Medical and Pediatric Oncology 36:593±600 (2001)
Echocardiographic Evaluation of Patients Cured of Childhood
Cancer: A Single Center Study of 117 Subjects Who
Received Anthracyclines
1
Grazia Bossi, MD, Luca Lanzarini, MD,2 Maria Luisa Laudisa, MD,2 Catherine Klersy,
Arturo Raisaro, MD,2 and Maurizio Arico, MD1*
Background. The risk of cardiomyopathy
following exposure to anthracycline in asymp-
tomatic long-term survivors of childhood cancer
is still hard to predict and precisely quantify.
To identify the impact of different cumulative
doses, even within a non-high dose range, and
the echocardiographic parameters suitable for
evaluating cardiac function, we studied diasto-
lic and systolic echocardiographic parameters in
a cohort of patients followed in a single center.
Procedure. A total of 117 subjects were studied
at a median time of 7 years after treatment
completion. A complete M-mode, two-dimen-
sional and Doppler echocardiographic study
was obtained at rest in all patients according to
the standard recommendations of the American
Society of Echocardiography. Results. Ninety-
nine patients (85%) had completely normal
cardiac function, while 18 had abnormal
echocardiographic ®ndings: 12 had one abnor-
Key words: anthracyclines; late cardiotoxicity; leukemia; childhood cancer
INTRODUCTION
Doxorubicin or daunorubicin are primary components
of most treatment regimens for childhood leukemia or
solid tumors, but their myocardial toxicity remains of
concern to pediatric oncologists. Overt myocardial dys-
function has been reported, albeit infrequently, in chil-
dren receiving high doses of anthracyclines. Besides
these most severe cases, even more worrisome is the risk
of late or very late myocardial damage impairing the
quality of life of subjects cured from childhood cancer
[1±7]. Questions have been raised as to whether these
subjects may form one large new group at risk of
premature cardiac death or various cardiac disturbances
[8,9]. For this reason identi®cation of safe cumulative
doses of anthracyclines or of sensitive tests for preclinical
screening for this complication have been pursued
by several investigators [10±12]. At present, the safe
cumulative dose threshold has not been de®ned; most
patients treated for childhood cancer remain under
evaluation and are being screened for myocardial dy-
sfunction.
The aim of our study was to assess the role of certain
echocardiographic parameters, including those describ-
ing the diastolic function, in evaluating cardiac function
ß 2001 Wiley-Liss, Inc.
MD,
3
mal value, 5 had two, and 1 had three abnormal
values. All the changes were in left ventricular
dimensions, wall thickness or indices of systolic
function; no alterations in left ventricular
diastolic function parameters were found. None
of the echocardiographic parameters correlated
signi®cantly with the cumulative dose of anthra-
cyclines administered either at univariate ana-
lysis or after adjusting for sex, body surface area
or considered risk factors. Conclusions. Subjects
exposed to a median cumulative dose of
214 mg/m2 had no echographic abnormalities
a median of 7 years later. We did not ®nd any
correlation between cumulative anthracycline
dose and the echocardiographic parameters
tested. We now offer echocardiographic fol-
low-up to patients with mildly reduced frac-
tional shortening and/or ejection fraction to rule
out late onset dysfunction. Med. Pediatr. Oncol.
36:593±600, 2001. ß 2001 Wiley-Liss, Inc.
in asymptomatic long-term survivors of childhood can-
cer. We hoped to achieve a clearer de®nition of the risk
associated with cumulative doses of anthracyclines in the
range of those employed in most current polychemother-
apy regimes for childhood malignancies.
PATIENTS AND METHODS
We studied 117 consecutive subjects (67 males) who
had been treated with anthracyclines for childhood mali-
gnancies at our center between 1974 and 1994. The initial
diagnoses had been acute lymphoblastic leukemia (ALL,
n ˆ 87, 74%), acute myeloid leukemia (AML, n ˆ 2, 2%),
non-Hodgkin lymphoma (NHL, n ˆ 7, 6%), Hodgkin
disease (HD, n ˆ 15, 13%), or other solid tumors (Ewing
sarcoma n ˆ 2, schwannoma, neuroblastoma, rhabdo-
Ð
1
ÐÐÐÐÐ
Department of Pediatrics, IRCCS Policlinico, S. Matteo, Pavia, Italy
2
Department of Cardiology, IRCCS Policlinico S. Matteo, Pavia, Italy
3
Biometry and Clinical Epidemiology Service, IRCCS Policlinico S.
