The new england
`journal of medicine
`
`established in 1812
`
`november 8, 2012
`
`vol. 367 no. 19
`
`Trastuzumab Emtansine for HER2-Positive Advanced
`Breast Cancer
`Sunil Verma, M.D., David Miles, M.D., Luca Gianni, M.D., Ian E. Krop, M.D., Ph.D., Manfred Welslau, M.D.,
`José Baselga, M.D., Ph.D., Mark Pegram, M.D., Do-Youn Oh, M.D., Ph.D., Véronique Diéras, M.D.,
`Ellie Guardino, M.D., Ph.D., Liang Fang, Ph.D., Michael W. Lu, Pharm.D., Steven Olsen, M.D., Ph.D.,
`and Kim Blackwell, M.D., for the EMILIA Study Group
`
`ABS TR ACT
`
`Background
`Trastuzumab emtansine (T-DM1) is an antibody–drug conjugate incorporating the
`human epidermal growth factor receptor 2 (HER2)–targeted antitumor properties of
`tras tuz u mab with the cytotoxic activity of the microtubule-inhibitory agent DM1. The
`antibody and the cytotoxic agent are conjugated by means of a stable linker.
`Methods
`We randomly assigned patients with HER2-positive advanced breast cancer, who had
`previously been treated with tras tuz u mab and a taxane, to T-DM1 or la pa ti nib plus
`cap e ci ta bine. The primary end points were progression-free survival (as assessed by
`independent review), overall survival, and safety. Secondary end points included
`progression-free survival (investigator-assessed), the objective response rate, and
`the time to symptom progression. Two interim analyses of overall survival were
`conducted.
`Results
`Among 991 randomly assigned patients, median progression-free survival as assessed
`by independent review was 9.6 months with T-DM1 versus 6.4 months with la pa ti nib
`plus cap e ci ta bine (hazard ratio for progression or death from any cause, 0.65; 95%
`confidence interval [CI], 0.55 to 0.77; P<0.001), and median overall survival at the
`second interim analysis crossed the stopping boundary for efficacy (30.9 months
`vs. 25.1 months; hazard ratio for death from any cause, 0.68; 95% CI, 0.55 to 0.85;
`P<0.001). The objective response rate was higher with T-DM1 (43.6%, vs. 30.8% with
`la pa ti nib plus cap e ci ta bine; P<0.001); results for all additional secondary end points
`favored T-DM1. Rates of adverse events of grade 3 or above were higher with la pa-
`ti nib plus cap e ci ta bine than with T-DM1 (57% vs. 41%). The incidences of throm-
`bocytopenia and increased serum aminotransferase levels were higher with T-DM1,
`whereas the incidences of diarrhea, nausea, vomiting, and palmar–plantar erythro-
`dysesthesia were higher with la pa ti nib plus cap e ci ta bine.
`Conclusions
`T-DM1 significantly prolonged progression-free and overall survival with less toxicity
`than la pa ti nib plus cap e ci ta bine in patients with HER2-positive advanced breast
`cancer previously treated with tras tuz u mab and a taxane. (Funded by F. Hoffmann–
`La Roche/Genentech; EMILIA ClinicalTrials.gov number, NCT00829166.)
`
`From Sunnybrook Odette Cancer Centre,
`Toronto (S.V.); Mount Vernon Cancer Cen-
`tre, Northwood, United Kingdom (D.M.);
`San Raffaele Hospital, Milan (L.G.); Dana–
`Farber Cancer Institute (I.E.K.) and Massa-
`chusetts General Hospital (J.B.) — both
`in Boston; Medical Office Hematology,
`Aschaffenburg, Germany (M.W.); Univer-
`sity of Miami Sylvester Comprehensive
`Cancer Center, Miami (M.P.); Seoul Na-
`tional University College of Medicine,
`Seoul, South Korea (D.-Y.O.); Institut
`Curie, Paris (V.D.); Genentech, South
`San Francisco, CA (E.G., L.F., M.W.L.,
`S.O.); and Duke University Medical Cen-
`ter, Durham, NC (K.B.). Address reprint
`requests to Dr. Verma at Sunnybrook
`Odette Cancer Centre, 2075 Bayview
`Ave., Toronto, ON M4N 3M5, Canada, or
`at sunil.verma@sunnybrook.ca.
`
`This article was published on October 1,
`2012, and updated on June 20, 2013, at
`NEJM.org.
`
`N Engl J Med 2012;367:1783-91.
