Journal Logo


Cardiovascular Testing Detects Underlying Dysfunction in Childhood Leukemia Survivors


Author Information
Medicine & Science in Sports & Exercise: March 2020 - Volume 52 - Issue 3 - p 525-534
doi: 10.1249/MSS.0000000000002168
  • Free


Concurrent to the introduction of anthracycline chemotherapy into treatment protocols for childhood leukemia there has been a vast increase in survival rates (1,2). However, while anthracyclines have proven successful in the treatment of hematological cancer, they are notorious for their toxicity to the cardiovascular system (1–7). Many of the resulting manifestations of anthracycline cardiovascular toxicity will lay latent for decades before becoming detectable—at which point they have often already developed into a clinically significant disease such as cardiomyopathy or congestive heart failure (2,8). This highlights the importance of life-long follow-up for all exposed survivors—even if they have been deemed healthy upon follow-up assessment in the years immediately after cessation of treatment.

Currently, resting echocardiogram is the standard follow-up procedure used to detect changes in cardiac structure and function after treatment with anthracyclines (9). However, it has been questioned whether this method is sensitive enough to detect subclinical cardiac injury, with no consensus on the most optimal screening tool (10–12). It has been suggested that the heart compensates for initial anthracycline injury during childhood via hypertrophy of the undamaged myocytes, therefore allowing for maintenance of normal cardiac output in the years after treatment (2,13). Although this may be beneficial in the short-term, it inevitably leads to long-term failure with the constant demands and stresses of everyday life (2,13). Based on this, it follows that putting the heart under the stress of exercise may naturally increase cardiovascular demand enough to help unmask initial abnormalities not seen at rest; this promotes the use of an exercise echocardiogram protocol (11,12,14,15). The current body of literature surrounding the use of exercise stress echocardiography in asymptomatic survivors of childhood cancer is inconsistent in its findings regarding efficacy (14–18). Further, the utilized study populations have, for the most part, been heterogeneous in cancer diagnosis and/or cancer treatment. Of note, existing studies tend to employ traditional echocardiographic measures of structure and function—for example, fractional shortening—to assess cardiac toxicity (14,15). Recent advances in imaging, such as speckle tracking strain echocardiography, allow for a more comprehensive analysis of cardiac structure and function than standard two-dimensional (2D) echocardiography alone (19,20). Thus far, strain imaging has proven to be effective at detecting early cardiac dysfunction at rest in combined groups of childhood cancer survivors and in breast cancer survivors (19,21). However, very few studies have attempted to measure strain at peak exercise in this cohort, with only one study successfully able to obtain analyzable images (17,18). To our knowledge, no one has utilized this state-of-the-art imaging technique in combination with regular 2D and color Doppler echocardiography to perform exercise stress echocardiograms in an isolated group of asymptomatic childhood leukemia survivors who have been treated with anthracyclines but no radiotherapy. We hypothesize that the combination of exercise-induced stress and more sensitive echocardiographic imaging will be successful at detecting early changes in cardiac structure and function indicative of future anthracycline cardiac toxicity in a cohort of asymptomatic, apparently healthy adolescent and young adult (AYA) survivors of acute childhood leukemia.

Endothelial dysfunction, overweight and obesity, and low cardiorespiratory fitness are all risk factors for cardiovascular disease (CVD) (22–24). In childhood leukemia survivors, these risk factors may develop as a result of poor lifestyle choices, or arise as secondary consequences of cancer and its treatment (25–27). Irrespective of the cause, the associated complications develop independently of, and supplementary to, anthracycline-mediated cardiovascular toxicity, resulting in complex and often multiple CVD in this population (28–31). Further to this, the presence of CVD risk factors may actually exacerbate or accelerate underlying anthracycline injury, leading to more advanced future disease that is difficult to treat (28–31). In the present study, we assessed endothelial function, body composition, and cardiorespiratory fitness in asymptomatic, anthracycline-treated survivors of childhood leukemia to determine overall CVD predisposition and to examine whether presentation of subclinical anthracycline injury is influenced by the presence of these risk factors.

Overall, our study will provide new insight into whether exercise stress echocardiography with strain imaging is effective at detecting underlying cardiac abnormality and dysfunction in apparently healthy, anthracycline-treated survivors of childhood acute lymphoblastic leukemia and acute myeloid leukemia (ALL and AML, respectively). Additionally, it will help to characterize cardiovascular health in this population, which will allow for a thorough examination of future CVD risk.



Long-term term (≥5 yr) AYA (15–25 yr) survivors of childhood ALL and AML were identified from the oncology database at Princess Margaret Hospital for Children (Western Australia). Survivors with any abnormal cardiac findings on their latest follow-up echocardiograms (eg, presentation of ventricular dilation, systolic and/or diastolic dysfunction, wall thickness or wall motion abnormalities, effusions, signs of disease or infection) were omitted from participation. As we were interested in examining late cardiac structure and function changes specific to treatment with anthracycline chemotherapy, those survivors who had received radiotherapy as part of their treatment were excluded from participation. All patients were treated at Princess Margaret Hospital for Children (ie, treatment was provided by a single center) using uniform treatment protocols for ALL and AML provided by the Children’s Oncology Group (9) (or by the Children’s Cancer Group if treatment was before formation of the Children’s Oncology Group). Cumulative anthracycline dosages were calculated based on guidelines provided by the Children’s Oncology Group (9), while calculations for determining cyclophosphamide equivalent dosages were based on Green et al. (32). Additional exclusion criteria included diagnosis of a CVD and/or presence of any cardiovascular-related signs and symptoms. Healthy individuals of a similar age were recruited from the community to act as a control group for the study.

