Cardiac Dysfunction in the Trastuzumab Clinical Trials
`Experience
`
`By Andrew Seidman, Clifford Hudis, Mary Kathryn Pierri, Steven Shak, Virginia Paton, Mark Ashby, Maureen Murphy,
`Stanford J. Stewart, and Deborah Keefe
`
`Purgos : This study sought to estimate cardiac dys-
`function (CD) risk for patients receiving trastuzumab; to
`characterize observed CD by severity, treatment, and
`clinical outcome; to assess effects of baseline clinical
`risk factors on CD; and to assess effects of cumulative
`doses of anthracyclines and trastuzumab on CD.
`Patients and Methods: A retrospective review of
`records for patients enrolled onto any of seven phase II
`and Ill trastuzumab clinical trials was performed. Pre-
`defined criteria were used for the diagnosis, and the
`New York Heart Association functional classification
`
`system was used to document CD severity. Product-limit
`estimates were used to summarize the cumulative an-
`
`thracycline and trastuzumab doses at the time of CD
`onset.
`
`Results: Patients treated with trastuzumab were
`found to be at an increased risk for CD. The incidence
`was greatest in patients receiving concomitant trastu-
`
`zumab and anthracycline plus cyclophosphamide
`(27%). The risk was substantially lower in patients
`receiving paclitaxel and trastuzumab (13%) or trastu-
`zumab alone (3% to 7%); however, most of these pa-
`tients had received prior anthracycline therapy. CD was
`noted in 8% of patients receiving anthracycline plus
`cyclophosphamide and 1% receiving paclitaxel alone.
`Most trastuzumab-treated patients developing CD were
`symptomatic (75%), and most improved with standard
`treatment for congestive heart failure (79%).
`Conclusion: Trastuzumab is associated with an in-
`
`creased risk of CD, which is greatest in patients receiv-
`ing concurrent anthracyclines.
`In most patients with
`metastatic breast cancer, the risk of CD can be iustified
`given the improvement in overall survival previously
`reported with trastuzumab.
`J Clin Oncol 20:1215-1221.
`Society of Clinical Oncology.
`
`[0 2002 by American
`
`RASTUZUMAB (HERCEPTIN; Genentech Inc, South
`San Francisco, CA) is a humanized monoclonal anti-
`body approved by the United States Food and Drug Admin-
`istration in September 1998 for the treatment of women
`with metastatic breast cancer whose tumors overexpress
`p1 SSHERZ. HER2, also known as c-erbB-2 or neu,
`is a
`proto-oncogene that codes for plSSHERZ. This protein is a
`transmembrane receptor tyrosine kinase that is a member of
`the epidermal growth factor family. The HER2 gene is
`amplified and the p185HER2 protein overexpressed in 20% to
`25% of human breast cancers.l Patients with tumors dem-
`
`onstrating this alteration have a poor prognosisl’2
`Trastuzumab is indicated in combination with paclitaxel
`for first-line treatment and as a single agent for second- or
`third-line treatment of metastatic breast cancer. When used
`
`.
`
`in combination with chemotherapy in the first-line treatment
`of metastatic breast cancer, trastuzumab improves response
`rate, response duration, time to progression, time to treat-
`ment failure,3 and median overall survival4 compared with
`treatment with chemotherapy alone. Single-agent
`trastu-
`zumab produces durable objective responses in patients
`whose tumors previously failed to respond to one or more
`chemotherapies for metastatic disease.5
`trials,
`During the course of the trastuzumab pivotal
`cardiac dysfunction (CD) similar to that previously de-
`scribed in patients treated with anthracyclines was observed
`in women treated with trastuzumab. Because of this obser-
`
`vation, a number of changes were instituted in the trials, in
`
`particular the establishment of an independent Cardiac
`Review and Evaluation Committee (CREC). The CREC
`was established to obtain independent and unbiased esti-
`mates of CD risk for patients receiving trastuzumab;
`to
`characterize the observed CD by severity, treatment, and
`clinical outcome; to assess the effects of baseline clinical
`risk factors on CD risk; and to assess the effects of
`cumulative doses of doxorubicin, epirubicin, and trastu-
`zumab on CD risk. This report details the findings of the
`CREC and describes the syndrome of CD associated with
`the administration of trastuzumab.
`
`PATIENTS AND METHODS
`
`Patient Selection and Event Definition
`
`Members of the CREC were Andrew D, Seidman, MD, Clifford A.
