Cohort studycited by the Tayebat archivefull text checked★ key study A-CHM-03 in-depth review

Adult survivors of childhood cancer have more heart disease than their siblings

Cardiac outcomes in a cohort of adult survivors of childhood and adolescent cancer: retrospective analysis of the Childhood Cancer Survivor Study cohort.

Overview

Mulrooney DA et al.. BMJ 2009 · PMID 19996459 · DOI 10.1136/bmj.b4606

Survivors reported heart failure about 6 times as often as siblings (1.7% vs 0.2%); by 30 years after diagnosis about 4% had heart failure. Risk rose with anthracycline dose and heart radiation.

Does the study support the specific claim? Yes · Methodological quality: moderate

Original publication

Plain-language summary

Study explained simply

What did the researchers want to find out?: Quantify late heart disease in survivors.
How was the study done?: Retrospective cohort with questionnaires.
How many people, animals or samples were studied?: 14,358 survivors, 3,899 siblings.
What was actually found?: HR 5.9 CHF, 5.0 MI, 6.3 pericardial, 4.8 valvular versus siblings; 30-year cumulative incidence of CHF 4.1%.
How large was the observed effect?: Relative hazards 5-6; absolute: CHF in 1.7% of survivors vs 0.2% of siblings.
What do the results mean in everyday life?: Lifelong heart follow-up for survivors.
What are the study's limitations?: Self-reported, unvalidated diagnoses; possible surveillance bias; few cardiovascular risk factors measured; treatments from 1970-1986.
What does it say about Elawady's specific claim?: Supports existence of late harm.
Do the results apply to Germany and Europe?: Moderate.
What can explicitly not be concluded from this study?: Absolute risk per patient; that treatment should have been avoided.

Scientific in-depth review

1. Full reference

Authors: Mulrooney DA, Yeazel MW, Kawashima T, Mertens AC, Mitby P, Stovall M, Donaldson SS, Green DM, Sklar CA, Robison LL, Leisenring WM
Journal: BMJ
Year: 2009
Identifier: PMID 19996459 DOI 10.1136/bmj.b4606 · PMC3266843
Verification: verified – PubMed E-utilities efetch, 2026-10-25
Retracted: no
Corrections: none
Extent checked: full text
Conflicts of interest and funding: Funding: National Cancer Institute grant U24 CA 55727, American Lebanese Syrian Associated Charities, NIH grant 1K12RR023247, Children's Cancer Research Fund; funders not involved in data collection, analysis, interpretation or writing. Competing interests: none declared.

2. Research question

Incidence and risk factors of CHF, MI, pericardial and valvular disease in adult survivors of childhood cancer.

3. Study design and population

Study design: Cohort study – Retrospective cohort (Childhood Cancer Survivor Study) with sibling comparison; self-reported outcomes
Population: Five-year survivors of childhood cancer diagnosed <21 years in 1970-1986
Country: USA/Canada (26 CCSS institutions)
Period: Diagnosis 1970-1986; follow-up up to 30 years
Sample size: 14,358 five-year survivors (81.8% of 17,564 contacted; 20,627 eligible) and 3,899 siblings (81.5% of eligible); 10,367 survivors completed the 2000-2002 follow-up questionnaire
Inclusion and exclusion criteria: Diagnosis <21 years of leukaemia, brain cancer, Hodgkin lymphoma, non-Hodgkin lymphoma, kidney cancer, neuroblastoma, soft tissue sarcoma or bone cancer at 26 institutions, 1 Jan 1970-31 Dec 1986, surviving >=5 years; only outcomes >=5 years after diagnosis (siblings: >=5 years after birth) counted. Siblings: nearest-age sibling nominated by a random sample of survivors.