Matteo, Pavia, Italy
*Correspondence to: Maurizio Arico, Clinica Pediatrica, Policlinico S.
Matteo, 27100 Pavia, Italy. E-mail aricom@unipv.it
Received 9 March 2000; Accepted 21 December 2000
594 Bossi et al.

myosarcoma, ganglioneuroblastoma; n ˆ 6, 5%). Median
age at diagnosis was 4.8 years (range 2.7±8.3 years) while
the median age at the time of the study was 14.8 years
(range 11.4±18.5 years). The median follow-up time after
treatment completion was 7 years. All subjects had been
treated with anthracyclines (doxorubicin and/or daunor-
ubicin) as part of their antiblastic chemotherapy;
anthracyclines were administered as weekly bolus doses
(30±50 mg/m2/dose over 30 min±24 h) to a mean
cumulative dose (SD) of 214  103 mg/m2.
All subjects had been treated with anthracyclines in
combination with other antiblastic agents; 12/117 (10%),
affected by solid tumors, had also received mediastinal
or thoracic radiotherapy (median dose 25.75 Gy, range
20±59.4 Gy) as a part of the current therapeutic
protocols.
Patients were considered eligible for the study if they
ful®lled the following criteria: free of malignant disease
(®rst or subsequent complete continuous remission); no
bone marrow transplantation; no exposure to antiblastic
agents or other known potentially cardiotoxic drugs for at
least 5 years.
Echocardiographic Study
A complete M-mode, two-dimensional and Doppler
echocardiographic study (Table I) was obtained at rest in
all patients according to the standard recommendations of
the American Society of Echocardiography [13]. All
evaluations have been made by the same two experts
(L. Lanzarini and M.L. Laudisa). Left ventricular (LV)
internal dimensions, diastolic (LVEDD) and systolic
(LVESD), and posterior wall thickness (LVPW) were
measured at the end of diastole (onset of the QRS) and
at the end of systole (maximal inward excursion of
the posterior wall endocardial echo). Percent systolic
thickening of LVPW was expressed as (LVPW systo-
le ÿ LVPW diastole)/LVPW diastole percent (LVPWTh).
Fractional shortening (FS) of the LV was calculated as the
percentage of (LVEDD ÿ LVESD)/LVEDD. Left ventri-
cular end diastolic (LVEDV) and end systolic volumes
(LVESV) were obtained from the four-chamber apical
view utilizing the area/length method. Ejection fraction
(EF) of the LV was calculated as a percentage from
(LVEDV ÿ LVESV)/LVEDV. The mean rate of circum-
ferential ®ber shortening was calculated as (LVEDD ÿ L-
VESD)/LVEDD  ejection time (circ/s). Left ventricular
ejection time was de®ned as the interval between the
opening and the closure of the aortic valve. Left
ventricular end systolic stress (ESS) was calculated
according to Grossman et al. [14]. Age-adjusted stress
velocity index (SVI), an index of contractility incorpor-
ating afterload, but independent of preload [15], was
calculated according to Lipshultz et al. [16]. Left
ventricular mass (LVM) was calculated according to
Devereux et al. [17]. Left ventricular peak systolic
pressure was assumed to be equal to maximal arterial
systolic pressure derived by brachial artery measurement

TABLE I. Echocardiographic and Doppler Parameters Analyzed in the 117 Patients

Parameter Abbreviation Unit
Left ventricular dimensions
Left ventricular end diastolic diameter LVEDD mm
Left ventricular and systolic diameter LVESD mm
Left ventricle shortening fraction FS %
Left ventricular wall thickness:
Septum thickness LVS mm
Septum systolic thickening LVSTh %
Posterior wall thickness LVPW mm
Posterior wall systolic thickening LVPWTh %
Left ventricular volumes and mass:
End-diastolic volume LVEDV ml
End-systolic volume LVESV ml
Left ventricular mass LVM g/m2
Left ventricular function:
Systolic parameters
Ejection fraction EF %
Mean rate of circumferential ®ber shortening MRCF Circumferences/s
End systolic stress ESS gr/cm2
Stress velocity index SVI
Diastolic parameters
Mitral valve E-wave velocity MV-E cm/s
Mitral valve A-wave velocity MV-A cm/s
Mitral valve E/A ratio E/A ratio
Mitral valve total time velocity integral MV-TVI M
A/MV-TVI ratio A-MV-TVI %
Peak ®lling rate normalized for stroke volume PFR/SV SV/s

utilizing a cuff sphygmomanometer [16]. For the as-
sessment of LV diastolic function, mitral pulsed Doppler
was analyzed; the sample volume was between the tips of