`Hudis, MD, and Deborah L. Keefe, MD. all of Memorial Sloan-
`Kettering Cancer Center, New York. Study patients reviewed by the
`CREC were identified through any one of several mechanisms, includ-
`ing the following: (1) independent CREC screening of the clinical trials
`
`From the Memorial Sloan-Kettering Cancer Center, New York, NY;
`and Genentech Inc, South San Francisco, CA.
`Submitted November 27, 2000; accepted November 2, 200].
`Address reprint requests to Andrew Seidman, MD, Memorial Sloan-
`Kettering Cancer Center, 1275 York Ave, New York, NY 10021; email:
`se idmana@mskcc. org.
`© 2002 by American Society of Clinical Oncology.
`0732-l83X/02/2005-1215/320. 00
`
`Journal of Clinical Oncology, Vol 20, No 5 (March 1), 2002: pp 1215-1221
`
`1215
`
`Information downloaded from jco.ascopubs.org and provided by at Reprints Desk on July 14, 2016 from 72.37.250.1899nenteCh 21 12
`Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
`Hospira v. Genentech
`|PR2017-OO737
`
`Genentech 2112
`Hospira v. Genentech
`IPR2017-00737
`
`

`

`1216
`
`Study
`
`H05519
`
`H0552g
`
`H0648g
`
`H0649g
`
`H06509
`
`H0659g
`
`H0693g
`
`Table 1. Trastuzumab and Concurrent Therapy Regimens in Trials Reviewed by the CREC
`Trastuzumab Loading (maintenance) Dose
`
`Phase
`
`II
`
`II
`
`Ill
`
`II
`
`||
`
`II
`
`||
`
`250 mg
`(100 mg weekly until PD)
`250 mg
`(100 mg weekly until PD)
`4 mg/kg
`(2 mg/kg weekly until PD)
`4 mg/kg
`(2 mg/kg weekly until PD)
`None
`(None)
`None
`(None)
`4 mg/kg
`(2 mg/kg weekly)
`4 or 8 mg/kg
`(2 or 4 mg/kg weekly until PD)
`4 mg/kg
`(2 mg/kg weekly)
`4 mg/kg
`(2 mg/kg weekly)
`
`SEIDMAN ET AL
`
`Concurrent Therapy
`
`None
`
`Cisplatin 75 mg/m2 lV every 3 weeks
`
`Doxorubicin‘ 60 mg/m2 IV +
`Cyclophosphamide 600 mg/m2 IV
`PocIitaxel 175 mg/m2 IV
`
`Doxorubicin‘ 60 mg/m2 IV
`600 mg/m2 IV
`PacIitaer 175 mg/m2 IV
`
`None before PD
`
`None
`
`None or investigator specified
`
`None or investigator specified
`
`Abbreviations: IV, intravenous; PD, progressive disease.
`‘Patients could receive other anthracyclines (eg, epirubicin).
`
`database by means of broad search criteria for signs and symptoms of
`CD; (2) adverse event reporting (serious or nonserious) suggestive of
`CD; (3) review of the drug safety and clinical trials databases by means
`of search criteria appropriate for CD (including concomitant medica-
`tions that might be used in the treatment of heart failure); and (4)
`review of cardiac left ventricular ejection fraction (LVEF) data for low
`absolute values (< 50%) or significant decreases (> 20% relative to
`baseline). Medical records for patients identified by the above pro-
`cesses were obtained and provided to the CREC for review. If medical
`records were not available, data collected from the clinical trials were
`provided. All
`identifying features of therapy were removed from
`records provided to the CREC. The CREC reviewed all information
`while remaining blinded to treatment assignment.
`The CREC established the following criteria to confirm or revise a
`preliminary diagnosis of CD:
`(1) cardiomyopathy characterized by a
`decrease in cardiac LVEF that was either global or more severe in the
`septum; (2) symptoms of congestive heart failure (Cl-IF); (3) associated
`signs of CHF, including but not limited to S3 gallop, tachycardia, or both;
`and (4) decline in LVEF of at least 5% to less than 55% with accompa-
`nying signs or symptoms of CHF, or a decline in LVEF of at least 10% to
`below 55% without accompanying signs or symptoms. Any one ofthe four
`criteria was sufficient to confirm a diagnosis of CD. Events were also
`categorized by means of the New York Heart Association (NYHA)
`functional classification system.6 In the NYHA system, patients are
`assigned to one of four functional classes on the basis of the degree of
`effort required to elicit symptoms of heart failure. Patients are classified as
`having symptoms at rest (class IV), at less than ordinary exertion (class
`III), with ordinary exertion (class II), and only at levels that would produce
`symptoms in normal people (class I). Data were collected until
`the
`trastuzumab pivotal trials were closed in mid-1999.
