Journal of the American College of Cardiology Vol. 36, No.

2, 2000
© 2000 by the American College of Cardiology ISSN 0735-1097/00/$20.00
Published by Elsevier Science Inc. PII S0735-1097(00)00748-8

Left Ventricular Dysfunction Predicted by Early

Troponin I Release After High-Dose Chemotherapy
Daniela Cardinale, MD, Maria Teresa Sandri, MD,† Alessandro Martinoni, MD, Alessio Tricca, LabTech,†
Maurizio Civelli, MD, Giuseppina Lamantia, MD, Saverio Cinieri, MD,* Giovanni Martinelli, MD,*
Carlo M. Cipolla, MD, Cesare Fiorentini, MD
Milan, Italy
OBJECTIVES We investigated the role of cardiac troponin I (cTnI) in patients with aggressive malignancies
treated with high-dose chemotherapy (HDC).
BACKGROUND High dose chemotherapy is potentially limited by cardiac toxicity. Considering the fact that
cardiac dysfunction may become clinically evident weeks or months after HDC, the
availability of an early marker of myocardial injury, able to predict late ventricular impairment,
is a current need.
METHODS We measured, in 204 patients (45 ⫾ 10 years) affected by cancer resistant to conventional
treatment, the cTnI plasma concentration after every single cycle of HDC. According to the
cTnI value (ⱕ or ⬎0.4 ng/ml), patients were divided into a troponin positive (cTnI⫹, n ⫽
65) and a troponin negative (cTnI⫺, n ⫽ 139) group. All patients underwent echocardio-
graphic examination during the following seven months.
RESULTS In the cTnI⫺ group, left ventricular ejection fraction (LVEF) progressively decreased after
HDC, reaching a maximal reduction after three months; however, myocardial depression was
transient and no longer detectable at later follow-up. By contrast, in the cTnI⫹ group LVEF
reduction was more marked and still evident at the end of the follow-up. In cTnI⫹ patients,
a close relationship between the short-term cTnI increment and the greatest LVEF reduction
was found (r ⫽ ⫺0.87, p ⬍ 0.0001).
CONCLUSIONS The elevation of cTnI in patients undergoing HDC for aggressive malignancies accurately
predicts the development of future LVEF depression. In this setting, cTnI can be considered
a sensitive and reliable marker of acute minor myocardial damage with relevant clinical and
prognostic implications. (J Am Coll Cardiol 2000;36:517–22) © 2000 by the American
College of Cardiology

High-dose chemotherapy (HDC) is a therapeutic approach oncological therapeutic adjustments and supporting cardio-
for several aggressive malignancies resistant to the tradi- logical treatment could be required.
tional chemotherapy schedules (1–3). This kind of treat- Cardiac troponin I (cTnI) is a new specific marker of
ment can favorably affect survival of patients with cancer minor myocardial damage released by cardiac cells in pro-
disease; however, its use is limited by considerable side portion to the degree of myocardial injury (11). While its
effects, in particular cardiotoxicity (4,5). Beyond early car- role in acute coronary syndromes is well appreciated, the
diotoxicity, which occurs during or soon after treatment (6), possible application of this peptide in the detection of
the development of heart failure many years after the last HDC-induced cardiac damage, as well as in the short- and
administration of chemotherapeutic drugs is increasingly long-term risk stratification of cancer patients undergoing
recognized (7,8). Cardiac involvement may become clini- this kind of treatment, has never been investigated before.
cally manifest late in the course of the natural history of the
disease and lead to overt heart failure. Moreover, with the METHODS
increasing availability of echocardiography, it has become
evident that chemotherapy-induced left ventricular impair- Study population. All consecutive patients undergoing
ment often occurs without symptoms, and, though it is HDC for aggressive malignancies who were evaluated at our
generally considered to be irreversible, some reports on Institute between August 1, 1997 and November 1, 1998
complete recovery of cardiac dysfunction have been reported were asked to participate in the study. Patients with a
(9,10). Hence, the possibility of identifying an early marker history of ischemic, valvular and hypertensive heart disease,
of cardiac injury, able to predict late ventricular dysfunction uncontrolled hypertension, left ventricular ejection fraction
after HDC, remains a stimulating incentive. This would (LVEF) ⬍50%, acute or chronic renal insufficiency (serum
permit clinicians to identify higher-risk patients needing a creatinine ⬎ 1.5 mg/dl) and liver disease (bilirubin ⬎ 2.0
close monitoring of cardiovascular function and in which mg/dl, AST ⬎ two times the upper limit of normal) were
excluded from the study.
Of the 232 patients screened, 204 patients (165 women,
From the Cardiology Unit, *Hematoncology Unit, and †Laboratory Medicine, 39 men, mean age 45 ⫾ 10 years) qualified for inclusion in
Istituto Europeo di Oncologia, University of Milan, Milan, Italy.
Manuscript received October 25, 1999; revised manuscript received February 11, the study. The indications for HDC were advanced or
2000, accepted March 30, 2000. primary resistant breast cancer in 133 cases, relapsed or
518 Cardinale et al. JACC Vol. 36, No. 2, 2000
Troponin I After High Dose Chemotherapy August 2000:517–22

