Prognostic Value of Troponin I in Cardiac Risk
Stratification of Cancer Patients Undergoing
Daniela Cardinale, MD; Maria T. Sandri, MD; Alessandro Colombo, MD; Nicola Colombo, MD;
Marina Boeri, MD; Giuseppina Lamantia, MD; Maurizio Civelli, MD; Fedro Peccatori, MD;
Giovanni Martinelli, MD; Cesare Fiorentini, MD; Carlo M. Cipolla, MD
Background—In patients with aggressive malignancies who are undergoing high-dose chemotherapy, even minimal
elevation of troponin I (TnI) is associated with late left ventricular dysfunction. The time course of the subclinical
myocardial damage and its impact on the clinical outcome have never been investigated previously.
Methods and Results—In 703 cancer patients, we measured TnI soon after chemotherapy (early TnI) and 1 month later
(late TnI). Troponin was considered positive for values ?0.08 ng/mL. Clinical and left ventricular ejection fraction
evaluation (echocardiography) were performed before chemotherapy, 1, 3, 6, and 12 months after the end of the
treatment, and again every 6 months afterward. Three different TnI patterns were identified, and patients were grouped
accordingly. In 495 patients, both early and late TnI values were ?0.08 ng/mL (TnI?/?group); in 145, there was only
an early increase (TnI?/?group); and in 63 patients, both values increased (TnI?/?group). In the TnI?/?group, no
significant reduction in ejection fraction was observed during the follow-up, and there was a very low incidence of
cardiac events (1%). In contrast, a greater incidence of cardiac events occurred in TnI-positive patients, particularly in
the TnI?/?group (84% versus 37% in the TnI?/?group; P?0.001).
Conclusions—TnI release pattern after high-dose chemotherapy identifies patients at different risks of cardiac events in the
3 years thereafter. This stratification allows us to differentiate the monitoring program and to plan, in selected patients,
preventive strategies aimed at improving clinical outcome. (Circulation. 2004;109:2749-2754.)
Key Words: troponin ? chemotherapy ? ventricles ? cardiac toxicity
chronic cardiotoxicity, can range from transient asymptom-
atic left ventricular dysfunction to cardiac death.1–3Its inci-
dence varies according to different clinical definitions but has
been reported to be as high as 65%.4The magnitude of the
problem is rising as a result of the increasing number of
long-term cancer survivors and because of the tendency to
use higher doses of anthracyclines (AC), as well as combined
treatments with synergistic cardiac toxic effects.1,5–7
Extensive and expensive monitoring programs are usually rec-
ommended to identify patients who develop cardiac toxicity.6,8–11
However, most methods used in clinical practice (echocardiogra-
phy, radionuclide angiocardiography, etc), have low sensitivity and
poor predictive value or, like endocardial biopsy, have specific
Troponin I (TnI) is a protein present exclusively in the
myocardial cells. The TnI plasma concentration is a well-
established specific and sensitive marker of myocardial in-
jury, with both high diagnostic and prognostic value.14
ardiotoxicity is a common complication of chemother-
apy. The clinical course of the more typical form,
In previous studies, we observed that in patients with
aggressive malignancies, TnI increased in ?33% of patients
soon after high-dose chemotherapy (HDC). This early incre-
ment was strongly associated with a reduction in left ventric-
ular ejection fraction (LVEF) that occurred during the fol-
lowing year.15,16However, data on TnI behavior after this
early increment, as well as on its potential impact on the
clinical outcome of cancer patients, are still lacking. Possibly,
a prolonged TnI follow-up study can provide us with infor-
mation on the time course of myocardial damage after HDC
and on the stratification of the cardiac risk of these patients.
The present prospective study was undertaken to investi-
gate whether TnI concentration, measured early after HDC as
well as 1 month later, is associated with the risk of future
We considered all consecutive cancer patients undergoing HDC in
our institute beginning from September 1, 1999. By protocol,
Received September 18, 2003; revision received January 9, 2004; accepted February 24, 2004.
From the Cardiology (D.C., A.C., N.C., M.B., G.L., M.C., C.F., C.M.C.) and Laboratory Medicine Units (M.T.S.), Hemato-Oncology Division (F.P.,
G.M.), Istituto Europeo di Oncologia, University of Milan, Milan, Italy.