Power analysis was conducted before commencing the study using G*Power (33). Two-tailed t tests were used to determine the sample size needed to detect changes occurring at an alpha level of 0.05 and a power of 80%, using V˙O2peak as the primary variable. The expected survivor mean and standard deviation for V˙O2peak (34.8 ± 9.3 mL·kg−1·min−1) was sourced from Järvelä et al. (34), who conducted cardiorespiratory fitness testing in 21 AYA survivors of childhood ALL. Input data for controls (46.6 ± 9.4 mL·kg−1·min−1) were taken from a previous article published by our group (35) which utilized the same testing protocol used here to measure cardiorespiratory fitness in 19 healthy, AYA individuals. Using these values, it was determined that 11 participants were needed in each group (ie, 22 participants in total), producing an effect size of 1.26. To increase reliability of our data and to account for participant drop-out, we aimed to recruit 15 individuals per group. Nineteen (9 male, 10 female) eligible survivors agreed to be involved in the study. Nineteen (9 male, 10 female) healthy individuals of a similar age were recruited from the community to act as a control group for the study. However, upon screening, one male and one female control had abnormal clinical findings on their echocardiograms (one with an enlarged aorta and one with supraventricular tachycardia), and so they were excluded from the study. As a result, we have only reported data from 17 (8 male, 9 female) healthy controls.

This study was granted ethical approval by the University of Western Australia Human Research Ethics Committee (reference number, RA/4/1/9090) and the Princess Margaret Hospital Human Research Ethics Committee (approval number, 2016108EP). All participants (or parents/guardians for those younger than 18 yr) were required to provide written informed consent before participation.

Experimental design

Testing for this study occurred over two sessions. The first testing session began with resting measurements of HR and blood pressure (BP), followed by assessment of conduit artery function. After this, anthropometrics were taken and body composition and bone mineral density (BMD) were assessed. The session concluded with measurement of peak cardiorespiratory fitness (V˙O2peak). Cardiac structure and function were assessed using exercise stress echocardiography. Due to limited access to cardiology facilities and equipment, all participants were scheduled to perform their exercise stress echocardiograms over the same weekend, which was within 1 month of the initial testing. Participants were instructed to avoid exercise for 24 h before each appointment.

HR, blood pressure, and endothelial function

Participants were instructed to fast for a minimum of 4 h before vascular assessment. An electronic cuff (Dinamap CARESCAPE V100; GE Healthcare, General Electric Company, Buckinghamshire, UK) was used to measure HR and BP after a 20-min supine rest period. The flow-mediated dilation (FMD) technique (36) was then used to assess endothelial function in the right brachial artery. Continuous (30 Hz) recordings of brachial artery diameter were captured using noninvasive, high-resolution ultrasound (uSmart 3300 NexGen; Terason, Burlington, MA). Measurements of brachial lumen cross-sectional area and Doppler velocity were used to calculate blood flow. A forearm cuff was placed distal to the olecranon process and inflated to 220 mm Hg for 5 min after baseline imaging (1 min). Recordings of diameter and blood flow recommenced 30 s before cuff deflation and continued for a further three min. Validated, custom-designed edge detection and wall-tracking software (37) was used to analyze all scans. Specific details are outlined by Thijssen et al. (36).

Cardiorespiratory fitness

A modified chronotropic treadmill protocol utilized previously by our group (35,38) and deemed as safe for clinical populations was used to measure V˙O2peak. Incremental stages were 3 min in length, with participants encouraged to continue until volitional exhaustion. Every participant started the test at stage 1, which was set at a speed of 3.2 km·h−1 and a gradient of 4%. Speed was increased by 0.8 km·h−1 for the following four stages. Gradient was increased by 1% for stages 2 and 3 and then by 2% for stages 4 and 5. Gradient was then kept the same (at 10%) for the remainder of the assessment. However, speed continued to increase from stage 6 by 1.6 km·h−1. HR (Polar H10; Polar Electro Oy, Professorintie Kempele, Finland) and ratings of perceived exertion (RPE; Borg scale ([6–20,39])) were recorded before exercise, at the end of each stage and at peak exertion to monitor effort. HR was also recorded at 10 min postcessation to ensure appropriate recovery.

A mouthpiece connected to a computerized metabolic cart was used to collect respired air throughout the assessment. Gas analyzers (Ametek Applied Electrochemistry S-3A/1 and CD-3A; AEI Technologies, Pittsburgh, PA) calculated percentages of oxygen and carbon dioxide in this respired air, whereas a ventilometer (Universal Ventilation Meter; VacuMed, Ventura, CA) calculated minute ventilation (E) and respiratory exchange ratio (RER). To ensure accuracy of measurement, calibration of the metabolic cart occurred before each assessment.