`DOI: 10.1056/NEJMoa1209124
`
`Copyright © 2012 Massachusetts Medical Society.
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`1783
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`T h e ne w e ngl a nd jou r na l o f m e dicine
`
`A mplification of human epidermal
`
`growth factor receptor 2 (HER2, also called
`ErbB2) occurs in approximately 20% of
`breast cancers and is associated with shortened
`survival.1-3 Combining HER2-targeted agents with
`standard chemotherapy is an effective therapeutic
`approach for patients with HER2-positive meta-
`static breast cancer. When combined with first-line
`chemotherapy, tras tuz u mab increases the time to
`progression and overall survival among patients
`with metastatic disease.4,5 The addition of la pa ti-
`nib to cap e ci ta bine increases the time to pro-
`gression in patients previously treated with tras-
`tuz u mab, an anthracycline, and a taxane,6 and
`this combination is a standard option for disease
`progression with tras tuz u mab.
`Tras tuz u mab emtansine (T-DM1) is an anti-
`body–drug conjugate that incorporates the HER2-
`targeted antitumor properties of tras tuz u mab with
`the cytotoxic activity of the microtubule-inhibitory
`agent DM1 (derivative of maytansine); the anti-
`body and the cytotoxic agent are conjugated by
`means of a stable linker.7,8 T-DM1 allows intra-
`cellular drug delivery specifically to HER2-over-
`expressing cells, thereby improving the therapeutic
`index and minimizing exposure of normal tissue.
`Phase 2 studies have shown the clinical activity of
`T-DM1 in patients with HER2-positive advanced
`breast cancer.9-11
`The EMILIA study, a phase 3 trial, assessed the
`efficacy and safety of T-DM1, as compared with
`la pa ti nib plus cap e ci ta bine, in patients with HER2-
`positive advanced breast cancer previously treated
`with tras tuz u mab and a taxane.
`
`Methods
`
`Study Design
`The EMILIA study is a randomized, open-label,
`international trial involving patients with HER2-
`positive, unresectable, locally advanced or meta-
`static breast cancer who were previously treated
`with tras tuz u mab and a taxane. The study was con-
`ducted in accordance with the International Con-
`ference on Harmonization Good Clinical Prac-
`tice standards and the Declaration of Helsinki.
`Patients provided written informed consent; the
`study was approved by the relevant institutional
`review board or independent ethics committee.
`Patients were randomly assigned in a 1:1 ratio
`to T-DM1 or la pa ti nib plus cap e ci ta bine with the
`use of a hierarchical, dynamic randomization
`scheme through an interactive voice-response
`
`system. Stratification factors were world region
`(United States, Western Europe, or other), the
`number of prior chemotherapy regimens for un-
`resectable, locally advanced or metastatic disease
`(0 or 1 vs. >1), and disease involvement (visceral
`vs. nonvisceral).
`The primary end points were progression-free
`survival assessed by independent review, overall
`survival, and safety. Progression-free survival
`was defined as the time from randomization to
`progression or death from any cause. Progression
`was assessed according to modified Response
`Evaluation Criteria in Solid Tumors (RECIST), ver-
`sion 1.012; the modified criteria are specified in
`the Supplementary Appendix, available with the
`full text of this article at NEJM.org. Overall sur-
`vival was defined as the time from randomization
`to death from any cause. Prespecified second-
`ary end points included progression-free sur-
`vival (investigator-assessed), the objective re-
`sponse rate, the duration of response, and the
`time to symptom progression. The objective re-
`sponse rate was determined according to modified
`RECIST on the basis of an independent review of
`patients with measurable disease at baseline; re-
`sponses were confirmed at least 28 days after the
`initial documentation of a response. The time to
`symptom progression was defined as the time
`from randomization to the first decrease of
`5 points or more from baseline scores on the Trial
`Outcome Index of the patient-reported Functional
`Assessment of Cancer Therapy–Breast (FACT-B
`TOI, on which scores range from 0 to 92, with
`higher scores indicating a better quality of life)13
`in women with a baseline score and at least one
`postbaseline score. Safety was monitored by an
`independent data monitoring committee and a
`cardiac review committee.
`
`Study Oversight
`The study was designed by the academic investi-
`gators, the trial steering committee, and represen-
`tatives of the sponsor, F. Hoffmann–La Roche/
`Genentech. The data were collected by the spon-
`sor and analyzed in collaboration with the authors,
`who vouch for the accuracy and completeness of
`the data and analysis and the fidelity of the study
`to the protocol, available at NEJM.org. The first
`author prepared the initial draft of the manuscript
`with support from a medical writer who was paid
`by Genentech. All the authors contributed to sub-
`sequent drafts and made the decision to submit
`the manuscript for publication.
