Recommendations for Cardiomyopathy Surveillance for Survivors of Childhood Cancer: A Report from the International Late Effects of Childhood Cancer Guideline Harmonization Group

Saro H Armenian; Melissa M Hudson; Renee L Mulder; Ming Hui Chen; Louis S Constine; Mary Dwyer; Paul C Nathan; Wim JE Tissing; Sadhna Shankar; Elske Sieswerda; Rod Skinner; Julia Steinberger; Elvira C van Dalen; Helena van der Pal; W Hamish Wallace; Gill Levitt; Leontien CM Kremer

doi:10.1016/S1470-2045(14)70409-7

. Author manuscript; available in PMC: 2016 Mar 1.

Published in final edited form as: Lancet Oncol. 2015 Mar;16(3):e123–e136. doi: 10.1016/S1470-2045(14)70409-7

Recommendations for Cardiomyopathy Surveillance for Survivors of Childhood Cancer: A Report from the International Late Effects of Childhood Cancer Guideline Harmonization Group

Saro H Armenian ¹, Melissa M Hudson ², Renee L Mulder ³, Ming Hui Chen ⁴, Louis S Constine ⁵, Mary Dwyer ⁶, Paul C Nathan ⁷, Wim JE Tissing ⁸, Sadhna Shankar ⁹, Elske Sieswerda ³, Rod Skinner ¹⁰, Julia Steinberger ¹¹, Elvira C van Dalen ³, Helena van der Pal ¹², W Hamish Wallace ¹³, Gill Levitt ¹⁴, Leontien CM Kremer ³

PMCID: PMC4485458 NIHMSID: NIHMS669480 PMID: 25752563

Abstract

Childhood cancer survivors treated with anthracycline chemotherapy or chest radiation are at an increased risk of developing congestive heart failure (CHF). In this population, CHF is well-recognized as a progressive disorder, with a variable period of asymptomatic cardiomyopathy which precedes signs and symptoms. As a result, a number of practice guidelines have been developed to facilitate detection and treatment of asymptomatic cardiomyopathy. These guidelines differ with regards to definitions of at risk populations, surveillance modality and frequency, and recommendations for interventions. These differences may hinder the effective implementation of these recommendations. We report on the results of an international collaboration to harmonize existing cardiomyopathy surveillance recommendations, using an evidence-based approach that relied on standardized definitions for outcomes of interest and transparent presentation of the quality of the evidence. The resultant recommendations were graded according to the quality of the evidence and the potential benefit gained from early detection and intervention.

INTRODUCTION

Advances in treatment strategies for childhood cancer have resulted in marked improvements in survival, with current 5-year survival rates approaching 80%.¹ However this improvement in outcome is has been compromised by the occurrence of long term morbidities of therapy. The cumulative incidence of severe or life-threatening chronic health conditions exceeds 40% for childhood cancer survivors surviving 30 years after primary diagnosis.^{2, 3} These conditions include second malignant neoplasms, endocrine disorders, cardiopulmonary dysfunction, renal dysfunction, and neurosensory impairment.^{2, 3}

Cardiovascular complications (such as coronary artery disease, and stroke, but especially congestive heart failure [CHF]) have emerged as a leading cause of morbidity and mortality in long-term survivors of childhood cancer.⁴ In fact, childhood cancer survivors are at a 15-fold increased risk of developing CHF² and are at 7-fold higher risk of premature death due to cardiac causes,⁵ when compared with the general population. There is a strong dosedependent relation between anthracycline chemotherapy exposure and CHF risk, and the risk is higher among those exposed to chest radiation.⁴ The incidence of CHF is <5% with cumulative anthracyclines exposure of <250 mg/m²; approaches 10% at doses between 250 and 600 mg/m²; and exceeds 30% for doses >600 mg/m².^{4, 6–8} Of note, nearly 60% of all childhood cancer survivors carry a history of prior anthracycline and/or chest radiation exposure.^{9, 10}

The American College of Cardiology/American Heart Association (ACC/AHA) guidelines for the diagnosis and management of CHF describe heart failure as a progressive disorder, with a variable period of asymptomatic cardiac dysfunction which precedes clinically overt signs and symptoms.¹¹ For anthracycline-exposed survivors, the asymptomatic stage is often characterized by thinning of the left ventricular (LV) wall, enlargement of LV diameter, and subsequent increase in LV wall stress, a clinical picture similar to dilated cardiomyopathy.^{4, 12} These subclinical changes can result in impairment of LV systolic function, manifesting as decreased ejection fraction (EF) and/or shortening fraction (SF).^{4, 12} It is important to recognize, however, that anthracycline-exposed survivors could, over time, also develop restrictive cardiomyopathy, resulting in abnormal E/A ratio (peak early atrial divided by peak late atrial velocities), or prolonged isovolumic relaxation time (IVRT) in the setting of preserved EF/SF.^{4, 12} Individuals who receive chest radiation may be at an especially high risk of developing combined dilated and restrictive cardiomyopathy that results from myocardial fibrosis primarily due to radiation effects on the supporting vasculature.^{4, 12}

In childhood cancer survivors, there is often a long latency between cardiotoxic exposure and clinically evident disease.^{4, 12} As a result, a number of clinical practice guidelines have been developed to facilitate early detection and treatment of asymptomatic cardiomyopathy.^13–16 These guidelines were developed by various North American and European groups and they differ with regards to definitions of at risk populations, surveillance modality and frequency, and recommendations for interventions. These differences may, in turn, hinder the effective implementation of screening across a wide spectrum of clinical settings. Recognizing the importance for collaboration, an international effort was organized to harmonize existing late effects screening recommendations for survivors of childhood cancer.¹⁷ The current effort represents the summary of the evidence and recommendations for cardiomyopathy surveillance in childhood cancer survivors treated with anthracyclines and/or chest radiation.

METHODS

A description of the international guideline harmonization effort and methodology has been provided elsewhere.¹⁷ The cardiomyopathy surveillance recommendations were prepared by representatives from the North American Children’s Oncology Group (COG),¹³ the Dutch Childhood Oncology Group (DCOG),¹⁴ the Scottish Intercollegiate Guidelines Network (SIGN),¹⁶ and the United Kingdom Children’s Cancer and Leukaemia Group (UKCCLG).¹⁵ The current effort encompassed published guidelines that were developed following systematic evaluation of the quality of the late effects literature, linking therapeutic interventions with adverse outcomes. The expert membership included pediatric and adult cardiologists, pediatric oncologists, radiation oncologists, epidemiologists, methodologists, nurses and other survivorship care providers.

The initial step of the cardiomyopathy harmonization effort involved identifying areas of concordance and discordance across the COG, DCOG, SIGN, and UKCCLG guidelines. In order to achieve consensus, clinical questions were devised to address areas of discordance for cardiomyopathy surveillance. Systematic literature searches were performed to update previous systematic searches for asymptomatic¹⁸ and symptomatic¹⁹ cardiomyopathy (search strategy through December 2012: Appendix 1), and evidence summaries were formed to address areas of discordance. When evidence was lacking for childhood cancer survivors, we extrapolated information from other populations. In the case of concordance, we extracted and evaluated the evidence cited by the guidelines.