Anthropometry, body composition, and bone mineral density

Body mass was measured using an electronic scale (Sauter Model EB60; FSE Scientific, New South Wales, Australia), whereas height was measured using a wall-mounted stadiometer (Seca 216 Measuring Pole, Birmingham, UK). From these measurements, body mass index (BMI) was calculated using formula published by The World Health Organization (WHO) (40). Additionally, height and weight were used to calculate body surface area, using formula published by Mosteller (41). A girth tape was used to measure waist and hip circumference. Waist-to-hip ratio was then calculated by dividing waist circumference by hip circumference. The anthropometric measurement protocols utilized were in accordance with The American College of Sports Medicine (42).

After measurement of anthropometry, body composition and BMD were assessed using dual x-ray absorptiometry (DEXA) Lunar iDXA; GE Healthcare Lunar, General Electric Company, Madison, WI). Total quantities of fat, lean body mass (LBM), and BMD were measured using whole body scans. Fat distribution in the central (trunk and android), peripheral (arms, legs, and gynoid), and visceral regions was also assessed.

Exercise stress echocardiography

Echocardiography was performed by an experienced sonographer using a commercially available ultrasound system (EPIQ-7; Koninklijke Philips N.V., Philips Electronics Australia Limited, New South Wales, Australia). An X5-1 xMATRIX array transducer was used to digitally capture images with participants lying in the left lateral decubitis position. A baseline echocardiogram with tissue harmonic imaging was performed for each participant. This included standard 2D echocardiographic imaging (parasternal long and short-axis, apical long axis, and apical two and four chambers) to assess chamber sizes, wall-motion thickness, aortic dimensions, and valvular structure and function. Left and right ventricular (LV and RV, respectively) measures were captured at baseline, stress (in the 2 min immediately after exercise) and recovery. Of note, stress images were obtained in the sequence of standard 2D imaging views (while the participant was still at peak HR), followed by tissue Doppler assessment (to minimize E/A fusion).

Participants were continuously monitored throughout the assessment using 12-lead ECG (Mortara Instrument, Milwaukee, WI) and manual BP readings. A cardiologist supervised all assessments and reviewed ECG, BP, and echocardiogram findings at each stage of the test to assure safety of participation. An accredited exercise physiologist ran participants through the same incremental treadmill protocol utilized in the V˙O2peak test. Participants were encouraged to achieve similar workloads and HR on both exercise tests, although they ultimately determined when they ceased the assessment (ie, continued until volitional exhaustion). Cine loops were saved in raw DICOM format for analysis.

Data analyses

Ejection fraction (EF) was measured using the Simpson’s biplane method. Indexed stroke volume, diastolic function, and LV longitudinal strain were measured using 2D speckle tracking in the standard apical views. Global longitudinal peak systolic strain was analyzed offline by an experienced technician using the Philips EPIQ 7 Automated Cardiac Motion Quantification (aCMQ) software. Using 2D speckle tracking, the software places a region of interest upon the selected apical views, generating measurements of both global and regional myocardial function. Global circumferential strain was analyzed using the same software. All strain measurements were analyzed by a separate technician. Diastolic function was assessed with transmitral flow velocities and myocardial tissue velocities at the mitral annulus. Right ventricular function was also quantitatively assessed using tricuspid annular plane systolic excursion, fractional area change, and tissue Doppler peak systolic velocity at the tricuspid annulus.

Statistical analyses

Statistical analysis was conducted using IBM SPSS Statistics version 20.0 for Windows (IBM Australia Ltd, New South Wales, Australia). Descriptive data were generated and presented as mean ± standard deviation. Independent samples t tests were used to determine differences between groups in vascular function, physical profile, cardiorespiratory fitness, and baseline cardiac health (resting echocardiography). Two-way mixed design ANOVA were then used to uncover time, group, and interaction effects for echocardiographic measurements; however, as time and group effects were not relevant for the primary questions raised in this study they were not reported in text. Significant interactions were further explored using simple effects testing. All post hoc values were adjusted for multiple comparisons using Fisher’s least significant difference. Significance was set at P ≤ 0.05 for all statistical procedures.


Participant Characteristics

Details regarding cancer treatment for the survivor group are presented in Table 1. The age of the leukemia survivors was 19 ± 3 yr, and the age of the controls was 22 ± 2 yr. Of the 19 leukemia survivors 12 had a diagnosis of ALL, and seven had a diagnosis of AML. The average age of the leukemia survivors at diagnosis was 7 ± 5 yr, resulting in an average time since diagnosis of 12 ± 4 yr. Mean time since final treatment was 12 ± 4 yr. One AML survivor experienced relapse 16 months after initial diagnosis. As this survivor was in adult care by the time he relapsed, we do not have specific details regarding his relapse treatment. We can confirm he did not receive radiotherapy but probably received alkylating agents. Eleven survivors had no history of acute or early-onset cardiac toxicity. Two survivors experienced acute LV dilatation (mild-moderate), LV wall thinning (mild), mitral, and tricuspid valve regurgitation (trivial-mild) and septal dyskinesia (mild) during treatment, and a further three experienced similar changes in the year after treatment. Finally, two survivors had a history of endocarditis and another one experienced a fungal infection of the heart while undergoing chemotherapy. All survivors with a history of acute or early-onset cardiac toxicity had recovered to normal cardiac structure and function at time of participation, as ascertained by their cardiologist.