`
`1784
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`trastuzumab Emtansine for Advanced Breast Cancer
`
`Patients
`Eligible patients had documented progression of
`unresectable, locally advanced or metastatic HER2-
`positive breast cancer previously treated with a tax-
`ane and tras tuz u mab. Inclusion criteria were pro-
`gression during or after the most recent treatment
`for locally advanced or metastatic disease or with-
`in 6 months after treatment for early-stage disease,
`and a centrally confirmed HER2-positive status,
`assessed by means of immunohistochemical analy-
`sis (with 3+ indicating positive status), fluores-
`cence in situ hybridization (with an amplification
`ratio ≥2.0 indicating positive status), or both. Pa-
`tients with measurable disease (according to mod-
`ified RECIST) and those with nonmeasurable dis-
`ease were included. Other eligibility criteria were
`a left ventricular ejection fraction of 50% or more
`(determined by echocardiography or multiple-gated
`acquisition [MUGA] scanning) and an Eastern Co-
`operative Oncology Group performance status of
`0 (asymptomatic) or 1 (restricted in strenuous ac-
`tivity but ambulatory and able to do light work).
`Major exclusion criteria were prior treatment
`with T-DM1, la pa ti nib, or cap e ci ta bine; peripheral
`neuropathy of grade 3 or higher (according to
`National Cancer Institute Common Terminology
`Criteria for Adverse Events [CTCAE], version 3.0)14;
`symptomatic central nervous system (CNS) me-
`tastases or treatment for these metastases with-
`in 2 months before randomization; a history of
`symptomatic congestive heart failure or serious
`cardiac arrhythmia requiring treatment; and a
`history of myocardial infarction or unstable an-
`gina within 6 months before randomization.
`
`Procedures
`Patients in the control group self-administered
`oral la pa ti nib at a dose of 1250 mg daily plus oral
`cap e ci ta bine at a dose of 1000 mg per square meter
`of body-surface area every 12 hours (maximum
`planned daily dose, 2000 mg per square meter)
`on days 1 through 14 of each 21-day treatment
`cycle and recorded their doses in a patient diary.
`Dose delays, reductions, and discontinuations ow-
`ing to toxic effects were defined in the protocol.
`For cap e ci ta bine, the first dose reduction was to
`75% of the total daily dose, and the second to
`50% of that dose (see Table 1 in the Supplemen-
`tary Appendix). For la pa ti nib, the first dose re-
`duction was to 1000 mg daily, and the second to
`750 mg daily. Patients could continue to take la pa-
`ti nib if cap e ci ta bine was discontinued and vice
`versa. If treatment with both drugs was delayed
`
`for more than 42 consecutive days, the drugs were
`discontinued.
`Patients randomly assigned to T-DM1 received
`3.6 mg per kilogram of body weight intravenous-
`ly every 21 days. Dose delays, reductions, and dis-
`continuations owing to toxic effects were defined
`in the protocol. The first dose reduction was to
`3.0 mg per kilogram and the second to 2.4 mg
`per kilogram (Table 1 in the Supplementary Ap-
`pendix). Dose escalation was not allowed after a
`dose reduction. If a toxic event did not resolve to
`a grade 1 level or to baseline status within 42 days
`after the most recent dose, the study treatment
`was discontinued. Patients continued to receive the
`study treatment until disease progression (inves-
`tigator-assessed) or the development of unmanage-
`able toxic effects.
`
`Assessments
`Tumor assessments were performed by the study
`investigators and by the independent review com-
`mittee at baseline and every 6 weeks thereafter
`until investigator-assessed disease progression; an
`additional assessment was required 6 weeks after
`progression. The left ventricular ejection fraction
`was measured by means of echocardiography (the
`preferred method) or MUGA scanning at baseline,
`week 6, week 12, and every 12 weeks thereafter
`until discontinuation of the study treatment; an
`additional assessment was performed 30 days af-
`ter the last dose of the study drug. Local labora-
`tory assessments were performed at baseline, on
`day 1 of each treatment cycle, on days 8 and 15 of
`cycles 1 through 4, and 30 days after the last dose
`of the study drug. Adverse events were monitored
`continuously and graded according to the CTCAE,
`version 3.0.
