Pediatr Blood Cancer 2011;57:227–230
Clinical Impact of the Baseline Echocardiogram in Children
With High-Risk Acute Lymphoblastic Leukemia
Taurino Avelar, MS, Linda B. Pauliks, MD,
Background. It is common practice to hold anthracycline induc-
tion chemotherapy in children with high-risk acute lymphoblastic
leukemia (HR-ALL) until an echocardiogram is performed and inter-
preted. It is unclear whether withholding therapy in HR-ALL children
is justified by echocardiogram findings. We reviewed the initial
echocardiograms in a cohort of children with HR-ALL to determine
the incidence of contraindications for anthracycline treatment.
Procedure. We identified 50 consecutive children (<21 years old)
with HR-ALL presenting at our institution over a 10-year period. One
didn’t have an initial echocardiogram, 39 had pre-therapy studies,
and 10 were studied within 6 days of beginning chemotherapy.
These 49 studies were reviewed to determine the incidence and
clinical significance of abnormalities. Results. All 49 patients had
normal cardiac function. Initial echocardiogram findings had no
impact on induction chemotherapy administration in any patient.
Key words: anthracycline; chemotherapy; child; echocardiography; leukemia
INTRODUCTION
The improved survival in childhood leukemia comes at the cost
of using cardiotoxic medications with possible long-term sequelae
for the heart [1]. This has become of increasing concern as the
majority of patients now survive long enough to experience those
long-term side effects during their adult life. Through improved
risk-specific therapies, the outcome for high-risk childhood acute
lymphoblastic leukemia has improved to a 4-year event-free sur-
vival rate of 76% [2]. Among the most effective drugs in treating
ALL have been the anthracyclines. All children with ALL receive
anthracyclines, usually adriamycin, at some point during therapy,
but certain groups receive it at the start of therapy. While they have
proven to be highly effective in the treatment of ALL, anthracy-
clines have been associated with myocardial toxicity, with cumu-
lative total dose being the best predictor of the risk [3]. Children
have been found to be particularly susceptible to this effect [4,5].
Guidelines published by the American Academy of Pediatrics
in 1992 recommend a baseline cardiac evaluation in all patients
expected to receive an anthracycline, then periodically during and
after therapy [6]. These evaluations include a conventional trans-
thoracic echocardiogram although the value of tissue Doppler
imaging is also increasingly recognized [7]. The purpose of the
initial screening is threefold. First, it has been shown that the
anthracycline toxicity may manifest itself in a rapid decline of
the fractional shortening, e.g., from 40 to 33%, when both numbers
are still within normal limits [1]. Therefore, a baseline study is
necessary to be able to detect such relative changes with preserved
fractional shortening on follow-up studies. Second, the patient may
have unrecognized heart disease. If the patient had dilated cardio-
myopathy at baseline, then anthracyclines should be avoided [1].
And third, the study will identify cardiac abnormalities associated
with the disease itself although cardiac involvement is rare in
pediatric patients.
Standard practice at some institutions is to hold the initial
anthracycline dose until a cardiologist provides a final reading of
the echocardiogram. In some cases, it may not be possible to obtain
ß 2011 Wiley-Liss, Inc.
DOI 10.1002/pbc.23066
Published online 25 February 2011 in Wiley Online Library
(wileyonlinelibrary.com).
MPH, and Andrew S. Freiberg, MD*
However, only 22(45%) of the studies were completely normal.
Echocardiographic abnormalities included pericardial effusion (17/
49), trivial or mild mitral or aortic insufficiency (13/49), left ventric-
ular enlargement (3/49), and structural heart disease (4/49). Twelve
percent of the children had a patent foramen ovale. None of the
cardiac findings required therapeutic intervention other than reposi-
tioning of indwelling lines (6/49) due to intracardiac positioning.
Conclusions. In our experience, findings on echocardiograms in
childhood HR-ALL did not impact anthracycline administration.
This study suggests that induction chemotherapy should not be
delayed for an echocardiogram. However, whenever possible, a
pre-therapy echocardiogram is still recommended for determining
baseline function and to identify associated problems like pericardial
effusions which were common in this study. Pediatr Blood Cancer
2011;57:227–230. ß 2011 Wiley-Liss, Inc.
an echocardiogram right away resulting in a delay in starting
induction chemotherapy.
It is well established that all patients with anthracycline chemo-
therapy require echocardiographic monitoring of cardiac function
but it is unclear whether treatment in children with high-risk ALL
should be delayed for the sake of a baseline study. We hypothesize
that echocardiographic abnormalities that would impact anthracy-
cline dosing are rare in newly diagnosed children with high-risk
ALL.