Correspondence to Daniela Cardinale, MD, Cardiology Unit, Istituto Europeo di Oncologia, Via Ripamonti 435, 20141 Milan, Italy. E-mail
© 2004 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.orgDOI: 10.1161/01.CIR.0000130926.51766.CC
contraindication for HDC was the presence of ischemic or valvular
heart disease, LVEF ?50%, age ?70 years, and abnormal renal or
hepatic function. Baseline clinical and echocardiographic evalua-
tions were performed within 1 week before HDC. We also excluded
hypertensive patients treated with ?-blocking agents and angioten-
sin-converting enzyme inhibitors because those substances can mask
a possible left ventricular impairment. Patients developing acute (?1
week) cardiotoxicity after HDC and those in whom a double TnI
evaluation (early and late) was not achieved were also excluded.1–3
A total of 703 patients (487 women; mean age, 47?12 years)
fulfilled the inclusion criteria and were enrolled in the study.
Clinical indications for HDC were advanced or primary resistant
breast cancer (n?326), high-grade non-Hodgkin’s lymphoma
(n?264), myeloma (n?44), poor-prognosis Hodgkin’s disease
(n?30), relapsed or refractory ovarian carcinoma (n?16), small-cell
lung cancer (n?10), germ-cell tumors (n?8), and Ewing’s sarcoma
All patients received HDC in different drug combinations, accord-
ing to our institute’s protocols (Table 1). All drugs were adminis-
tered via central venous catheters. Cycles were delivered at 28-day
intervals, and each cycle was supported by reinfusion of autologous
peripheral blood progenitor cells, with or without pretreatment with
high-dose cyclophosphamide, to accelerate hematopoietic recovery
and reduce supportive care requirement.17A total of 232 patients
(33%) received chest radiotherapy after HDC (100 patients to the left
side, 94 to the right side, and 38 to the mediastinum). The
radiotherapy was started at least 45 days after HDC.
The local ethics committee approved the protocol, and written
informed consent was obtained from all patients.
TnI was measured before and soon after each cycle of HDC (early
troponin; E-TnI). Determination of E-TnI consisted of a curve of
assays: a baseline and 5 successive samples during the 3 days after
chemotherapy infusion (immediately after and 12, 24, 36, and 72
hours later). This sequence was repeated after each cycle of therapy.
For each patient, only the highest TnI value was considered.
TnI was also determined 1 month after the last administration of
HDC (late troponin; L-TnI). The timing for L-TnI was chosen in a
such a way that all patients undergoing radiotherapy had L-TnI
determination before beginning radiotherapy to exclude interference
caused by a possible radiotherapy-induced cardiac injury.
Cardiac function was evaluated by physical examination, ECG,
and measurement of LVEF (biplane method, according to Simpson’s
rule) before HDC and 1, 3, 6, and 12 months after the end of HDC
and every 6 months thereafter or whenever required by the clinical
situation. Patients requiring new chemotherapy for the oncological
disease relapse were not considered in the following period of
observation. In these patients, as well as in those lost to follow-up or
those who died of oncological disease, measurements made at the
last follow-up check were considered as final measurements. All
patients’ management decisions and measurements were made
blindly, without the knowledge of the patient’s TnI results.
The primary end point of the study was the occurrence during the
follow-up of one of the following cardiac events: (1) death resulting
from a cardiac cause, (2) acute pulmonary edema, (3) overt heart
failure, (4) asymptomatic LVEF reduction (?25% from baseline), or
(5) life-threatening tachyarrhythmias and conduction disturbances
requiring a permanent pacemaker. All subjects were followed up in
our outpatient clinic until March 1, 2003. For each patient, event-free
survival was calculated from the end of HDC to the occurrence of the
first cardiac event, to the beginning of new chemotherapy for disease
relapse, or to the end of the study. When patient information was not
directly available, it was obtained from hospital records or from the
patient’s general practitioner and relatives.
Blood samples were collected into a Monovette containing a sodium
citrate solution (0.106 mol/L) with a dilution ratio after blood
collection of 1 to 9 and centrifuged at 1080g for 60 minutes, and the
plasma was then separated. TnI concentrations were determined by a
fluorometric enzyme immunoassay analyzer (Stratus CS, Dade
Behring) with a functional sensitivity of 0.03 ?g/L; the cutoff level
was 0.08 ng/mL.18All positive samples were immediately retested
TABLE 1.High-Dose Chemotherapeutic Schedules
ECB Epirubicin 200 mg/m2
Epirubicin 200 mg/m2
TECB Taxotere 85 mg/m2
Etoposide 1.2 g/m2
Carboplatin 1.2 g/m2
Cytarabine 1.6 g/m2
Cytarabine 16 g/m2
B, M, E, G, O, S
Ifosfamide 10 g/m2
Taxotere 85 mg/m2
Carmustine 300 mg/m2
Etoposide 1.2 g/m2
Carboplatin 1.2 g/m2
Ifosfamide 10 g/m2
Etoposide 0.8 g/m2
Carboplatin 1.2 g/m2
Etoposide 1.2 g/m2
Melphalan 140 mg/m2
Mitoxantrone 60 mg/m2
Melphalan 200 mg/m2
Idarubicin 45 mg/m2
Methotrexate 8 g/m2
Melphalan 180 mg/m2
HL, NHL, M, E
HL, NHL, M, E,
Melphalan 180 mg/m2
Etoposide 2 g/m2
Idarubicin 45 mg/m2
Melphalan 180 mg/m2
EC indicates epirubicin-cyclophosphamide; TEC, taxotere-epirubicin-cyclophosphamide; ICE, ifosfamide-carboplatin-etoposide; TICE, taxotere-isofamide-
carboplatin-etoposide; BEAM, BCUU(carmustine)-etoposide-ARA.C(cytarabine)-melphalan; ESAP, etoposide-solumedrol-ARA.C(cytarabine)-platinum; MITOX, mitox-
antrone; MEL, melphalan; IDA, idarubicin; SEQ, sequential; CTX, cyclophosphamide; B, breast cancer; E, Ewing’s sarcoma; G, germ-cell tumors; HL, Hodgkin’s disease;
M, myeloma; NHL, non-Hodgkin’s lymphoma; O, ovarian carcinoma; and S, small-cell lung cancer.
*Preceded by CTX.
†One drug for each cycle.
June 8, 2004
Data are expressed as mean?SD. Fisher’s exact or ?2tests were used
to compare categorical variables, as appropriate. Student’s t tests
were used to compare continuous variables. Comparison between
E-TnI and L-TnI values were made with the ANOVA method.
Kaplan-Meier analysis was used to compare the time-to-event rate
among the 3 groups. A probability value ?0.05 was considered
Three different TnI patterns were identified in our population,
and patients were grouped accordingly. In 495 patients
(70%), both E-TnI and L-TnI values were ?0.08 ng/mL
(TnI?/?group). In the remaining 208 patients, E-TnI was
?0.08 ng/mL (0.16?0.24 ng/mL; range, 0.08 to 1.98 ng/mL).
In 145 of them (70%), L-TnI was within the normal range
(TnI?/?group), whereas in 63 patients (30%), L-TnI levels
remained ?0.08 ng/mL, showing, in most cases, a further
increase (TnI?/?group) (Figure 1). The mean E-TnI value
was 0.12?0.11 ng/mL (range, 0.08 to 1.05 ng/mL) in the
TnI?/?group and 0.25?0.39 ng/mL (range, 0.08 to 1.98
ng/mL) in the TnI?/?group (P?0.001 versus TnI?/?). A wide
overlap between E-TnI values was observed between the 2
groups. Indeed, in only 4 TnI?/?patients was an E-TnI value
?1.05 ng/mL detected. The TnI peak value was observed
soon after the end of HDC in 33% of cases, after 12 hours in
22%, after 24 hours in 8%, after 36 hours in 24%, and after
72 hours in 13%. In patients with positive E-TnI values
(TnI?/?and TnI?/?groups; n?208), a relationship was found
between E-TnI value and LVEF maximal reduction during
follow-up (r?0.78; P?0.0001). This correlation improved
when only the TnI?/?patients (n?63) were considered
The clinical characteristics of the 3 groups are reported in
Table 2. A greater prevalence of women was observed in the
TnI?/?group, and breast cancer most common in TnI-positive
patients. More patients of the TnI?/?group had previous AC
therapy and non-Hodgkin’s lymphoma than in the other 2
groups. Finally, a greater incidence of TnI positivity was
observed after epirubicin-cyclophosphamide and taxotere-
The mean follow-up was 20?13 months (range, 1 to 42
months). Fifteen patients were lost to follow-up, and 180
patients started a new chemotherapy for relapse of the
oncological disease; among these, 80 subsequently died. No
patient died of extracardiac or oncological causes before a
new chemotherapy was attempted.
By protocol, in all patients, LVEF was normal at baseline
evaluation and was comparable in the 3 groups. A reduction
in LVEF was observed in most TnI-positive patients during
the follow-up. Figure 2 shows the percentage of patients with
different degrees of LVEF reduction in the 3 groups. For each
patient, the maximal LVEF reduction during the follow-up
The incidence of cardiac events in the 3 groups during the
follow-up is shown in Table 3. These were significantly more
frequent in TnI-positive patients, particularly in the TnI?/?
group. To calculate the positive and negative predictive
values of TnI, we defined a true-positive test as cardiac event
occurrence during follow-up in patients with both E-TnI and
L-TnI values ?0.08 ng/mL and a true-negative test as the
absence of cardiac events in patients with both E-TnI and
L-TnI normal value. Positive predictive value was 84%, and
negative predictive value was 99%.
The cumulative cardiac events rate, as a function of the
follow-up time based on Kaplan-Meier estimates in the 3
groups, is shown in Figure 3. Notably, all cardiac events were
registered during the first year of follow-up.
Our data clearly show that the TnI pattern after HDC allows
us to stratify the risk of cardiac events in cancer patients in the
3 years thereafter.
In previous studies, we demonstrated that TnI is a sensitive
and specific marker of myocardial injury after HDC and is
able to predict, in a very early phase, the development of
future ventricular dysfunction, as well as its severity.15,16In
the present study, this observation was extended to a larger
population with a longer follow-up in which a wide spectrum
of cardiac events was considered. A persistent TnI increase
allowed us to identify patients at diverse cardiac risk accord-
ing to 3 distinct TnI patterns. Patients without TnI elevation
after HDC had a good prognosis. Indeed, no significant
reduction in LVEF was observed in this group, and a very low
incidence of cardiac events (1%) occurred during the follow-
up. Hence, in consideration of the high negative predictive
value of troponin (99%), we can certainly identify low-risk
patients (70%) who do not require close cardiac surveillance
Figure 1. E-TnI and L-TnI values in 3 study groups. *P?0.05
versus E-TnI; **P?0.001 versus E-TnI; §P?0.001 vs TnI?/?;
#P?0.001 vs TnI?/?.
Cardinale et al Troponin I and High-Dose Chemotherapy
after HDC. In contrast, TnI-positive patients had a greater
incidence of adverse cardiac events. Careful cardiological
monitoring is mandatory in these patients, and prophylactic
strategies aimed at preventing clinical and subclinical cardio-
toxicity should be planned. Among TnI-positive patients, the
persistence of the TnI increase 1 month after HDC (TnI?/?
group) is consistent with a greater cardiac impairment and a
higher incidence of cardiac events than in patients showing
only a transient increase (TnI?/?group). Because of the wide
overlap of E-TnI value in TnI?/?and TnI?/?patients, a late
determination of TnI was necessary to accurately discrimi-
nate a patient’s given degree of risk.
Evidence of release of troponins after chemotherapy has
been demonstrated previously in animal studies, in children
undergoing AC chemotherapy, and in patients with hemato-
logical malignancies.19–22More recently, a prolonged rise of
troponin in 78 adult patients developing asymptomatic LVEF
decrease after chemotherapy has been reported.23
Our study is the first that clearly shows, in a wide adult
population, that the risk of cardiac events in cancer patients
can be predicted by TnI pattern after HDC. Most patients with
positive E-TnI values developed a significant LVEF reduc-
tion in the first year after HDC. In agreement with our
previous observations, a close relationship between the E-TnI
peak value and the degree of LVEF reduction during the
follow-up was found. In addition to anticipating the develop-
ment of cardiac dysfunction, TnI is also an important predic-
tor of cardiac events. This finding has important clinical
implications and provides an intriguing rationale for the
development of pharmacological strategies directed at coun-
teracting cardiac dysfunction and the occurrence of cardiac
complications. The benefit of such an early treatment is
expected to be particularly relevant in patients showing an
E-TnI rise and in whom the persistence, 1 month later, of a
myocardial cell damage, in terms of TnI positivity, is
The mechanism of troponin release after chemotherapy
needs further definition. We can reasonably exclude a minor
necrosis of myocardial cells because of chemotherapy-
induced ischemia. The low incidence of coronary risk factors,
the absence of coronary artery disease, the lack of any
symptoms associated with typical ECG changes, and the
similar incidence of anemia and hypotension among the 3
groups are all elements that support a nonischemic pathogen-
esis. Furthermore, persistence of TnI rise after 1 month
suggests that a release pattern different from ischemic injury
occurs. Indeed, in acute coronary syndromes, troponin eleva-
tion is classically exhausted within 10 days and is usually
associated with, not followed by, ventricular dysfunction.24
In our study, a greater incidence of TnI positivity in women
with breast cancer treated with epirubicin-cyclophosphamide
and taxotere-epirubicin-cyclophosphamide could be ex-
plained by the presence, in both groups, of epirubicin, an AC
with well-known cardiotoxic properties.25,26Conversely, the
lower incidence of TnI positivity among patients with non-
Hodgkin’s lymphoma and those previously treated with AC
could be a result of treatment that did not include ACs at all
or that included them at a lower dose.
According to TnI Results
Clinical Characteristics of the Study Population
Baseline LVEF, %
Previous AC therapy
Previous mediastinum RT†
Family history of CAD
In-hospital acute post-HDC
Anemia requiring blood transfusion
Small-cell lung cancer
Chest-wall RT (left)‡
Chest-wall RT (right)‡
Cumulative AC dose, mg/mq?
CAD indicates coronary artery disease; radiotherapy. For other abbreviations,
see Table 1. Values are mean?SD or n (%) of patients.
*P?0.05 vs TnI?/?group.
†Total dose 30 Gy.
‡Total dose 60 Gy.
§Total dose 36 Gy.
?Cumulative dose was calculated by adding the previous AC treatment to the
AC dose included in HDC and by converting different AC agents in terms of
June 8, 2004
In our population, the cumulative dose of AC (previous AC
dose plus that included in HDC) was similar in the 3 groups
(Table 2). However, it must be considered that patients who
had received previous AC treatment and developed cardio-
toxicity were excluded, by protocol, from HDC. Therefore,
patients having a greater propensity to AC-induced cardio-
toxicity may have been excluded by the preliminary selection,
and the greater incidence of TnI positivity among patients
receiving high-dose epirubicin can be explained by the
current epirubicin toxicity rather than by an AC cumulative
In our study, all cardiac events occurred within 1 year after
HDC. Consistent with the recent classification of cardiotox-
icity, they represent a clinical presentation of “early-onset”
chronic cardiotoxicity.2,3A longer follow-up may be neces-
sary to detect clinical manifestations of “late-onset” chronic
cardiotoxicity, which typically occurs ?1 year after chemo-
therapy. We cannot exclude the possibility that TnI?/?pa-
tients, characterized by a very low risk of cardiac events,
could experience cardiotoxicity during a longer follow-up.
Subclinical cardiotoxicity has been reported to play an
important role in the course of “late-onset” chronic cardio-
toxicity.1,4,27,28Several studies recommend that patients with
subclinical cardiotoxicity be monitored over a long period of
time to gain insight into the clinical consequences of this still
undefined condition.1,6,29,30However, even if many cases of
late abnormal cardiac function have been reported, only a
small percentage of these patients experienced late clinical
Several implications and speculations can be assumed from
our study. First, TnI, by revealing the presence as well as the
persistence of myocardial injury after HDC, is able to
discriminate patients at higher risk of developing a clinically
relevant cardiotoxicity from those with a good clinical out-
come. Second, TnI stratifies cardiac risk in a very early phase,
long before impairment in heart function and symptoms
develop and when many preventive therapeutic strategies are
likely to be effective. Third, TnI could be used to assess as
well as to monitor the safety and the effectiveness of different
antineoplastic treatments. Finally, we can suppose that car-
dioprotective therapies that might limit or prevent the TnI rise
after chemotherapy, as well as cardiological treatments that
interfere with TnI persistence, could improve cardiac prog-
nosis of these patients.20,31
Figure 2. Percentage of patient distribu-
tion for different threshold values with
different degrees of LVEF reduction.