creatine kinase, MB fraction (CK-MB) mass were measured

Abbreviations and Acronyms before, immediately after and then 12, 24, 36 and 72 h after
CK ⫽ creatine kinase every single cycle of HDC.
CK-MB ⫽ creatine kinase, MB fraction Cardiac function was assessed by echocardiography and
cTnI ⫽ cardiac troponin I
EDV ⫽ end-diastolic volume
electrocardiography. In all patients LVEF (biplane method,
ESV ⫽ end-systolic volume according to Simpson’s rule), end-diastolic (EDV) and
HDC ⫽ high-dose chemotherapy end-systolic (ESV) volumes were evaluated before begin-
LVEF ⫽ left ventricular ejection fraction ning HDC, one month, two months, three months, four
months and seven months after the end of the treatment.
Electrocardiogram was performed before and after each
refractory ovarian carcinoma in 12, small cell lung cancer in
HDC cycle and then at all follow-up checks. The duration
8, high grade non-Hodgkin’s lymphoma in 46 and relapsed
of the follow-up was nine months from the baseline evalu-
or refractory poor prognosis Hodgkin’s disease in 5. Previ-
ation for most patients and 10 months for patients under-
ous treatment with anthracyclines or radiotherapy was
going “sequential” therapy. During this period, no one was
present in 111 and 13 cases, respectively. Twenty-one
patients were receiving calcium antagonist agents for mild treated with other chemotherapeutic drugs or radiotherapy,
hypertension; none was receiving beta-adrenergic blocking and all patients’ management decisions were made without
agents, angiotensin-converting enzyme inhibitors or diuret- the knowledge of the patient’s cTnI results.
ics. Laboratory methods. The blood samples for cTnI deter-
Informed consent was obtained from all patients before mination were centrifuged within 60 min, and plasma was
participation in the study, and the protocol was approved by stored immediately at ⫺30°C. Cardiac TnI concentrations
the Ethical Committee of our Institution. were determined by an immunoenzymatic fluorescent assay
Study protocol (Fig. 1). Clinical examination, electrocar- (Stratus II, Dade International Inc., Miami, Florida) that
diogram, chest x-ray and echocardiogram were part of the uses two monoclonal antibodies specific for independent
preliminary evaluation. All 204 eligible patients who passed epitopes of cTnI (13) with a lower limit of detection of
the inclusion and exclusion criteria received HDC in dif- 0.35 ng/ml; the considered cut off level was 0.5 ng/ml. All
ferent drug combinations, according to our Institute’s med- blood samples were analyzed in duplicate with an interassay
ical oncology protocols (Table 1). All drugs were adminis- variability ⱕ0.1 ng/ml, and all “positive” samples were
tered intravenously via central venous catheters. In addition, immediately retested to confirm the result obtained. Total
all patients received reinfusion of autologous peripheral CK activity was determined by an enzymatic method, and
blood progenitor cells in order to accelerate hematopoietic CK-MB mass concentration was measured using a com-
recovery and reduce supportive care requirement (1,12). mercially available immunoabsorbent assay (Stratus II,
None of the patients received contemporary radiotherapy. Dade International Inc.). The upper limit of the reference
Plasma cTnI concentration, creatine kinase (CK) and interval is 190 U/L for CK and 5 ng/ml for CK-MB mass.