Given the heterogeneity in definitions used to describe relevant therapeutic exposures, surveillance strategies, and cardiovascular outcomes, we proposed standardized definitions which were incorporated into our literature review and final formulation of recommendations. Childhood cancer survivors included individuals treated for cancer up to 21 years of age, regardless of current age. Anthracyclines chemotherapy consisted of: doxorubicin, daunorubicin, epirubicin, idarubicin; the anthraquinone mitoxantrone was also included due to its similar cardiotoxic profile. Chest radiation included any radiation in which the heart was in the field of treatment (mediastinal, thoracic, spinal, left or whole upper abdominal or total body irradiation [TBI]). Asymptomatic cardiomyopathy was defined as a decline in LV systolic function (abnormal EF, SF, wall stress)^20–22 or diastolic dysfunction (abnormal E/A ratio, prolonged IVRT)^{22, 23} in the context of preserved EF, without corresponding symptoms of heart failure. CHF was defined per the ACC/AHA guidelines,¹¹ and corresponded to symptomatic cardiomyopathy with evidence of cardiac dysfunction on imaging studies. The current effort does not address screening for other known therapy-associated cardiovascular complications (coronary artery disease, carotid artery disease, pericardial fibrosis, conduction abnormalities, or valvular stenosis/insufficiency); these will be addressed by future collaborations.

The quality of the evidence and the strength of the recommendations were determined according to criteria that were based on modified Grading of Recommendations Assessment Development and Evaluation (GRADE) and the ACC/AHA classification for recommendations (Appendix 2).^{24, 25} Final recommendations relied on this scientific knowledge combined with other considerations such as clinical judgements, decisions about thresholds, costs, and potential harms from excessive screening. The harmonized cardiomyopathy surveillance recommendations were critically appraised by two external experts (K.O. and J.B.) in the field.

RESULTS

Discordances and concordances among the cardiomyopathy surveillance recommendations are provided in Table 1. There was concordance across guidelines for the following statements:

Childhood cancer survivors treated with anthracyclines (including mitoxantrone) or chest radiation are at increased risk of cardiomyopathy.
Surveillance using echocardiography should be lifelong and performed at a minimum of every five years.
Given the increased cardiometabolic demand on the heart of the mother during pregnancy, closer monitoring of survivors during pregnancy is warranted.
Survivors with documented asymptomatic cardiomyopathy should be referred to a cardiologist for further diagnostic work-up and possible treatment.
At risk cancer survivors should be regularly screened for traditional cardiovascular risk factors (i.e.: hypertension, diabetes, dyslipidemia, overweight/obesity) and should be counseled against smoking and physical inactivity.

Levels of evidence to support concordant areas are included in Table 2.

Table 1.

Concordances and discordances among cardiomyopathy surveillance recommendations

Who needs cardiomyopathy surveillance?
At risk
Anthracyclines	Yes	Yes	Yes	Yes	Concordance
Mitoxantrone	Yes	Yes	Yes	Yes	Concordance
Differing risk by anthracycline analogues	Yes	Not stated	Not stated	Not stated	Discordance
Chest Radiation^*	Yes	Yes	Yes	Yes	Concordance
CV risk factors	Yes	Yes	Yes	Yes	Concordance
Highest risk	≥300 mg/m² anthracyclines ≥30 Gy RT involving heart Anthracyclines + chest RT Younger age at treatment Pregnancy	≥300 mg/m² anthracyclines ≥30 Gy RT involving heart Anthracyclines + chest RT Pregnancy	>250 mg/m² anthracyclines Anthracyclines + chest RT Hx of transient cardiomyopathy during treatment Pregnancy	>250 mg/m² anthracyclines ≥30 Gy RT involving heart Anthracyclines + chest RT	Discordance
What surveillance modality should be used?
Screening for cardiomyopathy
Echocardiography	Yes	Yes	Yes	Yes	Concordance
Radionuclide angiography	Yes	Yes	No	No	Discordance
At what frequency and for how long should cardiomyopathy surveillance be performed?
Screening begins	≥2 yrs after treatment or ≥5 yrs after dx (whichever is first)	≥5 yrs after diagnosis	1–3 months after treatment	≥5 yrs after completion of treatment	Discordance
Screening frequency	Every 1 –5 yrs	Every 2–5 years	Every 3–5 yrs	Every 2–5 yrs	Discordance
Duration of screening	Lifelong	Lifelong	Not stated	Not stated	Discordance
Closer monitoring during pregnancy	Yes	Yes	Yes	Yes	Concordance
Refer to cardiologist	Yes	Yes	Yes	Yes	Concordance
Consider ACE-inhibitors	Not stated	Yes	Not stated	Yes	Discordance

Who needs cardiomyopathy surveillance?	Level of evidence
Risk by anthracycline dose

- Exponential increase in risk for symptomatic cardiomyopathy with increasing lifetime cumulative dose	Level A^{6, 19, 26, 27}
- Childhood cancer survivors treated with cumulative anthracycline dose ≥250 mg/m2 are at highest risk for symptomatic cardiomyopathy	Level A^{6, 19, 26, 27}
- Increased risk for asymptomatic cardiomyopathy with increasing cumulative dose	Level A^{18, 21, 38, 90}

Risk by age at anthracycline exposure

- Increased risk for symptomatic cardiomyopathy with younger age at exposure	Conflicting evidence^{6, 8, 26, 33}
- Increased risk for asymptomatic cardiomyopathy with younger age at exposure	Conflicting evidence^{18, 90, 91}

Risk by anthracycline derivatives (including mitoxantrone)

- Cardiomyopathy has been associated with all anthracycline derivatives	Level A⁹²
- Daunorubicin is as cardiotoxic as doxorubicin when given at an equieffective dose	Level C^{6, 26, 92}
- Epirubicin is less cardiotoxic than doxorubicin when given at an equieffective dose	No evidence
- Idarubicin is more cardiotoxic than doxorubicin when given at an equieffective dose	No evidence
- Mitoxantrone is more cardiotoxic than doxorubicin when given at an equieffective dose	No evidence

Risk by chest radiation dose

- Increased risk for symptomatic cardiomyopathy with increasing radiation dose to cardiac tissues	Level A^{6, 8, 26, 28, 29}
- Childhood cancer survivors treated with chest radiation dose ≥35 Gy are at highest risk for symptomatic cardiomyopathy	Level B^{6, 26}
- Increased risk for asymptomatic cardiomyopathy with increasing radiation dose to cardiac tissues	Level B^{90, 93, 94}

Risk following anthracycline and chest radiation exposure

- Increased risk after anthracycline and chest radiation exposure	Level A^{8, 19, 26}

Risk following conditioning with total body irradiation (TBI)

- There is no increased risk following conditioning with TBI	Level B^{31, 95, 96}

Risk due to modifiable cardiovascular risk factors

- Increased risk in anthracycline- and/or radiation- exposed survivors who develop modifiable cardiovascular risk factors (hypertension, diabetes, dyslipidemia, obesity)	Level B^{74, 97}

What surveillance modality should be used?

Diagnostic value of echocardiography

- Good diagnostic value of 2D echocardiography for detection of asymptomatic cardiomyopathy in childhood cancer survivors	Level B^{41, 98–100}

Diagnostic value of cardiac magnetic resonance imaging (CMR)

- Good diagnostic value of CMR for detection of asymptomatic cardiomyopathy in childhood cancer survivors	Level B⁴¹

Diagnostic value of radionuclide angiography

- Good diagnostic value for detection of asymptomatic cardiomyoathy in childhood cancer survivors	Level C^{101, 102}

Diagnostic value of blood biomarkers of cardiac injury and remodeling

- Poor diagnostic value of cardiac troponins (Troponin-T) for detection of asymptomatic cardiomyopathy in childhood cancer survivors	Level B^45–47

- Poor diagnostic value of cardiac troponins (Troponin-I) for detection of asymptomatic cardiomyopathy in childhood cancer survivors	Level C⁴⁸

- Poor diagnostic value of natriuretic peptides (ANP, BNP, NT Pro-BNP) for detection of asymptomatic cardiomyopathy in childhood cancer survivors	Level B^{45, 93, 103, 104}

Cost-benefit of surveillance in childhood cancer survivors

- Screening for asymptomatic cardiomyopathy using conventional imaging or blood biomarkers is cost-effective.	No evidence

Cost-benefit of surveillance in other populations

- Screening for asymptomatic cardiomyopathy using conventional imaging or blood biomarkers is cost-effective.	Level B⁵⁰

At what frequency and for how long should surveillance for cardiomyopathy be performed?