Cancer treatment details for the leukemia survivors.

HR, Blood Pressure, and Endothelial Function

There were no differences between groups in measures of resting HR or BP. Data on conduit artery function are presented in Table 2. The leukemia survivors had a significantly lower FMD than the controls. There were no significant differences between groups in arterial baseline diameter, peak diameter, delta diameter or in time to peak diameter.

Vascular function and cardiorespiratory fitness in leukemia survivors and controls.

Cardiorespiratory Fitness

Cardiorespiratory fitness data are presented in Table 2. Leukemia survivors had lower maximal and recovery (10 min postexercise cessation) HR compared with controls. Although, when converted to a percentage of HR maximum, recovery HR were not different between groups (survivors, 57.7% ± 7.0%; controls, 60.6% ± 5.2%; P = 0.132). The survivors also had lower V˙E and presented with reduced absolute and relative V˙O2peak. There were no differences in maximal RPE or RER between groups.

Anthropometry, Body Composition, and Bone Mineral Density

The leukemia survivors had a mean stature of 173.0 ± 7.8 cm, body mass of 76.16 ± 19.05 kg, BMI of 25.2 ± 5.1 and body surface area of 1.89 ± 0.24 m2. Average waist circumference was 84.1 ± 12.6 cm, whereas hip circumference was 98.7 ± 12.1 cm; this resulted in a waist-to-hip ratio of 0.85 ± 0.07. These measures were similar to those in the controls, who had an average height of 173.8 ± 9.1 cm (P = 0.796), body mass of 70.07 ± 13.96 kg (P = 0.287), BMI of 22.9 ± 2.7 (P = 0.109), body surface area of 1.82 ± 0.22 m2 (P = 0.396), waist circumference of 77.8 ± 10.0 cm (P = 0.107), hip circumference of 92.9 ± 7.1 cm (P = 0.093) and a waist-to-hip ratio of 0.84 ± 0.07 (P = 0.554).

Measures of DEXA-derived body fat are presented in Figure 1. Compared to controls, the leukemia survivors had higher total body fat (percentage, P = 0.034; mass, P = 0.024) and greater distributions of fat mass in the central (P = 0.050), peripheral (P = 0.039) and visceral (P < 0.001) regions. There were no differences between groups in percentage of central or peripheral fat (P = 0.065 and P = 0.067, respectively), LBM (survivors 46.52 ± 8.71 kg; controls, 48.83 ± 11.57 kg; P = 0.500) or BMD (survivors, 1.19 ± 0.11 g·cm−2; controls, 1.27 ± 0.14 g·cm−2; P = 0.053).

Body fat distribution for the leukemia survivors and the controls, presented as percentage and mass. Central fat encompasses the trunk and android region, whereas peripheral fat incorporates the arms, legs, and gynoid region. *P ≤ 0.05, •P ≤ 0.01.

Exercise Stress Echocardiography

Data from t-tests (used to determine whether groups were matched at rest) are presented in text, irrespective of significance. Full exercise stress echocardiogram measures are presented in Table 3.

Echocardiographic measures that responded similarly within and between groups over the course of the exercise stress echocardiogram.

Resting measures

With regard to cardiac structure and function, leukemia survivors, and controls were matched at rest (LV end diastolic volume, P = 0.946; LV end systolic volume, P = 0.622; LV stroke volume, P = 0.472; LV 2D EF, P = 0.593; RV end diastolic area, P = 0.320; RV end systolic area, P = 0.595; fractional area change P = 0.731).

Likewise, mitral valve (MV) inflow measures were similar between groups (MV peak flow velocity in early diastole [Peak E], P = 0.884; MV peak flow velocity in late diastole [Peak A], P = 0.136; MV deceleration time, P = 0.885; ratio of MV Peak E to Peak A velocity [E/A], P = 0.121).

Tissue Doppler imaging measures of the lateral wall were matched between groups (lateral mitral annulus Peak E’ tissue velocity, P = 0.183; ratio of MV Peak E flow velocity to Peak E’ lateral mitral annulus tissue velocity, P = 0.223). However, medial mitral annulus Peak E’ tissue velocity was reduced in survivors compared to controls (P = 0.006). As a result, the ratio of MV Peak E flow velocity to Peak E’ medial mitral annulus tissue velocity was also different between groups at rest (P = 0.042).

At rest, there were no differences in global longitudinal (P = 0.561) and global circumferential (P = 0.232) strain between leukemia survivors and controls.

Peak exercise measures

Mitral valve Peak A velocity had a significant interaction effect (Fig. 2; P = 0.007). Immediately after peak exercise MV Peak A velocities increased significantly from resting levels in survivors (P = 0.023) and controls (P = 0.020), such that there were no differences between groups in this measure at the exercise timepoint (P = 0.494).

Impact of exercise stress protocol on mitral valve peak flow velocity in late diastole (MV Peak A) for leukemia survivors and controls at rest, immediately after exercise and after recovery. Error bars represent standard deviation. *Significant (P ≤ 0.05) difference between groups for timepoint. °Significant (P ≤ 0.05) difference within group between rest and exercise. Significant (P ≤ 0.05) difference within group between exercise and recovery. •Significant (P ≤ 0.05) difference within group between rest and recovery.

Despite being reduced in the survivors at rest, medial mitral annulus Peak E’ tissue velocity and ratio of MV Peak E flow velocity to Peak E’ medial mitral annulus tissue velocity were similar between groups when measured immediately after peak exercise.

Recovery measures

There was a significant interaction effect for MV E/A ratio (Fig. 3; P < 0.001). Although, MV E/A responded similarly within and between groups at rest and exercise, the MV E/A ratio was significantly lower in the leukemia survivors at recovery compared to the controls (P < 0.001). The MV E/A ratio in the controls at recovery was higher than it was at both rest (P < 0.001) and exercise (P = 0.005). This was driven mainly by MV Peak A velocity, which did not fully recover in the leukemia survivors (Fig. 2; exercise-recovery, P = 0.784). As a result, MV Peak A velocity at recovery was higher than it was at rest (P = 0.012) in this group. In contrast, MV Peak A velocity in the control group dropped after recovery (Fig. 2) to below exercise (P = 0.001) and resting values (P = 0.033). When making comparisons between groups, the leukemia survivors had a significantly increased MV Peak A velocity at recovery compared to the controls (Fig. 2; P < 0.001).

Impact of exercise stress protocol on mitral valve (MV) Peak E to Peak A velocity (MV E/A) for leukemia survivors and controls at rest, immediately after exercise and after recovery. Error bars represent standard deviation. *Significant (P ≤ 0.05) difference between groups for timepoint. Significant (P ≤ 0.05) difference within group between exercise and recovery. •Significant (P ≤ 0.05) difference within group between rest and recovery.

Once again, despite being reduced at rest in the leukemia survivors, medial mitral annulus Peak E’ tissue velocity and ratio of MV Peak E flow velocity to Peak E’ medial mitral annulus tissue velocity were similar between groups at recovery.


We aimed to determine whether exercise-induced stress and the use of contemporary echocardiographic imaging techniques could detect early changes in cardiac structure and function indicative of late anthracycline cardiac toxicity in a cohort of asymptomatic AYA survivors of acute childhood leukemia with normal cardiac structure and function on resting echocardiography. Additionally, we characterized cardiovascular health by assessing recognized CVD risk factors, and used these combined findings to determine overall predisposition to future complications. Overall, we observed some interesting responses in the survivors at recovery which may indicate early cardiac diastolic dysfunction in this cohort. We also report evidence of vascular endothelial dysfunction, increased fat mass and reduced cardiorespiratory fitness that may exacerbate CVD outcomes in this population.

The exercise stress echocardiography protocol used here unmasked changes in atrial workload that may indicate the beginning of restrictive anthracycline-induced disease in the leukemia survivors. The leukemia survivors and controls had a comparable MV Peak A velocity at rest and both groups experienced a proportionate increase in this measure with exercise. However, the leukemia survivors failed to recover after a recuperation period. These findings may indicate reduced LV compliance after stress, with atrial contraction in late diastole possibly increasing to compensate for the lack of relaxed ventricular filling in early diastole. Notably, peak and recovery HR were actually lower in the leukemia survivors compared with the controls, and so MV Peak A velocity was not being influenced by increased HR and, hence, shorter filling times. Both Cifra et al. (17) and Ryerson et al. (18) concluded that exercise echocardiography was not a useful surveillance tool in childhood cancer survivors treated with anthracyclines as they observed normal responses in echocardiography measures with exercise. Interestingly, the survivors in our study also had a normal exercise response with regards to atrial contraction. However, our findings have brought to light the importance of monitoring cardiac structure and function after an extended recovery period as this is when abnormalities can actually become evident—even in the face of a preserved exercise response. To our knowledge, these are novel recovery findings in this cohort and so will need to be corroborated.

The assessment of secondary cardiovascular health measures is important for determining overall risk of disease and possible rate of anthracycline toxicity progression in this population. Endothelial injury promotes atherosclerosis and impairs vasodilation, which predisposes to coronary artery disease (22,36). Endothelial dysfunction may develop as a result of a poor lifestyle or be a consequence of injury from anthracyclines, with dysfunction evident in those survivors who eventually develop cardiomyopathy and congestive heart failure (25,31,43). In the present study, we observed endothelial dysfunction in the leukemia survivor cohort, indicating an increased risk of CVD. Additionally, the reduced FMD seen here indicates increased systemic vascular resistance, which may eventually influence afterload and cardiac volumes (25). This may lead to accelerated anthracycline-mediated disease in our leukemia survivors. Our results promote the use of FMD assessment in long-term follow-up to detect early cardiovascular changes indicative of disease concomitant to that caused by direct anthracycline injury.