`
`Statistical Considerations
`
`CREC determined diagnoses, assessed cardiac function at time of
`presentation, and summarized the management of CD, changes in
`
`NYHA class, and cardiac outcomes in all patients with CD. These data
`were summarized for seven different antineoplastic regimens (Table 1).
`Data regarding potential risk factors for CD were collected from the
`databases of the clinical trials. Major risk factors were total cumulative
`anthracycline dose, the presence of underlying cardiovascular disease
`(including hypertension), radiation to the left chest or mediastinum, and
`increasing age. For patients in the comparative chemotherapy pivotal
`trial, H0648g, product-limit estimates7 summarizing the distribution of
`cumulative anthracycline and trastuzumab doses at the time of onset of
`CD were calculated. The cumulative anthracycline dose was defined as
`the sum of the total doxorubicin administered before the study plus the
`total administered during the study, up to the date of CREC-defined
`onset of signs and symptoms or to the date of last patient contact.
`Similar definitions were used for cumulative trastuzumab doses.
`Cumulative trastuzumab and doxorubicin exposure was summarized
`by disease response status and antineoplastic therapy for patients in
`study H0648g. Overall response rate and time to disease progression
`were summarized by CD status and antineoplastic therapy.
`
`RESULTS
`
`The CREC screens were applied to 1,219 patient records,
`202 of which were identified for detailed review. CREC
`
`determined that 112 of these patients experienced CD, 83
`patients had clinical manifestations with explanations other
`than CD, and seven patients were not assessable. When
`conditions other than CD were cited in the CREC reviewed
`
`tamponade, cardiac
`they included pericardial
`materials,
`arrhythmia, malignant pleural effusion, cardiac ischemia,
`cardiac arrest, and sepsis, which, by careful review, were
`thought to be unrelated to cancer treatment.
`Table 2 summarizes the results of the CREC review by
`study and treatment regimen. Patients in study H0659g are
`
`lnfon'nation downloaded from jco.ascopubs.org and provided by at Reprints Desk on July 14, 2016 from 72.37.250.188
`Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
`
`

`

`CARDIAC DYSFUNCTION IN TRASTUZUMAB TRIALS
`
`1217
`
`Table 2. CREC Review Summary
`Prior
`Anthracycline
`Exposure (96 01
`patients)
`
`Study
`
`Treatment
`
`H0551g
`HO5529
`H0648g
`
`H0649g
`H0650
`H0659g
`H06939
`Total number of unique
`patients
`
`Trastuzumab alone
`Trastuzumab + cisplatin
`Trastuzumab + AC
`AC alone
`
`Trastuzumab + paclitaxel
`Paclitaxel alone
`Trastuzumab alone
`Trastuzumab alone
`Trastuzumab : other chemotherapy
`Trastuzumab : other chemotherapy
`
`Unknown
`Unknown
`1%
`1%
`
`91%
`97%
`Unknown
`Unknown
`Unknown
`Unknown
`Not
`applicable
`
`Treated‘
`(no. at
`patients)
`46
`39
`143
`135
`91
`95
`213
`1 14
`246
`343
`1,219‘l‘
`
`With CREC-Diagnosed
`Reviewed
`Cardiac Dystunction‘l’
`by CREC?
`(no. at —
`patients)
`No. of Patients
`%
`8
`3
`7
`4
`1
`3
`47
`39
`27
`22
`1 1
`8
`18
`1 2
`1 3
`9
`1
`1
`24
`1 1
`5
`6
`3
`3
`35
`16
`6
`29
`1 5
`A
`202
`1 12
`
`With NYHA Functional
`Class III/IV
`No. 01’ Patients
`2
`1
`23
`5
`2
`1
`8
`2
`8
`10
`
`%
`4
`3
`6
`A
`2
`1
`A
`2
`3
`3
`
`62
`
`‘The number at treated patients does not sum to the total number at unique patients because patients in study H0659g are a subset of patients enrolled in study
`H06489.
`J[For patients enrolled in both study H06489 and study H0659g, review and cardiac dystunction diagnosis are counted only For the trial in which the event first
`occurred .
