`
`Kesavan Shan, MD; A. Michael Lincoff, MD; and James B. Young, .MD
`
`£
`11
`:7 Ir!/.$-
`MARK RICHMAN, RPR
`
`Purpose: To review the current understanding of the
`clinical significance, detection, pathogenesis, and preven(cid:173)
`tion of anthracycline-induced cardiotoxicity.
`Data Sources: A MEDLINE search of the English-lan(cid:173)
`guage medical literature and a manual search of the bib(cid:173)
`liographies of relevant articles, including abstracts from
`national cardiology meetings.
`Study Selection: Pertinent clinical and experimental
`studies addressing the clinical relevance, pathogenesis,
`detection, and prevention of anthracycline cardiotoxicity
`were selected from peer-reviewed journals without judg(cid:173)
`ments about study design. A total of 137 original studies
`and 9 other articles were chosen.
`Data Extraction: Data quality and validity were assessed
`by each author independently. Statistical analysis of com(cid:173)
`bir;ied data was inappropriate given the differences in
`. ·patient selection, testing, and follow-up in the available
`studies.
`Data Synthesis: Anthracyline-induced cardiotoxicity lim(cid:173)
`its effective cancer chemotherapy by causing early cardio- ·
`myopathy, and it can produce late-onset ventricular dys(cid:173)
`function years after treatment has ceased. Detection of
`subclinical anthracyline-induced cardiomyopathy through
`resting left ventricular ejection fraction or echocardio(cid:173)
`graphic fractional shortening is suboptimal. Conventional
`doses of anthracycline often lead to permanent myocar(cid:173)
`dial damage and reduced functional reserve. Underlying
`pathogenetic mechanisms may include free-radical(cid:173)
`mediated myocyte damage, adrenergic dysfunction, in(cid:173)
`tracellular calcium overload, and the release of cardiotoxic
`cytokines. Dexrazoxane is the only cardioprotectant clini(cid:173)
`cally approved for use against anthracyclines, and it was
`only recently introduced for selected patients with breast
`cancer who are receiving anthracycline therapy.
`Conclusions: A rapidly growing number of persons, in(cid:173)
`cluding an alarming fraction of the 150 000 or more adults
`in the United States who have survived childhood cancer,
`will have substantial morbidity and mortality because of
`anthracycline-related cardiac disease. The development of
`effective protection against anthracycline-induced cardio(cid:173)
`toxicity will probably have· a significant effect on the over(cid:173)
`all survival of these patients.
`
`Ann Intern Med. 1996;125:47-58.
`
`From The Cleveland Clinic Foundation, Oeveland, Ohlo. For
`current author addresses, see end of text.
`·
`
`·Anthracyclines are well established as highly ef-
`ficacious antineoplastic agents for various he(cid:173)
`mopoietic (1) and solid tumors (2-4). A clear dose(cid:173)
`response relation for anthracyclines in several curative
`chemotherapeutic regimens has been shown; de(cid:173)
`creased doses result in inferior survival and remis(cid:173)
`sion rates (1, 4). However, the cardiotoxicity of
`these agents ( 4-6), which has been recognized for
`more than 20 years (7), continues to limit their
`therapeutic potential and threaten the cardiac func(cid:173)
`tion of many patients with cancer.
`Three distinct types of anthracycline-induced car(cid:173)
`diotoxicity have been described. First, acute or sub(cid:173)
`acute injury can occur immediately after treatment.
`This rare form of cardiotoxicity may cause transient
`arrhythmias (8, 9), a pericarditiS:-myocarditis syn(cid:173)
`drome, or acute failure of the left ventricle (10).
`Second, anthracyclines can induce chronic cardio(cid:173)
`toxicity resulting in cardiomyopathy. This a more
`common form of damage and is clinically the most
`·important (11-13). Finally, late-onset anthracydine
`cardiotoxicity causing late-onset ventricular dysfunc(cid:173)
`tion (14-16) and arrhythmias (17-19), which mani(cid:173)
`fest years to decades after anthracy~line ·treatment
`. has been completed, is increasingly recognized.
`Chronic anthracycline-induced cardiomyopathy
`characteristically presents within 1 year of treatment.
`In a series of more than 3900 patients treated with
`anthracycline, Von Hoff and associates (6) noted
`that congestive heart failure secondary to anthracy(cid:173)
`cline-induced chronic cardiomyopathy occurred 0 to
`231 days after the completion of anthracycline ther(cid:173)
`apy. In contrast, late-onset anthracycline-induced
`cardiac abnormalities have been reported to occur
`much later, after a prolonged asymptomatic period
`(14-16). Other cardiovascular risk factors predis(cid:173)
`posing to heart failure, such as occult hypertension
`and subclinical coronary artery disease, may have
`confounded interpretation of the exact contribution
`made by anthracyclines in these studies. However,
`anthracyclines are clearly an important independent
`risk factor leading to both early and delayed con(cid:173)
`gestive heart failure in survivors of cancer. There is
`no universally defined point · after the onset of
`chronic cardiomyopathy at which late-onset cardiac
`abnormalities appear. For the purposes of this re(cid:173)
`view, we broadly define "chronic cardiotoxicity" as
`cardiotoxicity occurring within 1 year of treatment
`
`©1996 American College of Physicians
`
`47
`
`
`
`1 of 12
`
`Celltrion, Inc. 1050
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`and "late-onset cardiotoxicity" as cardiotoxicity oc(cid:173)
`curring more than 1 year after the completion of
`anthracycline therapy.
`
`Clinical Significance
`
`Acute and Subacute Cardiotoxicity
`Acute and subacute cardiac toxicity, which occur
`immediately after a single dose of an anthracycline
`or a course of anthracycline therapy, are uncommon
`under current treatment protocols. Several distinct,
`early cardiotoxic effects of anthracyclines have been
`described. First, electrophysiologic abnormalities
`may result in nonspecific ST and T-wave changes,
`decreased QRS voltage, and prolongation of the QT
`interval. Sinus tachycardia is the most common
`rhythm disturbance, but arrhythmias: including ven(cid:173)
`tricular, supraventricular, and junctional tachycar(cid:173)
`dias, have been reported (8-11). Atrioventricular
`and bundle-branch bl.ock have also been seen (8).
`These electrophysiologic changes are seldom a seri(cid:173)
`ous clinical problem (11). Rare cases of subacute
`cardiotoxicity resulting in acute failure of the left
`ventricle, pericarditis, or a fatal pericarditis-myocar(cid:173)
`ditis syndrome have been reported (12).
`
`Chronic Cardiotoxicity
`The incidence of congestive heart failure second(cid:173)
`ary to doxorubicin-induced cardiomyopathy depends
`on the cumulative dose of the drug. At total doses
`of less than 400 mg/m2 body surface area, the inci-·
`dence of congestive heart failure is 0.14%; this in(cid:173)
`cidence increases to 7% at a dose of 550 mg!m2
`body surface area and to 18% at a dose of 700
`mg/m2 body surface area (6) (Figure 1). The rapid
`increase in clinical toxicity at doses greater than 550
`mg/m2 body surface area has made the 550-mg dose
`
`1.00
`
`0.90
`u.. 0.80
`:c
`u 0.70
`.... 0.60
`>.
`:c .,, 0.50
`..Q 0.40
`0 ....
`0.. 0.30
`0.20
`
`0.10
`
`00 100 200 300 400 500 600 700 800 900 1000
`Total Dose (mg/m2)
`Figure 1. Cumulative probability of developing doxorubicin·induced
`congestive heart failure (CHF) plotted against total cumulative dose
`of doxorubicin in all patients receiving the drug (3941 patients; 88
`cases of congestive heart failure). Reproduced from Von Hoff and col·
`leagues (6) with permission of Annals of Internal Medicine.
`
`the popular empiric limiting dose for doxorubicin(cid:173)
`induced cardiotoxicity. Mortality directly related to
`doxorubicin-induced cardiac failure is substantial;
`large series have reported rates of more than 20%
`(5, 6). However, recent reports have suggested a
`better prOSJ?.OSis (20-24), with up to 59% clinical
`recovery in patients with anthracycline-induced con(cid:173)
`gestive heart failure who are treated with digoxin ·
`and diuretics (14). Complete recovery of echocar(cid:173)
`diographic shortening fraction may also occur if an -
`thracycline therapy is discontinued at an early stage
`(24), but this does not exclude long-term reductions
`in functional reserve (23).
`Although reports conflict, proposed risk factors
`for chronic anthracycline cardiotoxicity
`include
`higher rates of drug administration (25), mediastinal
`radiation (10, 26), advanced age (3, 6), younger age
`(27, 28), female sex (29), pre-existing heart disease,
`and hypertension (6). Multivariate analysis of these
`factors by Torti and colleagues (30), based on his(cid:173)
`tologic evidence of anthracycline cardiotoxicity,
`showed that only higher rates of anthracycline ad(cid:173)
`ministration and previous cardiac irradiation were
`independent risk factors. An additional confounding
`factor in the identification of patients at highest risk
`for cardiotoxicity is the wide variation in individual
`sensitivity to anthracyclines (31-33). Doses in excess
`of 1000 mg/m2 body surface area can be well toler·
`ated by some patients (31, 33). In contrast, appre(cid:173)
`ciable decreases in left ventricular ejection fraction
`have been documented by multigated nuclear scans
`at doses as low as 300 mg/m2 body surface area (26,
`27). Furthermore, endomyocardial biopsy specimens
`may show histopathologic changes characteristic of
`doxorubicin-induced cardiotoxicity (Figure 2) at
`doses as low as 183 mg/m2 body surface area (13)(cid:173)
`less than one third of the conventional limiting
`dose. Thus, a substantial proportion of patients have
`anthracycline-induced cardiac damage while receiv(cid:173)
`ing standard treatment regimens, whereas others
`can tolerate cumulative doses twice as large as the
`conventional limiting dose.
`
`Late-Onset Cardiot oxicity
`Several recent studies have noted occult ventric(cid:173)
`ular dysfunction, heart failure, and arrhythmias oc(cid:173)
`curring in asymptomatic patients more than 1 year
`after anthracycline treatment (16-21, 34- 36). These
`initial findings suggest that survivors of cancer may
`have a previously unacknowledged increase in car(cid:173)
`diac morbidity and mortality due to anthracycline
`therapy. Steinherz and associates (16), who studied
`201 patients with solid tumors or leukemia, found
`an 18% incidence of reduced fractional shortening
`on resting echocardiograms in patients followed for
`4 to 10 years after completion of anthracycline ther(cid:173)
`apy. Even more troubling are the findings of Lip ..
`
`48
`
`1 July 1996 • Annals of Internal Medicine • Volume 125 • Number 1
`
`
`
`2 of 12
`
`Celltrion, Inc. 1050
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`shultz and coworkers ( 17), who noted that cumula(cid:173)
`tive doses of doxorubicin as low as 228 mg/m 2 body
`surface area increased afterload or decreased con(cid:173)
`tractility or both in 65% of patients with leukemia
`up to 15 years after treatment with anthracydines.
`These abnormalities appear to be progressive and
`reflect future clinical decompensation.
`Lipshultz and coworkers ( 17) noted both early
`and late congestive heart failure in 5 of I 15 patients
`within 11 years of the completion of anthracycline
`therapy. A similar incidence of late-onset heart fai l(cid:173)
`ure, occurring in 9 of 201 patients, was reported by
`Steinherz and associates (16). However, most pa(cid:173)
`tients in the study by Steinherz and associates were
`fol lowed for fewer than I 0 years after completion of
`therapy, and new-onset symptomatic ventricular dys(cid:173)
`function was not seen until I 2 to 14 years after
`treatment in the study by Lipshultz and coworkers
`( 17). Adtjitionally, the incidence of severe echocar(cid:173)
`diographic abnormalities increased with the dura(cid:173)
`tion of follow-up (Figure 3). These findings indicate
`that the full extent of the problem has yet to unfold
`in many asymptomatic patients after remote anthra(cid:173)
`cycline treatment. In addition, late-onset arrhyth(cid:173)
`mias and sudden death have been reported to have
`occurred in patients more than 15 years after an(cid:173)
`tluacycline treatment ( 19- 21 ). Incidences of non(cid:173)
`sustained ventricular tachycardia ranging between
`3% and 5% in <inthracycline-treated patients at
`long-term follow-up have been reported (17, 19).
`The natural history of these arrhythmias and
`whether they occur independently of ventricular
`dysfunction ( 19) are issues that remain to be clari(cid:173)
`fied.
`As with the form of anthracycline cardiotoxicity
`that manifests earlier, the incidence of late-onset
`c<irdiac decompensation increases with larger cumu(cid:173)
`lative doses ( l6, l 7), higher rates of anthracycline
`administration (37), and mediastinal radiotherapy
`(16). Young age (16) at the time of treatment and
`female sex (38) may be risk factors for late-onset
`anthracycline cardiotoxicity, but this is still contro(cid:173)
`versial (17, 39). However, recent evidence appears
`to support both characteristics as independent risk
`factors for late-onset ventricul<ir dysfunction (37).
`
`Pathogenesis
`
`Figure 2 . Changes characteristic of adriamycin·induced cardioto><·
`idty. Top. Li9ht microscopy. Section or left ventricle from a 57-year-old
`woman with ad11amycin-mduced card1omyopathy showing marked myofibril
`loss and vacuolar degeneration (arrow) (Hemato.xylm and eosin stain. Orig·
`inal magnification. X400). Bottom. Electron microscopy. Cardiac myocyte
`showing adriamycin-induced cardiotoxici ty with extensive loss or myolila(cid:173)
`merus (large arrows). Unaffected myocytes are shown m lower left. Small
`arrow denotes normal myocyte. (Original magnification, x 2800).
`
`at the cellular level include free-radica l- mediated
`myocardial injury (41 - 48), myocyte d<image from
`calcium overload (49-53), disturbances in myocar(cid:173)
`dial adrenergic function (54-56), release of vasoac(cid:173)
`tive amines (57, 58), and cellular toxicity from metab(cid:173)
`olites of doxorubicin (59, 60). Finally, elaboration of
`
`m*
`§ .~ 3(1
`J5 ~
`<t; ~
`~ ~ zc
`~~
`~ .g
`·~ ~ 10
`Q. u.
`
`• • • • • • • • • • • • • • • • • • · • • • • • • • •
`
`• • srl.?$.;. · · · · · · · · · · · ·
`
`(l'l·S8·
`
`··· ·· ·- ·· ····· ··· · ··· ·-··
`
`••
`
`Chronic Cardiomyopathy
`Anthracyclines cause the selective inhibition of
`cardiac muscle gene expression for a-actin, tropo(cid:173)
`nin, myosin light-chain 2, and the M isoform of
`creatine kinase in vivo (40), which may explain the
`myofibrillar loss (Figure 2, bouom) associated with
`anth racycline-induced cardiomyopathy. Hypotheses
`
`? 10
`
`7.9
`Follow-up. y
`Figure 3. Late-onset ventricular dysfunction over time. Percent(cid:173)
`age or patien ts with abnormal fract ional shortening at Jong·lerm follow-up
`over time a1ter completion of therapy. Adapted from Steinherz and col·
`leogues ( 16) with permission of The Journal of rhe American Medical Asso(cid:173)
`ciation. White bars "' mild reduction in fractional shor tening (25% to 28%);
`striped bars = moderate reduct ion in fractional shortening (2 1 % to 24%):
`black bars ,,,, severe reduction in fractional shortening (s 20%).
`
`l July 1996 • Annals of Internal Medicine • Volume 125 • Number 1
`
`49
`
`
`
`3 of 12
`
`Celltrion, Inc. 1050
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`Table 1. Abnormalities in Diastolic and Systolic Ventricular Function for the Early Detection of Anthracycline(cid:173)
`lnduced Cardiotoxicity*
`
`Anthracycline Dose
`
`Mediastlnal Radiotherapy
`
`Follow-up Period after
`Anthracydine Treatment
`
`Study (Reference)
`
`Patients
`
`Cottin et al. (107}
`
`Ewer et al. (t 06)
`
`Marchandise et al. (101)
`
`Stoddard et al. (103}
`
`Schmitt et al. (108)
`
`Lee et al. (102)
`
`n
`
`60
`
`41
`
`45
`
`26
`
`14
`
`12
`
`Age
`
`y
`
`Mean, 50
`Range, 23 to 72
`Range, 7 to 15.5
`
`Mean, 45
`
`mglm"
`
`Mean, 251
`Range. 75 to 550
`Mean, 119
`Range, 50 to 475
`Range. 200 to 450
`
`No
`
`No
`
`NS
`
`Mean,48
`
`Range. c:200
`
`Range, 3 to 12
`
`Mean, 53
`Range, 16 to 69
`
`Mean, 240
`Range, 200 to 300
`Mean, 193
`Range, 80 to 448
`
`None within 3 weeks of
`anthracycline
`treatment
`NS
`
`No
`
`Measurements during therapy
`
`Within 2 months
`
`3 to 9 months
`
`3 months (22 patients) and 3
`weeks (4 patients)
`
`Within 2 years
`
`7 to43 weeks
`
`• AIE • atiiaVearly mitral flow velocities; N.S = not stated.
`
`pro-inflammatory cytokines, which have been con(cid:173)
`sistently identified in other forms of ventricular dys(cid:173)
`function (61-64), may be directly relevant to an(cid:173)
`thracycline-induced cardiac injury.
`Although the cause of anthracycline-induced car(cid:173)
`diotoxicity is probably multifactorial, a large body of
`evidence points to free-radical-mediated myocyte
`damage (41-44). Increased oxygen radical activity
`generated through the semiquinone moiety of the
`doxorubicin molecule can cause lipid peroxidation
`and cell injury (41,. 42). Anthracycline-induced
`intracellular calcium overload may also lead to
`myocyte death (49-53). Doxorubicin activates the
`calcium-release channel across the sarcoplasmic
`reticulum (52) and causes calcium influx into the
`myocyte (50, 51, 53). Free-radical-induced cell
`membrane damage has also been seen with cal(cid:173)
`cium influx (65, 66), ·suggesting that these two
`putative anthracycline-induced cardiotoxic path(cid:173)
`ways may be linked. Adrenergic dysfunction, in(cid:173)
`cluding downregulation of myocardial p-adrener(cid:173)
`gic receptors (67, 68), may be present in evolving
`(69, 70) as well as established anthracycline-induced
`ventricular dysfunction (71, 72).
`Recent reports (61-64) that circulating pro-inflam(cid:173)
`matory cytokines may be intimately linked to the
`evolution of ventricular dysfunction and dilated car(cid:173)
`diomyopathies may provide further insight into the
`process by which anthracyclines produce cardiac in(cid:173)
`jury. Doxorubicin induces the release of tumor ne(cid:173)
`crosis factor-a from macrophages and of interleu(cid:173)
`kin-2 from monocytes (73-75). Interleukin-2 and
`tumor necrosis factor-a, which has functional
`myocardial ·receptors (76), have documented car(cid:173)
`diotoxicity that can result in dilated cardiomyopathy
`(77, 78). Varying degrees of cytokine liberation from
`different tumors (79-81) during anthracycline treat(cid:173)
`ment may provide another explanation for the dis-
`
`crepancy in the incidence of cardiotoxicity in differ(cid:173)
`ent populations with cancer (82, 83).
`
`Late-Onset Cardiotoxicity
`Progressive ventricular dysfunction after an initial
`myocardial insult probably underlies late-onset de(cid:173)
`compensation. Reductions in left ventricular mass.
`mass index, and compliance have been reported in
`anthracycline-treated survivors of childhood cancer
`followed for more than 7 years after completion of
`chemotherapy (34, 84). These patients appear to
`have a thin-walled left ventricle working against
`high systolic wall stress (34). Such a pattern of
`cardiac injury is concordant with the theory that
`occult late-onset anthracycline cardiac dysfunction
`manifests clinically in patients who remain in a com(cid:173)
`pensated state for many years. Acute viral infection
`(85) and cardiovascular stressors, such as weight
`lifting (20), pregnancy, and surgery, are possible
`triggers of late-onset anthracycline-induced cardiac
`dysfunction.
`
`Monitoring
`
`The lifelong cardiotoxic effects of conventional
`antbracycline therapy highlight the neeq for moni(cid:173)
`toring methods that are highly sensitive and capable
`of predicting cardiac dysfunction. In addition, the
`specificity of any test should allow for an accurate
`risk-benefit analysis in balancing the likelihood of
`cardiac dysfunction with greater drug doses against
`the harm that may result from withholding antitu(cid:173)
`mor therapy.