the mitral valve lea¯ets in the LV four-chamber apical
view. Measurements included peak early (E) and late (A)
diastolic velocity, their ratio (E/A), and time±velocity
integral (TVI) of the total in¯ow signal and of the A-
wave. The (A ÿTVI) (MV ÿ TVI) ratio was calculated to
correlate the atrial contribution to ventricular ®lling. Peak
®lling rate normalized to mitral stroke volume (PFR/SV)
was also calculated as an index of diastolic function,
independent of age, weight, body surface area, and heart
rate [18]. All examinations were performed in left lateral
decubitus after 5 min of rest by an ESAOTE Biomedics
CFM-Challenge Color Doppler, with 5 and 3.5 MHz
mechanical transducers. Observed echocardiographic left
ventricular measures (dimensions and wall thickness)
were related to body size and age according to Henry
et al. [19].
Statistical Analysis
Data are described as mean and standard deviation
(SD) or median and quartiles, if skewed, for continuous
variables and absolute and relative frequencies for
categorical variables. In order to assess the dependence
of the echocardiographic parameters on the dose of
anthracyclines administered, linear regression was per-
formed; the regression coef®cient and its standard error
(SE) are reported. To test for possible confounding
factors of the relationship, multiple regression was
performed by adding sex and body surface area to
the model. Bivariate plots were drawn to graphically
illustrate the relationship between the echocardiographic
parameters, corrected for sex and body surface area and
the cumulative anthracycline dose. The confounding
effect of a series of potential risk factors (type of cancer,
age at diagnosis, and duration of follow-up) was also
accounted for in a second multivariate model. Patients
were then grouped according to 75th percentile of
cumulative dose (280 mg/m2). Echocardiographic para-
meters were compared between these two groups by
means of the Student t test for continuous variables and
the Fisher exact test for categorical variables. Due to
multiplicity of tests, P-values are to be considered with
caution. Finally, the ranges of normality were computed
for those variables for which a formula or normality range
was available from the literature (for LV measurements:
LVEDD, LVESD, LVS, LVPW, SVI; for diastolic fun-
ction: E/A, A/MV-TVI%, PFR/SV) and cases above or
below the normal range (according to the variable) were
isolated. An appraisal of the departure of the observed
value from the expected one was made based on percent
and absolute variation from upper or lower limits. It
should be noted that systolic parameters quoted in the
Late Anthracycline Cardiotoxicity in Children 595
literature are adjusted for age and body surface area,
whereas the diastolic ones are not [20].
All computations were performed with Stata 6.0 for
Windows (Stata Corporation, College Station, TX). A
P-value <0.05 was considered as statistically signi®cant.
RESULTS
Ninety-nine of the 117 (85%) patients tested had
completely normal cardiac function. In detail, their LV
morphology (dimensions, wall thickness, volumes, and
mass) and function, either systolic or diastolic, were
comparable to those reported for normal controls, incl-
uding correction for age and body surface area (Table II).
Eighteen subjects (15%) had at least one echocardio-
graphic parameter which fell outside the normal range
(Table III). Twelve patients had one abnormal value, ®ve
had two, and one had three abnormal parameters. Eight
patients had increased left ventricular diameters. Five
of them had an increased LV end diastolic diameter
(LVEDD), with a mean absolute value increase of
0.6 mm, i.e., 4% (SD 6%) more than expected. Mean
LV end systolic diameter (LVESD) was 2 mm above the
normal range (i.e., 6%; SD 4%) in six patients; three out
of these six also had an increased LVEDD. As a
consequence of the above alterations, three patients
had a reduced LV fractional shortening below 29%. A
reduced LVS and/or LVPW diastolic thickness was found
in eight and ®ve patients, respectively. The mean
reduction compared to normal values was minimal for
LVS (0.5 mm) and LVPW thickness ( ˆ 0.3 mm); two
patients had both values below the normal range.