TABLE 3. Cardiac Events in the Study Groups
Acute pulmonary edema
Asymptomatic left ventricular dysfunction
Conduction disturbances requiring pacemaker
111 (16) 5 (1) 53 (37)*53(84)*†
Values are given as n (%).
*P?0.001 vs TnI?/?group; †P?0.001 vs TnI?/?group.
Cardinale et al Troponin I and High-Dose Chemotherapy
The TnI peak value was observed at different intervals of Download full-text
time after HDC, so that several samples for each patient were
needed to detect it. Furthermore, the best time point beyond
which a negative value can assure us that no further TnI
release will occur was not identified.15These represent
possible limitations to the use of this marker in clinical
practice. However, it must be emphasized that the cost of TnI
assay appears justified and cost-effective when a negative TnI
value allows for the exclusion of most patients from a long-term
monitoring program with expensive methods such as echocar-
diography and radionuclide angiocardiography.6,8–11
In conclusion, the TnI release pattern after HDC identifies
patients at different risks of cardiac events. This stratification
allows us to differentiate the monitoring program and to plan,
in selected patients, preventive strategies aimed at improving
We are indebted to Dr Paola Lasagna, Arnaldo Zanelotti, Antonietta
Scopigno, and Raffaella Balestrieri of the Cardiology Unit staff for
their help in the research protocol. We also acknowledge Dr Chiara
Mazzetta of the Division of Epidemiology and Biostatistics for
assistance with the statistical analysis.
1. Shan K, Lincoff AM, Young JB. Anthracycline-induced cardiotoxicity.
Ann Intern Med. 1996;125:47–58.
2. Gharib MI, Burnett AK. Chemotherapy-induced cardiotoxicity: current
practice and prospects of prophylaxis. Eur J Heart Fail. 2002;4:235–242.
3. Pai VB, Nahata MC. Cardiotoxicity of chemotherapeutic agents. Drug
4. Lipshultz SE, Colan SD, Gelber RD, et al. Late cardiac effects of doxo-
rubicin therapy for acute lymphoblastic leukemia in childhood. N Engl
J Med. 1991;324:808–815.
5. Bleyer WA. The impact of childhood cancer on the United States and the
world. CA Cancer J Clin. 1990;40:355–367.
6. Lipshultz SE, Colan SD. The use of echocardiography and Holter mon-
itoring in the assessment of anthracycline-treated patients. In: Bricker JT,
Green DM, D’Angio GJ, eds. Long Term Complications of Treatment of
Children and Adolescents for Cancer. Philadelphia, Pa: Wiley-Liss Inc;
7. de Forni M, Armand JP. Cardiotoxicity of chemotherapy. Curr Opin
8. Singal PK, Iliskovic N. Doxorubicin-induced cardiomyopathy. N Engl
J Med. 1998;3:900–905.
9. Steinherz LJ, Graham T, Hurwitz R, et al. Guidelines for cardiac moni-
toring of children during and after anthracycline therapy: report of the
cardiology committee of the Children Cancer Study Group. Pediatrics.
10. Meinardi MT, van der Graaf WTA, van Veldhuisen DJ, et al. Detection
of anthracycline-induced cardiotoxicity. Cancer Treat Rev. 1999;25:
11. Alexander J, Dainiak N, Berger HJ, et al. Serial assessment of doxo-
rubicin cardiotoxicity with quantitative radionuclide angiocardiography.
N Engl J Med. 1979;300:278–283.
12. Nielsen D, Jensen JB, Dombernowsky P, et al. Epirubicin cardiotoxicity:
a study of 135 patients with advanced breast cancer. J Clin Oncol.
