Paclitaxel Versus Doxorubicin as First-Line Single-Agent
`Chemotherapy for Metastatic Breast Cancer: A European
`Organization for Research and Treatment of Cancer
`Randomized Study With Cross-Over
`
`By R. Paridaens, L. Biganzoli, P. Bruning, J.G.M. Kliin, T. Gamucci, S. Houston, R. Coleman, J. Schachter,
`A. Van Vreckem, R. Sylvester, A. Awada, J. Wilcliers, and M. Piccart on behalf of the European Organization for Research
`and Treatment of Cancer-Investigational Drug Branch for Breast Cancer/ Early Clinical Studies Group
`
`Purpose: To compare the efficacy of paclitaxel ver-
`sus doxorubicin given as single agents in first-line ther-
`apy of advanced breast cancer (primary end point,
`progression-free survival [PFS]) and to explore the de-
`gree of cross-resistance between the two agents.
`Patients and Methods: Three hundred thirty-one pa-
`tients were randomized to receive either paclitaxel 200
`mg/mz, 3-hour infusion every 3 weeks, or doxorubicin
`75 mg/m”, intravenous bolus every 3 weeks. Seven
`courses were planned unless progression or unaccept-
`able toxicity occurred before the seven courses were
`finished. Patients who progressed within the seven
`courses underwent early cross-over to the alternative
`drug, while a delayed cross-over was optional for the
`remainder of patients at the time of disease progres-
`slon.
`
`Results: Obiective response in first-line therapy was
`significantly better (P = .003) for doxorubicin (response
`rate [RR], 41%) than for paclitaxel (RR, 25%), with
`
`doxorubicin achieving a longer median PFS (7.5 months
`for doxorubicin v 3.9 months for paclitaxel, P < .001).
`In second-line therapy, cross-over to doxorubicin (91
`patients) and to paclitaxel (77 patients) gave response
`rates of 30% and 16%, respectively. The median sur-
`vival durations of 18.3 months for doxorubicin and
`15.6 months for paclitaxel were not significantly differ-
`ent (P = .38). The doxorubicin arm had greater toxicity,
`but this was counterbalanced by better symptom con-
`trol.
`
`Conclusion: At the dosages and schedules used in
`the present study, doxorubicin achieves better disease
`and symptom control than paclitaxel in first-line treat-
`ment. Doxorubicin and paclitaxel are not totally cross-
`resistant, which supports further investigation of these
`drugs in combination or in sequence, both in advanced
`disease and in the adiuvant setting.
`J Clin Oncol 18:724-733. © 2000 by American
`Society of Clinical Oncology.
`
`LTHOUGH ADVANCED breast cancer may be con-
`sidered incurable, efficient palliation can be achieved
`with the optimal use of therapeutic resources, ie, radiother-
`apy, hormonotherapy, biphosphonates, and chemotherapy.
`In this setting, chemotherapy is generally prescribed after
`failure of endocrine therapy. However,
`it is prescribed as
`first-line treatment for patients with receptor-negative tu-
`mors or with life-threatening visceral metastases.
`
`From the University Hospital Gasthuisberg, Leuven; and Investiga-
`tional Drug Branchfor Breast Cancer; and European Organizationfor
`Research and Treatment of Cancer Data Center; and Institut Jules
`Bordet, Brussels, Belgium; and Netherlands Cancer Institute, Amster-
`dam; and Daniel den Hoed Klinik and Academic Hospital, Rotterdam,
`The Netherlands; and [stituto Regina Helena, Rome, Italy; and Guy ’s
`Hospital, London; and Weston Park Hospital, Sheflield, United King-
`dom: and Rabin Medical Centre, Tel Aviv, Israel.
`Submitted April 12, 1999; accepted October 29, 1999.
`Address reprint requests to Robert Paridaens, MD, Department of
`Oncology, University Hospital Gasthuisberg, Herestraat 49, B-3000
`Leuven, Belgium; email Robert.Paridaens@uz.kuleuvenoc. be.
`© 2000 by American Society of Clinical Oncology.