Increased LV dimensions in association with a decreased
LV wall thickness was observed in only one patient. No
alterations in the parameters of LV diastolic function
were observed.
None of the echocardiographic parameters evaluated
appears to be associated with the cumulative dose of
anthracyclines administered to the patients, either at
univariate analysis, after adjusting for sex and body
surface area (Fig. 1), or for sex, body surface area, and the
considered risk factors (Table IV). To investigate this
issue further, the patients were grouped according to the
75th percentile of cumulative dose (280 mg/m2): in this
setting we observed a 0.6 mm (i.e., 4%) and 2 mm (i.e.,
6%) increase in the mean values of systolic and diastolic
LV diameters, respectively, in the high dose group
(P ˆ 0:024 and ˆ 0:034, respectively); all the remaining
echocardiographic parameters studied were not different
in patients in the upper quartile (i.e., with higher cu-
mulative doses) from those in the remaining patients
(Table II). Finally, neither sex, nor body surface, type
of cancer, age at diagnosis, mediastinal irradiation or
duration of follow-up appeared to be different between
the two groups of patients.
596 Bossi et al.
TABLE II. Clinical and Echocardiographic Characteristics of the Study Population Overall and Divided According to Cumulative
Anthracycline Dose*
All Anthracyclines Anthracyclines
Characteristic patients ( < 280 mg/m2) (  280 mg/m2)
Number of patients 117 84 33
Age at diagnosis (years)a 4.8 (2.7±8.3) 4.6 (2.7±8.6) 4.9 (2.3±7.5)
Age at echocardiography (years)a 14.8 (11.4±18.5) 15.3 (11.3±18.5) 15.5 (11.8±18.2)
Gender (M/F) 67/50 47/37 20/13
Months from treatment completiona 82.9 (60.3±113.1) 84.8 (63.5±115.2) 74.6 (56.3±107.0)
Body surface area (m2) 1.52 (0.28) 1.51 (0.28) 1.54 (0.27)
Heart rate (beats/min)b 75.4 (14.6) 75.1 (13.0) 76.2 (18.1)
LVEDDb (mm) 44.4 (4.2) 43.8 (4.0) 45.8 (4.5)**
LVESDb (mm) 28.4 (3.8) 27.8 (3.6) 29.7 (3.9)***
FS%b (mm) 36.2 (6.0) 36.5 (6.5) 35.5 (4.5)
LVSb (mm) 7.6 (1.3) 7.6 (1.3) 7.6 (1.3)
LVS Th %b 38.8 (19.0) 39.2 (20.0) 37.8 (16.4)
LVPWb (mm) 8.0 (1.7) 8.0 (1.8) 7.8 (1.3)
LVPW Th %b 57.1 (24.0) 56.8 (23.5) 57.8 (25.7)
LVEDVb ml 91.2 (30.3) 89.4 (28.8) 95.8 (33.8)**
EF%b 62.2 (5.8) 62.1 (6.18) 62.5 (5.0)
LV Mass (gr/m2) 110.9 (35.5) 109.0 (34.7) 115.6 (37.5)
E/Ab 1.9 (0.8) 1.9 (0.9) 1.8 (0.4)
A/MV-TVI %b 0.25 (0.09) 0.25 (0.10) 0.23 (0.07)
PFR/SVb (sv/cm) 5.17 (0.96) 5.17 (0.10) 5.15 (0.19)
MRCF (circ/sec) 1.15 (0.29) 1.17 (0.33) 1.11 (0.14)
E.S.S. (gr/cm2) 61.2 (19.6) 59.6 (19.6) 64.2 (19.5)
*For abbreviations see Table 1.
a
Median and quartiles.
b
Mean and SD.