`
`Statistical Analysis
`The trial was originally designed with progression-
`free survival, as assessed by independent review,
`as the primary efficacy end point, with a planned
`sample of 580 patients. In October 2010, with all
`data still masked to the investigators, the proto-
`col was amended to add overall survival as a copri-
`mary efficacy end point, with an increase in the
`planned sample to 980 patients. The trial had
`90% power to detect a hazard ratio of 0.75 for
`progression or death from any cause with T-DM1
`as compared with la pa ti nib plus cap e ci ta bine and
`80% power to detect a hazard ratio of 0.80 for
`death from any cause, with a two-sided alpha
`level of 0.05.
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`1785
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`T h e ne w e ngl a nd jou r na l o f m e dicine
`
`The primary analysis of progression-free sur-
`vival was to be performed after 508 indepen-
`dently assessed events, and the final analysis of
`overall survival after 632 deaths. The first in-
`terim analysis of overall survival was to be per-
`formed at the time of the primary analysis of
`progression-free survival. A second interim anal-
`ysis was added to the statistical analysis plan
`after the completion of the first interim analysis
`and was conducted when 50% of the targeted
`events had occurred. Stopping boundaries for ef-
`ficacy were determined by means of the Lan–
`DeMets alpha-spending function with an O’Brien–
`Fleming boundary and the actual number of
`observed deaths. To adjust for multiple compari-
`sons, a fixed-sequence hypothesis-testing proce-
`dure was implemented. The hypothesis test for
`progression-free survival was conducted at a two-
`sided alpha level of 0.05. If the result was statisti-
`cally significant, overall survival was to be tested
`at a two-sided alpha level of 0.05, which was
`spent at the interim and final analyses according
`to the Lan–DeMets spending function. If both
`primary end points were statistically significant,
`the secondary end points were to be tested in a
`prespecified order. The statistical analysis plan is
`included in the protocol.
`The primary end points were assessed in the
`intention-to-treat population and tested by means
`of two-sided log-rank tests, with stratification ac-
`cording to the factors used for randomization. A
`sensitivity analysis was performed for progression-
`free survival among patients who received nonpro-
`tocol breast-cancer treatment before documented
`disease progression, with censoring of data at the
`last tumor assessment before the initiation of
`such therapy. Kaplan–Meier methods were used
`to estimate medians for the primary end points,
`1- and 2-year survival rates, and corresponding
`95% confidence intervals. Analyses of progression-
`free survival in 16 prespecified subgroups were
`performed. Ten post hoc analyses to assess po-
`tential effects of prior therapy were performed.
`All post hoc analyses had similar results, and
`the results of one representative analysis (line of
`therapy) are therefore reported here. We used a
`Cox proportional-hazards model, with the same
`stratification factors as those used for random-
`ization, to estimate hazard ratios and 95% con-
`fidence intervals for the primary efficacy end
`points and for subgroup analyses.
`
`Investigator-assessed progression-free survival
`and the time to symptom progression were ana-
`lyzed with the same methods as those used for
`independent review. The objective response rate
`was compared between groups with the use of
`the Mantel–Haenszel chi-square test, with strat-
`ification according to the factors used for random-
`ization. For patients with an objective response,
`the median duration of the response was estimated
`with the use of the Kaplan–Meier approach.
`
`R esults
`
`Study Population
`From February 2009 through October 2011, a to-
`tal of 991 patients were enrolled at 213 centers in
`26 countries; 496 patients were assigned to la pa-
`ti nib plus cap e ci ta bine, and 495 were assigned to
`T-DM1 (Fig. 1 in the Supplementary Appendix).
`Baseline demographic and disease characteristics
`were similar in the two groups (Table 1; see Ta-
`ble 2 in the Supplementary Appendix for addition-
`al baseline information). The first data-cutoff date
`of January 14, 2012 (median duration of follow-
`up, approximately 13 months), was used for all
`analyses in this report except the second interim
`analysis of overall survival, which had a data-
`cutoff date of July 31, 2012 (median duration of
`follow-up, approximately 19 months).