Figure 1. Study design. Open square ⫽ clinical and echocardiographic examination; Open circle ⫽ troponin I, creatine kinase (CK) and CK-MB mass
determination. EC ⫽ epirubicin and cyclophosphamide; ICE ⫽ ifosfamide, carboplatin and etoposide; SEQ ⫽ sequential therapy (see Table 1); TEC ⫽
taxotere, epirubicin and cyclophosphamide; TICE ⫽ taxotere, ifosfamide, carboplatin and etoposide.
JACC Vol. 36, No. 2, 2000 Cardinale et al. 519
August 2000:517–22 Troponin I After High Dose Chemotherapy

Statistical analysis. The values of LVEF were analyzed

Cycles

4*
using a repeated measures model taking into account the

3
3
3
3

*One drug for each cycle. EC ⫽ epirubicin-cyclophosphamide; ICE ⫽ ifosfamide-carboplatin-etoposide; SEQ ⫽ sequential; TEC ⫽ taxotere-epirubicin-cyclophosphamide; TICE ⫽ taxotere-ifosfamide-carboplatin-etoposide.
correlation among the time periods with an unstructured
covariance matrix. Time was treated as a factor, and the

45 mg/m2
1.2 g/m2
interaction between time and troponin value was included in
Dosage the model. We also included the linear effects of the baseline
troponin value and its interactions. All interactions were
tested using F tests based upon Type 3 sums on squares.
Least squares means and corresponding confidence intervals

Idarubicin
Etoposide of time by troponin value interactions were calculated. All
Drug

calculations were performed using PROC MIXED in SAS

(SAS for Windows, version 6.11; SAS Institute Inc., Cary,
North Carolina). Differences in LVEF values at baseline
1.2 g/m2
1.2 g/m2
Dosage

between the two groups were analyzed using a generalized

4 g/m2

2 g/m2

linear model. A generalized linear model was also used to

investigate differences in the maximum percentage change
in EDV, ESV and LVEF. A square root transformation was
used to achieve normality and homogeneity of variance of
Cyclophosphamide

the maximum percent changes.

Drug

Linear regression analysis was used for obtaining corre-

Carboplatin
Etoposide

Etoposide

lation coefficients. Significance was taken at the 5% level.

Results are presented as mean ⫾ standard deviation unless
otherwise specified.
200 mg/m2
1.2 g/m2

RESULTS
Dosage
2

10 g/m2
8 g/m2
4 g/m

All patients underwent HDC without clinically evident

acute cardiological side effects during or soon after the
drugs’ administration. Three patients developed overt heart
failure during the follow-up (four, six and seven months
Cyclophosphamide

after HDC).
Drug

Methotrexate

At the baseline evaluation, as well as before each cycle of

Carboplatin
Epirubicin

Ifosfamide

HDC, cTnI value was within the normal range in all cases.
Increment in circulating cTnI was detected in 65 (32%) of
the 204 treated patients and, when the total number of
HDC cycles was considered, in 112/661 (17%) cycles. In
particular, in the 112 cycles in which cTnI increased, in 59
2

mg/m2

mg/m2
mg/m
Dosage

g/m2

g/m2

(53%) cases the first abnormal cTnI value was observed soon
Table 1. High Dose Chemotherapeutic Schedules Utilized

after the end of drug(s) administration, in 10 (9%) after

200
85
10
85
7

12 h, in 21 (19%) after 24 h, in 8 (7%) after 36 h and in 14

(12%) after 72 h. When the whole population, as well as
every single HDC schedule, was considered, the percentage
Cyclophosphamide

of cTnI positivity progressively increased in parallel with the

Drug

increasing number of the cycles performed.