**P ˆ 0:024; ***P ˆ 0:03.

TABLE III. Characteristics of 18 Asymptomatic Long-Term Survivors After Childhood Cancer and Results of Their Echocardiographic
Evaluation*
Age at Follow-up
diagnosis time (years) Cumulative Chest LVEDDa LVESD FS LVS LVPW
Patient Sex (years) Diagnosis dose (mg/m2) anthracycline RT (mm) (mm) (%) (mm) (mm)
1 M 3 ALL 12 360 ÿ 52 (51.7) 37 (35.0) 5 (6.6)
2 M 13 HD 10 150 ‡ 52 (51.4) 37 (34.7) 26
3 F 2 ALL 3 240 ÿ 5 (5.3) 5 (5.2)
4 F 6 ALL 5 280 ÿ 6 (6.4) 6 (6.3)
5 M 2 ALL 15 380 ÿ 54 (47.0) 36 (31.9)
6 M 2 HD 9 150 ÿ 47 (45.1)
7 F 4 ALL 5 280 ÿ 47 (46.9)
8 M 8 ALL 5 280 ÿ 38 (36.1) 25

9 M 1 AML 13 585 ÿ 37 (34.5) 27

10 M 9 ALL 5 180 ÿ 36 (35.5)

11 M 6 HD 9 150 ‡ 6 (6.7)
12 M 4 GNB 11 215 ÿ 6 (6.7)
13 M 1 ALL 6 180 ÿ 6 (6.1)
14 F 4 ALL 12 60 ÿ 6 (6.5)
15 F 9 HD 6 40 ‡ 6 (7.1)
16 F 2 ALL 6 180 ÿ 5 (5.7)
17 M 1 AML 8 490 ÿ 5 (5.2)
18 M 2 ALL 4 240 ÿ 5 (5.7)
*Only abnormal values are scored, with normal reference values in brackets when available.
 Late Anthracycline Cardiotoxicity in Children 597

Fig. 1. Bivariate plots of echo parameters (y-axis) and cumulative anthracycline dose (x-axis) controlled
for demographic characteristics.
598 Bossi et al.
TABLE IV. Association of Echocadiographic Results and Cumulative Dose of Anthracyclines
Controlled for age and Controlled for age, body surface
Raw estimate body surface area area, and risk factors
Echo parametera b (SE) P-value bb (SE) P-value bb (SE) P-value
LV Mass ÿ 0.012 (0.030) 0.668 0.005 (0.020) 0.818 ÿ 0.005 (0.024) 0.823
LVEDD 0.004 (0.005) 0.370 0.006 (0.0049) 0.095 0.007 (0.005) 0.182
LVESD 0.005 (0.004) 0.251 0.006 (0.004) 0.095 0.007 (0.005) 0.168
FS% ÿ 0.004 (0.005) 0.387 ÿ 0.004 (0.005) 0.375 ÿ 0.006 (0.006) 0.264
LVS ÿ 0.001 (0.001) 0.401 < ÿ 0.001 0.649 ÿ 0.001 (0.001) 0.198
LVS Th % ÿ 0.008 (0.015) 0.611 ÿ 0.008 (0.015) 0.584 0.005 (0.018) 0.781
LVPW ÿ 0.002 (0.001) 0.081 ÿ 0.002 (0.001) 0.117 ÿ 0.002 (0.001) 0.126
LVPW Th % ÿ 0.028 (0.019) 0.148 ÿ 0.031 (0.019) 0.093 ÿ 0.011 (0.019) 0.577
LVEDV 0.010 (0.029) 0.735 ÿ 0.008 (0.015) 0.584 0.024 (0.029) 0.412
EP % ÿ 0.002 (0.005) 0.646 ÿ 0.002 (0.006) 0.662 0.001 (0.006) 0.900
E/A ÿ 0.001 (0.001) 0.248 ÿ 0.001 (0.001) 0.274 ÿ 0.001 (0.001) 0.084
A/MV-TVI % < ÿ 0.001 0.898 < ÿ 0.001 0.836 < 0.001 0.698
PFR/SV < ÿ 0.001 0.974 < ÿ 0.001 0.861 < 0.001 0.990
MRCF < ÿ 0.001 0.359 < ÿ 0.001 0.350 < ÿ 0.001 0.574
E.S.S. 0.047 (0.025) 0.061 0.047 (0.025) 0.068 0.042 (0.024) 0.086
a
For abbreviations see Table 1.
b
Regression coef®cient for cumulative anthracyclines dose (SE).