`
`Primary Analysis
`Treatment with T-DM1 significantly improved pro-
`gression-free survival as assessed by independent
`review (median survival, 9.6 months, vs. 6.4 months
`with la pa ti nib plus cap e ci ta bine; stratified haz-
`ard ratio for progression or death from any cause,
`0.65; 95% confidence interval [CI], 0.55 to 0.77;
`P<0.001) (Fig. 1) and in the sensitivity analysis
`with censoring for nonprotocol therapy (Table 3
`in the Supplementary Appendix). This benefit was
`consistently observed across clinically relevant
`subgroups, with a less definitive benefit among
`patients 75 years of age or older and those with
`nonvisceral or nonmeasurable disease (Fig. 2 in
`the Supplementary Appendix).
`At the first interim analysis of overall survival
`(223 deaths), the stratified hazard ratio for death
`from any cause with T-DM1 versus la pa ti nib plus
`cap e ci ta bine was 0.62 (95% CI, 0.48 to 0.81;
`P = 0.0005) and did not cross the predefined
`O’Brien–Fleming stopping boundary (P = 0.0003).
`
`1786
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`trastuzumab Emtansine for Advanced Breast Cancer
`
`Table 1. Selected Demographic and Baseline Characteristics of the Patients.*
`
`Characteristic
`
`Age — yr
`
`Median
`
`Range
`
`Race — no. (%)†
`
`White
`
`Asian
`
`Black
`
`Other
`
`Not available
`
`World region — no. (%)
`
`United States
`
`Western Europe
`
`Asia
`
`Other
`
`ECOG performance status — no. (%)‡
`
`0
`
`1
`
`Not available
`
`Site of disease involvement — no. (%)
`
`Visceral
`
`Nonvisceral
`
`Hormone-receptor status — no. (%)
`
`ER-positive, PR-positive, or both
`
`ER-negative and PR-negative
`
`Unknown
`
`Prior systemic therapy — no. (%)§
`
`Anthracycline
`
`Other chemotherapy
`
`Biologic agent other than trastuzumab or pertuzumab
`
`Endocrine therapy
`
`Prior chemotherapy regimens for locally advanced or metastatic
`disease — no. (%)
`
`0 or 1
`
`>1
`
`Prior trastuzumab treatment — no. (%)§
`
`For metastatic breast cancer, early breast cancer, or both
`
`For early breast cancer only
`
`Lapatinib plus Capecitabine
`(N = 496)
`
`T-DM1
`(N = 495)
`
`53
`
`24–83
`
`374 (75)
`
`86 (17)
`
`21 (4)
`
`10 (2)
`
`5 (1)
`
`136 (27)
`
`160 (32)
`
`76 (15)
`
`124 (25)
`
`312 (63)
`
`176 (35)
`
`8 (2)
`
`335 (68)
`
`161 (32)
`
`263 (53)
`
`224 (45)
`
`9 (2)
`
`302 (61)
`
`382 (77)
`
`21 (4)
`
`204 (41)
`
`305 (61)
`
`191 (39)
`
`419 (84)
`
`77 (16)
`
`53
`
`25–84
`
`358 (72)
`
`94 (19)
`
`29 (6)
`
`7 (1)
`
`7 (1)
`
`134 (27)
`
`157 (32)
`
`82 (17)
`
`122 (25)
`
`299 (60)
`
`194 (39)
`
`2 (<1)
`
`334 (67)
`
`161 (33)
`
`282 (57)
`
`202 (41)
`
`11 (2)
`
`303 (61)
`
`385 (78)
`
`13 (3)
`
`205 (41)
`
`304 (61)
`
`191 (39)
`
`417 (84)
`
`78 (16)
`
`* ER denotes estrogen receptor, PR progesterone receptor, and T-DM1 trastuzumab emtansine.
`† Race was self-reported.
`‡ An Eastern Cooperative Oncology Group (ECOG) performance status of 0 indicates that the patient is asymptomatic,
`and a status of 1 indicates that the patient is restricted in strenuous activity but ambulatory and able to do light work.
`§ The study protocol specified that previous treatment with a taxane and trastuzumab was required for enrollment.
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`1787
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`T h e ne w e ngl a nd jou r na l o f m e dicine
`
`score was delayed in the T-DM1 group (7.1
`months, vs. 4.6 months with la pa ti nib plus cap e-
`ci ta bine; hazard ratio, 0.80; 95% CI, 0.67 to 0.95;
`P = 0.012).