- High risk childhood cancer survivors have a more rapid rate of deterioration in cardiac function when compared to moderate/low-risk survivors	No evidence

- There is a more rapid rate of deterioration in cardiac function during puberty	No evidence

- Female childhood cancer survivors who have asymptomatic cardiomyopathy at the time of becoming pregnant are at risk for symptomatic cardiomyopathy during pregnancy/delivery	Level C⁵⁶

- Female childhood cancer survivors treated with anthracyclines or radiation who have normal LV systolic function at the time of becoming pregnant are not at increased risk for deterioration in cardiac function during pregnancy/delivery	Level C^{56, 57}

- The risk for deterioration in cardiac function continues to increase with longer follow-up	Level B^{6, 8, 19, 26, 90}

What should be done when abnormalities are detected during surveillance?

Utility of medical interventions in childhood cancer survivors

- ACE-inhibitors are effective for improving cardiac function in survivors with asymptomatic cardiomyopathy	No evidence¹⁰⁵

- Beta-blockers are effective for improving cardiac function in survivors with asymptomatic cardiomyopathy	No evidence¹⁰⁵

- Other interventions such as angiotensin II receptor blockers or placement of ICD can be effective for improving cardiac function for prevention of sudden arrhythmic cardiac death in survivors with asymptomatic cardiomyopathy	No evidence¹⁰⁵

Utility of medical interventions in other populations

- ACE-inhibitors are effective for improving cardiac function in individuals with asymptomatic cardiomyopathy	Level A^{60, 80–82}

- Beta-blockers are effective for improving cardiac function in individuals with asymptomatic cardiomyopathy	Level C^{60, 106–109}

- Other interventions such as angiotensin II receptor blockers or placement of ICD can be effective for improving cardiac function or for prevention of arrhythmic cardiac death in survivors with asymptomatic cardiomyopathy	Level C^{60, 109, 110}

What are the limitations for physical activity?

Role of physical activity in childhood cancer survivors

- Regular physical exercise, as recommended by the AHA and ESC, is beneficial for childhood cancer survivors with normal LV systolic function	Level C⁶⁶

- Regular physical exercise, as recommended by the AHA and ESC, is beneficial for childhood cancer survivors with asymptomatic cardiomyopathy	No evidence

- Participation in high intensity exercise increases the risk for cardiac functional deterioration in childhood cancer survivors	No evidence

Role of physical activity in other populations

- Regular physical exercise, as recommended by the AHA and ESC, is beneficial for individuals who have normal cardiac function	Level A^{62, 63}

- Regular physical exercise, as recommended by the AHA and ESC, is beneficial for individuals who have normal cardiac function, but at risk for cardiomyopathy due to genetic susceptibility	Level B^{67, 68}

- Participation in high intensity exercise increases the risk for cardiac functional deterioration in individuals with asymptomatic cardiomyopathy	Level B⁶³

Risk Group	Anthracycline dose (mg/m²)	Chest radiation dose (Gy)	Anthracycline (mg/m²) + Chest radiation (Gy)
High	≥ 250	≥ 35	≥ 100 (Anthracycline) + ≥ 15 (Radiation)
Moderate	100 to < 250	≥ 15 to < 35	--
Low	< 100	--	--

Conclusions of evidence	Study quality	Study findings	Wording in conclusions
A High level of evidence	Evidence from well performed and high quality studies or systematic reviews (low risk of bias, direct,^* consistent, precise)	If a risk factor is significantly associated with the outcome in ≥95% of studies	‘There is evidence that…’
B Moderate/Low level of evidence	Evidence from studies or systematic reviews with few important limitations	If a risk factor is significantly associated with the outcome in ≥50% of the studies reporting on this risk factor, and in the remaining studies this association is not significant	‘Evidence suggests that…’
C Very low level of evidence	Evidence from studies with serious flaws (high risk of bias, inconsistent, indirect^*, imprecise)	If a risk factor is significantly associated with the outcome in 1 study	‘Some evidence suggests that…’
		If a risk factor is significantly associated with the outcome in <50% of the studies, while in the remaining studies this association is not significant
		If a risk factor is significantly (either positively or negatively) associated with the outcome in >50% of the studies, while the remaining studies show the opposite association of the risk factor and outcome.
Conflicting evidence	N/A	If a risk factor is significantly (both positively and negatively) associated with the outcome in the same number of studies of comparable quality.	‘There is conflicting evidence…’
No evidence	N/A	If no studies reported on a risk factor	‘No studies reported on…’

Intensity Level	HCM†	LQTS†	Marfan Syndrome‡	ARVC	Brugada Syndrome
High
Basketball
Full court	0	0	2	1	2
Half court	0	0	2	1	2
Body building§	1	1	0	1	1
Ice hockey§	0	0	1	0	0
Racquetball/squash	0	2	2	0	2
Rock climbing§	1	1	1	1	1
Running (sprinting)	0	0	2	0	2
Skllng (downhill)§	2	2	2	1	1
Skllng (cross-country)	2	3	2	1	4
Soccer	0	0	2	0	2
Tennis (singles)	0	0	3	0	2
Touch (flag) football	1	1	3	1	3
Windsurfing‖	1	0	1	1	1
Moderate
Baseball/softball	2	2	2	2	4
Biking	4	4	3	2	5
Modest hiking	4	5	5	2	4
Motorcycling§	3	1	2	2	2
Jogging	3	3	3	2	5
Sailing‖	3	3	2	2	4
Surfing‖	2	0	1	1	1
Swimming (lap)‖	5	0	3	3	4
Tennis (doubles)	4	4	4	3	4
Treadmill/stationary bicycle	5	5	4	3	5
Weightlifting (free weights)§¶	1	1	0	1	1
Hiking	3	3	3	2	4
Low
Bowling	5	5	5	4	5
Golf	5	5	5	4	5
Horseback riding§	3	3	3	3	3
Scuba diving‖	0	0	0	0	0
Skating#	5	5	5	4	5
Snorkeling‖	5	0	5	4	4
Weights (non–free weights)	4	4	0	4	4
Brisk walking	5	5	5	5	5