DISCUSSION these abnormalities involved LV parameters (LVEDD,

In this series, 99 of 117 (85%) asymptomatic long-
term survivors of childhood malignancies had completely
normal cardiac function evaluated by thorough echocar-
diographic examination performed at a median time of
about 7 years after treatment completion, when their
median age was 14.8 years. This ®nding is quite reas-
suring in contrast to previous reports suggesting that up to
57% of patients treated for acute lymphoblastic leukemia
have late myocardial abnormalities (1±5). In particular,
Lipshultz et al. (1) described abnormalities of LV after-
load (measured as ESS), due to reduced LVPW diastolic
thickness, and not due to hypertension or dilation, or
reduced contractility (measured as SVI). Their main
conclusion was that myocardial late damage was
signi®cantly dose-dependent, while younger age at trea-
tment was predictive for afterload increase. Lipshultz's
report was considered as a warning and prompted us, as
many others, to follow such patients carefully [21]. The
present ®ndings show a decidedly lower incidence of
myocardial dysfunction. Although the median cumula-
tive dose given to the patients in our study was lower than
that administered to the children in Lipshultz's series
(214  103 mg/m2, compared to 360 mg/m2), this dif-
ference does not explain the discrepancy in the results; in
fact, according to their proportions (i.e., 65% of patients
receiving more than 228 mg/m2) one-third of our patients
would have been expected to have echocardiographic
changes, whereas the observed proportion was 15%. Nor
can the discrepancy be explained by potential biases such
as modality of echo study, sample size, or duration of
patients' follow-up.
Fifteen percent of our patients cured from childhood
cancer had abnormal echocardiographic ®ndings; all
LVESD, LVS, and LVPW) while no impairment of LV
diastolic function (E/A, PFR/SV) was observed in any
case, in contrast to some previous reports suggesting an
early as well as late involvement of diastolic function
[22±25]. In particular, no signi®cant departure from nor-
mality was observed for PFR/SV, the only echocardio-
graphic parameter of diastolic function considered not to
be in¯uenced by age, body surface area, or heart rate [18].
Our ®nding of signi®cantly increased LV dimension,
measured as both LVEDD and/or LVESD, in a subset of
patients, is not unexpected [6]. It must, however, be
considered that such increases were based on 0.6 and 2
mm differences, accounting for 4 and 6% of the expected
values, respectively. Although statistically signi®cant,
such alterations should be regarded with caution. In fact,
all of these data come from cross-sectional studies with
single observations. Furthermore, one must be critical
with regard to their true clinical relevance. If these
alterations are to be considered true defects (which we
do not believe), although minor and preclinical, these
patients deserve tailored follow-up studies with thorough
medical and cardiological counseling. Clinical decisions
should be made for each individual patient by an expert
cardiologist, preferably one also experienced in this ®eld,
who takes into account all the available echocardio-
graphic parameters. Caution should be exercised when
considering reduced FS% in a patient whose EF%
remains normal. In fact, in our series only one patient
with reduced FS% also had a reduced EF%; this patient
had received the highest cumulative dose of anthracy-
clines (585 mg/m2).
It has been suggested that anthracyclines cause pro-
gressive cardiac damage with inadequate cardiac growth
manifested as reduced wall thickness and/or reduced LV

mass. This process imposes an excess of systolic wall
stress on the LV that may induce a decrease of
contractility and LV global systolic function [1]. Only
one patient in our series had abnormally increased LV
dimensions associated with a reduction in LV wall
thickness. Global systolic function in this patient was at
the lower limit of the normal range. It is not clear why in
some cases the cardiotoxic effect of anthracyclines is
limited to myocytes with no impairment of cardiac fun-
ction while in other apparently similar cases, dilation
and/or systolic cardiac dysfunction with no myocardial
cellular loss may ensue. We were not able to ®nd
correlation between cumulative dose and any of the
echocardiographic parameters tested. A correlation with
ventricular diameters reached statistical signi®cance
only when the analysis was focused on patients included
in the quartile receiving the highest cumulative dose
( > 280 mg/m2). Thus, at present we are not able to
predict dose-dependent late cardiac defects, and studies
of potential markers for sensitivity to anthracyclines are
thus warranted.