`
`Treatment Exposure
`More patients in the la pa ti nib–cap e ci ta bine group
`than in the T-DM1 group required a dose reduc-
`tion (la pa ti nib, 27.3% of patients; cap e ci ta bine,
`53.4%; T-DM1, 16.3%). As a result, the median
`daily dose received was 1250.0 mg per day (range,
`250.0 to 1332.3) for la pa ti nib, 1729.8 mg per square
`meter per day (range, 781.6 to 2338.4) for cap e ci-
`ta bine, and 3.5 mg per kilogram every 21 days
`(range, 2.7 to 4.0) for T-DM1. In the safety popula-
`tion, 37 of 488 patients (7.6%) discontinued treat-
`ment with la pa ti nib, 46 of 488 patients (9.4%) dis-
`continued treatment with cap e ci ta bine, and 29 of
`490 patients (5.9%) discontinued treatment with
`T-DM1 because of adverse events (Fig. 1 in the
`Supplementary Appendix).
`
`Safety
`Serious adverse events in the safety population were
`reported for 88 patients (18.0%) in the la pa ti nib–
`cap e ci ta bine group and for 76 patients (15.5%) in
`the T-DM1 group. The incidence rates of adverse
`events of grade 3 or above were higher in the la pa-
`ti nib–cap e ci ta bine group than in the T-DM1 group
`(57.0% vs. 40.8%) (Table 3). Diarrhea and palmar–
`plantar erythrodysesthesia were the most com-
`monly reported grade 3 or 4 events in the la pa ti-
`nib–cap e ci ta bine group, affecting 20.7% and 16.4%
`of patients, respectively. The most commonly re-
`ported grade 3 or 4 events with T-DM1 were
`thrombocytopenia (12.9%) and elevated serum
`concentrations of aspartate aminotransferase
`(4.3%) and alanine aminotransferase (2.9%).
`For most patients, the first occurrence of grade
`3 or 4 thrombocytopenia was reported during the
`first two cycles of T-DM1 treatment; with dose
`modifications, the majority of these patients
`were able to continue treatment (10 patients
`[2.0%] discontinued T-DM1 because of thrombo-
`cytopenia). The overall incidence of bleeding
`events was higher with T-DM1 (29.8%, vs. 15.8%
`with la pa ti nib plus cap e ci ta bine); rates of grade
`3 or 4 bleeding events were low in both groups
`(1.4% and 0.8%, respectively). The only grade 4
`bleeding event was a gastrointestinal hemorrhage
`in a patient treated with T-DM1 whose platelet
`counts were within the normal range during the
`
`No.of
`MedianNo.
`ofMonths
`Events
`Lapatinib–Capecitabine
`6.4
`304
`9.6
`265
`T-DM1
`Stratified hazard ratio, 0.65
`(95% CI, 0.55–0.77)
`P<0.001
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`12
`
`T-DM1
`
`Lapatinib–capecitabine
`
`14
`18
`16
`Months
`
`20
`
`22
`
`24
`
`26
`
`28
`
`30
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`Progression-freeSurvival(%)
`
`0 0
`
`0 1
`
`1 3
`
`5 9
`
`496
`
`404
`
`310
`
`176
`
`129
`
`73
`
`53
`
`35
`
`25
`
`14
`
`9
`
`8
`
`495
`
`419
`
`341
`
`236
`
`183
`
`130
`
`101
`
`72
`
`54
`
`44
`
`30
`
`18
`
`No.atRisk
`Lapatinib–
`capecitabine
`T-DM1
`
`Figure 1. Progression-free Survival, as Assessed by an Independent Review
`Committee.
`Shown are Kaplan–Meier estimates of progression-free survival in the
`intention-to-treat population, stratified according to world region, number
`of prior chemotherapy regimens (0 or 1 vs. >1), and site of disease involvement
`(visceral vs. nonvisceral). Median progression-free survival was 3.2 months
`longer in the tras tuz u mab emtansine (T-DM1) group than in the la pa ti nib–
`cap e ci ta bine group. CI denotes confidence interval.
`
`At the second interim analysis of overall survival
`(331 deaths), T-DM1 significantly increased me-
`dian overall survival (30.9 months, vs. 25.1 months
`with la pa ti nib plus cap e ci ta bine; hazard ratio for
`death from any cause, 0.68; 95% CI, 0.55 to 0.85;
`P<0.001) (Fig. 2). Estimated 1-year survival rates
`were 85.2% (95% CI, 82.0 to 88.5) in the T-DM1
`group and 78.4% (95% CI, 74.6 to 82.3) in the
`la pa ti nib–cap e ci ta bine group; rates at 2 years
`were 64.7% (95% CI, 59.3 to 70.2) and 51.8%
`(95% CI, 45.9 to 57.7), respectively.
