VOLUME 26
`
`NUMBER 10
`
`APRIL 1
`
`2008
`
`JOURNAL OF CLINICAL ONCOLOGY
`
`ORIGINAL REPORT
`
`Randomized Phase III Trial Of Weekly Compared With
`Every—S—Weeks Paclitaxel for Metastatic Breast Cancer, With
`Trastuzumab for all HER—2 Overexpressors and Random
`Assignment to Trastuzumab or Not in HER—2
`Nonoverexpressors: Final Results Of Cancer and Leukemia
`Group B Protocol 9840
`Andrew D. Seia'nran, Donald Berry, Constance Cirrincinne, Lyndsay Harris, Hyman Muss, P. Kelly Marconi,
`Graridelia Grpson, Harold Birrstein, Diana Lake. Charles L. Shapiro, Peter Ungaro, Larry Norton, Eric Witter,
`and Clifl'nrd Hrrdis
`
`ABSTRACT
`
`Pur use
`trials suggested that weekly paclitaxel might be more effective and less toxic than
`Phapse II
`every—3—weeks administration for metastatic breast cancer (MBC). Cancer and Leukemia Group B
`(CALGB) protocol 9840 was initiated tO address this question. Subsequently trastuzumab was
`demonstrated to improve outcomes of paclitaxei therapy for human epidermal growth factor
`receptor72 (HERQFpOSitive patients, and was therefore incorporated. Because inhibition of
`HER—family Signaling had potential efficacy even without HER—2 overexpressmn, we randomly
`assrgned for trastuzumab in this population.
`Patients and Methods
`Patients were randomly assrgned to paclitaxel 175 mg/m2 every 3 weeks or 80 mg/m2 weekly.
`After the first 171 patients, all HER727positive patients received trastuzumab; HER-2 nonoverex—
`pressors were randomly assrgned for trastuzumab, in addition to paclitaxel schedule. A total of 577
`patients were treated on 9840. An additional 158 patients were included in analyses, for combined
`sample of 735. The primary and paint was response rate (HR): secondary and pomts were time to
`progressron (Tl'P), overall survrval, and toxicrty. Primary comparisons were between weekly versus
`every-3-weeks paclitaxel, and trastuzumab versus no trastuzumab in HER—2 nonoverexpressors.
`Results
`In the combined sample. weekly paclitaxel was superior to every-S—weeks administration: RR
`(42% v29%, unadjusted odds ratio [OR] = 1.75; P = .0004). TIP (median, 9 v5 months; adjusted
`HR : 1.43; P< .0001), and surVIval (median, 24 v 12 months; adjusted HR = 1.28; P -
`.0092).
`For HER-2 nonoverexpressors, trastuzumab did not improve efficacy. Grade 3 neuropathy was
`more common wrth weekly dosing (24% v12%;
`= .0003).
`Conclusion
`Weekly paclitaxel is more effective than every-3Aweeks administration for MBC. Trastuzumab did
`not improve efficacy for HER-2 nonoverexpressors. NeurOtOXIOity is a treatmentAlimiting tOXIcity
`for weekly paclitaxel.
`
`J Clin Oncol 26:1642-1649. @ 2008 by American Society of Clinical Oncology
`,_ __..
`._
`
`
`Fewsingle chemotherapeutic agents have been stud-
`ied as rigorously as the taxanes regarding dose and
`schedule for breast cancer. Three—hour infusions
`were found to be similarly effective and more con-
`venient than 241 and 96-hour2 infusions. Dose esca-
`lation Ofpaclitaxel from 175 to 210 and 250 mglm2
`Offered no improvement in efficacy, but increased
`
`trials of weekly
`neurotoxicity.3 Phase II clinical
`l-hour paclitaxel in metastatic breast cancer (MBC)
`have demonstrated promising eflicacy and favorable
`tolerability, including with trastuzumab."'m
`In 1998, Cancer and Leukemia Group B
`(CALGB) protocol 9840 began as a prospective ran-
`domized comparison of weekly and every-S-weeks
`(3-weeldy) priclitaxel. This trial began in the pretrzis-
`tuzumab em,” and the first 171 patients enrolled
`EXHIBIT
`
`From the Memorial Sloan—Kettering
`Cancer Center, New York, NY. Cancer
`and Leukemia Group B Statistical
`Center. Duke Unrversrry Medical Center,
`Durham: UnNers-ty oi North Carolina at
`Chapel Hill. Chapel Hill. NC, Dana Faiber
`Cancer Institute. Boston, MA. Vermont
`Cancer Center, Burlington, VT, and The
`Ohio State Unuvarsrty Medical Center
`Columbus, OH
`Submitted May 23, 2007. accepted
`November 27. 2007.