`0732—183X/00/I804-724
`
`Among the presently available cytotoxic drugs, doxoru-
`bicin is generally considered as the most active agent in
`advanced breast cancer.l Although the best chemotherapy
`regimen remains a matter of debate, the results of several
`studies indicate that there is a dose-response relationship for
`this agent and that, at the maximum-tolerated dose of 75
`mg/m2, the results obtained with monotherapy are compa-
`rable with those achieved with a standard combination such
`
`as cyclophosphamide, methotrexate, and fluorouracil?‘5
`When progression occurs after first-line chemotherapy, the
`response to salvage treatment is generally disappointing,
`indicating that similar mechanisms of resistance affect most
`conventional regimens. In this regard, paclitaxel represents
`a major advance. This cytotoxic compound, originally
`extracted from the bark of the Pacific yew (Taxus brevifo-
`lia), stabilizes polymerized microtubules, thereby disrupting
`111itosis.°‘7 It was selected in 1977 by the National Cancer
`Institute (NCI) for further development as an antitumor drug
`and proved, subsequently, to be active in a broad panel of
`tumors,
`including ovarian and breast cancer.8 In breast
`cancer, paclitaxel seemed particularly promising because it
`elicited impressive response rates in phase II studies as
`second- or third-line therapy in patients with demonstrated
`resistance to anthracyclinesf)‘20
`
`724
`
`Journal of Clinical Oncology, Vol 18, No 4 (February), 2000: pp 724733
`Genentech 2104
`
`Downloaded from ascopubsorg by Reprints Desk on November 27. 2017 from 216185156028
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`Hospira v. Genentech
`|PR2017-OO737
`
`Genentech 2104
`Hospira v. Genentech
`IPR2017-00737
`
`

`

`PACLITAXEL v DOXORUBIClN IN BREAST CANCER
`
`725
`
`The purpose of the present phase III trial was to compare
`the efficacy of paclitaxel and doxorubicin given as single
`agents in first-line therapy of advanced breast cancer. A
`cross-over applied systematically on demonstrated progres-
`sion during first-line treatment, unless contraindicated or
`refused by the patient, intended to explore further the degree
`of cross-resistance between the two agents.
`
`PATIENTS AND METHODS
`
`Eligibility Criteria
`
`To be eligible, patients were required to have histologically or
`cytologically proven adenocarcinoma of the breast and metastatic
`disease in overt progression with uni- or bidimensionally measurable
`lesions. Prior hormonotherapy, radiotherapy, or immunotherapy for
`adjuvant or neoadjuvant purposes or for advanced disease was allowed
`but had to be stopped at study entry. No prior chemotherapy for
`advanced disease was permitted. Prior adjuvant chemotherapy with
`cyclophosphamide, methotrexate, and fluorouracil or comparable reg-
`imens was authorized on the provision that patients had never been
`exposed to anthracyclines or to taxanes and that all cytotoxic drugs
`were stopped at least 3 months before study entry. Other requirements
`for eligibility were an age greater than 18 years, a World Health
`Organization performance status of 0 to 2, a life expectancy of at least
`3 months, adequate hematologic parameters (absolute neutrophil
`count > 1.5 X 10%, platelets > 100 X 10%), adequate renal and
`hepatic functions (total bilirubin and creatinine < 1.25 X upper normal
`limit), a normal cardiac fimction, and absence of cardiac disease or
`significant arythmia as demonstrated by ECG and left ventricular
`ejection fraction (LVEF) measurement (echocardiography or multiple
`gated aquisition scan). Men and pregnant or lactating women were
`excluded. The protocol was approved by the European Organization for
`Research and Treatment of Cancer (EORTC) Protocol Review Com-
`mittee and by the ethics committee of each participating center; all
`patients gave their written informed consent to participate in the trial.
`
`(grade 3 or 4) nonhematologic toxicity persisting at the scheduled
`retreatrnent date, treatment was delayed until recovery. If treatment
`could not be administered on day 43, patients went off study. Three
`dose reduction levels were foreseen for each regimen (60, 50, and 40
`mg/m2 for doxorubicin; 175, 135, and 110 mg/ml for paclitaxel) if the
`patient experienced severe toxicity during the previous cycle. This
`included grade 3 nonhematologic toxicity (other than nausea, vomiting,
`alopecia, or diarrhea), prolonged myelosuppression (ANC < 0.5 X
`109/L or platelets < 50 X 109/L for Z 7 days), febrile neutropenia, or
`documented infection during neutropenia. No subsequent dose re-
`escalation was permitted. Patients who could not tolerate the lowest
`dose level went off study.
`In first-line treatment, a maximum of seven cycles of doxorubicin
`were planned (cumulative dose 525 mg/mz), whereas paclitaxel therapy
`could be either stopped after seven cycles or pursued in responding
`patients at the discretion of the treating physician. Patients progressing
`during their seven courses were to undergo early cross-over to the
`alternative drug, whereas a delayed cross-over was optional for the
`remainder at the time of disease progression.