Epirubicin

Ifosfamide

No change in CK serum levels was detected, whereas

Taxotere

Taxotere

CK-MB concentration increased in three cases (5.9, 5.7 and

6.3 ng/ml, respectively). In these three patients, the maxi-
mal value of cTnI detected was 1.9, 1.9 and 2.0 ng/ml,
Patients (n)

respectively.
In all patients no significant electrocardiographic changes
50
41
48
16
49

were observed, both after HDC and at the follow-up

checks.
Patients were allocated to two subgroups according to the
Acronym

maximal cTnI value detected after HDC: the troponin

TICE
TEC

SEQ
ICE

negative group (cTnI⫺; n ⫽ 139) and the troponin positive

EC

group (cTnI⫹, n ⫽ 65; mean value 1.0 ⫾ 0.5 ng/ml; range

520 Cardinale et al. JACC Vol. 36, No. 2, 2000
Troponin I After High Dose Chemotherapy August 2000:517–22

Table 2. Clinical Charateristics in Troponin Positive (cTnI⫹)

and Troponin Negative (cTnI⫺) Groups
cTnIⴙ cTnIⴚ
(n ⴝ 65) (n ⴝ 139)
Age (yrs) 44 ⫾ 10 45 ⫾ 11
Gender (male/female) 14/51 25/114
Prior anthyracyclines therapy 46 (71%) 65 (46%)*
Prior radiotherapy 4 (6%) 9 (6%)
Hypertension 6 (9%) 15 (11%)
Disease
Breast cancer 42 (65%) 91 (65%)
Ovarian carcinoma 4 (6%) 8 (6%)
Small cell lung cancer 3 (5%) 5 (4%)
Non-Hodgkin’s lymphoma 15 (23%) 31 (22%)
Hodgkin’s disease 1 (2%) 4 (3%)
Treatment
EC 17 (26%) 33 (24%)
Figure 2. Left ventricular ejection fraction (LVEF) at baseline and during
TEC 14 (22%) 27 (19%) the seven months of follow-up of troponin I positive (cTnI⫹; solid circle)
ICE 14 (22%) 34 (24%) and negative (cTnI⫺; solid square) patients. *p ⬍ 0.001 vs. baseline
TICE 5 (8%) 11 (8%) (month 0); §p ⬍ 0.001 vs. cTnI⫺ group. Data are shown as mean ⫾ 95%
SEQ 15 (23%) 34 (24%) confidence interval.
Data are mean ⫾ SD. *Chi-square p ⬍ 0.05.
EC ⫽ epirubicin-cyclophosphamide; ICE ⫽ ifosfamide-carboplatin-etoposide; in EDV and in ESV observed during the follow-up in the
SEQ ⫽ sequential; TEC ⫽ taxotere-epirubicin-cyclophosphamide; TICE ⫽ two groups, regardless of the moment in which they were
taxotere-ifosfamide-carboplatin-etoposide.
detected. The changes from the baseline value of these
parameters were significant in both groups; however, the
0.5–2.0). The cTnI⫹ group was defined by a value equal to variations were significantly greater in the cTnI⫹ than in
or greater than 0.5 ng/ml at least at one of the points of the cTnI⫺ group.
measurement considered, while the cTnI⫺ group was de- When the incidence of cTnI⫹ patients was considered
fined by a value ⬍0.5 ng/ml in every determination. Table according to the degree of left ventricular depression ob-