`Supported In part by National Cancer
`Institute Grants NO. CA77651,
`CA33601. CA32291. CA77406
`CA47577, CA77658, and CA47559
`Presented at the 40th Annual Meeting
`of the American Society 01 C-‘inical
`Oncology. June 58, 2004, New
`Orleans. LA
`The content 01 this manuscript Is solely
`the responsibility Oi the authors and
`does not necessarily represent the offi-
`cial Views 121' the Nat anal Cancer insti-
`tute
`Authors' disclosures oi potential can
`firms of Interest and author coniribu
`irons are found at the end of this
`article.
`Corresponding author Andrew D
`Seidman, MD, Memorial S‘oarir
`Kettering Can:er Center, 1275 York
`Avenue. New York. NY 10021: e mail
`seidmanafimserorg
`© 2008 by Amer-can Society at Clin cal
`Oncology
`0732-183X/DEI26ID-1642/SZO 00
`DOI' 10 lZDO/JCO 2007 11 6699
`
`1642
`
`© 2008 by American Scenery of Clinical Oncology
`
`Downloaded from nscopubsprg by 38.98774 on November 19, 2017 from 038.098.007.074
`Copyright .r;. 2017 American Society qulinical Oncology. All rights reserved.
`
`i 1023
`
`(cid:43)(cid:82)(cid:86)(cid:83)(cid:76)(cid:85)(cid:68)(cid:3)(cid:89)(cid:17)(cid:3)(cid:42)(cid:72)(cid:81)(cid:72)(cid:81)(cid:87)(cid:72)(cid:70)(cid:75)
`Hospira v. Genentech
`IPR2017-00805
`(cid:44)(cid:51)(cid:53)(cid:21)(cid:19)(cid:20)(cid:26)(cid:16)(cid:19)(cid:19)(cid:27)(cid:19)(cid:24)(cid:3)
`Genentech Exhibit 2027
`(cid:42)(cid:72)(cid:81)(cid:72)(cid:81)(cid:87)(cid:72)(cid:70)(cid:75)(cid:3)(cid:40)(cid:91)(cid:75)(cid:76)(cid:69)(cid:76)(cid:87)(cid:3)(cid:21)(cid:19)(cid:21)(cid:26)
`
`

`

`Weekly v Every-3-Weeks Paclitaxel for Breast Cancer
`
`had tumors of unknown human epidermal growth factor receptor 2
`(HER-2) status. The protocol was subsequently revised, and all pa-
`tients with HERvZ—overexpressing breast cancer received trastu-
`zumab. Recognizing that no assessment ofHER-2 status is perfect and
`that there were many unanswered questions about the role of trastu-
`zumab in breast cancer."‘"" we also tested the potential value of
`trastuzumab in patients with HER-Z—nonoverexpressing tumors.
`
`r
`
`4-“. “.,.__,___,1'
`
`Patients
`The CALGB 9840 study population consisted of women with measur-
`able, histologically confirmed MBC. Up to one line of prior chemotherapy for
`locally advanced or metastatic disease was allowed. Boneionly, CNS, lymi
`phangitic pulmonary metastases. and previously irradiated tumors without
`subsequent progression were considered nonmeasurable. As per the protocol
`stipulated design, data from MBC patients who received paclitaxel at
`l?5
`mg/m2 on CALGB 9342 were incorporated into the analysis." Random assign-
`ment was weighted (60:40), favoring weekly paclitaxel.
`A total of 585 patients were accrued to CALGB 9840; 577 patients began
`protocol therapy and [58 patients from the 175 mg/m2 arm of CALGB 9342
`are included in this analysis. for a total of735 patients (Fig l).
`Prior taxane was allowed as adjuvant therapy, provided that 1 year had
`transpired from its completion to protocol entry. Adequate renal, hepatic, and
`hematologic parameters were required (Appendix, online only). After trastu-
`Lumab was added to protocol therapy, a normal baseline left ventricular
`ejection fraction {LVEF) was required. Patients with CNS metastases were
`eligible if asymptomatic, not receiving corticosteroids, and more than 6
`months from cranial irradiation.
`
`Methods
`Baseline imaging was performed within 30 days of registration, and an
`ECG within 42 days. Women of child—bearing potential required a negative
`serum B-human chorionic gonadotropin test. Each participant signed an
`institutional review board—approved, protocol-specific informed consent in
`accordance with federal and institutional guidelines.
`Patients were stratified by prior chemotherapy: (I) no chemotherapy in
`the metastatic setting or recurrence more than 6 months of completion of
`
`adjuvant therapy and (2) one prior regimen in the metastatic setting, or no
`prior chemotherapy for metastases but recurred less than 6 months from
`completing adjuvant therapy. Subsequent to the amendment requiringHER-Z
`testing, patients were also stratified by HER-2 status. The first l7l patients were
`not required to have HER—2 testing and were randomly assigned to paclitaxel
`80 mg/m2 weekly via 1-hour infusion, or to paclitaxel 175 mg/m“ every 3 weeks
`via 3-hour infusion. For the first six infusions, weekly paclitaxel was dosed at
`100 mg/m2 but subsequently continued at 80 mg/m2. A 30% incidence of
`grade 3 peripheral sensory neu ropathy resulted in an amended starting dose of
`80 nig/m= weekly. Paclitaxel was to be continued until disease progression or
`limiting toxicity. CBC was obtained every 3 weeks. Prophylactic hematopoietic
`growth factor support could be used as required for treatment—limiting neu-
`tropenia or anemia. A 21-day cycle of therapy could be initiated in weekly
`paclitaxel patients provided that the absolute granulocyte count {AGC} was at
`least [,000iuL; for 3-weekly paclitaxel, the AGC was to be at least 1,500/ttL.