`
`Pretreatment and Follow- Up Studies
`
`At baseline, standard evaluations including physical examination,
`performance status, routine hematology, and biochemistry parameters,
`ECG and tumor assessments were performed. Toxicity was assessed in
`each cycle of therapy using the NCI common toxicity criteria (NCI
`CTC)?‘ Tumor response was assessed after cycles 3, 5, and 7, and
`every 2 months thereafter. A full re-evaluation of disease extension was
`performed within 2 weeks of cross-over. A multiple gated aquisition
`scan or an echocardiography for the evaluation of the LVEF was
`mandatory at study entry and at completion of the fifth and seventh
`course of treatment during doxorubicin administration. Quality of life
`(QOL) was assessed by the EORTC Quality-of-Life Questionnaire C30
`(EORTC QLQ-C30) and the Rotterdam Symptom Checklist22 at
`baseline, at the completion of cycles 3, 5, and 7 of both first-line and
`cross-over therapy, and during follow-up at every 2 months until
`disease progression.
`
`Treatment Allocation and Administration
`
`Criteria of Evaluation
`
`All participating institutions were affiliated with the EORTC and
`were members of the lnvestigational Drug Branch for Breast Cancer
`(IDBBC) of the Breast Cancer Cooperative Group and/or the Early
`Clinical Studies Group (ECSG). Randomization was performed cen-
`trally at the EORTC Data Center located in Brussels (Belgium) by
`telephone, fax, or computer. Alter stratification for the institution and
`for prior adjuvant chemotherapy (no; yes with relapse earlier than 1
`year after completion; or yes with relapse 1 year or more afier
`completion), patients were randomly assigned to receive either doxo-
`rubicin (75 mg/m2, administered as a short infusion over 5 to 15
`minutes) or paclitaxel
`(200 mg/mz, administered as a continuous
`infusion during 3 hours) as first-line therapy. For patients receiving
`paclitaxel, a premedication of dexamethasone (20 mg orally, 12 and 6
`hours before paclitaxel), diphenhydramine (50 mg intravenous 30
`minutes before paclitaxel), and cimetidine (300 mg intravenous 30
`minutes before paclitaxel) was systematically given. For patients
`receiving doxorubicin, a premedication of dexamethasone and a 5-hy-
`droxytryptamine—3 antagonist was administered.
`Cycles were repeated every 3 weeks, with weekly monitoring of
`hematology. For patients who did not achieve hematologic recovery
`(ANC 2 1.5 X 109/L and platelets _> 100 X 109/L) or in case of severe
`
`Tumor response was assessed according to the Union Internationale
`Contre le Cancer criteria.23 Complete remission (CR) was defined as
`the disappearance of all known disease determined by two observations
`not less than 4 weeks apart. Partial remission (PR) was defined, for
`bidimensionally measurable disease, as the decrease of at least 50% of
`the sum of the products of the largest perpendicular diameters of all
`measurable lesions as determined by two observations not less than 4
`weeks apart or, for unidimensionally measurable disease, a decrease of
`at least 50% of the sum of the largest diameters of all lesions. It was not
`necessary for all
`lesions to have regressed to qualify for a partial
`response, but no lesion should have progressed and no new lesion
`should have appeared. Serial evidence of appreciable change docu-
`mented by radiography or photography was obtained for external
`review. The stable disease category was defined as a less than 50%
`decrease and a less than 25% increase either in the sum of the products
`of the largest perpendicular diameters of all measurable lesions, in case
`of bidimensionally measurable disease, or in the sum of the diameters
`of all lesions for unidimensionally measurable disease, maintained at
`least 2 months after treatment initiation. In addition, no lesion should
`have progressed and no new lesion should have appeared. Progressive
`disease (PD) was defined as a 25% or greater increase in the size of at
`
`Downloaded from ascopubsorg by Reprints Desk on November 27, 2017 from 216.185.156.028
`Convright © 2017 American Societv of Clinical Oncology. All rights reserved.
`
`

`

`726
`
`least one bidimensionally measurable lesion or the appearance of a new
`lesion.
`All case report forms were regularly reviewed and source verified. A
`large proportion of the objective remissions (CR or PR) (85% and 90%
`for first- and second-line treatment, respectively) and a sample of
`disease stabilizations (22% of first-line stable disease) were reviewed.
`The cases were presented by the local investigators to two independent
`radiologists who were blinded to treatment arm.
`
`Study End Points and Statistical Methods
`
`The primary end points of the study were progression-free survival
`(PFS, calculated from the day of randomization until the date of PD or
`death, if it occurred before documentation of PD) in first-line chemo-
`therapy and response rate (RR) in second-line chemotherapy. RR to
`first-line therapy, QOL, and overall survival (08) were secondary end
`points.
`Results were analyzed according to the intent-to-treat principle. The
`X2 test and Fisher’s exact test were used to compare response rate and
`toxicity. The duration of response was computed from the date of
`randomization until documentation of PD. OS and PFS were estimated
`
`using the Kaplan-Meier method24 and compared using a two-sided
`log-rank test.25 Logistic regression and Cox’s proportional hazards
`regression model was used to assess treatment effect after adjustment
`for important prognostic variables.26 When a patient was crossed over
`to the alternative drug or started a new therapy without documented
`progression, this patient was counted as a treatment failure at the start
`of this unauthorized therapy.
`
`RESULTS
`
`Treatment Delivery
`
`Between August 1993 and May 1996, 331 patients from
`14 institutions were randomized to receive either paclitaxel
`(166 patients) or doxorubicin (165 patients) as first-line
`chemotherapy. Sixteen patients were considered not eligible
`because of the absence of measurable lesions (n = 9),
`inadequate organ function (n = 4), the presence of brain
`metastases (n = 2), or prior chemotherapy for metastatic
`disease (n = 1). Patient characteristics were similar in the
`two groups, and, as
`listed in Table 1,
`there was no
`significant imbalance in classical prognostic factors. The
`details of treatment delivery in both first- and second-line
`treatments are given in Table 2. Two patients in each group
`never started first-line treatment,
`leaving 164 and 163
`patients who received at least one course of paclitaxel and
`doxorubicin, respectively.
`First-line treatment. A median of seven cycles was
`administered per patient in both the paclitaxel arm (range,
`one to 22 cycles) and the doxorubicin arm (range, one to
`nine cycles). Twenty-one patients versus one patient re-
`ceived more than seven cycles of therapy in the paclitaxel
`and doxorubicin arms, respectively. However, because dose
`reductions and treatment delays were more frequent in the
`doxorubicin arm, the average dose-intensity achieved with
`doxorubicin (93% of planned dose-intensity; range, 53% to
`
`PARIDAENS ET AL
`
`Table 1. Patient Characteristics
`Patients Treated
`Vlfith Paclituxel
`(n ‘- 166)
`No.
`
`%
`
`Patients Treated
`With Doxorubicin
`[n = 165)
`No.
`
`%
`
`Age, years
`Median
`Range
`WHO performance status
`0
`1
`2
`DFI, months
`Median
`
`Range
`Prior radiotherapy
`Prior hormonotherapy
`Prior adiuvant chemotherapy
`Dominant site of disease
`SoFt tissue only
`Bone +/- soft tissue
`Single visceral
`Multiple visceral
`No. of metastatic sites
`1
`2
`3 or more
`Hormone receptor status
`ER=positive
`ER = negative
`ER unknown
`
`54
`31 -74
`
`24
`
`0-212
`
`69
`83
`14
`
`130
`126
`53
`
`10
`26
`105
`25
`
`47
`73
`46
`
`45
`64
`57
`
`42
`5O
`8
`
`78
`76
`32
`
`6
`16
`63
`15
`
`28
`44
`28
`
`27
`39
`34
`
`55
`26-75
`
`33
`
`0-383
`
`68
`82
`15
`
`124
`122
`54
`
`18
`23
`93
`31
`
`43
`7O
`52
`
`40
`64
`61
`
`41
`50
`9
`
`75
`74
`33
`
`1 1
`14
`56
`19
`
`26
`42
`32
`
`24
`39
`37
`
`Abbreviation: DFI, disease-Free interval; WHO, World Health Organization.
`
`103%) was lower than with paclitaxel (99% of planned
`dose-intensity; range, 72% to 107%). In patients receiving
`paclitaxel, the most frequent reasons for delaying treatment
`were neutropenia, neutropenic fever, and neurotoxicity, and
`infection was the most frequent reason for dose reduction.
`In the doxorubicin arm, neutropenia and stomatitis were the
`principal causes of treatment delay, and neutropenia, sto-
`matitis, and neutropenic fever were the principal causes of
`dose reduction. The most frequent
`toxicities leading to
`treatment interruption were gastrointestinal (four patients),
`hematologic (three patients), infection (two patients), and
`cardiac function (17 patients) with doxorubicin, and neuro-
`toxicity (four patients) with paclitaxel.
`The most frequent reasons for stopping first-line pacli-
`taxel were disease progression, including death caused by
`malignancy (90 patients; 55%), treatment completion (57
`patients; 35%), and excessive toxicity and/or patient refusal
`(12 patients; 7%). Eighty-one doxorubicin patients (50%)
`ended first-line therapy because of treatment completion, 37
`(23%) because of excessive toxicity and/or patient refusal,
`and 36 (22%) because of PD.
`
`Downloaded from ascopubsorg by Reprints Desk on November 27, 2017 from 216.185.156.028
`Convright © 2017 American Society of Clinical Oncologv. All rights reserved.
`
`

`

`PACLITAXEL v DOXORUBICIN IN BREAST CANCER
`
`727
`
`Table 2. Drug Administration in First- and Second-Line Treatment
`First-Line Treatment
`Second'Line Treatment
`Paclitaxel Doxorubicin
`Paclitaxel Doxorubicin
`
`Total no. 01 patients
`Total no. at cycles
`No. of cycles per patient
`Median
`Range
`No. of patients > 7 cycles
`Relative DI, %
`Median
`Range
`Dose reduction, % of cycles
`Not reduced
`Reduced by 1 dose level
`Reduced by 2 dose levels
`Reduced by 3 dose levels
`Dose reduction, % oF patients
`'Not requiring dose
`reduction
`
`Delay, % at cycles
`Delay, % of patients
`
`164
`949
`
`7
`1-22
`21
`
`163
`908
`
`7
`1-9
`1
`
`77
`382
`
`5
`1-15
`7
`
`91
`485
`
`6
`1-8
`1
`
`99
`72-107
`
`93
`53-103
`
`97
`67-104
`
`92
`68-106
`
`94
`5
`1
`-
`
`90
`
`5
`20
`
`80
`17
`2
`1
`
`68
`
`19
`44
`
`89
`10
`1
`-
`
`85
`
`5
`16
`
`87
`1 1
`2
`-
`
`75
`
`18
`42
`
`Abbreviation: D1, dose-intensity.
`
`Second-line treatment. Among those patients eligible
`for early cross-over, 65 (76%) out of 86 paclitaxel-resistant
`patients received doxorubicin, whereas 24 (75%) out of 32
`doxorubicin-resistant patients received paclitaxel. Overall,
`168 patients have been crossed to the alternate drug, for a
`total of 77 patients receiving paclitaxel and 91 patients
`receiving doxorubicin. Late cross-over was performed in 26
`(46%) out of 57 patients previously exposed to paclitaxel
`and in 53 (65%) out of 81 patients first
`treated with
`doxorubicin.
`
`cycles of paclitaxel and 18% of doxorubicin cycles were
`delayed, mainly for neutropenia and neutropenic fever.
`Approximately 10% of the cycles were dose-reduced in the
`two treatment arms. The most frequent reasons for stopping
`second-line therapy were disease progression (61% of
`patients) under paclitaxel and treatment completion (47% of
`patients) under doxorubicin.
`Data on second-line therapy are available for 95% and
`91% of the first-line paclitaxel and first-line doxorubicin
`patients, respectively, who did not cross over. Fifty-five
`percent of paclitaxel first-line patients received a further
`systemic treatment (hormonotherapy 20% and chemother-
`apy 80%); 60% of the second-line regimens were anthracy-
`cline-based. A second-line chemotherapy was received by
`62% of the patients previously treated with doxorubicin. A
`taxane-based regimen was received by 21% of these pa-
`tients.
`
`Response and Efi‘icacy Data
`
`In first-line therapy, 15 paclitaxel patients and 17 doxo-
`rubicin patients were not assessable for response, but all
`were included in the analyses of response rate,
`time to
`progression, and survival according to the intent-to-treat
`principle. As previously stated, four patients received no
`trial
`therapy. The main reasons for failure of response
`evaluation in the eligible patients were insufficient docu-
`mentation of response under treatment (paclitaxel, n = 2;
`doxorubicin, n = 3), drop out from study for toxicity/refusal
`(four patients in each treatment arm), and early death from
`nonmalignant conditions (paclitaxel, n = 3; doxorubicin,
`n = I) or from toxicity (two patients on doxorubicin).
`The patterns of response achieved under paclitaxel or
`doxorubicin in first— and in second-line treatment are de-
`
`Overall, a median of five cycles (range, one to 15 cycles)
`were administered per patient in the paclitaxel arm, and a
`median of six cycles (range, one to eight cycles) were
`administered in the doxorubicin arm. Five percent of the
`
`picted in Table 3. The objective remission rate (CR + PR)
`was significantly better for doxorubicin than for paclitaxel
`in first-line treatment (41%; 95% confidence interval, 33.1%
`to 48.1% v 25%; 95% confidence interval, 18.7 to 31.9%,
`
`Table 3. Best Response in First and Second-Line Treatment
`Second-Line Treatment
`
`Late Cross-Over
`Early Cross-Over
`First-Line Treatment
`Patients Treated
`Patients Treated
`Patients Treated
`Patients Treated
`Patients Treated
`Patients Treated
`VWth Paclitaxel
`Vtfith Doxorubicin
`With Paclitaxel
`\Mth Doxorubicin
`With Doxarubicin
`With Paclitaxel
`(n = 166)
`(n ‘ 165)
`(n 1* 24)
`(n
`65)
`(n - 53)
`[n ' 26)
`No.
`%
`No.
`%
`P
`No.
`%
`No.
`%
`No.
`%
`No.
`%
`
`
`0
`0
`0
`0
`O
`0
`0
`6
`10
`2
`3
`CR
`38
`10
`17
`9
`26
`17
`13
`35
`57
`23
`39
`PR
`38
`10
`38
`20
`52
`34
`42
`36
`60
`35
`58
`SD
`12
`3
`43
`23
`15
`10
`38
`13
`21
`31
`51
`PD
`
`Not assessable 12 15 9 17 10 8 4 7 l 2 3
`
`
`
`
`
`
`
`
`
`
`Abbreviation: SD, stable disease.
`
`'003
`
`10
`
`Downloaded from ascopubsorg by Reprints Desk on November 27, 2017 from 216.185.156.028
`Copvright © 2017 American Societv of Clinical Oncologv. All rights reserved.
`
`

`

`728
`
`100
`
`90
`
`80
`
`70
`
`60
`
`50
`
`20
`
`10
`
`
`
`Logrank: P=0.0001
`
`40 _
`30
`
`
`
`
`
`0
`
`‘\
`
`0
`
`160
`
`166
`
`156
`
`165
`
`1
`
`2
`
`Number of patients at risk:
`
`11
`
`29
`
`4
`
`2
`
`3
`
`3
`
`1
`
`I (years)
`4
`
`Treatment
`
`— Paclitaxel
`
`Doxorubicin
`
`PARIDAENS ET AL
`
`PFS on first-line treat-
`Fig l.
`ment with paclitaxel (solid line) or
`with doxorubicin (dashed line).
`
`the
`In second line treatment,
`.003).
`respectively; P
`response rates were 30% for doxorubicin and 16% for
`paclitaxel. Response by disease site (number of responses/
`number of sites) for first-line doxorubicin and paclitaxel,
`respectively, was as follows: soft tissue, 67% (76 of 114)
`versus 38% (36 of95); liver, 42% (32 of 77) versus 19% (14
`of 72); lung, 34% (12 of 35) versus 43% (20 of 46); and
`bone, 0% (zero of 14) versus 14% (two of 14). Response by
`disease site for second-line doxorubicin and paclitaxel,
`respectively, was as follows: soft tissue, 33% (15 of 45)
`versus 40% (14 of 35); liver, 38% (17 of 45) versus 18%
`(seven of 39); lung, 26% (six of 23) versus 25% (three of
`12); and bone, 43% (three of seven) versus 0% (zero of
`eight).
`The median duration of objective response in first—line
`treatment was 9.7 months (range, 3.4 to 458+ months) for
`doxorubicin and 7.7 months (range, 3.8 to 555+ months)
`for paclitaxel. In second-line treatment, median duration of
`response was 8.3 months (range, 3.0 to 25.3 months) for
`doxorubicin, and 5.6 months (range, 4.4 to l 1.5 months) for
`paclitaxel. PFS curves for first-line therapy are presented in
`Fig 1. PFS in first-line therapy was significantly longer for
`doxorubicin than for paclitaxel (median, 7.5 months v 3.9
`months, respectively; P = .0001). We did not compare the
`time to disease progression in patients who stopped after
`seven cycles with those who continued beyond seven cycles
`because this is a subgroup analysis of two groups of patients
`that were not determined by randomization and, therefore,
`
`are almost certainly not comparable. There was no signifi-
`cant difference in OS between the two study arms (P = .38),
`with a median survival of 18.3 months in the doxorubicin
`
`arm and 15.6 months in the paclitaxel arm (Fig 2).
`The relationship between treatment and prognostic fac-
`tors on response, PFS, and OS has been examined. The
`prognostic factors selected by univariate analyses were age,
`
`performance status, dominant site, number of sites, prior
`adjuvant chemotherapy, and disease-free interval. After
`adjusting for these factors, doxorubicin remained more
`active than paclitaxel in terms of RR and PFS; again, no
`difference in OS was demonstrated (Table 4).
`
`Toxicity
`
`Toxicities and clinically relevant adverse events encoun-
`tered during treatment in first— and second-line therapy are
`presented in Tables 5 and 6, respectively. As expected, the
`toxicity profiles of the two drugs were different. Doxorubi-
`cin was more toxic than paclitaxel in terms of hematologic,
`gastrointestinal, and cardiac side effects. In contrast, pacli-
`taxel elicited more neurotoxicity (mainly sensory) and
`arthralgia/myalgia. Neutropenia,
`febrile neutropenia (fe-
`ver 2 385°C, associated with an ANC of < 0.5 X 109/L,
`and requiring hospitalization and intravenous antibiotic
`treatment), documented infections, and hospital admission
`for serious adverse events were more frequently observed in
`the doxorubicin arm. No severe hypersensitivity reaction
`
`Downloaded from ascopubs.org by Reprints Desk on November 27, 2017 from 216.185.156.028
`Copyright © 2017 American Society of Clinical Oncolozv. All rights reserved.
`
`

`

`PACLITAXEL v DOXORUBICIN IN BREAST CANCER
`
`729
`
`100
`
`90 80
`
`70
`
`60
`
`50
`
`40
`
`30
`
`20
`
`10
`
`Logrank: P=0.3813
`
` Fig 2. OS in patients treated in
`
`first-line treatment with paclitaxel
`(solid line) or with doxorubicin
`(dashed line).
`
`0
`
`N
`
`0
`
`130
`
`166
`
`124
`
`165
`
`1
`
`2
`
`Number of patients at risk :
`
`105
`
`111
`
`39
`
`43
`
`3
`
`11
`
`ll
`
`Treatment
`
`— Paclitaxel
`
`Doxorubicin
`
`necessitating prolongation of the 3-hour paclitaxel infusion
`was reported.
`Six toxic deaths were encountered: three after doxorubi-
`
`cin in first—line therapy (two patients died as a result of a
`neutropenic sepsis and one patient suffered an episode of
`acute dyspnea and died at home 2 weeks after the first cycle
`of chemotherapy); two in second-line therapy (two myocar-
`diopathies after a cumulative dose of 525 mg/m2 of doxo-
`rubicin); and one after paclitaxel was encountered in sec-
`ond-line treatment (neutropenic sepsis).
`Regarding cardiotoxicity, a total of 12 patients developed
`congestive heart failure, six patients (4%) on first-line
`doxorubicin,
`two on second-line paclitaxel, and four on
`second-line doxorubicin. Of all these patients, the median
`
`Table 4. Cox Proportional Hazards Regression Analysis of Prognostic
`Factors for Response, Progression, and Survival: Unadjusted and Adjusted
` Results
`Risk Ratio
`(doxorubicin/paclitaxel)
`
`95% Cl
`
`P
`
`Response
`Unadjusted
`Adjusted
`PFS
`
`Unadjusted
`Adjusted
`OS
`
`0.49
`0.50
`
`0.60
`0.59
`
`0.92
`Unadjusted
`0.86
`Adjusted
`Abbreviation: 05, overall survival.
`
`0.31-0.79
`0.31-0.81
`
`0.48-0.75
`0.47-0.75
`
`'
`
`0.72-1 .19
`0.67-l .l l
`
`.003
`.005
`
`.001
`.001
`
`.38
`.25
`
`age was 60 years (range, 42 to 69 years), and the median
`cumulative dose of doxorubicin was 480 mg/m2 (range, 300
`to 675 mg/mz). Twenty-nine patients (17 on first-line
`doxorubicin and 12 on second-line doxorubicin) ceased
`therapy because of a drop in LVEF of 2 20% of the
`baseline value. Myocardial ischemic events were reported in
`an additional three patients (two on doxorubicin and one on
`paclitaxel).
`
`QOL
`
`Two QOL questionnaires (EORTC QLQ-C30 and the
`Rotterdam Symptom Checklist) were completed by patients
`at fixed intervals during first-line and cross-over therapy
`until disease progression. The results will be the subject of
`a separate paper, but,
`in brief, 176 patients completed a
`baseline QOL questionnaire (85 patients receiving first-line
`paclitaxel and 91 patients receiving first-line doxorubicin).
`Baseline QOL was strongly associated with survival. An
`analysis based on the assessment afier the third cycle on
`first-line treatment showed no difference in global health
`status/QOL between the two treatment groups. A more
`detailed analysis showed that the greater toxicity of doxo-
`rubicin at
`this time point was compensated by better
`symptom control, particularly pain, compared with that
`achieved with paclitaxel.
`
`DISCUSSION
`
`This study demonstrates that, in advanced breast cancer,
`monotherapy with doxorubicin given at
`the dose of 75
`
`Downloaded from ascopubsorg by Reprints Desk on November 27, 2017 from 216.185.156.028
`Copyright © 2017 American Societv of Clinical Oncolozv. All rights reserved.