`
`Prespecified Secondary Efficacy End Points
`Treatment with T-DM1 improved investigator-
`assessed progression-free survival (median, 9.4
`months with T-DM1 vs. 5.8 months with la pa ti nib
`plus cap e ci ta bine; hazard ratio for progression or
`death from any cause, 0.66; 95% CI, 0.56 to 0.77;
`P<0.001). The objective-response rate was higher
`in the T-DM1 group (43.6%; 95% CI, 38.6 to 48.6)
`than in the la pa ti nib–cap e ci ta bine group (30.8%;
`95% CI, 26.3 to 35.7; P<0.001), and the median
`duration of response was longer (12.6 months vs.
`6.5 months) (Table 2). The median time to a de-
`crease of 5 points or more in the FACT-B TOI
`
`1788
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`trastuzumab Emtansine for Advanced Breast Cancer
`
`MedianNo.
`ofMonths
`25.1
`30.9
`
`No.of
`Events
`182
`149
`
`64.7% (95% CI, 59.3–70.2)
`
`T-DM1
`
`Lapatinib–Capecitabine
`T-DM1
`Stratified hazard ratio, 0.68
`(95% CI, 0.55–0.85)
`P<0.001
`Efficacy stopping boundary,
`P=0.0037 or hazard ratio, 0.73
`
`85.2% (95% CI, 82.0–88.5)
`
`78.4% (95% CI, 74.6–82.3)
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`OverallSurvival(%)
`
`51.8% (95% CI, 45.9–57.7)
`Lapatinib–capecitabine
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`12
`
`14
`
`18
`16
`20
`Months
`
`22
`
`24
`
`26
`
`28
`
`30
`
`32
`
`34
`
`36
`
`4 5
`
`45
`
`62
`
`27
`
`38
`
`17
`
`28
`
`7
`
`13
`
`496
`
`471
`
`453
`
`435
`
`403
`
`368
`
`297
`
`240
`
`204
`
`159
`
`133
`
`110
`
`86
`
`63
`
`495
`
`485
`
`474
`
`457
`
`439
`
`418
`
`349
`
`293
`
`242
`
`197
`
`164
`
`136
`
`111
`
`86
`
`No.atRisk
`Lapatinib–
`capecitabine
`T-DM1
`
`Figure 2. Second Interim Analysis of Overall Survival.
`Shown are Kaplan–Meier estimates of overall survival in the intention-to-treat population, stratified according to
`world region, number of prior chemotherapy regimens (0 or 1 vs. >1), and site of disease involvement (visceral vs.
`nonvisceral). The second interim analysis was conducted on the basis of 331 deaths and met the predefined
`O’Brien–Fleming stopping boundary. The data-cutoff date was July 31, 2012. Median follow-up was 18.6 months
`(range, 0 to 41) in the la pa ti nib–cap e ci ta bine group and 19.1 months (range, 0 to 40) in the T-DM1 group.
`
`study treatment. Reports of hyperbilirubinemia
`of any grade were more frequent in the la pa ti nib–
`cap e ci ta bine group than in the T-DM1 group (8.2%
`vs. 1.2%). With appropriate dose modifications,
`the majority of patients with grade 3 or 4 eleva-
`tions in serum aminotransferase levels were able
`to continue treatment (3 patients discontinued
`T-DM1 because of grade 3 elevations in aspartate
`aminotransferase levels), and no patients met Hy’s
`law criteria for drug-induced liver injury.15
`In the majority of patients, a left ventricular
`ejection fraction of 45% or more was maintained
`during the study treatment (in 97.1% of patients
`in the T-DM1 group and 93.0% of patients in the
`la pa ti nib–cap e ci ta bine group). Three patients in
`each group had a decrease from baseline to less
`than 40%. Of 481 patients in the T-DM1 group
`and 445 in the la pa ti nib–cap e ci ta bine group who
`could be evaluated, 8 patients (1.7%) and 7 pa-
`tients (1.6%), respectively, had an ejection fraction
`that was less than 50% and at least 15 percentage
`points below the baseline value. To date, grade
`3 left ventricular systolic dysfunction has devel-
`oped in 1 patient in the T-DM1 group and in no
`patients in the la pa ti nib–cap e ci ta bine group.
`
`Most of the deaths that occurred during the
`study period were attributed to disease progression
`(123 deaths [96.1%] in the la pa ti nib–cap e ci ta bine
`group and 91 deaths [96.8%] in the T-DM1 group).