`Platelets had to be at least 100,000r'p.L for both schedules on day l ofeach cycle.
`Standard premedication with dexamethasone 10 mg, diphenhydramine 50
`mg, and either cimetidine 300 mg or ranitidine 50 mg intravenously 30 to 60
`minutes before paciitaxel infusion was required. ‘7
`The protocol was amended on March 15, 2000, to require HERJ status
`assessment by immunohistochemistry I lHC) or fluorescence in situ hybrid-
`ization (FISH ). Those patients with HER—Z- positive disease (IHC 3 + or
`FISH +) received weekly trastuzumab 2 mg/kg via 30-minute infusion follow
`ing a 4-mg/kg loading dose administered over 90 minutes. HER—2 normal
`patients were randomly assigned 50:50 to receive or not receive trasturumab.
`Patients were to receive a minimum oftwo cycles of therapy unless there was
`rapid disease progression. Prior trastuzumab became an exclusion criterion,
`and patients were stratified by HER-Z-status for paclitaxel schedule.
`Treatment Schedule and Dose Modification
`Hematologic toxicity. Filgrastim was prescribed for febrile neutropenia
`or severe neutropenia [absolute neutropllil count [ANC] ‘1 500r'mm1 or WBC
`count r-' 1,000er m l for 27 days) as a S—ptglkg injection daily beginning on day
`2, continuing until the ANC was more than 10,000fmm'l (Ii-weekly arml or
`subcutaneously daily from day 2 to 5, or until the ANC was more than
`l0,000rmml, whichever came first (weekly arm). Filgrastim was continued in
`subsequent cycles. A paclitaxel dose reduction with the next course of therapy
`was to occur if grade 4 thrombocytopenia lpiatelets <. 25,000r'mm‘) occurred,
`with further dose reductions possible {Table I}, with no re-escalation. Dose-
`level reductions are described in Table l.
`
`60%
`
`mm
`N - 11a
`Never-RX. 3
`Assessable. 116
`
`Fig 1. Consort diagram. CALGB, Cancer
`and Leukemia Group B; 3-weekly. every 3
`weeks; Rx. treatment; T. trastuzumab
`
`Original design- '
`July 1993
`__
`
`RANDOMZE
`
`--
`
`'
`
`' came 9342
`3;”mm
`Never began Rx. 3 +
`Assessable, 153
`
`mitt!
`N-56
`NeverRx.1
`Assessabth-‘a _
`
`~ a?
`..
`..
`Revused desrgn
`March 2000
`i
`
`“ma...
`HER2 E
`ran
`
`l
`
`I
`
`_._ c_ ____
`
`.. Nuts
`.
`i NeverRxJ.
`.3 Navarfix,2
`Never-R)“
`rumpbAflu
`Wham r_melsr.§§_,
`
`' N978
`Newman!)
`--A§mbb-7F .
`
`N-‘ioz
`Nmrfixo
`. Assemble. ‘02 .-
`
`wmv.jm.urg
`
`© 2008 by American Socrety of CllnICal Oncorogv
`
`1643
`
`Downloaded from ascopubsorg by 38.98174 on November 19, 20 | 7 from 038.098.007.074
`Copyright El- 2017 American Society ofClirrica! Oncology. All rights reserved.
`
`

`

`Seidman at al
`
`Table 1. Paclrtaxel Dose Reductions
`
`Startung Dose
`lmg/m‘l
`175 over 3 hours
`80 over 1 hour
`
`Dose Reduction (mg/m2)
`Level 2
`Level 3
`125
`105
`50
`
`50
`
`Nonhcnmtologic toxicity. Dosing guidelines for nonhematologic toxic-
`ity are described in detail in the Appendix (online only).
`Statistical Methods
`The primary study end point was tumor response. Secondary end points
`were overall survival (05). time to disease progression (11?), and treatment-
`related toxicity. 05 was measured from date of study entry until date of death
`resulting from any cause. Surviving patients were censored at the date they
`were last known to be alive. TTP was measured from date ofstudy entry until
`date of first disease progression in any site or death resulting from any cause,
`whichever occurred first. Surviving patients without disease progression were
`censored at the date last known to be progression-free.
`Power was based on the two primary study objectives: ( l) to determine
`whether weekly paclitaxel results in a significantly higher response rate (RR)
`than 3‘weckly paclitaxel, regardless of HER-2 status and assignment to trastu-
`zumah, and (2) to determine whether trastuzumab significantly increases RR
`among HER—2 nonoverexpressors, regardless of paclitaxel schedule. Regard-
`ing Peretz et al, ' 95% power to test a 50% increase in response incidence from
`25% on standard paclitaxel to 37.5% on weekly paclitaxel required 700 pa-
`tients randomly assigned equally to standard or weekly schedule. Regarding
`Holmes et al,z 85% power to detect a 50% increase in response incidence from
`25% to 37.5% with trastuzumab required 490 patients randomly assigned to
`receive or not receive trastuzumab. Both calculations assumed a two-sided a of
`.05. Because protocol therapy for the standard arm of CALGB 9342, a precur-
`sor to the current study, was identical to the standard arm of the current study,
`the former group of patients was included in the current study analysis to
`conserve patient resources, reducing accrual from 700 to 580. Inclusion ofthe
`CALGB 9342 patients necessitated a 60:40 weighted random assignment of
`
`treatment assignment (weekly:3-weekly). The CALGB Statistical Center per-
`formed all random assignment using a permuted block scheme.
`
`Data Analyses
`Primary analyses were performed separately using both the CALGB
`9840 + CALGB 9342 sample (combined) and the CALGB 9840 sample (lim-
`itedl. The primary analysis used multivariate logistic regression to relate treat-
`ment schedule with response. Secondary analyses used proportional hazards
`regression and Wald X: tests to model and assess the relationship of treatment
`with OS and TIP. Multivariate models for each end point were built using
`variables of known prognostic importance in MBC: number of metastatic
`sites, ER status, performance status, prior adjuvant chemotherapy, and prior
`radiotherapy. Also included were line of therapy, HER-2 status, and trastu-
`zumab use. Two or more proportions were compared using contingency table
`analysis; their 95% CIs used exact binomial methods. 08 and TY? distribu-
`tions were plotted using the Kaplan-Meier method. Estimates of treatment
`effect and their corresponding significance levels were derived using multivar-
`iate models that adjust for prognostic variables were labeled “adjusted"; those
`derived from univariate models were labeled “unadjusted.” All P values are
`two-sided. Because the study was not powered to address therapeutic effect
`within subsets of patients, comparisons within HER-2 subsets are exploratory
`only, P values are provided as descriptive measures only.
`CALGB 9840 was monitored biannually by an independent data safety
`monitoring board beginning within 6 months of activation and continuing
`until November 2003. Formal interim analyses on tumor response used two
`sided bounds constructed from the O’Brien-Fleming approach” and the
`Lan-DeMets” spending function. As part of CALGB’s quality-assurance pro—
`gram, study data were reviewed by the study chair and randomly selected
`patient charts were audited on site at least once every 3 years. CALGB study
`statisticians performed statistical analyses using SAS 9.0 (SAS Institute. Cary.
`NC) on data extracted from the CALGB database in February 2006.
`
`
`Patient Characteristics
`
`Table 2 presents patient characteristics in the combined sample
`
`Table 2. Patient Characteristics
`
`CALGB 9342
`l3-Weekly)
`
`3-Weeklv
`
`CALGB 9840
`
`3-Weekly +
`Trastuzumab
`
`Weekly +
`Trastuzumab
`
`Characteristic
`Total patlents treated
`No.
`%
`Study stratlflers, %
`2nd line of therapy
`HER-2 nonoverexpressor
`Demographics. %
`Age < 50 years
`Premenopausal
`AfriCan Amerlcan race
`Clmicopathologlc characteristlcs. %
`ER negatlve
`PgFl negative
`Performance score of O
`1 measurable Involved Slte at study entry
`Prlor hormone therapy
`Prior chemotherapy
`Prior radiotherapy
`
`
`
`158
`100
`
`75
`0
`
`32
`13
`25
`
`39
`44
`42
`54
`62
`82
`56
`
`104
`100
`
`20
`47
`
`23
`15
`13
`
`38
`46
`52
`63
`51
`57
`54
`
`20
`36
`
`31
`16
`19
`
`37
`42
`45
`60
`57
`65
`56
`
`123
`100
`
`14
`38
`
`42
`26
`15
`
`44
`56
`50
`50
`48
`54
`43
`
`168
`100
`
`12
`39
`
`31
`24
`15
`
`44
`52
`43
`51
`49
`54
`49
`
`Abbrevratlons: CALGB, Cancer and Leukemia Group B; 3<weekly, every 3 weeks; HER-2. human epldermal growth factor receptor 2; ER, estrogen receptor; PgR,
`progesterone receptor
`
`
`1644 © 2008 by American Socratv of Clinical Oncology
`
`JOURNAL or CLINICAL ONCOLOGY
`
`Downloaded from aseopubsorg by 38.98774 on November 19. 2017 from 038.098.007.074
`Copyright it? 2017 American Society ofClinical Oncology. All rights reserved.
`
`

`

`Weekly v Every-3«Weeks Paclitaxel for Breast Cancer
`
`by stratification factors (HER-2 status and line of therapy), demo—
`graphics (age, race), and pretreatment clinical characteristics.
`
`>
`
`
`
`Efficacy
`Primary end point and tumor response. Table 3 summarizes the
`RR for the combined and limited samples. For the combined sample,
`the RR ofweekly paclitaxel is 42% versus 29% for 3-weekly paclitaxel,
`with an unadjusted odds ratio (OR) of 1.75 (P = .0004). In the limited
`sample, trastuzumab in HER—2wnonoverexpressing tumors did not
`significantly improve RR (38% v 32%; P = .28).
`'ITP. Figure 2A shows TI‘P by paclitaxel schedule (combined
`sample) and by trastuzumab use in HER«2 nonoverexpressors (lim-
`ited sample). An early and persistent advantage for weekly paclitaxel
`over standard paclitaxel was observed. Median TFP for patients re—
`ceiving weekly paclitaxel was prolonged by 4 months {9 v 5 months;
`adjusted hazard ratio [HR] = 1.43; P < .0001). Because patients
`enrolled onto CALGB 9342 were more likely to have been in the
`second-line setting compared with CALGB 9840, we adjusted for line
`of therapy. Treatment outcomes were similar in the two studies. Fig—
`ures 3A and 3B demonstrate the comparability of the CALGB 9342
`and CALGB 9840 populations when adjusting for line oftherapy. The
`addition of trastuzumab to paclitaxel in patients with HER-Z—normal
`breast cancer was not associated with significantly longer TTP (7 v 6
`months; P = .28; Figure 28).
`Overall survival. Appendix Figure AEA {online only) shows
`overall survival (OS) by paclitaxel schedule in the combined sample.
`After adjusting for line of therapy, HER-2 status, trastuzumab, tu-
`moral estrogen—receptor status, and performance score, the HR of
`3—weeldy to weekly paclitaxel was 1.28 (95% CI, 1.06 to 1.54; P =
`.0092); in the limited sample, the HR was 1.17 {95% CI, 0.95 to 1.44;
`P : .14). The addition of trastuzumab in HER-2 nonoverexpressors
`did not have a significant impact on 08 (Fig A 1 B, online only).
`Toxicity. Adverse event data are presented for the limited sam-
`ple only. Hematologic toxicity was generally mild, and is summarized
`in Tables 4 and 5. Although grade 3 or worse granulocytopenia was
`more frequent with standard versus weekly paclitaxel (15% V 9%;
`P = .017), febrile neutropenia requiring hospitalization was infre-
`quent with either schedule (4% v 3%). Trastuzumab did not contrib—
`ute to hematologic toxicity.
`Nonhematologic toxicities occurring with at least 5% incidence
`are summarized in Tables 6 and 7. Grade 2 and 3 sensory neuropathy
`was encountered in 21% and 24% of patients receiving weekly pacli-
`
`
`
`
`
`ProportionAlive81Progression-Free
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0
`
`
`
`Adj HR = 1.43
`P .0001
`
`l
`
`2
`
`3
`
`4
`
`5
`
`6
`
`"‘0‘-
`
`Time From Study Entry (years)
`3W N-—385
`Events = 360
`Median a 0.43
`1W N—350
`Events _. 300
`Median : 0.72
`
`
`
`
`
`
`
`ProportionAlive81ProgressionvFree
`
`
`Time From Study Entry (years)
`
`Median = 0.46
`--'-— NOT N=115
`Events = 99
`
`
`
`T
`N—‘l13
`Events = 98
`Median = 0.54
`
`
`
`
`Fig 2.
`(A) Time to progressron by pacntaxel schedule (combined sample)
`Adjusted hazard ratio
`1 43, P-
`0001 [B] Time to progressron by trastuzumab
`use in HER 2 nonoverexpressors (limited samp'el. 1W. weekly, 3W, every 3
`weeks; HER 2. human epidermal growth factor receptor 2, T, trastuzumab
`
`taxel versus 21% and 12% receiving standard paclitaxel, respectively
`(comparison of grades 2 and worse; P = .0046). Grade 2 and 3 motor
`neuropathy was noted in 8% and 9% of weekly versus 5% and 4% of
`
`Patient Populatron
`All patients (combined)
`
`All patients (limited)
`
`HERv2 negative (limited)
`
`HER-2 negative (limited)
`
`HER-2 positive (limited)
`
`Table 3. Tumor Response
`Response 1%)
`No of Patients
`Comparison
`95% Ct for Response
`25 to 34
`29
`383
`3~weekly
`37 to 47
`42
`346
`Weekly
`28 to 41
`35
`225
`3 weekly
`37 to 47
`42
`346
`Weekly
`16 to 34
`24
`94
`3~weekly
`34 to 51
`42
`132
`Weekly
`32
`1 14
`No trastuzumab
`23 to 41
`29 to 4B
`38
`1 12
`Trastuzumab
`46 to 69
`58
`76
`3-weekly
`45 to 65
`55
`98
`Weekly
`Abbrevrations. OR, odds ratio; 3-weekly, every 3 weeks; HER-2, human epidermal growth factor receptor 2.
`'Unadjusted OR; the ratio of the odds of tumor response in the second group to the first group (eg, weekly v 3-weekly)
`
`
`
`OFl'
`1 75
`
`1 36
`
`2 28
`
`l 35
`
`O 89
`
`95% Cl for OH
`1.28 to 2 37
`
`Unadjusted A} P
`0004
`
`0 96 to 1 93
`
`1.2710 4.08
`
`0.78 to 2 34
`
`049t0163
`
`083
`
`0053
`
`28
`
`7'l
`
`wwnzjcrl.org
`
`e 2008 by American Socrety of Clinica- Oncology
`
`1645
`
`Downloaded from ascopubsorg by 33.98.7714 on November 19. 2017 from 038.098.007.074
`Copyright (‘3 2017 American Society of Clinical Oncology. All rights reserved.
`
`

`

`Seidman at al
`
`3,
`
`Table 4. Grade 34 Hematologic Tommy by Paclitaxel Dosrng
`Schedule (n = 572)
`
`Measure
`
`Treatment Arm
`
`Toxicity Grade
`4 (life
`threatening)
`%
`
`3 (severe)
`59
`No
`
`
`
`WBC
`
`Platelets
`
`Hemoglobin
`
`3—weekly
`Weekly
`3—weekly
`Weekly
`3-weekly
`Weekly
`3-weekly
`Weekly
`3-weekly
`Weekly
`Abbrewation: 3-weekly, every 3 weeks.
`
`Granulocytes/bands
`
`Lymphocytes
`
`17
`21
`4
`3
`6
`17
`22
`19
`19
`
`mmmomwammm
`
`.4
`
`/\Abwm—AOAoM-I
`
`"n—u-
`—-—-
`
`Time From Study Entry (years)
`9342, HER unknown N - 40 Events - 40 Median = 0.35
`9840, HER unknown N = 42 Events 2 40 Median = 0.48
`9840, HER-neg
`N = 79 Events = 70 Median = 0.35
`
`
`
`ProportionAlive8iProgression-Free
`
`
`
`ProportionAlive8iProgression-Free
`
`Events = 14 Median = 0.5
`
`2
`
`3
`
`4
`
`Time From Study Entry (years)
`9342, HER unknown N - 118 Events -113 Median = 0.32
`9340. HER unknown N x 13
`Events = 13 Median a: 0.37
`9840, HER»neg
`N - 17
`
`Fig 3. Time to progression by study. (A) First-line patients With unknown HER-2
`status receivmg 3-weekly paclitaxel.
`(B) Second-line patients With unknown
`HER»2 status receiving 3-weekly paclitaxel. HER-2. human epidermal growth
`factor receptor 2, 3-weekly, every 3 weeks.
`
`conventionally dosed patients, resepctively (comparison of grades 2
`and worse; P = .013). The incidence of neurosensory toxicity with
`weekly paclitaxel is inflated as a result of the excess neuropathy en-
`countered in the first 116 patients who received 100 mg/m2 dosing for
`the first six infusions; for these patients, the incidence of grade 3
`neuropathy was 30%, compared with 21% for the subsequent 232
`patients who received constant dosing of paclitaxel at 80 mg/m2.
`Slightly more patients receiving 3-weekly paclitaxel experienced grade
`3 or worse myalgia and arthralgia, whereas slightly more weekly pac-
`litaxel patients experienced grade 3 or worse dyspnea. Other grade 3
`and 4 nonhematologic toxicities were rare, including serious hyper—
`sensitivity reactions, as a result of the all-parenteral premedication
`regimen employed. '7
`The use of trastuzumab was associated with a 2.7% incidence of
`
`National Cancer Institute Common Toxicity Criteria grade 3 cardiac
`dysfunction, versus 0% among patients not receiving trastuzumab.
`Clinically significant cardiac events prompting serious adverse event
`reporting and hospitalization occurred in four patients receiving tras—
`
`1546
`
`© 2008 by American Socraty of Clinical Oncology
`
`tuzumab and in one patient not receiving trastuzumab; there were no
`deaths attributable to cardiac toxicity.
`Two treatment-related deaths occurred. attributable to pneumo-
`nia, in patients randomly assigned to weekly paclitaxel alone. Two
`secondary malignancies occurred, both renal cell carcinomas, one in
`each paclitaxel schedule, both without trastuzumab.
`
`_.__L:..
` -
`:J-7ma\fino-;a_'-E_Lwi)—'rmi;_filulanl__hr}mui
`-_ _-._'_.e=___.-'_ .
`
`:
`
`Weekly paclitaxel was superior to 3-weekly paclitaxel as treatment of
`MBC in R and TTP. Importantly, this study also demonstrated a lack
`of therapeutic effect for trastuzumab in HER-Z—nonoverexpressing
`breast cancer.
`
`Weekly paclitaxel improved RR over standard paclitaxel (42%
`v 29%), and nearly doubled T'l'P, from 5 to 9 months. This im—
`proved efficacy was accompanied by increased neurotoxicity, but
`did not influence overall quality-of—life scores (assessed prospec-
`tively and reported previously). '“ Strategies to prevent cumulative
`neurotoxicity with weekly paclitaxel administration are needed
`
`Table 5. Grade 3—4 Hematologic TOXlClN by Trastuzumab Use in = 572l
`ToxiCIty Grade
`4 (life
`threatening)
`"/1:
`
`Measure
`
`WBC
`
`Platelets
`
`Hemoglobin
`
`Granulocytes/bands
`
`Lymphocytes
`
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`
`
`ommuwmommco
`
`3
`
`3 (severe)
`BE
`No
`23
`15
`6
`‘I
`15
`8
`19
`22
`43
`
`.4...
`
`JUL’RNAt. OF CLINICAL ONCOLOGY
`
`Downloaded from ascopubsorg by 38.98774 on November 19. 2017 from 038.098.007.074
`Copyright a.“ 2017 Amencan Society ofClinical Oncology. All rights reserved.
`
`

`

`Weekly v Every-3-Weeks Paclitaxel for Breast Cancer
`
`it uses historical controls in addition to concurrently randomized
`controls. As a result, we were able to conduct the trial with fewer
`patients and less expense, and in less time. A disadvantage is the
`possibility that patients in the concurrent and historical settings could
`have fundamentally different characteristics. As part of “borrowing”
`from historical controls, we demonstrated that that they had out~
`comes similar to those in concurrent controls, once we adjusted for
`line of therapy.
`When trastuzumab was approved for the treatment of HER-2—
`positive MBC, we faced a critical choice that ultimately made the study
`more complex. We could either exclude HER-Z—positive patients
`from the trial or treat them with trastuzumab in addition to paclitaxel,
`recognizing that they would likely have better outcomes than the
`previous patients with HER-2—positive disease who were treated with
`paclitaxel alone, generally without knowledge of the HER—2 status. We
`chose to include trastuzumab for patients with HER-Z—positive tu-
`mors. At the same time, we took advantage ofthe opportunity to assess
`the benefit oftrastuzumab in tumors assessed to be HER-2 normal. In
`
`effect, this decision created a subtrial with a factorial design. There was
`the possibility of an interaction between trastuzumab and paclitaxel
`schedule in the patients with HER-Z—normal tumors. Without in—
`creasing the sample size. we also had limited ability to assess whether
`the benefit ofweekly paclitaxel was in HER—Z—positive tumors than in
`HER—2—normal or HER-Z— unknown tumors. However, in the inter—
`est of obtaining timely answers, we deliberately did not increase the
`sample size and accepted our limited ability to look for interactions or
`address subsets.
`
`All ofthe historical controls [3-weekly paclitaxcl) and the first 55
`concurrent controls had unknown HER~2 status. These patients also
`differed from the remainder of the population in terms ofthe proper
`tion of patients treated in the first— versus the second-line setting (25%
`and 76%, respectively). After adjusting for line of therapy, RR,
`TTP, and OS were similar among these two groups of patients. In
`addition, after adjusting for other relevant covariates in multivar-
`iate analysis, there were no major differences in our conclusions
`whether we used the combined sample or just the patients ran—
`domly assigned on this study.
`Other studies have demonstrated similar results. Weekly pac-
`litaxel yielded more pathologic complete responses in the neo—
`adjuvant setting compared with 3-weekly scheduling.“ In the
`adjuvant setting (ECOG 1199), weekly paclitaxel
`improved
`disease—free survival over 3—weekly {HR = 1.27; 95% CI, 1.07 to
`1.51; P = .006) after four cycles ofdoxorubicin and cyclophosph-
`amide," although this was not a planned, protocol-specified anal-
`ysis. Recently, the Anglo~Celtic 1V trial comparing weekly with
`3-weekly paclitaxel reported 42% and 27% RRs (P = .002), respec-
`tively, in 560 randomly assigned patients with advanced disease}5
`As opposed to CALGB 9840, where treatment lasted until disease
`progression, in the Anglo-Celtic IV trial, treatment lasted for 6
`cycles ( 18 weeks) in the standard paclitaxel arm and 12 weeks in the
`weekly arm. It is possible that this asymmetry explains the lack of
`advantage in TTP observed in the latter trial, despite the higher RR
`for weekly paclitaxel.
`Our study showed no benefit for the addition of trastuzumab in
`patients whose tumors lacked HER—2 overexpression or gene amplifi-
`cation. Although this result was expected, we know of no other pro-
`spective demonstration of this observation. Further, this observation
`
`© 2008 by American Societv of Cllnlcal Oncology
`
`1647
`
`Table 6. Grade 34 Nonhematologic Tuxrcnty by Paciltaxel Dosmg
`Schedule in = 5721
`
`Toxmrty Grade
`
`'There was one lethal Infection
`LON—'OOO—‘ONOQWOONC
`
`Tommy
`Infection
`
`Diarrhea
`
`Dvspnea
`
`Edema
`
`Neurosensory
`
`Neuromotor
`
`Malaise/fatigue
`
`Hyperglycemia
`
`Treatment
`3-weekly
`Weekly“
`3Aweekly
`Weekly
`3~weekly
`Weekly
`3-weekly
`Weekly
`3-weekly
`Weekly
`3rweekly
`Weekly
`3-weekly
`Weekly
`3-weeka
`Weekly
`
`Abbrewation. 3 weekly. every 3 weeks
`
`No
`10
`16
`6
`1 6
`7
`18
`2
`18
`27
`84
`9
`30
`l 1
`20
`15
`
`3 lseverel
`32
`
`4 (life
`threatening)
`No
`%
`
`-‘ON'-‘OCD-'O
`—-—-—-ooc:I—-D
`
`r'-.
`
`,H.
`
`MU’l—JUIQU'IMC'l4P
`aqmmwb
`
`NA
`
`and might include the “intermittent weekly” approach used in
`other studies (ie, Eastern Cooperative Oncology Group [ECOG]
`protocol 2100).”
`Five previously reported randomized trials (l 3' 2['22) have used
`paclitaxel at 175 mg/m2 via 3—hour infusion 3—weekly in more than
`1,000 MBC patients. The primary analysis in this trial is unique in that
`
`Table 7. Grade 3-4 Nonhematologic Toxicny by Trastuzumab Use in - 572)
`Toxccnty Grade
`4 (life
`threatening)
`No
`%
`
`‘There was one lethal Infection.
`Ao—too—ao—-coo¢cmmoo—-o
`
`A
`
`12O065D D(
`
`J0l
`
`3201 001 0
`
`Trastuzumab
`No trastuzumab‘
`Trastuzurnab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzurnab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`No trastuzumab
`Trastuzumab
`
`3 (severel
`5?
`No
`‘11
`1 5
`8
`14
`8
`
`Nm-AONOODMU‘WU'IA
`
`NN
`
`mmmmxi
`
`Tommy
`infection
`
`Diarrhea
`
`Dyspnea
`
`Cardiac function
`
`Other Heart
`
`Edema
`
`Neurosensory
`
`Neuromotor
`
`Malaise/fatigue
`
`wmvjm. org
`
`Downloaded from ascopubsnrg by 38.98774 on November 19, 2017 from 038.098.007.074
`Copyright © 20l7 American Society ofClinical Oncology. All rights reserved.
`
`

`

`Seidman at al
`
`addresses the concern that substantial numbers of patients with HER-
`2—dependent breast cancers might have been mislabeled as “negative”
`for this receptor. In our study, HER-2 assessment was performed
`locally. Patients with either IHC 3+ or IHC 2+ and FISH-amplified
`tumors were considered HER-2 positive and assigned to trastuzumab;
`all others were considered HER-2 normal and randomly assigned to
`trastuzumab versus no trastuzumab. Our findings provide a counter-
`balance to the recently reported results from National Surgical Adju-
`vant Breast and Bowel Project (NSABP) protocol B-31 suggesting an
`apparent benefit for adjuvant trastuzumab in patients whose tumors
`tested negative at a central laboratory by both irnmunohistochemistry
`and FISH.” Weekly paclitaxel served as a foundation for the North
`Central Cancer Treatment Group (NCCTG)/US Intergroup trial
`N983] examining the role of adjuvant trastuzumab}6 Our study con-
`firms the appropriateness ofthis approach.
`Weekly therapy may be preferable for other taxanes as well as for
`paclitaxel. A recently reported randomized trial comparing weekly
`nanoparticle albumin-bound paclitaxel with 3-weekly dosing demon-
`strated a higher RR and longer "IT? in favor of weekly dosing.27 For
`paclitaxel, our study establishes the appropriateness of basing future
`studies, and standard practice, on weekly administration. Compari-
`sons with other schedules, such as every-Z-weeks full-dose (dose-
`dense) paclitaxel, as well as with newer taxanes and formulations,
`are appropriate.
`
`
`
`Although all authors completed the disclosure declaration, thefollowing
`author(s) indicated a financial or other interest that is relevant to the subject
`matter under consideration in this article. Certain relationships marked
`with a ”U" are those for which no compensation was received; those
`relationships marked with a “C" were compensate