`
`‘l‘lncludes all patients who received any treatment (including placebo) in any at these seven studies.
`
`a subset of patients from study H0648g who elected to
`participate in an extension study that provided an opportu-
`nity to receive trastuzumab alone or with chemotherapy, as
`directed by the treating physician. CD designations were
`made only for the study in which the event first occurred.
`This process limited to some degree the conclusions that
`could be drawn regarding the risks associated with any
`particular treatment group.
`The experience in the pivotal comparative chemotherapy
`trial, H0648g, is particularly noteworthy. In this trial, patients
`treated with concurrent anthracyclines and trastuzumab expe-
`rienced CD at a rate of 27% compared with a rate of 8% in
`patients treated with the anthracycline combination alone. An
`increased incidence of CD was also seen in patients treated
`with paclitaxel plus trastuzumab versus those treated with
`paclitaxel alone (13% compared with 1%, respectively).
`CD was also observed in trials in which patients were
`treated with trastuzumab as a single agent or in combination
`with a variety of chemotherapeutic regimens. In the H0650g
`trial, patients were treated with trastuzumab as a single
`agent as the first therapy for metastatic disease. The ob-
`served rate of CD was 3%. In the H0649g trial, patients
`were treated with trastuzumab as a single agent as the
`second or third treatment for metastatic disease. The ob-
`served rate of CD in this trial was 5%.
`
`Most patients who developed CD presented with symptoms
`(83 of l 10 patients had their symptom status noted). Eight-two
`of these 83 patients received treatment for CHF. Treatments
`included diuretics (in 78% of treated patients), angiotensin-
`converting enzyme inhibitors (in 58% of patients), cardiac
`
`glycosides (in 58% of patients), and other inotropic agents (in
`10% of patients). Other treatments reported included beta
`blockers, nitrates, and supplemental oxygen. Most patients
`received multiple treatments. When asked to assess response to
`initial
`therapy,
`the CREC concluded that the majority of
`patients (79%) improved with treatment.
`It
`is of note that clinical presentation and treatment
`outcome differed markedly in patients presenting with CD
`in the trastuzumab and anthracycline plus cyclophospha-
`mide (AC) and the trastuzumab plus paclitaxel subgroups of
`H0648g. Sixty-four percent of patients treated with concur-
`rent trastuzumab plus AC presented with significant time-
`tional
`impairment (NYHA class III or IV), whereas only
`20% of patients treated with trastuzumab plus paclitaxel
`presented with significant functional impairment (all class
`III). Eight of thirty-nine patients treated with trastuzumab
`plus AC continued to exhibit significant functional impair-
`ment (NYHA class III) after treatment, whereas none of 12
`patients treated with trastuzumab plus paclitaxel exhibited
`significant functional
`impairment after treatment.
`In the
`heavily pretreated patients who received trastuzumab as a
`single agent (studies H0649g and H0551 g), less impressive
`improvement in functional status was noted. This refracto-
`riness to therapy might have related to more advanced
`disease, reduced functional capacity, and rigorous prior
`therapies. The relationship between treatment
`regimen,
`clinical presentation, and treatment outcome in patients
`presenting with CD is provided in Table 3 for patients who
`had both pretreatment and posttreatment evaluations.
`
`14, 2016 from 72.37.250.188
`Information downloaded from jcoascopubsorg and provided by at Reprints Desk on July
`Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
`
`

`

`1218
`
`Before
`Treatment
`
`|
`
`ll
`
`III
`
`IV
`
`Alter
`Treatment
`I
`II
`|
`ll
`Ill
`|
`H
`ill
`|
`ll
`III
`IV
`
`Trastuzumab Alone
`in *— 14)
`A
`0
`0
`0
`0
`O
`2
`0
`i
`2
`2
`3
`
`
`Table 3. NYHA Functional Classification of CD Cases Before and After Treatment for Dysfunction: Selected Antineoplastic Regimens‘
`NYHA Classification
`No. ol Subiects
`Trostuzumob + ACt
`[n
`34)
`
`SEIDMAN ET AL
`
`Trastuzumab + Paclitaxeli=
`l"
`10)
`
`wuhwa—‘NNflV
`
`0
`
`OOOOOONO—‘NOUI
`
`‘Not all patients underwent both pretreatment and posttreatment evaluations. This table includes data only tor patients with both pretreatment and posttreatment
`evaluations.