In the meanwhile, the issue of a safe cumulative
threshold of anthracyclines is relevant in as much as most
patients enrolled in current, non-high risk treatment
regimes for acute lymphoblastic leukemia will receive
moderate doses, i.e., less than 300 mg/m2. In our series,
only 8 of 58 patients falling in the half of patients with
lower cumulative dose (i.e., inferior to the median value
of <214 mg/m2) had any echocardiographic abnorm-
ality, at a median time of 7 years after completion of
treatment. Three of these patients had also been exposed
to chest irradiation. Our ®ndings are in keeping with a
recent observation from the Boston group, suggesting
that a cumulative dose of <300 mg/m2 may be con-
sidered safe regarding the preservation of cardiac
function [6]. This may have implications in designing
future treatment protocols, since in some cases further
reduction of one or a few doses of anthracyclines
(resulting in a cumulative reduction in the range of 30
to 80 mg/m2) could affect treatment results without
avoiding the risk of cardiac sequelae. The current BFM-
based cooperative treatment regime AIEOP-ALL95 [26]
incorporates a cumulative anthracycline dose of 120 mg/
m2 for the standard and 240 mg/m2 for the intermediate
risk patient groups, accounting for about 10 and 80% of
the patients, respectively. Re®ned, early identi®cation of
10% of patients at very high risk allowed us to con®ne
more aggressive chemotherapy, in which the use of
higher doses of anthracyclines (370 mg/m2) seems jus-
ti®ed, to this group.
CONCLUSION
Despite measurement of several diastolic and systolic
parameters, our echocardiographic study did not reveal
Late Anthracycline Cardiotoxicity in Children 599
any clinically relevant myocardial changes in these
asymptomatic adolescents or young adults exposed to
anthracyclines for treatment of childhood cancer. There
was evidence of some minimal modi®cation of ventri-
cular diameter. Although statistically signi®cant, the cli-
nical relevance of these ®ndings is unclear. We cannot
exclude that the reassuring ®ndings observed at about
7 years follow-up de®nitely rule out development of later
dysfunction. Thus, we shall offer echocardiographic
follow-up to patients with reduced fractional shortening
and/or ejection fraction.
REFERENCES
1. Lipshultz SE, Colan SD, Gelber RD, et al. Late cardiac effects of
doxorubicin therapy for acute lymphoblastic leukemia in child-
hood. N Engl J Med 1991;324:808±815.
2. Yeung ST, Yoong C, Spink J, et al. Functional myocardial
impairment in children treated with anthracyclines for cancer.
Lancet 1991;337:816±818.
3. Steinherz LJ, Steinherz PG, Tan CT, et al. Cardiac toxicity 4 to
20 years after completing anthracycline therapy. JAMA 1991;
266:1672±1677.
4. Larsen RL, Jakacki RI, Vetter VL, et al. Electrocardiographic
changes and arrhythmias after cancer therapy in children and
young adults. Am J Cardiol 1992;70:73±77.
5. Leandro J, Dyck J, Poppe D, et al. Cardiac dysfunction late after
cardiotoxic therapy for childhood cancer. Am J Cardiol 1994;
74:1152±1156.
6. Nysom K, Holm K, Lipsitz SR, et al. Relationship between
cumulative anthracycline dose and late cardiotoxicity in child-
hood acute lymphoblastic leukemia. J Clin Oncol 1998;16:
545±550.
7. Johnson GL, Moffett CB, Geil JD, et al. Late echocardiographic
®ndings following childhood chemotherapy with normal serial
cardiac monitoring. J Pediatr Hematol Oncol 1996;18:
72±75.
8. Lipshultz SE, Colan SD. The use of echocardiography and Holter
monitoring in the assessment of anthracycline-treated patients. In:
Bricker JT, Green DM, D'Angio GJ, (editors). Cardiac toxicity
after treatment for childhood cancer. New York, Wiley-Liss,
1993, p 45±62.
9. Shan K, Lincoff AM, Young JB. Anthracycline-induced cardio-
toxicity. Ann Intern Med 1996;125:47±58.
10. Hale JP, Lewis IJ. Anthracyclines: cardiotoxicity and its
prevention. Arch Dis Child 1994; 71:457±462.
11. Sorensen K, Levitt G, Bull C, et al. Anthracycline dose in
childhood acute lymphoblastic leukemia: issues of early survival
versus late cardiotoxity. J Clin Oncol 1997;15:61±68.
12. Lipshultz SE, Lipsitz SR, Mone SM, et al. Female sex and higher
drug dose as risk factors for late cardiotoxic effects of doxo-
rubicin therapy for childhood cancer. N Engl J Med 1995;332:
1738±1743.
13. Sahn DJ, DeMaria A, Kisslo J, Weyman A. The Committee on
M-Mode Standardisation of the American Society of Echocardio-
graphy. Recommendations regarding quantitation in M-Mode
echocardiography: results of a survey of echocardiographic
measurements. Circulation 1978;58:1072±1083.
14. Grossman W, Jones D, McLaurin LP. Wall stress and patterns
of hypertrophy in the human left ventricle. J Clin Invest 1975;56:
56±64.
15. Colan SD, Parness IA, Spevak PJ, Sanders SP. Developmental
modulation of myocardial mechanics: age and growth-related
600 Bossi et al.
alterations in afterload and contractility. J Am Coll Cardiol
1992;19:619±629.
16. Lipshultz SE, Easley KA, Orav EJ, et al. The Pediatric pulmonary
and cardiac complications of vertically transmitted HIV infection
(P2C2HIV) Study Group: Left ventricular structure and function
in children infected with human immunode®ciency virus. The
prospective P2C2 HIV multicenter study. Circulation 1998;97:
1246±1256.
17. Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic
assessment of left ventricular hypertrophy: comparison to nec-
ropsy ®ndings. Am J Cardiol 1986;57:450±458.
18. Bowman LK, Lee FA, Jaffe CC, et al. Peak ®lling rate normalized
to mitral stroke volume: a new Doppler echocardiographic ®lling
index validated by radionuclide angiographic techniques. J Am
Coll Cardiol 1988;12:937±943.
19. Henry WL, Ware J, Gardin JM, et al. Echocardiographic measure-
ments in normal subjects: growth-related changes that occur be-
tween infancy and early adulthood. Circulation 1978; 57:278±285.
20. Feingenbaum, H, editor Echocardiography. Fifth Edn. Philadel-
phia, PA: Lea and Febiger, p 665±683.
21. Steinherz LJ, Graham T, Hurwitz R, et al. Guidelines for cardiac
monitoring of children during and after anthracycline therapy-
report of the cardiology committee of the Children's Cancer Study
Group. Pediatrics 1992;89:942±949.
22. Lee BH, Goodenday LS, Muswick GJ, et al. Alterations in left
ventricular diastolic function with doxorubicin therapy. J Am Coll
Cardiol 1987;9:184±188.
23. Schmitt K, Tulzer G, Merl M, et al. Early detection of doxorubicin
and daunorubicin cardiotoxicity by echocardiography: Diastolic
versus systolic parameters. Eur J Pediatr 1995;154:201±
204.
24. Cottin Y, Touzery C, Coudert B, et al. Impairment of diastolic
function during short-term anthracycline chemotherapy. Br Heart
J 1995,73:61±64.
25. Hausdorf G, Morf G, Beron G, et al. Long-term doxorubicin
cardiotoxicity in childhood: non-invasive evaluation of the
contractile state and diastolic ®lling. Br Heart J 1988;60:309±
315.
26. AricoÁ M, Conter V, Valsecchi MG et al. Associazione Italiana di
Ematologia Oncologia Pediatrica (AIEOP): Treatment intensi®-
cation with repeated BFM-protocol II may prevent early relapse
in very high risk ALL. AIEOP-ALL 9503 study. 39th Meeting of
the American Society of Hematology. S.Diego 5±9 December,
1997. Blood 1997, 263b, 90, (Abstract).