1. what is the evidence behind the conversion score for different derivates for anthracyclines (including mitoxantrone)
First Author Year	Study Design Treatment era Years of follow-up	Participants	Treatment	Main outcomes		Addt’l remarks
van der Pal¹ 2012	Retrospective cohort 1966–1996 22.2 yrs (5.0–44.5)	5-yr survivors (N=1362)		Conversion score: Doxorubicin : 1.0 Daunorubicin: 1.0 Epirubicin: 0.67		Refs: Mertens (2008): late mortality Le Deley (2003): SMN after solid CA Perez (1991): Breast CA (epi vs.dox)
Mulrooney² 2009	Retrospective cohort 1970–1986 27.0 yrs (8–51)	5-yr Survivors (N=14, 358) Siblings (N=3899)		Conversion score: Doxorubicin = Daunorubicin Idarubicin = 3× doxorubicin		Conversion score based on a review paper recommendations (Pai Nahata 2000)
Blanco³ 2012	Case-Control 1966–2008 Cases: 9.2 (0.1–35.1) Controls: 12.3 (0.4–40)	Case (CHF) – N=170 Control (none) – N=317		Conversion score: Guidelines Doxorubicin: 1.0 Daunorubicin: 0.75 0.83 Epirubicin: 0.75 Idarubicin: 3 Mitoxantrone: 3	COG LTFU Doxorubicin: 1.0 Daunorubicin: Epirubicin: 0.67 Idarubicin: 5 Mitoxantrone: 4	Conversion score based on: Lehmann (2000), which is based on sited review literature with 1 in vivo model of acute toxicity
Temming⁴ 2011	Retrospective cohort N=124, 86 1987–2004 7.3 yrs (0–21.7)	124/158 available for Cardiotox analysis 86 data for late cardiotox		AML 10 and 12 trials Anthracyclines: Dauno and Mitox (1:5 conversion) 550–610 mg/m2		Anthracycline dose range similar across AML 10 and 12, unable to assess dose-association No discussion on conversion factor
Creutzig⁵ 2007	Retrospective cohort 1993–2003 BFM98: 3.6ys (0.8–7.0) BFM93: 7.5ys (1.1–11)	Eligible: N=1207 Late Cartox eval: N=547 (45%) 76% of echo w/in first 5yrs		AML BFM 93 98 Dauno : Ida 1:5 Dauno : Mitox 1:5
van Dalen⁶ 2010	Systematic review Meta-analysis 1966–2009 RCT’s: children, adults	Different anthracycline derivatives	Dox Epi Lipo-Dox	Epi vs. Dox (5 RCTs) = 1036 pts Clinical: RR=0.36, NS Lipo- vs. Dox (2 RCTs) = 521 pts Clinical: RR=0.2 (0.02–0.75) Subclinical: RR=0.38 (0.24–0.59)		For other possible combinations of different anthracycline derivatives, only 1 RCT or no RCT was identified Inconclusive evidence for children
Le Deley⁷ 2003	Case-control 1980–1999	Secondary leukemias after treatment of solid ca in childhood		Doxorubicin 50 mg/m2 = 75 mg/m2 epirubicin 60 mg/m2 dauno 12.5 mg/m2 mitox		Conversion based on leukemogenic potential of anthracyclines -NO ref for basis of anthracycline dose calculation
Neri⁸ 1989	Observational ?Tx era: 1980’s	Doxorubicin N=9 Epirubicin N=13 Authors propose: - Epi less concentrated in heart - Epi inhibits less of the Na/Ca exchange in heart sarcomeres - Epi produces less oxidative mitochondrial damage than dox	Dox 60 mg/m2 (Max 540) Vs. Epi 60 mg/m2 (Max 720)	Blood biomarker measurements, Echo’s Epirubicin less CK-MB elevation VO2 changes: Dox vs. Epi: 44% vs. 13% reduction Incidence of CHF: Dox vs. Epi: 67% vs. 23% Conclusion: “Epi-related cardiotoxicity 40% less than that produced by doxorubicin..”		Small numbers, not controlled for risk factors, older treatment era Non-random assignment tBreast CA, non-pediatric Acute cardiotoxicity