METHODS
Patients
The medical records of Penn State Milton S. Hershey Medical
Center for the time period December 1, 2000 and July 31, 2009
were searched for patients up to 21 years of age admitted with
high-risk acute lymphoblastic leukemia (HR-ALL). Patients were
identified as high risk if they were
10 years at time of diagnosis or
presented with a WBC
50  109/L. To qualify for the study,
patients had to have been diagnosed at our institution and treated
with induction therapy that included an anthracycline. With a high-
risk ALL diagnosis, all patients received a ‘‘4-drug’’ induction
consisting of an anthracycline (daunomycin or daunorubicin), vin-
cristine, L-asparaginase or PEG-asparaginase, and a steroid (pre-
dnisone or dexamethasone) according to frontline Children’s
Cancer Group or Children’s Oncology Group protocols.
1
Department of Pediatrics, Penn State College of Medicine, Hershey,
Pennsylvania
Conflict of Interest: Nothing to report.
*Correspondence to: Andrew S. Freiberg, MD, 500 University Drive,
Hershey, PA 17033. E-mail: afreiberg@hmc.psu.edu
Received 27 July 2010; Accepted 10 January 2011
228 Avelar et al.
ECHOCARDIOGRAPHIC METHOD
A standard transthoracic echocardiogram was performed with
the Acouson Sequoia Ultrasound System (Siemens, Mountain
View, CA) using 4, 7, or 8 MHz ultrasound transducer as appro-
priate for patient size. A simultaneous electrocardiogram was
recorded. The standardized imaging protocol included 2D, M-
mode and Doppler echocardiography, and the study reports and
images were reviewed for this study. To assess left ventricular
systolic function, fractional shortening was determined from M-
mode images obtain from parasternal short axis views and calcu-
lated as follows:
ðLeft ventricular enddiastolic diameter
Left ventricular systolic diameterÞ
Left ventricular diastolic diameter
Fractional shortening values between 28 and 42% were considered
indicative of normal left ventricular systolic function. In the age
group covered by this study, fractional shortening is age independ-
ent. In contrast, left ventricular enlargement has to be evaluated in
relation to the child’s age and somatic size. For this study, the
presence of left ventricular enlargement was assessed by plotting
the left ventricular diastolic diameter (LVEDD) against body sur-
face area using published normal values; the result is expressed as a
z-score [8]. In determining the severity of the echocardiographic
findings, patients were placed into one of four following categories:
normal, minor findings not requiring follow up, findings requiring a
cardiology follow-up evaluation per routine, and findings requiring
a change in therapy.
STATISTICS
Patient characteristics were reported as means and standard
deviation using commercially available software (MS Excel,
Microsoft Co., Seattle, WA). The z-scores for the left ventricular
dimensions were calculated using an online tool based on
reference 8.
ETHICS
This study was reviewed by the Penn State Hershey Institu-
tional Review Board which waived the requirement for informed
consent.
RESULTS
Patient Characteristics
Initially, 60 children with high-risk ALL were identified. Of
these, 10 were excluded because they had initiated therapy at
another institution and 1 who did not have an identified echocardio-
gram until 5 months after induction therapy. Ten other patients had
an echocardiogram after starting therapy. In total, the cohort of 49
patients consisted of 31 males and 18 females. In this retrospective
series, 80% of the patients had their initial echocardiogram per-
formed prior to the first anthracycline dose. For the remaining 20%
the echocardiogram was performed for 3 within 1 day, 4 within 2
days, 2 within 5 days, and 1 within 6 days. The median age at
diagnosis was 10.9 years (range: 4 months–21 years) and the
median WBC was 69  109/L. One patient had trisomy 21. None
Pediatr Blood Cancer DOI 10.1002/pbc
of the patients were on a ventilator or inotropic support during the
pre-therapy echocardiogram.
ECHOCARDIOGRAPHIC DATA
An echocardiogram was performed in all 49 patients. All
patients had normal left ventricular systolic function (fractional
shortening values of at least 28%). In 12% (6/49), the indwelling
intravenous line tip was seen inside the heart requiring reposition-
ing of the catheter. With this exception, no therapeutic changes
were based on the echocardiogram. Specifically, anthracycline
chemotherapy was administered as planned for all children in this
study. None of the patients required pericardiocentesis.
While all children had normal cardiac function, only 45% of the
studies were completely normal, with minor abnormalities preva-
lent in the remaining children. Findings are summarized in
Figure 1. The most common abnormality was the presence of a
pericardial effusion, found in 35% of studies. While pericardial
effusions were common only two were characterized as moderate.
None were hemodynamically significant or required pericardio-
centesis. Trivial or mild mitral insufficiency was found in 24% of
patients and 2% had aortic insufficiency. Left ventricular enlarge-
ment was also prevalent with three patients (6%) having an
LVEDD z-score greater than 2, with two having a score of 3, and
one having a score of 4. In addition, the z-score for the group was
greater relative to the normal by 0.4 standard deviations. Congen-
ital heart disease was found in 8% of patients, including 2 cases of
atrial septal defect secundum, one surgically corrected ventricular
septal defect, and one aortic valve abnormality. Furthermore, 8% of
the patients were found to have the incidental finding of a patent
foramen ovale. Overall, small atrial left to right shunts were ident-
ified in 12% of patients.
The therapeutic impact based on echocardiographic findings is
shown in Figure 2. The only treatment changes were adjustments
of indwelling lines, necessary in 12% of all patients. The majority
of studies were either normal (45%) or showed minor findings not
requiring Cardiology follow up (47%). In 8%, the recommendation
was a complete cardiac evaluation but none of these were deemed
urgent. None impacted chemotherapy itself.
DISCUSSION
This study investigated the range of echocardiographic abnor-
malities found in a cohort of children with high-risk ALL and
identified no findings that would have required a change in the
anthracycline therapy. None of these 49 patients over a 10-year
period had evidence of cardiac involvement in the malignant proc-
ess itself. Pericardial effusions were common but none were hemo-
dynamically significant. There were a surprising number of studies
(8%) with coincidental congenital heart defects (including one
known ventricular septal defect, two atrial septal defects of the
secundum type and one dysplastic aortic valve). The incidence of
congenital heart disease in the general population is 0.8% and one
would expect that most cases would be diagnosed well before a
child is 10 years of age [9]. In some cases, the interpreting pediatric
cardiologist may have a tendency to overemphasize a questionable
finding since clinical correlation is not available. In addition to this
study, others have also reported higher incidences of congenital
heart disease in oncology patients [10]. The number of small atrial
left to right shunts is comparable to what others have found on

Fig. 1. Echocardiographic findings. This figure shows the frequen-
cies of abnormal echocardiographic findings. Although the majority of
patients had an abnormality present, none required alteration of the
induction chemotherapy. Note that the total exceeds 100% due to
multiple findings in some patients.
transthoracic echocardiograms [11]. In this study 12% of the HR-
ALL children had the echocardiographic finding of patent foramen
ovale or small atrial septal defect secundum, which is comparable
to the general population.
Overall, approximately half of the echocardiograms (27/49)
showed some abnormal finding but there were no cases where they
had an impact on treatment. In addition to structural heart disease,
pericardial effusion, mitral insufficiency, and left ventricular
enlargement were also identified. These non-specific findings could
be explained by the underlying illness. It is well recognized that
HR-ALL patients frequently present with pericardial and pleural
effusions. Previous proposals have described pericardial effusions
as being part of the leukemic process, arising secondary to hem-
Fig. 2. Impact of the echocardiogram on therapy This figure shows
how the echocardiogram affected cardiac monitoring. While the
majority of studies were abnormal, none required additional studies
that would have delayed the induction of therapy. Note that the total
exceeds 100% since some patients with positive findings required
adjustment of line position as well.
Pediatr Blood Cancer DOI 10.1002/pbc
Initial Echocardiogram in High Risk ALL 229
orrhagic, infectious, or leukemic infiltrates [12]. Others have dem-
onstrated that survival does not differ based on the presence of an
effusion [13], presumably because such effusions responded to
treatment of the primary disease. Left heart enlargement may be
in response to acute or chronic anemia. In this study all children
were hemodynamically stable with normal left ventricular systolic
function.
The fact that 12% of patients required repositioning of the
indwelling central line is noteworthy. These were as follows:
2000 (0/1), 2001 (0/6), 2002 (1/5), 2003 (0/3), 2004 (1/3), 2005
(1/5), 2006 (2/7), 2007 (0/10), 2008 (0/5), and 2009 (1/4). For this
study, we did not determine how many patients had central lines
placed prior to the echocardiogram nor how many echocardio-
grams were able to document a correctly placed line, as this was
not our objective. However, the study spans a 10-year period and it
is felt that with newer insertion techniques intracardiac placement
is now avoided in most.
In a similar retrospective study, Porea et al. [14] found that there
were no alterations in therapy made as a result of initial echocar-
diograms performed in 128 newly diagnosed children with ALL
lacking a previous cardiac history. The authors suggested that the
initial echocardiogram was therefore unnecessary. However, this
would deprive patients of the other benefits of the echocardiogram
including the establishment of a baseline for their fractional short-
ening, identification of other abnormalities and rarely identification
of cardiac involvement in the primary disease process. Therefore,
we would disagree with the recommendation to forgo the study
entirely.