2 gives the clinical characteristics of the two populations. No served at the follow-up, the percentage of cTnI⫹ patients
difference was observed in regards to age, gender, kind of increased in parallel with the degree of LVEF reduction, up
neoplasm, HDC schedule or other clinical characteristics to a 100% incidence for a decrease in LVEF greater than
between the two groups, except for previous treatment with 30%.
anthracyclines that was more frequent in the cTnI⫹ group. In the cTnI⫹ group, a strong relationship between the
The time elapsed from previous anthracycline treatment to cTnI maximal value and the LVEF maximal reduction was
the enrollment in the study was similar in the two groups found (r ⫽ ⫺0.87; p ⬍ 0.0001; Fig. 4).
(range: 2– 6 months).
At the baseline evaluation, LVEF, EDV and ESV were DISCUSSION
similar in the two groups and in all cases within the normal
limits. After HDC, there was a significantly different Many recent studies have elucidated the value of both
pattern of LVEF in the cTnI⫹ compared with the cTnI⫺ troponin T and I in the diagnosis and in the risk assessment
group (p ⬍ 0.0001) (Fig. 2). Among the patients in the of acute coronary syndromes (14 –17). Our study addressed
cTnI⫹ group, there was evidence of a significant reduction
in LVEF from three months onwards. Indeed, LVEF
impairment was still evident at the end of the follow-up. In
the cTnI⫺ group there was also a significant reduction in
LVEF at three months, which was not as great as the
reduction in the cTnI⫹ group. This transient decrease was
followed by a recovery to baseline levels at four and seven
months. In particular, during the entire follow-up, an
LVEF value less than 50% was observed in 19/65 (29%)
cTnI⫹ and in 0/139 cTnI⫺ patients, respectively (chi-
square p ⬍ 0.001). The three patients developing symptoms
of heart failure during the follow-up have had positive value
of both cTnI and CK-MB after HDC and an LVEF ⬍30% Figure 3. Maximal percent changes in end-diastolic volume (EDV),
at the last evaluation before symptom onset. Cardiovascular end-systolic volume (ESV) and left ventricular ejection fraction (LVEF)
observed during the follow-up in the two groups of patients. cTnI⫹ ⫽
treatment was required only in these three patients. cardiac troponin I ⬎ 0.4 ng/ml; cTnI⫺ ⫽ cardiac troponin I ⱕ 0.4 ng/ml.
Figure 3 shows the maximal percent of changes in LVEF, Data are shown as mean ⫾ 95% confidence interval.
JACC Vol. 36, No. 2, 2000 Cardinale et al. 521
August 2000:517–22 Troponin I After High Dose Chemotherapy