4. What is the risk for different cardiac RT doses for developing (a)symptomatic cardiac systolic dysfunction in childhood and young adult cancer survivors?
Author Year	Study Design Treatment era Years of follow-up	Participants	Treatment	Main outcomes	Addt’l remarks
Symptomatic cardiomyopathy and radiation dose
van der Pal¹ 2012	Retrospective cohort 1966–1996 22.2 yrs (5.0–44.5)	5-yr survivors (N=1362) Age at Dx: 5.9 (0–18)	Anthracyclines: 33.6% Anth+XRT: 7.9% Median Anth: 250 mg/m2 (25–775) Cardiac irradiation: None (80.4%) Any (19.5%) Localization of XRT: Thorax (31.6%) Abdomen (24.4%) Spine (33.5%) TBI (10.5%) Cardiac XRT (EQD2): Thorax: 24 (9.5–88.5) Abd: 26.9 (3.7–57) Spine: 30.14 (8–50) TBI: 15.8 (14–21.6)	Symptomatic cardiac events (CE) Grading: CTCAE v 3.0 50 CEs in 42 CS (CHF in 27/50) Median time to event: 18.6 yrs CI of CHF: Radiotherapy only: 0.7% at 30-yrs XRT + Anth: 7.9% at 30yrs Multivariate regression (Model 1) Radiotherapy (per 10 Gy) HR 1.4 (1.1–2.0) Multivariate regression (Model 2) Radiotherapy (Yes vs. No) HR 6.6 (0.6–73), p=0.13 Anth + Radiotherapy (Yes vs. No) HR 55.9 (6.6–470), p<0.001	Clinically validated outcomes Long follow-up, large cohort XRT dose conversion: Fractions of 2 Gy (EQD2) – includes both fractionation size and total dose Model 2 removes mutually exclusive cardiotoxic treatments. Radiotherapy alone not significant for CHF, but is predictive of other cardiac events
Schellong³⁰ 2010	Prospective cohort 1978–1995 15.1 yrs (3.1–29.4)	Hodgkin lymphoma: All pts. treated on German HD-78 to HD90 studies XRT field/dose reduction Uniform anth. dose Age at Dx:12.8 (2.5–17.9) Cardiac screening recs: Every 2–3 yrs up to 10 yrs Every 5 years thereafter In person +questionnaire	1132 eligible survivors Anthracyclines: 160mg/m2 everyone Mediastinal XRT: Median 25Gy (8–50) Mediast RT (MedRT) ≥36 Gy: 248 (21.9%) 30 Gy: 133 (11.7%) 25 Gy: 282 (24.9%) 20 Gy: 171 (15.1%) None: 298 (26.3%)	Cardiac grading per ACC/AHA 50/1132 (4.4%) w/ cardiac dz 14/1132 (1.2%) w/myocardial dz. 10/14 (71%) – MedRD-36 3/14 – MedRD20–30 25-yr CI of non-valvular cards dz ≥36 Gy: 4%, 30 Gy: 9%, 25 Gy: 4%, 20 Gy: 5%, None: 3%; p=0.2 Cox-regression: MedRD only predictor	Low prevalence/ incidence of myocardial disease likely due to low dose of anthracycline. Large study, long f/up, XRT is the only modified cardiotoxic exposure Unable to look at anth+XRT Non-valvular card dz includes CADz, valvular, conduction Homogeneous patient pop (age)
Mulrooney² 2009	Retrospective cohort 1970–1986 27.0 yrs (8–51)	5-yr Survivors (N=14, 358) Age at Dx: 0–4 yrs: 40.1% 5–9 yrs: 22.3% 10–14 yrs: 20.3% 15–20 yrs: 17.3% Siblings (N=3899)	Anthracyclines: 33.1% No Cardiac XRT: 29% <5 Gy: 34% 5–15 Gy: 5.8% 15–35Gy: 9.7% >=35Gy: 6.9%	CV outcomes Graded per: CTCAE v. 3.0 CHF (N=248) – HR 5.9 (3.4–9.6) Multivariate (CHF): No cardiac radiation (Ref) <5 Gy: HR 0.9 (0.6–1.4) 5–15 Gy: HR 1.3 (0.7–2.5) 15–35Gy: HR 2.2 (1.4–3.5) ≥35Gy: HR (4.5 (2.8–7.2) Dose-dependent increase in cumulative incidence of CHF	Self-reported Large sample size Long-term follow-up Cardiac XRT dosimetry calculations (Stovall et al.) Significance emerges at 15–35Gy XRT data not mutually exclusive of anthracycline exposure.
Blanco³ 2012	Case-Control 1966–2008	Case (CHF) – N=170 Control (none) – N=317 Matching criteria: Diagnosis Year of Dx (+/−5 yrs) Race/ethnicity Follow-up (controls)	Cases vs. controls: Anthracyclines 291 vs. 168, p<0.01 Chest XRT 25% vs. 14%, p<0.01	Clinically validated DCM, CHF Genetic susceptibility Multivariate (CHF): Chest radiation None (Ref) Any: OR 4.29 (1.9–9.6), p<0.001	Largest pop of clinically validated DCM, CHF XRT prevalence difference, but no info on dosimetry.
Aleman³¹ 2007	Retrospective cohort 1965–1995 8.7 yrs (28 669 person-years for cohort)	5-year survivors of HL Age at treatment: <20 yo (21.3%) 20–35 yo (63.4%) >35 yo (15.3%) Age at f/up: <35 yo (16.6%) >55 yo (20.1%)	RT only 27.5% Chemo (CT) only 4.8% RT + CT, anth 29.5% RT + CT, no anth 38% Unknown 0.2% 17% recent smokers 10% HTN 5% diabetes 8.5% Dyslipidemia	Cumulative incidence of CHF 25y: No RT 0.4% Mediastinal RT only 6.8% Mediast RT + CT, no anth 4.9% Mediast RT + CT, anth 7.9% Multivariate regression (CHF): Model 2 Mediastinal RT only (Ref) Med. RT + CT, no anthracycline: RR 1.3 (0.79–2.24) Med. RT + CT, anthracycline: RR 2.81 (1.44–5.49)	Large pop of adult lymphoma survivors (most <35 yo at Dx) Very long follow-up Critical role of cardiovascular risk factors Suggest that RT alone no inc. risk for CHF? Ref group is RT No dosimetry for cardiac XRT Includes older treatment era
van Dalen¹⁸ 2006	Retrospective cohort 1976–2001 8.5 yrs (0.01–28.4) F/up on prev 2001 JCO study	830 Children treated with anthracyclines Age at Anth exposure: <2 - 9.2% 2–6 – 30.9% 7–11 – 27% 12–16 – 30.2% >16 – 2.7%	Anthracyclines: Mean – 288 (15–900) Chest XRT: Any 21.2% None 78.7% Unknown 0.1%	CI and risk factors for A-CHF Univariate (CHF): RT on heart: RR 0.67 (0.2–2.3), NS Multivariate (CHF): No association with chest RT reported.	Not limited to long-term survivors No XRT dosimetry reported
Guldner³² 2006	Retrospective cohort Cross-sectional eval 1968–1985 5.4 yrs	447 eligible based on anthracycline exposure No XRT alone pop. 245 (N=55%) participated in study Age at Dx: 6.2 (0–21 yrs)	Anthracyclines: Median: 300 mg/m2 Entire cohort XRT heart dose: Mean 8.1 (15.6)	140 examined and healthy 24 with cardiac failure 65 with other cardiac disorders Heart radiation dose: Healthy vs. heart failure: 0.6 Gy vs. 17.8 Gy, p<0.001 Dose-dependent increase in HF risk by radiation dose	No XRT heart dosimetry, dosing estimated
Pein¹⁹ 2004	Retrospective cohort 1968–1982 18 yrs	Original cohort: 447 218 (48.8%) not evaluated 229 (51.2%) echo’s 15+year survivors Age at treatment: 6.2 yrs (0–21)	Anthracycline: 344 mg/m2 (40–600) Radiotherapy: 245 (55%) XRT dose to heart: Mean 6.7 Gy (0–91) Max 31.3 Gy (0–125)	Clear increase incidence w/time Multivariate regression: Cardiac failure, FS<25, EF<50, or ESWS>100 (not limited to CHF) Avg. XRT dose to heart, p<0.001 0 No XRT (Ref) >0–5 Gy: 1.63 (0.82–3.26) >5–20 Gy: 6.48 (2.76–15.20) >20 Gy: 4.40 (1.11–17.48)	High proportion treated with chest radiation Very long term follow-up One of the earlier studies to demonstrate dose-resposne with XRT
Adams³³ 2004	Cross-sectional 1970–1991 14.3 (5.9–27.5)	Hodgkin Lymphoma 24% participation rate Age at diagnosis: Median 16.5 (6.3–25.0) Age at study visit: Median 31.9 (18–49)	Anthracycline: 4/48 (8.3%) Mediastinal XRT dose: Median 40 Gy (27–52)	Comprehensive echo evaluation and stress testing No discussion of CHF Very few had systolic dysfunction Most with indices of diastolic dysfunction	Very long-term follow-up One of few studies to evaluate XRT without anthracyclines Homogeneous population with not much variance in XRT dose Poor participation rate
Green²⁰ 2001	Retrospective cohort Case-Control Through 1998	NWTS 1–4 Cohort 1: 1–4 received dox N=2,843 Cohort 2: 1–3, dox as part of salvage only (N=228) Age at Dx: 80% <8 y.o.	Anthracyclines Chest XRT – mostly due to lung XRT	CI and risk factors for CHF Risk of CHF est. to increase by factor of 1.6 for every 10 Gy of lung XRT, 1.8 for every 10Gy of left abd. XRT (no effect for Right) Multivariate regression (inclanth) Lung XRT: None (Ref) 10–19.9 Gy: RR 1.5 (0.6–3.9), p-0.4 ≥20 Gy: 4.3 (0.8–24), p=0.1 L. Abd XRT: None or right (Ref) Left: RR 4.0 (1.4–11.6)	Homogeneous population due to diagnosis, the vast majority were exposed before 7 yo Results approach sig at high dose lung XRT
Van der Pal³⁴ 2005	Systematic review of risk of morbidity and mortality from cardiovascular disease for childhood cancer Lit Review: 1966–2002	Criteria for review: Original report English, Dutch, French, German Study pop.: >50 pts. Childhood CA: <=18 y. XRT involving heart region Outcome: Clinical cardiovascular event (CVE) or cardiovascular mortality	Many studies include arterial events (ie: MI) and CHF as CVE. For CVE: 9 studies selected based on validity and inclusion criteria. 8/9 studies, outcome well-defined 3/9 risk estimation well-defined and adequate	Relative Risk for CVE: Cardiac event, matched for anthracycline, time at risk, cohort Continuous tx. Variables (RR): Female/Male: 4.5, p<0.01 Anth, 100 mg/m2: 3.2, p<0.01 Lung RT, 10 Gy: 1.6, p=0.06 Left abd, 10 Gy: 1.8, p=0.02 Right abd. 10 Gy: 0.94, p=0.77 Categorical tx. Variables (RR): Female/Male: 3.7, p<0.01 Anth,>300 mg/m2: 5.0, p<0.02 Lung RT >20Gy: 3.1, p=0.21 Left abd. RT: 3.5, p=0.02	Older treatment eras For many, no clear delineation between RT-related systolic heart failure vs. CHF due to coronary artery disease, or MI alone. Dose-dependent Risk
Kremer²¹ 2002	Review of Frequency and Risk Factors of anthracycline- induced clinical heart failure Medline: 1966–2000	71 articles reviewed Limitations in many: issing info Lack of RF analysis Non-rep. populations	Assess RR of possible Risk factors in 10 studies	Univariate (CHF): Risk with XRT reported in 4 out of 10 studies (3 out of 4 significant) Gilladoga (1976) N=50 XRT to heart: RR 5.2 (1.6–16.8) Dearth (1984) N=116 XRT to heat: RR13.5 (3.4–53.3) Bu’Lock (1996) N=226 XRT to heart: 11.1 (3.7–33.5) Krischer (1997) N=6493 XRT to heart: RR 0.7 (0.3–1.9)	Review is driven by anthracycline exposure Few with XRT dose quantification and none with careful heart dosimetry calculation
Asymptomatic cardiomyopathy and radiation dose (Abnormal EF, SF).
Brouwer²² 2011	Cross-sectional 1976–1999 17.7 years	5-yr survivors 401 eligible 277 (69%) participated 8 (3%) on cardiac meds for CHF/renal	Anthracycline Median: 183 (50–600) Radiation 63%??	No breakdown by dose Multivariate LogisticRegression SF<29% Anthracycline ≥183: OR 2.2, 1.25–3.8, p<0.01 Mediast RT: 3.0, 1.4– 6.7,p<0.01 TBI: 1.9, 0.6–5.6	Good participation rates Comprehensive echo screen Long term follow-up Handful with clinical HF included in analysis
van der Pal²³ 2010	Prospective cohort-Survivorship clinic 1966–1997 15.4 yrs (5.1–4.3)	5-yr survivors 735 anthracycline-treated 601 Eligible for study 525 Had echocardiogram Age at Dx: 8.9 (0.1–17.8)	Anthracycline: Med – 250 (33–720) Chest XRT: 36.4% Cumm. XRT dose: ≤30 Gy 10.8% >30 Gy 23.2%	Asymptomatic cardiac dysf. Graded per CTCAE LVSF as primary outcome (1^st echo) LVSF<30% XRT ≤30 vs. >30 Gy: 12.5% vs. 31% Multivariate regression (SF<30%): No Radiotherapy (Ref) Odds Ratio Thorax: 3.49 (1.6–7.6) Abdomen: 2.66 (1.0–7.05) Spine: 0.64 (0.23–1.74) TBI: 0.53 (0.10–2.87)
Abosoudah²⁴ 2011	Prospective cohort -Survivorship clinic 1995–2003 3.0 yrs (1–10)	4-year survivors 896 anthracycline-treated 603 eligible for study 469 >=1 screening echo Age at Dx: 7.7 (SD 4.6)	Anthracycline: Mean – 205 (114.7) Chest XRT: 34% No dose in modelField involving heart	Screening echo per COG LTFU GuidelinesNot limited to abn EF/FS Multivariate regression: No radiation (Ref) RT to heart: HR 1.7 (1.1–2.8)	Time to first abnormal echocardiogram Screening frequency driven by age, anthracycline dose, and XRT so unclear implication
Hudson²⁵ 2007	Cross-sectional 9.0 (3.0–18.0)	223 anthracycline-treated Vs. 55 – not at risk Age at Dx: 5.5 (0–23.6)	Anthracycline (AR) Med: 202 (25–510) Anth + XRT: 26.9% Chest XRT: 2.7%	Screening echo. LVSF, Wall stress Univariate regression (SF<28%): No Cardiac RT (Ref) RT: OR 0.9 (0.4–2.05)	Asymptomatic One time-point No cardiac dose quantification
Kremer²⁹ 2002	Review of Frequency and Risk Factors of anthracycline-induced subclinical cardiotoxicity Medline: 1966–2001 >50 children/study	58 articles reviewed Limitations in many: Missing info Non-rep. populations Non-original research Validity evaluated in 25 studies 10 studies w/RF analyses 6 studies which defined an abnormal SF with validity score>5	Risk Factor analysis: Steinherz (1991) Lipshutz (1991) Silber (1993) Sorensen (1995) Lipshultz (1995) Pihkala (1996) Sorensen (1997) Nysom (1998) Lanzarini (2000) Bossi (2001)	1 Study with chest radiation dose as predictor (limited to FS or EF abn) Risk Factor analysis: Steinherz (1991), N=201 >cumulative anth dose × f/up >mediastinal radiation No dose-effect calculations	Not all 10 studies had populations that would have received chest radiation (ie: ALL, AML)