`
`

`

`730
`
`PARIDAENS ET AL
`
`Table 5. Main Clinically Relevant Adverse Events and Grade 3 to 4 NCI
`CTC Toxicity Observed Per Patient During First-Line Treatment
`Patients
`Patients
`Treated With
`Treated Wit’n
`Paclitaxel
`Doxorubicin
`[n = 164)
`(n = 163)
`No.
`%
`No.
`
`%
`
`P
`
`Toxicity
`
`toxicity profile and the observation that the efficacy of
`doxorubicin is sufficiently maintained in second-line ther-
`apy and that the paclitaxel followed by doxorubicin se-
`quence does not adversely affect survival.
`Another randomized trial conducted by the American
`Intergroup has compared paclitaxel with doxorubicin in
`first-line treatment for metastatic disease. In that three-arm
`
`study, reported by Sledge eta
`l.27 at the American Society of
`Clinical Oncology meeting in 1997, 739 patients were
`randomized to receive either doxorubicin (60 mg/m2) or
`paclitaxel (175 mg/mz) continuous infusion over 24 hours
`or the combination of doxorubicin (50 mg/m2) and pacli-
`taxel (150 mg/mz) over 24 hours plus granulocyte colony-
`stimulating factor support. Patients receiving single-agent
`chemotherapy were crossed to the alternate drug at the time
`of progression. RRs and median durations of PFS in
`first-line therapy were almost identical for paclitaxel (33%
`RR; 5.9 months) and doxorubicin (34% RR; 6.2 months) but
`were significantly better for the combination of doxorubicin
`and paclitaxel (46% RR; 8 months). There was no signifi-
`cant difference in OS between the three arms of the study.
`Cross-over from doxorubicin to paclitaxel yielded a 20%
`RR, whereas the alternate sequence resulted in a 16% R
`(P = .06).
`The results of the monotherapy arms of the American
`trial are difficult to compare with those of the present study
`because the Intergroup used a lower dose of doxorubicin (60
`instead of 75 mg/m2) and a different dose and schedule of
`administration of paclitaxel
`(175 mg/m2 over 24 hours
`instead of 200 mg/m2 over 3 hours).
`The ideal dose and schedule of paclitaxel are yet to be
`defined. Using the 3-hour infusion, increasing the dose of
`paclitaxel from 175 up to 250 mg/m2 does not significantly
`improve efficacy but does increase toxicity.28 Prolonging
`the duration of infiision of paclitaxel, at a fixed dose of 250
`mg/mz, from 3 to 24 hours results in higher RRs but does
`not improve PFS or OS.29 A randomized trial comparing
`paclitaxel 250 mg/m2 as a 3-hour infusion with paclitaxel
`140 mg/m2 as a 96-hour infusion did not demonstrate any
`difference between the schedules in terms of RR, median
`
`duration of response, and OS.30 However, prolonged infu-
`sion of 120 to 140 mg/m2 over 96 hours may still elicit
`responses in patients who have become refi'actory to higher
`doses given over 3 hours.“
`Finally, new strategies, such as the weekly administration
`of lower doses of paclitaxel, are under investigation.32
`However, the above-mentioned data would suggest that the
`3-hour infiision and the dose of 175 mg/m2 are an adequate
`way of administering paclitaxel in advanced breast cancer
`patients. Should we assume that the superiority of doxoru-
`bicin over paclitaxel in our study is because of the use of a
`
`Neutropenia grade 4
`Febrile neutropenia
`Documented infection
`
`Vomiting
`Stomatitis
`
`Arthralgia/myalgia
`Sensory neurotoxicity
`CHF
`
`Myocardial infarction
`Hospitalization (any cause)
`Toxic death
`
`66
`1 1
`6
`
`4
`2
`
`7
`14
`0
`
`1
`46
`0
`
`40
`7
`4
`
`2
`1
`
`4
`9
`0
`
`1
`28
`0
`
`139
`33
`10
`
`22
`25
`
`0
`0
`6
`
`2
`61
`3
`
`85
`2O
`6
`
`13
`15
`
`0
`O
`4
`
`2
`37
`3
`
`<.OO1
`<.001
`.319
`
`< .001
`< .001
`
`.015
`<.001
`.015
`
`.62
`.07
`.123
`
`Abbreviation: CHF, congestive heart failure.
`
`mg/m2 every 3 weeks achieves better disease control than
`paclitaxel 200 mg/m2 given as a 3-hour infusion every 3
`weeks.
`In first-line treatment,
`therapy with doxorubicin
`results in a higher objective RR and almost
`twice the
`median PFS achiev