13. McKillop JH, Bristow MR, Goris ML, et al. Sensitivity and specificity of
radionuclide ejection fraction in doxorubicin cardiotoxicity. Am Heart J.
14. The Joint European Society of Cardiology/American College of Cardiol-
ogy Committee. Myocardial infarction redefined: a consensus document
of the Joint European Society of Cardiology/American College of Car-
diology for the redefinition of myocardial infarction. Eur Heart J. 2000;
15. Cardinale D, Sandri MT, Martinoni A, et al. Left ventricular dysfunction
predicted by early troponin I release after high-dose chemotherapy. J Am
Coll Cardiol. 2000;36:517–522.
16. Cardinale D, Sandri MT, Martinoni A, et al. Myocardial injury revealed
by plasma troponin I in breast cancer treated with high-dose chemo-
therapy. Ann Oncol. 2002;13:710–715.
17. Bregni M, Siena S, Di Nicola M, et al. Comparative effects of
granulocyte-macrophage colony-stimulating factor after high-dose cyclo-
phosphamide. J Clin Oncol. 1996;14:628–635.
18. Sandri MT, Cardinale D, Zorzino L, et al. Minor increases in plasma
troponin I predict decreased left ventricular ejection fraction after
high-dose chemotherapy. Clin Chem. 2003;49:248–252.
19. Herman EH, Zhang J, Lipshultz, et al. Correlation between serum levels
of cardiac troponin-T and the severity of the chronic cardiomyopathy
induced by doxorubicin. J Clin Oncol. 1999;17:2237–2243.
20. Herman EH, Zhang J, Rifai N, et al. The use of serum levels of cardiac
troponin T to compare the protective activity of dexrazosane against
doxorubicin- and mitoxantrone-induced cardiotoxicity. Cancer Che-
mother Pharmacol. 2001;48:297–304.
21. Lipshultz SE, Rifai N, Sallan SE, et al. Predictive value of cardiac
troponin T in pediatric patients at risk for myocardial injury. Circulation.
22. Missov E, Calzolari C, Davy JM, et al. Cardiac troponin I in patients with
hematologic malignancies. Coron Artery Dis. 1997;8:537–541.
23. Auner HW, Tinchon C, Linkesch W, et al. Prolonged monitoring of
troponin T for detection of anthracycline cardiotoxicity in adults with
hematological malignancies. Ann Hematol. 2003;82:218–222.
24. Newby LK, Ohman EM, Christenson RH. The role of troponin and other
markers for myocardial necrosis in risk stratification. In: Topol EJ, ed.
Acute Coronary Syndromes. New York, NY: Marcel Dekker Inc; 1998:
25. Ryberg M, Nielsen D, Skovsgaard T, et al. Epirubicin cardiotoxicity: an
analysis of 469 patients with metastatic breast cancer. J Clin Oncol.
26. Keefe DL. Anthracycline-induced cardiomyopathy. Semin Oncol. 2001;
27. Steinherz LJ, Steinherz PG, Tan CTC, et al. Cardiac toxicity 4 to 20 years
after completing anthracycline therapy. JAMA. 1991;266:1672–1677.
28. Kremer LCM, van Dalen EC, Offringa M, et al. Anthracycline-induced
clinical heart failure in a cohort of 607 children: long-term follow-up
study. J Clin Oncol. 2001;16:191–196.
29. Kremer LCM, van Dalen EC, Offringa M, et al. Frequency and risk
factors of anthracycline-induced clinical heart failure in children: a sys-
tematic review. Ann Oncol. 2002;13:503–512.
30. Green DM, Hyland A, Chung CS, et al. Cancer and cardiac mortality
among 15-years survivors of cancer diagnosed during childhood or ado-
lescence. J Clin Oncol. 1999;17:3207–3215.
31. Ali MM, Frei E, Straub J, et al. Induction of metallothionein by zinc
protects from daunorubicin toxicity in rats. Toxicol. 2002;179:85–93.
Figure 3. Cumulative cardiac events rate in 3 study groups.
P?0.001 for TnI?/?vs TnI?/?and TnI?/?, and for TnI?/?vs TnI?/?.
June 8, 2004