3. What is the diagnostic value (i.e. sensitivity and/or specificity) of biomarker ANP, BNP, NT-pro-BNP to detect asymptomatic cardiac systolic dysfunction as measured by echocardiography in adult non-cancer populations?
First Author Year	Study Design	Participants	Diagnostic tests	Main outcomes^#	Addt’l remarks
Hill⁴⁷ 2008	Systematic review of RCTs and observational studies (published between 1989 and February 2005). For screening studies general populations with no known symptomatic heart failure were included. 6 studies were addressing our question^* (n=2 cross sectional study, n=4 cohort study).	Setting: population-based cohort study (n=1; males and females reported separately), GP sample (n=1), population samples (n=3), cohort with stable coronary artery disease (n=1). Sample size: range 293–2042 participants (1 study presented males (1470) en females (1707) separately: 3177 in total). Males: range 43–49.6% (n=3), results presented for males and females separately (46.3% males) (n=1), nm (n=2). Age: range mean age 58–75 years (n=3), >45 years (n=1), range 50–90 years (n=1), nm (n=1). Prevalence cardiac dysfunction: 1–16%.	Index test: BNP (n=5) or NT-pro-BNP (n=2)^¶. Reference standard: LVSD based on LVEF (n=5) or a combination of LV mass, LVEF<50% and moderate to severe LVSD (LVEF<40%) (n=1). Time between tests: Nm. Cutoff points: BNP: range 21->115 pg/mL. NTproBNP: range >338–850 pg/mL. Reference test: LVEF range 35–55%.	BNP: Sensitivity: range 26–93% Specificity: range 47–89% NT-pro-BNP: Sensitivity: range 70–80% Specificity: range 63–85%	Risk of bias assessment of included studies: nm.
Ewald⁴⁸ 2008	Systematic review of prospective studies (published up to June 2005). 7 studies were addressing our question^*.	Setting: population-based cohort studies (n=2; 1 study reporting males and females separately), GP samples (n=2), population samples (n=3). Sample size: range 203–1997 participants (1 study presented males (1470) and females (1707) separately: 3177 in total). Males: range 43–56% (n=6), results presented for males and females separately (46.3% males) (n=1). Median/average age: range 58–75 years. Prevalence cardiac dysfunction: 0.6–6.9%.	Index test: BNP (n=5) or NT-pro-BNP (n=3)^¶. Reference standard: LVSD based on LVSF (n=1), LVEF (n=4), wall motion index (n=2). Time between tests: nm for each study separately, but it was stated that the quality of studies was generally adequate, except for 1 study with delays up to one year between both tests. Cutoff points: BNP: range 6.9–19.2 pM/L (n=4); >54.5 pg/ml (n=1). NTproBNP: range 37.7–48.9 pM/L (n=2), nm (n=1). Reference test: LVSF: 28% (n=1); LVEF: range 40–50% (n=4); wall motion index: >2 (n=1) and < 1.7 (equates LVEF < 40%) (n=1).	BNP: Sensitivity: range 55–90%^~ Specificity: range 77–90%^~ NT-pro-BNP: Sensitivity: range 76–92% Specificity: range 67–81%	Risk of bias assessment of included studies was based on (1) blinding of outcome assessor for other test result, (2) detailed description of methods and criteria for both tests, and (3) performance of both tests on same day. The quality of included studies was generally adequate, but in 1 study delays of up to 1 year occurred between the echocardiography and the peptide estimation (no further information provided); a sensitivity analysis taking into account the quality score was done, but not presented in the paper
Wang⁴⁹ 2003	Systematic review of studies of patients with asymptomatic LVSD (published between 1975 and November 2002). 13 studies were addressing our question^* (n=5 community based studies, n=6 referral series).	Setting: population-based cohort studies (n=3; 1 study reporting males and females separately), GP sample (n=1), population sample (n=1), referral series (not further specified) (n=6). Sample size: Community based: range 126–1707 participants (1 study presented males (1470) and females (1707) separately: 3177 in total); Referral series: range 75–466 participants. Males: Community based: only men (n=1), results presented for males and females separately (46.3% males) (n=1), nm (n=3). Referral series: nm (n=6) Age: Nm. Prevalence cardiac dysfunction: Nm.	Index test^#: Community based: BNP (n=3), NT-ANP (n=2). Referral series: BNP (n=5), NT-ANP (n=1). Reference standard: Community based: LVSD based on LVSF (n=1), LVSF or mild or greater reduction in LVEF on visual estimation (n=1) or LVEF (n=3). Referral series: LVSD based on LVEF alone (n=4), LVEF in rest or exercise (n=1) or LVEF or wall-motion abnormalities (n=1) Time between tests: Nm. Cutoff points: Community based: BNP: range 17.9–34 ng/L. NT-ANP: range 398–800 pmol/L. Reference test: LVSF: range 0.28–0.29 (no further information provided on combination with LVEF reduction); LVEF: range 0.30–0.45. Referral series: BNP: range 13.8–87 ng/L. NT-ANP: 54 pmol/L Reference test: LVEF: range 0.35–0.55 (LVEF at rest or during exercise: resting LVEF<0.45 or exercise LVEF<0.55; no further information provided on combination with wall motion abnormalities).	Community based: BNP: Sensitivity: range 26–77% Specificity: range 84–89% NT-ANP: Sensitivity: range 43–86% Specificity: range 75–89% Referral series: BNP: Sensitivity: range 58–100% Specificity: range 58–81% NT-ANP: Sensitivity: 90% Specificity: 92%	Risk of bias assessment of included studies: nm.