Our study had several limitations. This study is retrospective
with a relatively small sample size since we selected only high-risk
patients who required anthracycline therapy in induction. As
expected from its known low incidence of about 1.13/100,000
children [15], none of the 49 children we examined had pre-exist-
ing cardiomyopathy. Certainly our small retrospective study was
not intended to determine the incidence. We recommend that when
possible, the initial study be performed prior to the anthracycline,
but the low incidence of cardiomyopathy should allow the drug to
be administered safely without waiting for the final reading.
Another important limitation is that 10 of our 49 patients had the
initial study just after induction of chemotherapy instead of prior to
treatment. This practice defeats the purpose of screening to identify
cardiac abnormalities associated with the disease, identifying pre-
viously unrecognized heart disease and accurately accounting for
changes to the fractional shortening due to anthracycline toxicity. It
is possible that this delay changed results, for instance, reduction of
effusions from initial chemotherapy or acute volume loading with
intravenous hydration increasing the degree of atrioventricular
valve insufficiency. However, there was no obvious difference
between these 10 and the other 39 studies. Although acute chemo-
therapy cardiotoxicity is well described there were no cases of
abnormal function among our patients regardless of the timing of
the study. Also, chemotherapy administration can lead to antihist-
amine and epinephrine release and in theory mask decreases in
cardiac function but the clinical impact of this effect is unknown
and probably minor in this population. It is also worth mentioning
that the echocardiographic studies were interpreted by several
different pediatric cardiologists over a 10-year period which may
have led to some inconsistencies.
In summary, in children with HR-ALL requiring anthracycline
chemotherapy at the beginning of induction therapy, the
230 Avelar et al.
anthracycline should be able to safely be given before the baseline
echocardiogram. We still recommend an echocardiogram prior
to treatment to provide a reference for possible future changes
of cardiac function and to identify other abnormalities such as
pericardial effusions.
REFERENCES
1. Steinherz LJ, Steinherz PG, Tan CT, et al. Cardiac toxicity 4 to 20
years after completing anthracycline therapy. JAMA 1991;266:
1672–1677.
2. Carroll JB, Hunger SP, Borowitz MJ, et al. Risk-adapted therapy
for children with acute lymphoblastic leukemia (ALL): The
Children’s Oncology Group (COG) approach. Ann Hematol
2008;87: S42–S44.
3. Singal PK, Iliskovic N. Doxorubicin-induced cardiomyopathy.
N Engl J Med 1998;339:900–905.
4. Lipshultz SE, Colan SD, Gelber RD, et al. Late cardiac
effects of doxorubicin (Adriamycin) therapy for childhood
acute lymphoblastic leukemia. N Engl J Med 1991;324:808–
915.
5. Kremer LC, Caron HN. Anthracycline cardiotoxicity in children.
N Engl J Med 2004;351:120–121.
6. 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.
Pediatr Blood Cancer DOI 10.1002/pbc
7. Kapusta L, Groot-Loonen J, Thijssen JM, et al. Regional cardiac
wall motion abnormalities during and shortly after anthracyclines
therapy. Med Pediatr Oncol 2003;41:426–435.
8. Kampmann C, Wiethoff C, Wenzel A, et al. Normal values of M
mode echocardiographic measurements of more than 2000
healthy infants and children in central Europe. Heart 2000;
83:667–672.
9. Ferencz C, Rubin JD, McCarter RJ, et al. Congenital heart dis-
ease: Prevalence at livebirth. The Baltimore–Washington Infant
Study. Am J Epidemiol 1985;121:31–36.
10. George RE, Lipshultz SE, Lipsitz SR, et al. Association between
congenital cardiovascular malformations and neuroblastoma.
Pediatrics 2004;144:444–448.
11. Di Tullio MR. Patent foramen ovale: Echocardiographic detection
and clinical relevance in stroke. J Am Soc Echocardiogr 2010;23:
144–155.
12. Bierman HR, Perkins EK, Ortega P. Pericarditis in patients with
leukemia. Am Heart J 1952;43:413–422.
13. Sampat K, Rossi A, Garcia-Gutierrez V, et al. Characteristics of
pericardial effusions in patients with leukemia. Cancer 2010;116:
2366–2371.
14. Porea TJ, Dreyer ZE, Bricker JT, et al. Evaluation of left ventric-
ular function in asymptomatic children about to undergo anthra-
cycline-based chemotherapy for acute leukemia: An outcome
study. J Ped Heme/Onc 2001;23:420–423.
15. Lipshultz SE, Sleeper LA, Towbin JA, et al. The incidence of
pediatric cardiomyopathy in two regions of the United States.
N Engl J Med 2003;348:1647–1655.