Atherosclerosis may further be accelerated by an excessive accumulation of fat in the body (23). In particular, central and visceral fat function as endocrine organs, releasing cytokines that result in a plethora of metabolic complications that encourage plaque formation and the development of chronic disease (23). In the current study, the leukemia survivors had greater amounts of total body fat compared to the controls, with significant proportions of this distributed in the central and visceral compartments. This is a noteworthy finding as we have demonstrated that the vascular endothelium in our survivors is already injured and is therefore more prone to fatty lesions. Moreover, obesity has been shown to worsen severity of cardiac toxicity (44) and can independently result in cardiomyopathy (28,29). Hence, our survivors may be at risk of multiple and enhanced CVD.

Like obesity, cardiorespiratory fitness is a strong and independent predictor of cardiovascular mortality, with lower levels of fitness inversely correlating with a higher risk of death (24,30,45). In a cohort already at risk of anthracycline-mediated cardiac dysfunction, low fitness is a detrimental adjunct complication with the possibility to advance progression of disease. Unfortunately, we observed a decreased V˙O2peak in the leukemia survivors used here, indicating a heightened risk of cardiovascular death. Additionally, such deconditioning implies that these survivors may experience difficulties upon physical exertion in everyday life.

At rest, medial mitral annular tissue velocity was reduced, and the ratio of MV flow velocity to medial mitral annular tissue velocity was increased in the leukemia survivors. It has been found that changes in tissue velocity often succeed those in strain in the progression of cardiomyopathy and heart failure (46). Additionally, the ratio of early diastolic mitral inflow to early diastolic mitral annular tissue velocity has a good correlation with LV filling pressure, making it useful for the diagnosis of diastolic dysfunction (46,47). Studies show that the medial annulus provides improved sensitivity and specificity for the diagnosis of diastolic dysfunction compared with the lateral annulus, which is quite dynamic (47). This may explain why we saw no differences in resting measures of the lateral wall. Notably, although we observed these medial tissue velocity abnormalities at rest, the leukemia survivors responded similarly to the controls at exercise and recovery, indicating that exercise actually ameliorates resting diastolic dysfunction. Similar findings have been reported by Ryerson et al. (18), who found reduced MV E/A at rest that was augmented with exercise in survivors of childhood cancer who were deemed “high risk” based on increased anthracycline dosages and young age at exposure. Interestingly though, in the study by Ryerson et al. (18), the “high risk” survivors achieved normal V˙O2peak, leading the authors to conclude that the improved MV E/A at exercise allowed for normal cardiorespiratory function. Conversely, our survivors had reduced V˙O2peak in the face of normal exercise Doppler velocities and normal exercise cardiac structure and function (eg, matched LVEF and strain). This indicates that reduced cardiorespiratory fitness is actually more attributable to deficiencies in peripheral circulation rather than central abnormalities. This theory aligns with our FMD data. The increase in systemic vascular resistance caused by impaired vasodilation of the arteries may have contributed to the lower V˙O2peak reported here. Further investigation into these findings is warranted.

The current body of literature surrounding the use of exercise stress echocardiography for the detection of subclinical disease in asymptomatic survivors of childhood cancer is inconsistent regarding outcomes (14–18). Overall, we report positive findings that support the use of an exercise echocardiogram protocol. We have provided some interesting and novel evidence that demonstrates that the recovery period after exercise may be important for unmasking subclinical cardiac toxicity that may not be evident at rest or peak exercise. Although we did not see any changes in strain, the abnormalities in resting tissue velocities are an early indicator of future dysfunction, and so this measure should not be discounted. Our supplementary findings regarding endothelial dysfunction, increased fat mass, and cardiorespiratory insufficiency in the leukemia survivors demonstrate that resting echocardiography alone may not be the most optimal screening tool in this population for assessing future CVD risk. Specifically, they highlight that, although the heart may initially be able to adequately compensate for underlying cardiac injury caused by anthracyclines, peripheral abnormalities progress, greatly predisposing to additional CVD and possibly accelerating anthracycline toxicity. Including these measures into long-term follow-up will allow for a more comprehensive analysis of cardiovascular health in this population and may even allow for earlier detection and prevention of subclinical abnormalities.

Although the sample size used here was relatively small, and we cannot determine the prognostic significance of our findings—especially in the long-term—we believe our study provides a strong foundation for future research studies utilizing larger populations of leukemia survivors treated explicitly with anthracycline chemotherapy. We acknowledge that stress and strain measures can be difficult to capture at peak exercise. However, we followed a strict imaging protocol that has previously been validated and proven as feasible (48,49).

This study has demonstrated that exercise echocardiography is useful for unmasking early dysfunction that may lead to future anthracycline-mediated disease in apparently healthy AYA survivors of acute childhood leukemia. Further to this, we have demonstrated that the inclusion of other cardiovascular screening measures into long-term follow-up is integral – not only for providing a more in-depth analysis of future CVD risk but for allowing for earlier detection of cardiovascular abnormalities.

We would like to acknowledge Dr. Denitza Mironova for her assistance with screening cancer survivors for eligibility and for aiding with recruitment. We would also like to thank Mrs. Lauren C Chasland for her help with data collection and Miss Melissa Sale who was the sonographer for the study. This work was supported by The Heart Foundation Vanguard Grant (LHN) award number 101412.

The results of the present study do not constitute endorsement by ACSM and are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. There are no other conflicts of interest to acknowledge.