5. What is the effect of treatment with beta-blockers in non-oncology populations with asymptomatic cardiomyopathy?
First Author Year	Study Designh Treatment era Years of follow-up	Participants	Treatment	Main outcomes	Addt’l remarks
Dargie⁶⁴ 2001	Double-blind, placebo-controlled RCT 1.3 years clinical follow-up	1959 patients with MI 3–21 days before randomization, EF≤ 40% or wall-motion score index ≤ 1.3 and at least 24 hours on a stable dose of ACE-inhibitor treatment.	Carvedilol: N=975 Placebo: N=984	All-cause mortality: Carvedilol: 116 (12%) Placebo: 141 (15%) HR: 0.77 (0.60 – 0.98) Hospitalization for heart failure: Carvedilol: 118 (12%) Placebo: (138 (14%) HR 0.86 (0.67 – 1.09)	Eligible patients had LV dysfunction with or without heart failure, but patients with severe heart failure were excluded. EF was determined by echocardiography, RNA or ventriculography
Exner⁶⁵ 1999	Retrospective analysis of RCT Mean followup 35 months	4228 patients participating in the SOLVD prevention trial	Patients that used a beta blocker at the start of the trial, in addition to study medication: N=1015 (24%) Patients that did not use a beta blocker at the start of the trial, in addition to study medication: N=3213 (76%)	All-cause mortality: Using a beta blocker: IR 4.3/100 person-years No beta blocker: IR 5.6/100 person-years Multivariate model, using a beta blocker in addition to ACE inhibitor allocation: * All-cause mortality RR 0.70 * All-cause mortality or hospitalization for CHF: RR 0.64 (0.49 – 0.83)
Vantrimpont⁶⁶ 1997	Retrospective analysis of RCT Mean clinical follow-up of surviving patients: 42 months (+/−10 months)	2231 patients participating in the SAVE trial	Patients that used captopril at the start of the trial, in addition to study medication: N=789 (35%) Patients that did not use captopril at the start of the trial, in addition to study medication: N=1442 (65%)	Cardiovascular mortality: Captopril: 13.1% No captopril: 22.1% (RR 0.58, 0.43 – 0.79) Severe heart failure: Captopril: 16.5% No captopril: 22.6% (RR 0.68, 0.55 – 0.83) Multivariate model (including captopril use): * CV mortality RR 0.70 * Severe CHF RR 0.79
Hunt^61,62 AHA/ACC Guideline (2005 and 2009)	Beta-blockers are indicated in all patients without a history of MI who have a reduced LVEF with no HF symptoms	Stage B*		Class of recommendation I Level of evidence C

2. What is the risk of (a)symptomatic cardiac systolic dysfunction in childhood and young adult cancer survivors of TBI that is above and beyond the risk due to pre-HCT anthracycline and chest radiation?
First Author Year	Study Design Treatment era Years of follow-up	Participants	Treatment	Main outcomes	Addt’l remarks
Uderzo⁹ 2007	Prospective cohort 1994–1997 5 yrs.	N= 162, Age: 0–18 y.o. at HCT	Allogeneic HCT 67% anthracyclines 58% TBI 80% HCT for malignancy	Decline in FS over time Univariate: TBI alone, p=0.04 TBI + Anthracyclines, p=0.004 Multivariate No association with TBI and FS decline	In addition, no differences seen by gender or age at HCT. TBI fractionated (12Gy) in nearly all except 2.
Lonnerholm¹⁰ 1999	Prospective cohort 1985–1996 1–10 years (median 5)	N= 45, Age: 1.2–16.2 at dx	Autologous HCT 53% TBI Pre-HCT anthr: 150–450	Standard echo: 1y-, 3y-and 5-post LVDD/SD, EF, FS No difference in LV dimensions by TBI No discussion of anthracycline dose and changes in LV parameters
Eames¹¹ 1997	Cross-sectional 1994–1995 Mean f-up 4.1 yrs	N=63 Age: 2y–32 y at partic.	Allo HCT: 82% Auto HCT: 18% TBI: 65% HD-Cy: 95% Anth: 63.5% Anth dose: 308 (60–450)	Comprehensive cardiac echo: NYHA grading of all participants Normal FS (>=29%): 98% No regression analysis for risk factors for abn EF/FS TBI (fractionated or not) NOT predictive of cardiotoxicity	Selection bias 22% of HCT population included Treadmill exercise testing Abnormal: 48.4%
Armenian¹² 2011	Retrospective cohort 1970–1986 CCSS 1974–1998 BMTSS CCSS: 16 yrs (+/−5) BMTSS: 13 yrs (+/−5.6)	Heme malign CCSS: N=7207Age: 8.9 yrs at dx 25 yrs at partic. BMTSS: N=145 Age: 10.9 yrs at dx 24 yrs at partic. Sibling N=4020 Age: 26. yrs at partic.	BMTSS Chemo + TBI: 76.6% Autologous HCT: 28% Anthracycline: None −8.3% 1–249 – 50.3% >=250 – 41.4% Chest Radiation: 5.5% CCSS Anthracycline: None – 61.0% 1–249 – 19.3% >=250 – 19.7% Chest radiation: 23.1%	CTCAE graded chronic health conditions Grade 3–5 cardiac disease Multivariate regression adjusting for: Age, gender, race, insurance, treatment era, time from dx, diagnosis, chest radiation, anthracycline dose BMTSS vs. siblings: RR 12.7 p<0.01 BMTSS vs. CCSS: RR 0.5, p=NS	After adjusting for pre-HCT treatment-related exposures, no differences in CV outcomes seen, Sub-analysis of specific HCT-related exposures (TBI, HD Cytoxan) did not reveal a difference
Armenian 2008¹³	Case-control 1981–2003 6.4 yrs (1.3–22.1)	1+year survivors Allo and auto HCT Case (CHF): 60 Control: 166 Age 43 yrs (+/−13)	Mean Anthracycline: 261 vs. 171 mg/m2 Chest XRT: 10% vs. 8% TBI: 65.0% vs. 65.7% HD-Cy: 75.0% vs. 75.3%	Clinical CHF per AHA/ACC def. Anthracyclines as the only treatment-related predictor of post-HCT CHF. TBI, HD-Cy not significant in univariate or multivariate models.	Mostly adults, only included late-occurring events.
Armenian 2011¹⁴	Retrospective cohort Nested case-control 1988–2002 5.3 yrs (0.1–20.5 yrs)	Autologous HCT Cohort: N=1244 CHF: N=88 peds + adults 7200 person-yrs	TBI (12 Gy Frax): 59.2% (60% vs. 59%) HD-CY: 85.9% (87% vs. 86%) Anthracycline mg/m2: 309 vs. 237, p<0.01	Clinical CHF per AHA/ACC def. Multivariate Condit. regression: Female: RR 2.4, p<0.01 Lymphoma dx: 1.5, p=0.05 Age: RR↑ wth age TBI, HD-Cy NOT associated with risk	Pre-HCT anthracycline dose, and post-HCT CV risk factors, gender, most significant predictors of post-HCT risk. CI of CHF 15% at 15 yrs in female lymphoma survivors.
Chow¹⁵ 2011	Retrospective cohort 1985–2006	2+year survivors Allo and auto HCT N=1491 Gen pop (by age) matching N=4352	Autologous: 43.7% Allogeneic: 56.3% TBI: 76.7% HD-Cy: 48.1%	CV outcomes, ICD-9 coding, hospital records: MI, DCM, CHF, stroke, other vascular dz. Multivariate regression Risk of DCM, CHF: Post HCT relapse: RR 1.9 (1.1–3.3) TBI: RR 1.0 (0.6–1.8) Allo HCT: 0.8 (0.5–1.4)	No anthracycline in models Hosp ICD-9 codes, not validated outcomes Post-HCT CV risk factors as significant predictors of DCM or CHF.
Tichelli¹⁶2008	Retrospective cohort 1990–1995 9 yrs (1–16 yrs)	1+-year survivors Allogeneic HCT Adult HCT N=548	Hem. Malign: 85% TBI: 58%	Limited to clinically validated arterial events TBI: 70% (arterial dz), 57% (no dz), NS Multivariate model: Older age at HCT and CVRFs as the only independent predictors of dz.	No anthracycline in models Post-HCT risk factors as predictors of post-HCT CV outcomes