sensitive, as well as a simple and low-cost, method for the

identification of early, and possibly reversible, cardiotoxicity.
Hypothetical mechanism(s) of cardiotoxicity. Because of
the combination of different drugs, the mechanisms by
which HDC may generate cardiotoxicity are multifactorial
and complex. They include production of free oxygen
radicals, disturbance of the mitochondrial energy metabo-
lism, intracellular calcium overloading, increase in lipid
peroxidation, etc. (27–29). In addition, cardiac toxic effects
deriving from the combination of different drugs, some of
which are not classically considered to have cardiotoxic
properties, such as ifosfamide and etoposide, cannot be
excluded. Indeed, a similar dose-dependent cTnI release
was observed in our study whatever the drug composition
Figure 4. Scatterplot of left ventricular ejection fraction (LVEF) changes
considered and, in previous reports, congestive heart failure
against troponin I value in cTnI⫹ patients. cTnI ⫽ cardiac troponin I. has been described as occurring when intermediate cumu-
lative doses with multiagent chemotherapy were utilized
the question of whether cTnI measurement gives significant (30).
information in a subgroup of patients with aggressive Comparison with previous studies. Previous animal stud-
malignancies resistant to conventional chemotherapic regi- ies have consistently shown that troponin T is released into
mens and undergoing HDC, in relation to the possible the circulation after anthracycline administration and that
direct cardiotoxic effect (4,5). This study shows that cTnI is the elevated troponin levels were correlated with clinical
a risk marker for future development of significant and toxicity (31). Similar observations were reported in children
prolonged left ventricular dysfunction. This issue is partic- receiving anthracycline chemotherapy (32). In addition,
ularly important in patients with cancer disease in which Missov et al. (33) described cTnI increase during the course
onset of cardiac dysfunction, even asymptomatic, impor- of anthracycline chemotherapy in patients with hematolog-
ical malignancies. The small release of cTnI indicates that
tantly limits the therapeutical opportunities and negatively
only a minimal acute necrosis occurs during HDC, as
influences the prognosis of these patients (18,19). Indeed,
compared with that observed in acute coronary syndromes.
HDC has contributed considerably to the improvement of
However, the clinical interest of this cTnI increment is quite
survival of patients with poor-prognosis cancer (2,3,20), and
relevant. Indeed, in addition to the impairment in systolic
cardiotoxicity represents a major limitation to the total dose
cardiac function predicted by cTnI elevation, a close rela-
that can safely be administered (4 –5). Hence, the possibility
tionship between the cTnI maximal value observed after
of identifying patients at higher risk of developing late
HDC and the degree of late LVEF reduction was observed
myocardial function depression could permit clinicians to
(Fig. 4). This finding strongly amplifies the clinical signif-
modify or to discontinue the oncologic regimen, to support icance of cTnI in order to individuate patients who will
cardiac function with cardiovascular therapy or cardiopro- develop late cardiac impairment and, to a greater extent, to
tective agents (21,22) and to accurately monitor the pro- predict the degree of the future left ventricular dysfunction.
gression of cardiac damage. This last point is particularly Clinical implications. As our group of patients was less
relevant, considering that a progressive and cumulative symptomatic, the detection of elevated cTnI levels would
cardiotoxicity occurs during HDC and that the risk of seem to substantially increase the information regarding the
developing cardiac dysfunction, as revealed by troponin risk of heart failure. This information could not be obtained
positivity, increases in parallel with the number of cycles of by conventional criteria such as symptoms, electrocardio-
HDC completed. Our data confirm previous studies report- graphic and echocardiographic changes. The rise in cTnI
ing that cardiotoxicity, due to anthracyclines, is cumulative indicates minor myocardial damage, which seems to precede
and dose-related (6,8,18,19); further evidence derives from left ventricular systolic impairment. The time course of this
the observation that, in our population, previous treatment cardiac damage is unclear, and further studies are needed to
with anthracyclines was more frequently associated with clarify whether patients with acute minor myocardial injury,
cTnI positivity (Table 2). Furthermore, increasing evidence attested by cTnI increment and prolonged LVEF reduction,
that cardiotoxicity can be reversible, at least in its early will develop an irreversible dilated cardiomyopathy. On the
phases, exist (9,10). Accordingly, several methods of iden- other hand, normal cTnI values, after HDC, seem to
tifying cardiotoxicity before it causes irreversible depression identify patients at lower risk in which no cardiac damage,
of myocardial function had been proposed. Among these or only transient subclinical dysfunction, occurs. Finally, the
were: monoclonal antimyosin antibodies imaging (23,24), three patients who developed overt heart failure during the
study of cardiac autonomic function (25) and endomyocar- follow-up of our study had, in addition to cTnI elevation,
dial biopsy (26). In comparison, cTnI appears to be a also CK-MB positivity. This finding emphasizes the clinical
522 Cardinale et al. JACC Vol. 36, No. 2, 2000
Troponin I After High Dose Chemotherapy August 2000:517–22