`
`‘l'Studies Hossig, H0649g, and H06509.
`*Study H06489.
`
`In the entire trastuzumab developmental clinical trials expe-
`rience of 1,219 patients, 10 heart-related deaths were reported.
`Nine of the 10 were in patients who had received trastuzumab
`in the clinical trials. Analysis of potential risk factors associ-
`ated with trastuzumab-associated CD was performed. Table 4
`demonstrates that of suspected factors, only age had a statisti-
`
`cally significant effect on CD risk and reached statistical
`significance only in the trastuzumab plus AC subgroup.
`The relationship of CD risk to cumulative anthracycline
`exposure was
`investigated for patients
`in the pivotal
`H0648g study. Among patients in the trastuzumab plus AC
`subgroup, the cumulative dysfunction risk began to increase
`
`Risk Factor
`
`Mean age, years
`Risk ratio, per lO-year increase
`95% Cl
`
`Preexisting hypertension
`No.
`%
`
`Risk ratio, hypertension v no hypertension
`95% Cl
`
`Prior radiotherapy
`No.
`%
`
`Risk ratio, radiotherapy v no radiotherapy
`95% Cl
`
`Mean cumulative doxorubicin dose, mg/mz‘l'
`Risk ratio, < 400 v 2 400 mg/m2
`95% Cl
`Mean baseline cardiac EF, %
`Risk ratio, < 60 v 2 60%
`95% Cl
`
`Table 4. Baseline Clinical Risk Factors for CD in Study H06489
`Trasruzumab + AC‘ (n = 143)
`Cases
`Noncases
`(n :102)
`[n
`39)
`53
`
`56
`
`7
`l8
`
`16
`Al
`
`NA
`
`62
`
`1.56
`1.12-2.17
`
`i.0l
`0.44-2.3l
`
`0.64
`0.3A-l .23
`
`NA
`
`l.46
`0.56-3.79
`
`19
`l9
`
`52
`51
`
`NA
`
`63
`
`Trastuzumab — Paclitaxel‘ (n = 91)
`Cases
`Noncases
`(n =12)
`(n' 77)
`54
`50
`
`l
`9
`
`8
`73
`
`243
`
`60
`
`l.43
`0.82-2.50
`
`0.69
`0.08-5.7i
`
`l.l6
`0.31-4.58
`
`0.71
`0.08-6.26
`
`0.69
`0.05-9.97
`
`l l
`14
`
`5i
`66
`
`261
`
`61
`
`Abbreviations: EF, eiection traction; NA, not applicable; CI, confidence interval.
`‘Two patients in each treatment group were not assessable by CREC.
`i'Patients receiving epirubicin are not included.
`
`Information downloaded from jco.ascopubs.org and provided by at Reprints Desk on July 14, 2016 from 72.37.250.188
`Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
`
`

`

`CARDIAC DYSFUNCTION IN TRASTUZUMAB TRIALS
`
`1219
`
`10
`
`.0a:
`
`— Trastuzamab + AC
`-
`AC Alone
`
`0.0
`
`
`
`
`
`CumulativeEventProbability
`
`0.6 -
`
`0.4 -
`
`0.2
`
`020 60100140180 220 260 300 340 380 420 460 500 540 580
`
`Cumulative Doxorublcin Dose (mg/m2)
`
`Fig 1. Product-limit estimates of cumulative doxorubicin dose to CD
`onset. CREC-diagnosed CD (nonossessable patients excluded) for study
`H0648g.
`
`with doxorubicin exposure of approximately 300 mg/m2 (Fig
`l). The product-limit estimate demonstrated a dramatic in-
`crease in risk at 360 mg/mz, but this effect was largely an
`artifact of the large fraction of patients who discontinued AC
`therapy after receiving six cycles of doxorubicin (per study
`protocol) and subsequently experienced CD. Approximately
`60% of patients in both the trastuzumab plus AC and the AC
`alone subgroups discontinued AC therapy after six cycles.
`Because relatively few events occurred at cumulative doxoru-
`bicin doses less than 300 mg/mz, this trial provides limited
`information about CD risk below this exposure level.
`In study H0648g, the curve representing cumulative CD risk
`in patients receiving trastuzumab plus AC was similar in shape
`to that of patients receiving AC alone; however,
`the risk
`function was shifted to the left, indicating a greater risk in the
`trastuzumab plus AC subgroup. The magnitude of the shift
`suggests that trastuzumab may increase the risk of CD by a
`degree equivalent to approximately 100 mg/m2 of doxorubicin
`exposure. Product-line estimates assume that a patient whose
`data has been censored has a nonzero risk after censoring equal
`to that of a patient remaining at
`risk. The product-limit
`estimates can be expected to overestimate the true CD risk
`because patients whose data have been censored on average
`received lower cumulative doses of chemotherapy and trastu-
`zumab, and their risk is likely to be less informative than the
`relative positions ofthe two curves, indicating that trastuzumab
`increases the risk of CD more than that induced by treatment
`with anthracyclines alone.
`An exploratory analysis was conducted to evaluate the
`overall benefit of trastuzumab therapy to patients with
`metastatic breast cancer.8 Time to treatment failure was
`
`defined as the time to progressive disease or CD, whichever
`occurred first. We found that trastuzumab-containing regi-
`mens produced an improvement in time to treatment failure
`
`Table 5. Time to Treatment Failure (progressive disease or CD) by
`Treatment
`
`Treatment
`
`Time to Treatment Failure
`(months)
`
`Trastuzumab plus AC
`AC alone
`Trastuzumab plus paclilaxel
`Paclitaxel alone
`Trastuzumab plus chemotherapy
`Chemotherapy alone
`'P = .2381 VAC alone.
`
`l’P = .0001 v paclitoxel alone.
`iiP = .0001 v chemotherapy alone.
`
`6.6
`6.0
`6.6
`2.7
`6.6
`4.6
`
`95% CI
`
`5.5-7.3'
`4.8-6.9
`3.5-1 1 .21"
`2.0-4.3
`5.8-7.0‘1=
`4.4-5.3
`
`compared with regimens that did not include trastuzumab
`(Table 5). A similar difference was observed in each of the
`treatment substrata. CD-free survival, defined as the time to
`symptomatic CHF (NYHA functional class III or IV) or
`death, followed a similar pattern, with longer survival noted
`in trastuzumab-treated groups.8
`
`DISCUSSION
`
`This article details the unexpected observation of increased
`rates of CD in the trastuzumab clinical trials program and the
`steps that were taken to document and characterize the course
`of the observed syndrome. CHF was detected as a safety signal
`late in the development of the drug. Although CHF was seen
`in some patients participating in phase II trials, the rate of
`occurrence was sufficiently low and patients at sufficiently
`increased risk because of anthracycline pre-exposure that the
`increased risk with trastuzumab treatment was not appreciated.
`Concomitant use of anthracyclines and trastuzumab had not
`been used in the phase II program that focused on single-agent
`therapy and the interaction of trastuzumab with cisplatin.
`Therefore, it was not until the occurrence of several cases in
`the pivotal phase III combination chemotherapy trial that a
`safety signal was detected. It was soon recognized that the
`syndrome of trastuzumab-associated CD resembled the
`cardiomyopathy associated with previously described
`anthracycline cardiotoxicity9’1° and that
`the rates of
`occurrence were greatest when trastuzumab was used
`concomitantly with anthracyclines.
`Caution should be exercised in interpreting the results of
`the CREC experience detailed here. After detection of the
`safety signal, several communications to the trial investiga-
`tors detailed concerns related to the use of trastuzumab in
`
`this open-label trial. These might have resulted in biased
`overreporting of cardiac events in the trastuzumab-contain-
`ing subgroups within the trial.“ Furthermore, although the
`CREC was established to provide independent analysis of
`the events observed, the results are necessarily limited by
`
`Information downloaded from jco.ascopubs.org and provided by at Reprints Desk on July 14, 2016 from 72.37.250.188
`Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
`
`

`

`1220
`
`SEIDMAN ET AL
`
`the retrospective nature of the analysis. Finally, the conser-
`vative criteria applied by CREC to confirm or revise a
`preliminary diagnosis of CD to include “an absolute decline
`in lefi ventricular ejection fraction of at least 5% to below
`55%” could conceivably have biased the analysis toward
`overestimation of the incidence of CD.
`
`The pathophysiologic basis oftrastuzumab-associated CD is
`poorly understood. Four endomyocardial biopsies obtained
`during the conduct of the trials failed to clarify this issue (data
`not provided). One biopsy was consistent with anthracycline
`toxicity, and a second had mild changes consistent with
`anthracycline-induced myocyte damage. However, two speci-
`mens did not demonstrate changes consistent with anthracy-
`cline damage and also had no evidence of inflammation,
`necrosis, or fibrosis.
`lmmunohistochemical
`staining from
`p185HER2 was consistently negative in the specimens. Two of
`four specimens were positive for HER2 mRNA by in situ
`hybridization, but only three of four specimens demonstrated
`control binding of an actin probe. It remains unknown whether
`trastuzumab exacerbates previous damage caused by anthra-
`cyclines or acts through an independent mechanism to directly
`affect the cardiac myocyte.
`It has been suggested that pl85HER2 may play a role in
`myocyte survival.12 However, the molecular mechanisms un-
`derlying trastuzumab-associated cardiomyopathy are obscure.
`Knockout ofthe c-erbB—2 gene results in embryonic lethality in
`mice and is associated with profound developmental defects in
`the heart and central nervous systems.13 Recent discovery of
`the association of p185HERZ with gp130 and gp130-inducible
`myocyte survival pathwaysI4 suggests that identification of
`myocyte survival factors could lead to new therapeutic ap-
`proaches for this and other cardiomyopathies.15
`Risk factors for trastuzumab-associated CD are poorly
`delineated. Only age was associated with increased risk, and
`only in the AC substratum. A better understanding of risk
`factors is needed. Multigated blood pool imaging of the heart
`(MUGA scanning) is currently recommended for baseline
`assessment and on-treatment evaluation oftrastuzumab-treated
`
`patients. MUGA scanning seems unable to identify early
`evidence of CD, and once impaired systolic function is
`detected by MUGA, significant cardiac damage has already
`occurred. It would be optimal to have a sensitive, noninvasive
`test to detect minor cardiac damage in time for an intervention
`to preserve myocardium. Troponin-T16 and pro-BNPl7 levels
`are being examined in clinical trials to define their utility in
`detecting early myocardial damage. In addition, echocardiog-
`raphy is being compared with MUGA scanning to determine if
`this modality might be more sensitive to changes associated
`with trastuzumab exposure.
`In the phase III pivotal trial, all patients had been or were
`exposed to anthracyclines, the paclitaxel substratum having
`
`received anthracyclines in the adjuvant setting and anthra-
`cycline-naive patients entering the AC substratum receiving
`anthracyclines during the trial by protocol design. We found
`incidence rates for development of CD to be greatest when
`trastuzumab and anthracyclines are used concomitantly.
`Although epirubicin was used much less frequently in these
`studies than doxorubicin, the incidence of cardiac toxicity
`seemed to be similar regardless of the anthracycline used.
`Rates of CD seem substantially lower and the syndrome less
`severe when the use of trastuzumab is temporally separated
`from anthracycline exposure, as demonstrated by the rates
`in the trastuzumab treatment arm of the paclitaxel substra-
`tum and rates in the single-agent trials. Results in other
`clinical trials with prospective frequent monitoring of left
`ventricular function have suggested that the safety profile of
`concomitant treatment of patients with trastuzumab and
`weekly paclitaxel is acceptable.‘8
`Observations of CD in the trastuzumab pivotal trials and the
`experience with trastuzumab and weekly paclitaxel have influ-
`enced the design of trastuzumab adjuvant trials undertaken
`recently by the major United States cooperative groups. The
`National Surgical Adjuvant Breast and Bowel Project B-3l
`trial has been designed with a formal safety evaluation phase
`and an efficacy evaluation phase. The first 1,000 patients
`entered onto the trial will participate in a meticulous evaluation
`of cardiac safety. Patients will undergo baseline and critical
`time MUGA scanning through the course of treatment with
`AC for four cycles, then paclitaxel for four cycles with or
`without trastuzumab for 1 year. The United States Inter-
`group trial uses similar monitoring as patients receive AC
`for four cycles,
`then paclitaxel weekly for 12 weeks
`without, with, or followed by trastuzumab for 1 year.
`Neither trial uses concomitant use of trastuzumab and
`
`anthracyclines, and the lntergroup trial examines the
`importance of further temporal separation of trastuzumab
`administration from anthracycline administration.
`Trastuzumab is an important new treatment for women
`with breast cancers that overexpress the HERZ protein; it is
`one of only a few drugs that has produced a survival
`advantage for patients with metastatic disease. The note-
`worthy efficacy of trastuzumab has been detailed else-
`where.3'5 In this report, we detail the experience with the
`major side effect noted in the trastuzumab clinical trials:
`CD. Patients treated with trastuzumab seem to be at
`in-
`
`creased risk for the development of a cardiomyopathy
`reminiscent of the cardiomyopathy associated with anthra-
`cyclines. All clinical trials reviewed by CREC were con-
`ducted on a background of anthracycline exposure because
`anthracyclines are the foundation of many adjuvant treat—
`ments for early-stage breast cancer, and anthracyclines are
`believed to be of special
`importance in the treatment of
`
`Information downloaded from jco.ascopubs.org and provided by at Reprints Desk on July 14, 2016 from 72.37.250.188
`Copyright © 2002 American Society of Clinical Oncology. All rights reserved.
`
`

`

`CARDIAC DYSFUNCTION IN TRASTUZUMAB TRIALS
`
`1221
`
`risk for
`HER2-positive breast cancer.”‘20 The greatest
`development of CD was seen in patients receiving concom-
`itant
`trastuzumab and anthracycline. Risks for patients
`treated with concurrent paclitaxel and trastuzumab and
`patients treated with trastuzumab as a single agent seem to
`be substantially lower than those associated with the con-
`current administration of anthracycline and trastuzumab.
`The risk in anthracycline-naive patients is not well defined
`but is apparently greater than zero because at
`least one
`patient with preexisting cardiac disease in the H0650g study
`experienced CD during the trial.
`As with all decisions in clinical medicine, the decision to
`use trastuzumab involves an analysis of risk and benefit.
`The risk of CD associated with the use of trastuzumab can
`
`usually be justified, given the 25% improvement in overall
`survival associated with the use of trastuzumab in the
`
`pivotal comparative chemotherapy trial, H0648g.5 Trastu-
`zumab provided improved time to treatment failure in this
`study, an end point that incorporates time to progressive
`disease or cardiac event. This conclusion of benefit
`is
`
`further bolstered by the poor prognosis associated with
`
`HER2-positive breast cancer and the likelihood of func-
`tional improvement with appropriate treatment of the CD.
`Because of the important antitumor activity of trastu-
`zumab when combined with doxorubicin and cyclophosph-
`amide, strategies to prevent or limit cardiotoxicity are being
`investigated. Clinical
`trials that use Iiposomal anthracy-
`clines as well as the cardioprotective agent dexrazoxane in
`combination with trastuzumab are ongoing. For example,
`TLC D-99, a Iiposomal doxorubicin formulation with less
`cardiotoxicity than conventional doxorubicin, has been
`preliminarily indicated to be well tolerated and to have good
`antitumor efficacy in patients with HER2-positive tumors
`treated with TLC D-99 in combination with trastuzumab. In
`
`this phase I/II study, no cardiotoxicity was reported in 20
`women with locally advanced or metastatic breast cancer.”
`Important goals of future trastuzumab trials should be the
`identification of risk factors for CD, as well as identification
`of appropriate strategies for prevention or treatment of CD.
`It is likely that meeting these goals will require a better
`understanding of the molecular mechanisms of trastuzumab
`action on the myocardial cell.
`
`REFERENCES
`
`1. Slamon DJ, Clark GM, Wong SG. et al: Human breast cancer:
`Correlation of relapse and survival with amplification of the HER-2/
`neu oncogene. Science 235:177-182, 1987
`2. Ravdin PM, Chamness GC: The c-erbB-Z prom-oncogene as a
`prognostic and predictive marker in breast cancer: A paradigm for the
`development of other macromolecular markers—A review. Gene
`159:19-27, 1995
`3. Slamon DJ, Leyland-Jones B, Shak S, et al: Use of chemotherapy
`plus a monoclonal antibody against HER2 for metastatic breast cancer
`that overexpresses HER2. N Engl J Med 344:783-792, 2001
`4. Norton L, Slamon D, Leyland—Jones B, et al: Overall survival
`(OS) advantage to simultaneous chemotherapy (CRX) plus the human-
`ized anti-HER2 monoclonal antibody Herceptin (H) in HER2-overex-
`pressing (HER2+) metastatic breast cancer (MBC). Proc Am Soc Clin
`Oncol 18:127a, 1999 (abstr 483)
`5. Cobleigh MA, Vogel CL, Tripathy D, et al: Multinational study
`of the efficacy and safety of humanized anti-HER2 monoclonal
`antibody in women who have HER2-overexpressing metastatic breast
`cancer that ha