`Five deaths were attributed to adverse events that
`occurred within 30 days after the last dose of a
`study drug: 4 deaths in the la pa ti nib–cap e ci ta bine
`group (due to coronary artery disease, multiorgan
`failure, coma, and hydrocephalus) and 1 death
`in the T-DM1 group (due to metabolic encepha-
`lopathy after CNS progression).
`
`Discussion
`
`In this phase 3 study, the antibody–drug conjugate
`T-DM1, as compared with la pa ti nib plus cap e ci ta-
`bine, significantly improved progression-free and
`overall survival among patients with HER2-posi-
`tive metastatic breast cancer who had previously
`received tras tuz u mab and a taxane. The benefit was
`observed regardless of the line of therapy in pa-
`tients with metastatic disease and was seen in
`patients with a disease-free interval of less than
`6 months after completion of tras tuz u mab-based
`therapy in the adjuvant or neoadjuvant setting.
`
`n engl j med 367;19 nejm.org november 8, 2012
`
`1789
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org on May 20, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2012 Massachusetts Medical Society. All rights reserved.
`
`IMMUNOGEN 2121, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`
`T h e ne w e ngl a nd jou r na l o f m e dicine
`
`Table 2. Objective-Response Rate and Duration of Response, as Assessed by the Independent Review Committee.*
`
`Variable
`
`Complete or partial response
`
`No. of patients
`
`Percent (95% CI)
`
`Complete response — no. (%)
`
`Partial response — no. (%)
`
`Duration of complete or partial
` response — mo
`
`Median
`
`95% CI
`
`Lapatinib plus Capecitabine
`(N = 389)
`
`T-DM1
`(N = 397)
`
`Difference
`
`P Value
`
`120
`
`173
`
`30.8 (26.3–35.7)
`
`43.6 (38.6–48.6)
`
`12.7 (6.0–19.4)
`
`<0.001
`
`2 (0.5)
`
`118 (30.3)
`
`6.5
`
`5.5–7.2
`
`4 (1.0)
`
`169 (42.6)
`
`12.6
`
`8.4–20.8
`
`* The total number of patients in each group is the number with measurable disease at baseline, as determined by inde-
`pendent review. CI denotes confidence interval.
`
`Table 3. Adverse Events in the Safety Population.*
`
`Adverse Event
`
`Lapatinib plus Capecitabine
`(N = 488)
`
`T-DM1
`(N = 490)
`
`Any event
`
`Specific events†
`
`Diarrhea
`
`Palmar–plantar erythrodysesthesia
`
`Events of Any
`Grade
`
`Events of Grade
`3 or Above
`
`Events of Any
`Grade
`
`Events of Grade 3
`or Above
`
`number of patients (percent)
`
`477 (97.7)
`
`278 (57.0)
`
`470 (95.9)
`
`389 (79.7)
`
`283 (58.0)
`
`101 (20.7)
`
`80 (16.4)
`
`114 (23.3)
`
`6 (1.2)
`
`200 (40.8)
`
`8 (1.6)
`
`0
`
`Vomiting
`
`Neutropenia
`
`Hypokalemia
`
`Fatigue
`
`Nausea
`
`Mucosal inflammation
`
`Anemia
`
`Elevated ALT
`
`Elevated AST
`
`Thrombocytopenia
`
`143 (29.3)
`
`42 (8.6)
`
`42 (8.6)
`
`136 (27.9)
`
`218 (44.7)
`
`93 (19.1)
`
`39 (8.0)
`
`43 (8.8)
`
`46 (9.4)
`
`12 (2.5)
`
`22 (4.5)
`
`21 (4.3)
`
`20 (4.1)
`
`17 (3.5)
`
`12 (2.5)
`
`11 (2.3)
`
`8 (1.6)
`
`7 (1.4)
`
`4 (0.8)
`
`1 (0.2)
`
`93 (19.0)
`
`29 (5.9)
`
`42 (8.6)
`
`172 (35.1)
`
`192 (39.2)
`
`33 (6.7)
`
`51 (10.4)
`
`83 (16.9)
`
`110 (22.4)
`
`137 (28.0)
`
`4 (0.8)
`
`10 (2.0)
`
`11 (2.2)
`
`12 (2.4)
`
`4 (0.8)
`
`1 (0.2)
`
`13 (2.7)
`
`14 (2.9)
`
`21 (4.3)
`
`63 (12.9)
`
`* The safety population included all patients who received at least one d