`
`Det ection of Chronic Anthracycline-lnduced
`Cardiomyopathy
`Billingham and colleagues (86) have developed a
`semiquantitative histologic scoring system for endo-
`
`50
`
`1 July 1996 • Annals of lntemaf Medicine • Volume 125 • Number 1
`
`
`
`4 of 12
`
`Celltrion, Inc. 1050
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`Table 1. Continued
`
`Method of Evaluation
`
`Major Abnormality in Diastolic
`Function
`
`Major Abnormality in Systolic
`Function
`
`Relation of Systolic to Diastolic
`Abnormalities
`
`Radionuclide angiocardiography
`
`Decreased peak filling rate
`
`Resting echocardiography
`
`Decreased peak A wave
`
`Decreased left ventricular ejection
`fraction
`Decreased fractional shortening
`
`Resting echocardiography
`
`Resting echocardiography
`
`Increased isovolumic relaxation period; No significant decrease in fractional
`reduced early peak flow velocity and
`shortening
`deceleration rate
`Increased isovolumic relaxation period
`
`Decreased.left ventricular ejection
`fraction
`
`Simultaneous diastolic and systolic
`dysfunction
`Systolic dysfunction p<ecedes
`diastolic dysfunction
`Diastolic dysfunction precedes
`systolic dysfunction
`
`Diastolic dysfunction precedes and
`predicts systolic dysfunction
`
`Resting echocardiography
`
`Increased A/E velocity
`
`Radionuclide angiography
`
`Decreased rapid ventricular filling rate
`
`Inconsistent decrease in fractional
`Diastolic dysfunction; no significant
`shortening
`systolic dysfunction
`No significant decrease in left ventricular Diastolic dysfunction; no significant
`ejection fraction
`systolic dysfunction
`
`myocardial biopsy specimens that correlates well
`with ·cumulative anthracycline dose. Biopsy grade is
`predictive of the rate of early progression around
`the time of therapy and is currently considered the
`most sensitive indicator of chronic anthracycline(cid:173)
`induced cardiotoxicity (87). Underestimation of car(cid:173)
`diac damage with right ventricular biopsy may occur
`because of scattered cardiomyopathic changes (88)
`or the predominance of left ventricular injury (89).
`Also, expertise in obtaining and interpreting biopsy
`specimens is not widely available, and concerns re(cid:173)
`main about the safety of repeated testing, particu(cid:173)
`larly in children (90). Thus, the use of endomyocar(cid:173)
`dial biopsy for the routine monitoring of early
`anthracycline-induced cardiotoxicity has been limited.
`Radionuclide angiocardiography
`is extensively
`used in monitoring for early anthracycline-induced
`cardiotoxicity on the basis of its proven value in
`reducing the incidence of cardiac failure from early
`anthracycline cardiotoxicity (91-93). Having used
`data on serial left ventricular ejection fraction in
`patients with doxorubicin-induced cardiomyopathy,
`Alexander and colleagues (91) suggested that a 15%
`decline in left ventricular ejection fraction or a
`baseline left ventricular ejection fraction less than
`40% identifies patients at high risk for cardiac de(cid:173)
`compensation. Similar findings by others (92, 93)
`have led to the proposal of broad guidelines for
`basing administration of anthracycline therapy on
`RNA findings (94). Unfortunately, resting left ven(cid:173)
`tricular ejection fraction measurements obtained
`through RNA findings are relatively insensitive in
`detecting early anthracycline cardiotoxicity. This is
`largely because no appreciable change in left ven(cid:173)
`tricular ejection fraction occurs until a critieal
`amount of morphologic damage has been done. Af(cid:173)
`ter this point, functional deterioration proceeds rap(cid:173)
`idly. Exercise radibnuclide studies may increase the
`chances of detecting subclinical early anthracycline
`
`cardiotoxicity (26, 95). McKillop and coworkers (96)
`have suggested that the failure to increase ejection
`fraction by 5% over the resting value, as measured
`by stress radionuclide angiocardiography,
`is a
`marker of high risk for the development of early
`anthracycline-induced ventricular dysfunction. How(cid:173)
`ever, the specifh:ity of exercise radionuclide angio(cid:173)
`cardiography is low without serial testing, and max(cid:173)
`imal exercise may be difficult for debilitated patients
`with cancer.
`Two-dimensional echocardiography is the other
`primary noninvasive technique used to monitor an(cid:173)
`thracycline cardiotoxicity, particularly in children
`(97-100). Resting left ventricular ejection fraction
`and fractional shortening are the most commonly
`used echocardiographic parameters. However, as
`with radionuclide measurements, cardiac compensa(cid:173)
`tion in the face of substantial anthracycline-induced
`cardiac injury often maintains normal left ventricu(cid:173)
`lar ejection fraction until the cardiomyopathic changes
`are relatively well established.
`Parameters of diastolic function have also been
`examined for their usefulness in detecting early
`anthracycline-induced cardiac
`injury
`(101-108).
`Chronic anthracycline cardiotoxicity can present
`with substantial fibrous thickening of the endocar(cid:173)
`dium (89), suggesting that diastolic dysfunction is
`probably an impairment co-existing with the disease.
`Marchandise and coworkers (101), in an echocar(cid:173)
`diographic study, found a prolongation of the iso(cid:173)
`volumic relaxation period of 32% (from 65 ms to 86
`ms) an.d a reduction in early peak flow velocity of
`18% (from 60 ms to 49 ms) in anthracycline-treated
`adults before detectable decreases in left ventricular
`shortening fraction, which remained at about 40%
`(101 ). In addition, Stoddard and colleagues (103)
`have reported an increase in the mean isovolumic
`relaxation time from 66 to 84 ms after cumulative
`doxorubicin doses as low as 100 mg/m2 body surface
`
`1 July 1996 • Annals of /nJemal Medicine • Volume 125 • Number 1
`
`51
`
`
`
`5 of 12
`
`Celltrion, Inc. 1050
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`Table 2.
`
`late-Onset Ventricular Function Abnormalities with Anthracydine Treatment•
`
`Study (Reference}
`
`Patients
`
`Anthracycline Dose
`
`Mediastinal
`Radiotherapy
`
`FoOow-up Period after
`Anthracydine
`Treatment
`
`n
`
`201
`
`115
`
`226
`
`57
`
`so
`29
`
`21
`
`19
`
`19
`
`10
`
`65
`
`Steinherz et al. ( 16}
`
`lipshultz et al. (17)
`
`Bu' lock et al. ( 112)
`
`Jakacki et al. (35)
`
`Hausdorf et al. (105)
`
`Leandro et al. (34)
`
`Klev.-er et al. (109)
`
`Yeung et al. (18)
`
`LaMonte et al. (109)
`
`Weesner et al. (110}
`
`lipshultz et al. (84)
`
`• NS • no1 staled.
`t Study incklded a t'Olllrol group.
`
`mglm2
`
`Mean, 450
`Range, 200 to 1275
`Mean,334
`Range, 45 to 550
`
`Range, so to 750
`
`Yes (56 patients)
`
`No
`
`Mean, 7 y
`Range, 4 to 20 y
`Mean, 6.5y
`Range, 1to 15 y
`
`Yes (22 patients)
`
`6.5 mo to 17 mot
`
`Range, :s350
`
`Yes (16 patients)
`
`Range, 2.1 to 20.7 y
`
`Mean, 273
`Range, 31to656
`Range, 200 to 290
`
`Mean, 196
`Range, 27 to 532
`
`Mean,230
`
`Range, 83 to 229
`
`Range, 90 to 498
`
`Range, > 228
`
`No
`
`Yes (8 patients)
`
`NS
`
`NS
`
`Yes (19 patients)
`
`NS
`
`NS
`
`Range, 2.1 to 10.4 y
`
`Mean, 7.2 y
`Range, 1.4 to 16 y
`Mean, 6.0y
`Range, 1.6 to 14.6 y
`
`Mean, 7 y
`Range, 4 to 13 y
`Mean, 3.3 y
`Range, 1.3 to 5.9 y
`Mean, 7y
`Range, 4 to 13 y
`
`Mean, 11.1 y
`
`tomatic children (108). Table 1 summarizes studies
`examining diastolic and systolic abnormalities in
`early anthracycline-induced cardiotoxicity. Despite
`inherent limitations in sensitivity, resting left ven(cid:173)
`tricular ejection fraction based on radionuclide an(cid:173)
`giocardiography or echocardiography remains .the
`most widely used monitoring method for early an(cid:173)
`thracycline-induced cardiotoxicity. The limited sizes
`of these studies (the largest of which involved 60
`patients) underscore the need for larger prospective
`studies to establish the value of diastolic abnormal(cid:173)
`ities in the detection of early anthracycline-induced
`cardiotoxici ty.
`
`area in a prospective study of 26 anthracycline-treated
`adults. These authors have suggested that an increase
`in the isovolurnic relaxation time of more than 37%
`reliably predicts anthracycline-induced systolic dys-
`. function (defined as a decrease in ejection fraction
`of s 10% to <55% ). Other echocardiographic diastolic
`parameters, such as rapid and slow filling velocities
`(103), have also been reported to precede anthracy(cid:173)
`cline-induced systolic dysfunction in adults. Reductions
`in peak filling rate, a diastolic parameter measured
`by multigated angiocardiography in adults, occur be(cid:173)
`fore anthracycline-induced decrements in left ven(cid:173)
`tricular ejection fraction (104). However, a recent
`report on atithracycline-treated adults (107) suggests
`that reductions in peak filling rate and ejection frac-
`Detectio n of Late-Onset Anthracycline-lnduced
`Cardiotoxicity
`tion measured by radionuclide studies appear simul-
`Recognition of late-onset cardiac dysfunction and
`taneously. The utility of diastolic filling abnorrnali-
`failure has highlighted the need for monitoring
`ties to unmask early anthracycline cardiotoxicity in
`children also needs to be confirmed (105-108). methods that will help detect and predict long-term
`cardiac impairment in asymptomatic patients. Cur-
`Fractional shortening has been reported to be more
`sensitive than diastolic parameters in children with
`rently, there are few universalJy accepted guidelines
`early anthracycline-induced cardiomyopathy (106).