importance of cardiac enzymes in revealing the extension of 12. Elias AD, Ayash L, Anderson KC, et al. Mobilization of peripheral
blood progenitor cells by chemotherapy and granulocyte-macrophage
acute myocardial necrosis during HDC. Therefore, the
colony-stimulating factor for hematologic support after high-dose
relationship between extension of enzyme elevation, both intensification for breast cancer. Blood 1992;79:3036 – 44.
cTnI and CK-MB, and cardiac damage allows us to 13. Bodor GS, Porter S, Landt Y, Ladenson JH. Development of
anticipate a continuum spectrum of future clinical events, monoclonal antibodies for an assay of cardiac troponin I and prelim-
inary results in suspected cases of myocardial infarction. Clin Chem
namely only transient left ventricular impairment in cTnI⫺ 1992;38:2203–14.
patients, prolonged and more marked cardiac dysfunction in 14. Hamm CW, Ravkilde J, Gerhardt W, et al. The prognostic value of
cTnI⫹ patients and symptomatic heart failure in patients serum troponin T in unstable angina. N Engl J Med 1992;327:146 –
with positivity of both cTnI and CK-MB. 50.
15. Lindahl B, Venge P, Wallentin L. Relation between troponin T and
Conclusions. Elevation of cTnI in cancer patients under- the risk of subsequent cardiac events in unstable coronary artery
going HDC for aggressive malignancies accurately predicts disease. Circulation 1996;93:1651–7.
the development of ventricular systolic dysfunction in the 16. Ohman EM, Armstrong PW, Christenson RH, et al. for the
months following. In such a patient population, cTnI can be GUSTO-IIa Investigators. Cardiac troponin T level for risk stratifi-
cation in acute myocardial ischemia. N Engl J Med 1996;335:1333–
considered a sensitive and reliable marker of acute minor 41.
cardiac damage with relevant clinical and prognostic impli- 17. Luscher MS, Thygesen K, Ravkilde J, Heickendorff L, for the TRIM
cations. Study Group. Applicability of cardiac troponin T and I for early risk
stratification in unstable coronary artery disease. Circulation 1997;96:
2578 – 85.
Acknowledgments 18. De Forni M, Armand JP. Cardiotoxicity of chemotherapy. Curr
We would like to express our gratitude to Dr. Sara Gandini Opinion Oncol 1994;6:340 – 4.
of the Division of Epidemiology and Biostatistics for 19. Feenstra J, Grobbee DE, Remme WJ, Stricker BHC. Drug-induced
heart failure. J Am Coll Cardiol 1999;33:1152– 62.
assistance with the statistical analysis and to Mrs. Angela
20. Gianni AM, Siena S, Bregni M, et al. Prolonged disease-free survival
Cocquio and the nursing staff of the Hematoncology Unit after high-dose sequential chemoradiotherapy and hematopoietic au-
for their help in the research protocol. tologous transplantation in poor prognosis Hodgkin’s disease. Ann
Oncol 1991;2:645–53.
21. Speyer JL, Green MD, Kramer E, et al. Protective effect of the
Reprint requests and correspondence: Dr. Daniela Cardinale,
bispiperazinedione ICRF-187 against doxorubicin-induced cardiac
Cardiology Unit, Istituto Europeo di Oncologia, Via Ripamonti toxicity in women with advanced breast cancer. N Engl J Med
435, 20141 Milan, Italy. E-mail: daniela.cardinale@ieo.it. 1988;319:745–52.
22. Schmitt-Graff A, Scheulen ME. Prevention of adriamycin cardiotox-
icity by niacin, isocitrate or N-acetyl-cysteine in mice. A morpholog-
REFERENCES ical study. Pathol Res Pract 1986;181:168 –74.
23. Carrio I, Estorch M, Berna L, et al. Assessment of anthracycline-
1. Coiffier B, Philip T, Burnett AK, Symann M. Consensus conference induced myocardial damage by quantitative indium-111 myosin-
on intensive chemotherapy plus hematopoietic stem-cell transplanta- specific monoclonal antibody studies. Eur J Nucl Med 1991;18:806 –
tion in malignancies: Lyon, France, June 4 – 6, 1993. J Clin Oncol 12.
1994;12:226 –31. 24. Yamada T, Matsumori A, Tamaki N, et al. Detection of adriamycin
2. Ayash LJ, Wheeler C, Fairclough D, et al. Prognostic factors for cardiotoxicity with indium-111 labeled antimyosin monoclonal anti-
prolonged progression-free survival with high-dose chemotherapy
body imaging. Jpn Circ J 1991;55:377– 83.
with autologous stem-cell support for advanced breast cancer. J Clin
25. Viniegra M, Marchetti M, Losso M, et al. Cardiovascular autonomic
Oncol 1995;13:2043–9.
3. Bezwoda WR, Seymour L, Dansey RD. High-dose chemotherapy function in anthracycline-treated breast cancer patients. Cancer Che-
with hematopoietic rescue as primary treatment for metastatic breast mother Pharmacol 1990;23:227–31.
cancer: a randomized trial. J Clin Oncol 1995;13:2483–9. 26. Billingham ME, Bristow MR. Evaluation of anthracycline cardiotox-
4. Basser RL, Abraham R, Bik To L, Fox RM, Green MD. Cardiac icity: predictive ability and functional correlation of endomyocardial
effects of high-dose epirubicin and cyclophosphamide in women with biopsy. Cancer Treat Symp 1984;3:71– 6.
poor prognosis breast cancer. Ann Oncol 1999;10:53– 8. 27. Fu LX, Waagstein F, Hjalmarson A. A new insight into adriamycin-
5. Shapiro CL, Ervin T, Welles L, Azarnia N, Keating J, Hayes DF, for induced cardiotoxicity. Int J Cardiol 1990;29:15–20.
the TLC D-99 Study Group. Phase II trial of high dose liposome- 28. Galaris D, Georgellis A, Rydstrom J. Toxic effects of daunorubicin on
encapsulated doxorubicin with granulocyte colony-stimulating factor isolated and cultured heart cells from neonatal rats. Biochem Pharma-
in metastatic breast cancer. J Clin Oncol 1999;17:1435– 41. col 1985;34:989 –95.
6. Rhoden W, Hasleton P, Brooks N. Anthracyclines and the heart. Br 29. Ganey PE, Carter LS, Mueller RA, Thurman RG. Doxorubicin
Heart J 1993;70:499 –502. toxicity in perfused rat heart. Decrease cell death at low oxygen
7. Freter CE, Lee TC, Billingham ME, Chak L, Bristow MR. Doxo- tension. Circ Res 1991;68:1610 –3.
rubicin cardiac toxicity manifesting seven years after treatment. Case 30. Watts RG. Severe and fatal anthracycline cardiotoxicity at cumulative
report and review. Am J Med 1986;80:483–5. doses below 400 mg/m: evidence for enhanced toxicity with multiagent
8. Steinherz LJ, Steinherz PG, Tan CTC, Heller G, Murphy ML. chemotherapy. Am J Hematol 1991;36:217– 8.
Cardiac toxicity 4 to 20 years after completing anthracycline therapy.
31. Herman EH, Lipshultz SE, Rifai N, et al. Use of cardiac troponin T
JAMA 1991;266:1672–7.
levels as an indicator of doxorubicin-induced cardiotoxicity. Cancer
9. Saini J, Rich MW, Lyss AP. Reversibility of severe left ventricular
dysfunction due to doxorubicin cardiotoxicity. Report of three cases. Res 1998;58:195–7.
Ann Intern Med 1987;106:814 – 6. 32. Lipshultz SE, Rifai N, Sallan SE, et al. Predictive value of cardiac
10. Cohen M, Kronzon I, Lebowitz A. Reversible doxorubicin-induced troponin T in pediatric patients at risk for myocardial injury. Circu-
congestive heart failure. Arch Intern Med 1982;142:1570 –1. lation 1997;96:2641– 8.
11. Mair J, Wagner I, Morass B, et al. Cardiac troponin I release correlates 33. Missov E, Calzolari C, Davy JM, Leclercq F, Rossi M, Pau B. Cardiac
with myocardial infarction size. Eur J Clin Chem Clin Biochem troponin I in patients with hematologic malignancies. Coron Artery
1995;33:869 –72. Dis 1997;8:537– 41.