1. Lipshultz SE, Adams MJ. Cardiotoxicity after childhood cancer: beginning with the end in mind. Am Soc Clin Oncol. 2010;28:1276–81.
2. Lipshultz SE, Alvarez JA, Scully RE. Anthracycline associated cardiotoxicity in survivors of childhood cancer. Heart. 2008;94(4):525–33.
3. Giantris A, Abdurrahman L, Hinkle A, Asselin B, Lipshultz SE. Anthracycline-induced cardiotoxicity in children and young adults. Crit Rev Oncol Hematol. 1998;27(1):53–68.
4. Menna P, Gonzalez Paz O, Chello M, Covino E, Salvatorelli E, Minotti G. Anthracycline cardiotoxicity. Expert Opin Drug Saf. 2012;11(1 Suppl):S21–36.
5. Nysom K, Holm K, Lipsitz SR, et al. Relationship between cumulative anthracycline dose and late cardiotoxicity in childhood acute lymphoblastic leukemia. J Clin Oncol. 1998;16(2):545–50.
6. Sorensen K, Levitt GA, Bull C, Dorup I, Sullivan ID. Late anthracycline cardiotoxicity after childhood cancer. Cancer. 2003;97(8):1991–8.
7. Lipshultz SE, Colan SD, Gelber RD, Perez-Atayde AR, Sallan SE, Sanders SP. Late cardiac effects of doxorubicin therapy for acute lymphoblastic leukemia in childhood. N Engl J Med. 1991;324(12):808–15.
8. Carver JR, Shapiro CL, Ng A, et al. American Society of Clinical Oncology clinical evidence review on the ongoing care of adult cancer survivors: cardiac and pulmonary late effects. J Clin Oncol. 2007;25(25):3991–4008.
9. Children's Oncology Group. Long-Term Follow-Up Guidelines for Survivors of Childhood, Adolescent, and Young Adult Cancers, Version 5.0. Monrovia, CA: Children's Oncology Group; 2018.
10. Armstrong GT, Plana JC, Zhang N, et al. Screening adult survivors of childhood cancer for cardiomyopathy: comparison of echocardiography and cardiac magnetic resonance imaging. J Clin Oncol. 2012;30(23):2876–84.
11. De Souza AM, Potts JE, Potts MT, et al. A stress echocardiography study of cardiac function during progressive exercise in pediatric oncology patients treated with anthracyclines. Pediatr Blood Cancer. 2007;49(1):56–64.
12. Sieswerda E, Kremer LC, Vidmar S, et al. Exercise echocardiography in asymptomatic survivors of childhood cancer treated with anthracyclines: a prospective follow-up study. Pediatr Blood Cancer. 2010;54(4):579–84.
13. Adams MJ, Lipshultz SE. Pathophysiology of anthracycline-and radiation-associated cardiomyopathies: implications for screening and prevention. Pediatr Blood Cancer. 2005;44(7):600–6.
14. Jarfelt M, Kujacic V, Holmgren D, Bjarnason R, Lannering B. Exercise echocardiography reveals subclinical cardiac dysfunction in young adult survivors of childhood acute lymphoblastic leukemia. Pediatr Blood Cancer. 2007;49(6):835–40.
15. Smibert E, Carlin JB, Vidmar S, Wilkinson LC, Newton M, Weintraub RG. Exercise echocardiography reflects cumulative anthracycline exposure during childhood. Pediatr Blood Cancer. 2004;42(7):556–62.
16. Guimaraes-Filho FV, Tan DM, Braga JC, Rodrigues A, Waib PH, Matsubara BB. Ventricular systolic reserve in asymptomatic children previously treated with low doses of anthracyclines: a longitudinal, prospective exercise echocardiography study. Pediatr Blood Cancer. 2012;59(3):548–52.
17. Cifra B, Chen CK, Fan CS, et al. Dynamic myocardial response to exercise in childhood cancer survivors treated with anthracyclines. J Am Soc Echocardiogr. 2018;31(8):933–42.
18. Ryerson AB, Border WL, Wasilewski-Masker K, et al. Assessing anthracycline-treated childhood cancer survivors with advanced stress echocardiography. Pediatr Blood Cancer. 2015;62(3):502–8.
19. Armstrong GT, Joshi VM, Ness KK, et al. Comprehensive echocardiographic detection of treatment-related cardiac dysfunction in adult survivors of childhood cancer: results from the St. Jude Lifetime Cohort Study. J Am Coll Cardiol. 2015;65(23):2511–22.
20. Venneri L, Zoppellaro G, Khattar RS. Cardio-oncology: the role of advanced echocardiography in cancer patients. Expert Rev Cardiovasc Ther. 2018;16(4):249–58.
21. Zhang KW, Finkelman BS, Gulati G, et al. Abnormalities in 3-dimensional left ventricular mechanics with anthracycline chemotherapy are associated with systolic and diastolic dysfunction. JACC Cardiovasc Imaging. 2018;11(8):1059–68.
22. Celermajer DS, Sorensen KE, Gooch VM, et al. Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. The Lancet. 1992;340(8828):1111–5.
23. Fukuhara A, Matsuda M, Nishizawa M, et al. Visfatin: a protein secreted by visceral fat that mimics the effects of insulin. Science. 2005;307(5708):426–30.
24. Ekelund LG, Haskell WL, Johnson JL, et al. Physical fitness as a predictor of cardiovascular mortality in asymptomatic North American men. N Engl J Med. 1988;319(21):1379–84.
25. Di Lisi D, Madonna R, Zito C, et al. Anticancer therapy-induced vascular toxicity: VEGF inhibition and beyond. Int J Cardiol. 2017;227:11–7.
26. Florin TA, Fryer GE, Miyoshi T, et al. Physical inactivity in adult survivors of childhood acute lymphoblastic leukemia: a report from the childhood cancer survivor study. Cancer Epidemiol Biomarkers Prev. 2007;16(7):1356–63.
27. Gurney JG, Ness KK, Sibley SD, et al. Metabolic syndrome and growth hormone deficiency in adult survivors of childhood acute lymphoblastic leukemia. Cancer. 2006;107(6):1303–12.
28. Bhatheja S, Panchal HB, Ventura H, Paul TK. Obesity cardiomyopathy: pathophysiologic factors and nosologic reevaluation. Am J Med Sci. 2016;352(2):219–22.
29. Dela Cruz CS, Matthay RA. Role of obesity in cardiomyopathy and pulmonary hypertension. Clin Chest Med. 2009;30(3):509–23.
30. Sandvik L, Erikssen J, Thaulow E, Erikssen G, Mundal R, Rodahl K. Physical fitness as a predictor of mortality among healthy, middle-aged Norwegian men. N Engl J Med. 1993;328(8):533–7.
31. Zsary A, Szucs S, Keltai K, et al. Endothelins: a possible mechanism of cytostatics-induced cardiomyopathy. Leuk Lymphoma. 2004;45(2):351–5.
32. Green DM, Nolan VG, Goodman PJ, et al. The cyclophosphamide equivalent dose as an approach for quantifying alkylating agent exposure: a report from the Childhood Cancer Survivor Study. Pediatr Blood Cancer. 2014;61(1):53–67.
33. Faul F, Erdfelder E, Lang AG, Buchner A. G* Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007;39(2):175–91.
34. Järvelä LS, Niinikoski H, Lähteenmäki P, et al. Physical activity and fitness in adolescent and young adult long-term survivors of childhood acute lymphoblastic leukaemia. J Cancer Surviv. 2010;4(4):339–45.
35. Long TM, Rath SR, Maroni TD, et al. Fitness, body composition and vascular health in adolescent and young adult survivors of paediatric brain cancer and cranial radiotherapy. Int J Adolesc Med Health. 2017.
36. Thijssen DH, Black MA, Pyke KE, et al. Assessment of flow-mediated dilation in humans: a methodological and physiological guideline. Am J Physiol Heart Circ Physiol. 2011;300(1):H2–12.
37. Woodman R, Playford D, Watts G, et al. Improved analysis of brachial artery ultrasound using a novel edge-detection software system. J Appl Physiol. 2001;91(2):929–37.
38. Long TM, Rath SR, Wallman KE, et al. Exercise training improves vascular function and secondary health measures in survivors of pediatric oncology related cerebral insult. PLoS One. 2018;13(8):e0201449.
39. Borg GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc. 1982;14(5):377–81.
40. Global Database on Body Mass Index [Internet]. 2017. Available from:
41. Mosteller R. Simplified calculation of body surface area. N Engl J Med. 1987;317:1098.
42. The American College of Sports Medicine. ACSM’s Guidelines for Exercise Testing and Prescription. Ninth ed. 2014. 480 p.
43. Wolf MB, Baynes JW. The anti-cancer drug, doxorubicin, causes oxidant stress-induced endothelial dysfunction. Biochimica et Biophysica Acta (BBA)-General Subjects. 2006;1760(2):267–71.
44. Guenancia C, Lefebvre A, Cardinale D, et al. Obesity as a risk factor for anthracyclines and trastuzumab cardiotoxicity in breast cancer: a systematic review and meta-analysis. J Clin Oncol. 2016;34(26):3157–65.
45. Blair SN, Kampert JB, Kohl HW 3rd, et al. Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women. JAMA. 1996;276(3):205–10.
46. Abraham TP, Dimaano VL, Liang HY. Role of tissue Doppler and strain echocardiography in current clinical practice. Circulation. 2007;116(22):2597–609.
47. Srivastava PM, Burrell LM, Calafiore P. Lateral vs medial mitral annular tissue Doppler in the echocardiographic assessment of diastolic function and filling pressures: which should we use? Eur J Echocardiogr. 2005;6(2):97–106.
48. Caniggia CC, Amor M, Lowenstein D, et al. Feasibility and contribution of global and regional 2D strain during exercise stress echocardiography. Argentine Journal of Cardiology. 2014;82(2):110–7.
49. Goebel B, Arnold R, Koletzki E, et al. Exercise tissue Doppler echocardiography with strain rate imaging in healthy young individuals: feasibility, normal values and reproducibility. Int J Cardiovasc Imaging. 2007;23(2):149–55.


© 2020 American College of Sports Medicine