`on which tests of ventricular function should di-
`rectly influence long-term monitoring for late-on-
`In contrast, a recent smaller prospective study has
`set cardiotoxicity in asymptomatic adults, although the
`shown the value of diastolic indices in detecting
`issue has been extensively debated in children (97-99).
`early anthracycline-induced cardiomyopathy in asymp-
`
`52
`
`1 July 1996 • Annals of Internal Medicine • Volume 125 • Number 1
`
`
`
`6 of 12
`
`Celltrion, Inc. 1050
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`Table 2. Continued
`
`Method of Evaluation
`
`Diastolic Abnormalities
`
`Systolic Abnormalities
`
`Resting echocardiography, cardiac
`biopsy
`Resting echocardiography
`
`NS
`
`NS
`
`Resting echocardiography
`
`Peak E deceleration and acceleration;
`increased isovolumic relaxation period
`
`Resting echocardiography;
`multigated acquisition nudear
`scant
`
`NS
`
`Fractional shortening decrease (47 patients)
`Abnormal endomyocardial biopsy (9 patients)
`End-systolic wall stress increase or decreased left
`ventricular posterior wall thickness or both
`(60 patients)
`Fractional shortening decrease (32 patients)
`Abnormal fractional shortening (<30%)
`(51 patients)
`Borderline fractional shortening (30%-34%)
`(85 patients)
`Fractional shortening decrease (5 of 54 patients)
`Abnormal multigated acquisition nuclear scan
`(9 of 30 patients)
`
`Resting echocardiographyt
`
`Decreased rapid diastolic filling
`
`End-systolic wall stress increase
`
`Resting echocardiographyt
`
`Decreased early peak filling velocity
`
`End-systolic wall stress increase
`
`Dobutamine stress
`echocardiographyt
`
`Exercise stress echocardiographyt
`
`Resting echocardiography;
`radionudide angiocardiography
`
`Exercise stress echocardiography
`
`Resting echocardiography
`
`None
`
`NS
`
`NS
`
`NS
`
`" Features of restrictive cardiomyopathy";
`decreased left ventricular mass:
`volume ratio and left ventricular
`end-diastolic diameter
`
`Poor increase in end-systolic wall stress;
`fractional shortening increase with inotropic
`stress ·
`
`Poor increase in fractional shortening with
`exercise
`Decrease in resting left ventricular ejection
`fraction (3 patients)
`Poor increase in left ventricular ejection fraction
`with stress (2 patients)
`Poor increase in fractional shortening with
`exercise
`Afterload increase (1 patient)
`Fractional shortening decrease (2 patients)
`
`Total Patients with
`Ventricular Abnormalities
`
`n("lo)
`
`47 (23)
`
`66 (57)
`
`Systolic: 51 (22) to 136 (60)
`Diastolic: NS
`
`NS
`
`NS
`
`NS
`
`NS
`
`NS
`
`5 (26)
`
`NS
`
`NS
`
`Llpshultz and coworkers (17) have reported that frac(cid:173)
`tional shortening as assessed by echocardiography has
`a sensitivity of 64% and a specificity of 81 % for either
`abnormal contractility or abnormal afterload at long(cid:173)
`term follow-up (17). Congruent with these findings are
`the results of Steinherz and colleagues (16). In a study
`of 201 children treated with anthracycline, they found
`that only 29% of patients with mildly abnormal or
`worse fractional shortening (:528%) on an "end-ther(cid:173)
`apy'' echocardiogram had normal fractional shortening
`(;::::29%) at late follow-up. Thus, abnormal fractional
`shortening after therapy may help predict long-term
`cardiac dysfunction. In contrast, 87% of patients with
`normal fractional shortening during the first year after
`therapy have maintained normal fractional shortening
`at late follow-up (16). However, most late-onset clin(cid:173)
`ical decompen8ations occur after more than 10 years
`of follow-up, and data from beyond that time point
`are limited. Ec