8. Is there evidence that exercise increases the risk of deterioration of cardiac systolic function in adult-onset cancer survivors and non- oncology at-risk populations?
First Author Year	Study Design Treatment era Years of follow-up	Participants	Treatment	Main outcomes	Addt’l remarks
Schmitz⁷⁰ 2010	Guideline-expert opinion- American College of Sports Medicine	Only ADULT cancer studies reviewed		Physical activity is strongly recommended with the exception of activities resulting in rapid BP elevation (eg isometric exercise)
Pellicia⁷¹ 2006	Guideline-expert opinion- European Society of Cardiology			Recommendation is for physical activity in individuals with genetic susceptibility to CHF, but with normal systolic function.
Dickstein⁶³ 2008	Guideline – review of published evidence, expert panel; European Society of Cardiology			Recommendations – Weight reduction should be considered in obese persons with heart failure In moderate to severe heart failure, weight reduction should not be recommended routinely	No supporting evidence supplied Level of evidence C
Maron⁷² 2004	Consensus document; expert international panel of clinical cardiovascular specialists and molecular biologists; American Heart Association	Young people (<40 years age) with genetic cardiovascular diseases including hypertrophic cardiomyopathy but not specifically including dilated cardiomyopathy.	Not specifically considered. Considered recommendations for physical activity and recreational sports participation. Childhood cancer survivors (CCS) not included.	Recommendations: Can safely participate in most low or moderate-intensity recreational exercise Some activities should be avoided, eg burst exertion, extremely adverse environmental conditions, exercise programmes with systematic / progressive levels of exertion and aiming at higher levels of conditioning, intense isometric exertion, extreme sports, performance-enhancing substances
Riegel⁷³ 2009	Review / scientific statement; expert panel; American Heart Association	Persons with heart failure	Not specifically considered. CCS not mentioned specifically.	Statements In moderate heart failure, exercise improves certain physiological parameters including V_O2max, ventilatory response, heart rate variability. Can also reduce depression. Effect on mortality not clear. Cites Pina et al 2003. Individually tailored exercise programme based on results of formal exercise testing may benefit patients with severe symptomatic LV dysfunction. Cites Fletcher et al 2001. Exercise is a beneficial adjunctive treatment in patients with current or prior heart failure symptoms and reduced LVEF. Cites Hunt et al 2005 (states this is level 1B evidence). Modest benefit in HF-Action RCT (Flynn et al, 2009, see below)
Flynn⁷⁴ 2009	HF-Action Randomised controlled trial Randomised 2003-7 Median FU 2.5 years	2331 stable out-patients with heart failure (LVEF ≤35%) 82 centres in USA, Canada, France	Randomised to Usual care + aerobic exercise training (initially supervised, subsequently home-based) vs usual care + recommendation for regular physical activity. Usual care included optimal medical therapy.	At 3 months, usual care + exercise training group showed statistically greater improvement in Kansas City Cardiomyopathy Questionnaire (KCCQ – a 23 item disease-specific questionnaire) score than usual care group. Improvement was maintained. Also modest but significant improvement in quality of life and non-significant reduction in all-cause mortality and hospitalisation in usual care + exercise training group.
Piepoli⁷⁵ 2004	Meta-analysis (individual patient data) 1990–2002 Individual median F/U 5–75mths, overall 23mths	9 studies, total 395 training to 406 control 87% males, 59% with IHD, mean LVEF <28%, 73% on ACE inhibitors	All RCTs, usual care vs addition of exercise training (mostly supervised)	Outcome of mortality in favour of exercise – 0.65 (0.46–0.92) Outcome of death or admission to hospital also in favour of exercise – 0.72 (0.56–0.93)	Intensity generally set at 60–80% peak oxygen consumption. These trials are designed to be “safe” first and foremost. Question of whether differing aetiologies of systolic dysfunction/heart failure have differing responses to physical activity not yet answered.
Davies⁷⁶ 2010	Meta-analysis (publication data) 2001-Jan2008 Individual median F/U 5 mths-60mths., overall 11mths	19 trials, total 3647 patients (HF-ACTION trial contributed 60%) Only one trial 57% femaies, others 72–100% male; age 58	All RCTs, usual care vs addition of exercise training (mostly supervised) Only 4 trials F/U longer than 12 mths.	All cause mortality <12 mth F/U outcome in favour of usual care – 1.03 (0.70–1.53), but >12mth F/U favoured exercise – 0.91 (0.78–1.06) All hospital admissions both < and >12 mths favoured exercise. HRQoL measurements also favoured exercise.	If HF-ACTION trial excluded, significant reduction longer-term mortality seen (0.62 (0.39–0.98). Issues of mix of endurance and resistance training starting to be addressed.

PERMALINK

Recommendations for Cardiomyopathy Surveillance for Survivors of Childhood Cancer: A Report from the International Late Effects of Childhood Cancer Guideline Harmonization Group

Saro H Armenian

Melissa M Hudson

Renee L Mulder

Ming Hui Chen

Louis S Constine

Mary Dwyer

Paul C Nathan

Wim JE Tissing

Sadhna Shankar

Elske Sieswerda

Rod Skinner

Julia Steinberger

Elvira C van Dalen

Helena van der Pal

W Hamish Wallace

Gill Levitt

Leontien CM Kremer

Abstract

INTRODUCTION

METHODS

RESULTS

Table 1.

Table 2.

Table 4.

Who needs cardiomyopathy surveillance?

Figure 1. Risk of cardiomyopathy and CHF by cumulative lifetime anthracycline (A and B) and radiotherapy dose (C).

Table 3.

What surveillance modality should be used?

At what frequency and for how long should surveillance be performed?

What should be done when abnormalities are identified?

What are the limitations for physical activity?

Role of modifiable cardiovascular risk factors and cardiomyopathy risk

DISCUSSION

Table 5.

Supplementary Material

Acknowledgements

Appendix 1

Appendix 2

Appendix 3

Appendix 3.

Appendix 4: Working Group Evidence Summaries

References

Footnotes

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases