
`95% CI
`
`Tumor response
`
`40–51
`
`<0.001
`
`42–54
`
`<0.001
`
`26–37
`
`7.7
`
`8.2
`
`7.8
`
`7.1–8.6
`
`6.3–11.5
`
`5.4–10.5
`
`No. who could be evaluated
`
`141
`
`Response (%)
`
`Rate
`
`95% CI
`
`P value
`
`Quality of life
`
`12
`
`7–19
`
`0.11
`
`No. who could be evaluated
`
`278
`
`Response (%)§
`
`Rate
`
`95% CI
`
`P value
`
`22
`
`17–27
`
`0.009
`
`134
`
`8
`
`4–14
`
`0.59
`
`270
`
`23
`
`18–28
`
`0.005
`
`137
`
`7
`
`3–12
`
`267
`
`13
`
`9–18
`
`* P values are for comparisons with the mitoxantrone group. CI denotes confi-
`dence interval.
`† A pain response was defined as a two-point reduction in the Present Pain In-
`tensity (PPI) score without an increase in the analgesic score or a reduction of
`at least 50 percent in the analgesic score without an increase in the PPI score,
`which was maintained for at least three weeks.
`‡ Data on 54 percent and 63 percent of patients were censored in the Kaplan–
`Meier analysis of the median duration of pain and PSA response, respectively.
`The chief reason for data censoring was further antitumor therapy after pro-
`gression of disease as defined by other criteria.
`§ A response was defined by a 16-point improvement from baseline in the Func-
`tional Assessment of Cancer Therapy–Prostate (FACT-P) score on two mea-
`surements obtained at least three weeks apart.
`
`1508
`
`n engl j med
`
`351;15
`
`www.nejm.org october
`
`,
`
`7
`
`2004
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on August 10, 2015. For personal use only. No other uses without permission.
`
` Copyright © 2004 Massachusetts Medical Society. All rights reserved.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`docetaxel versus mitoxantrone for prostate cancer
`
`adverse events
`The incidence of grade 3 and 4 neutropenia was
`relatively low, and febrile neutropenia was rare (Ta-
`ble 4). Two patients died from sepsis during treat-
`ment, one in the docetaxel group and one in the
`mitoxantrone group. There was a higher incidence
`of cardiac events among patients who received mi-
`toxantrone (Table 4). Most other types of adverse
`events were more frequent among patients receiv-
`ing docetaxel, and there was no trend toward a low-
`er frequency with weekly docetaxel than with doce-
`taxel given every three weeks. Low-grade adverse
`events that occurred in at least 15 percent of pa-
`tients in one of the groups included fatigue, nausea
`or vomiting or both, alopecia, diarrhea, nail chang-
`es, sensory neuropathy, anorexia, changes in taste,
`stomatitis, dyspnea, tearing, peripheral edema, and
`epistaxis (Table 4). More patients in the docetaxel
`groups than in the mitoxantrone group had at least
`one serious adverse event, with rates of 26 percent
`among those in the group given docetaxel every
`three weeks, 29 percent among those given weekly
`docetaxel, and 20 percent among those given mito-
`xantrone. Five deaths were probably related to treat-
`ment, three of them in the mitoxantrone group.
`More patients in the mitoxantrone group stopped
`treatment because of disease progression than was
`the case in the docetaxel groups, and more stopped
`treatment because of completion of treatment or
`adverse events in the docetaxel groups (Table 2).
`Adverse events that led to the discontinuation of
`treatment included fatigue, musculoskel