4. Methods and endpoints

Exposure / intervention: Cumulative anthracycline dose (doxorubicin + daunorubicin + 3x idarubicin, from medical records) and mean cardiac radiation dose (estimated by MD Anderson dosimetry); other agents (cisplatin, vincristine, bleomycin, cyclophosphamide) entered if univariate P<0.2. Treatment data missing for 1,111 participants.
Comparator: Siblings; unexposed survivors
Endpoints: Self- or proxy-reported first occurrence of congestive heart failure/cardiomyopathy, myocardial infarction, pericardial disease (pericarditis/constriction) and valvular abnormalities ('told by a doctor'); medical-record validation attempted but abandoned as unfeasible
Statistical methods: Cox proportional hazards with age as time scale, adjusted for gender, race, household income, education, tobacco use; sandwich variance for within-family correlation; multiple imputation of missing age at onset (9% of survivors, 14% of siblings with an event); multivariable treatment models among survivors; cumulative incidence curves. Authors note P values 0.01-0.05 are less conclusive because of many comparisons.

5. Actual results

Results: Versus siblings (adjusted): CHF HR 5.9 (95% CI 3.4-9.6), MI HR 5.0 (2.3-10.4), pericardial disease HR 6.3 (3.3-11.9), valvular abnormalities HR 4.8 (3.0-7.6); highest in Hodgkin lymphoma (MI HR 12.2, 5.2-28.2). Among survivors (multivariable): anthracycline <250 mg/m2 CHF HR 2.4 (1.5-3.9); >=250 mg/m2 CHF HR 5.2 (3.6-7.4), pericardial HR 1.8 (1.1-3.0), valvular HR 2.3 (1.6-3.3), MI HR 1.1 (0.5-2.1, not significant); cardiac radiation 1500-<3500 cGy HR 2.2-3.3 and >=3500 cGy HR 3.6-5.5 across all four outcomes; <1500 cGy not significant. Female sex CHF HR 1.4 (1.1-1.9). Cumulative incidence rose without plateau (except possibly MI).
Absolute risks: Prevalence in survivors vs siblings: CHF 248 (1.7%) vs 7 (0.2%); MI 101 (0.7%) vs 6 (0.2%); pericardial disease 181 (1.3%) vs 13 (0.3%); valvular abnormalities 238 (1.6%) vs 21 (0.5%). Age-adjusted rates per 10,000 person-years in survivors: CHF 9.7 (9.4-10.1), MI 2.8 (2.4-3.3), pericardial 5.8 (5.4-6.4), valvular 6.4 (5.9-7.1). 30-year cumulative incidence: CHF 4.1% (3.2-5.0%), valvular 4.0% (3.1-4.9%), pericardial 3.0% (2.1-3.9%).
Authors' conclusion: Survivors of childhood cancer are at substantial risk for cardiovascular disease; clinicians must be aware.

6. Strengths

Very large multi-institutional cohort (14,358 survivors)
Sibling comparison group, adjusted for socioeconomic factors and smoking
Long follow-up (mean about 20 years; up to 30 years)
Treatment doses abstracted from medical records; individual cardiac radiation dosimetry
Clear dose-response for anthracyclines and cardiac radiation

7. Weaknesses

Self/proxy-reported outcomes; medical-record validation attempted but abandoned (e.g. records obtainable for only 103 of 292 CHF reporters)
Possible surveillance bias (exposed survivors monitored more) inflating relative hazards (authors)
No data on cardiovascular risk factors other than sex and smoking (blood pressure, lipids unknown)
Treatment data missing for 1,111 participants
Non-participants had higher mortality (18% vs 12%); possible underestimation
Subclinical cardiomyopathy not captured
Historical treatments (1970-1986); many comparisons (authors flag P 0.01-0.05 as less conclusive)
Few sibling events (e.g. 7 CHF); wide CIs for sibling comparisons

8. Risk of bias

Orienting tool: ROBINS-E
Type of assessment: structured assessment based on the full text by a single reviewer (no formal double review)
Full formal assessment: no
Overall risk of bias: some concerns
Rationale: Outcome measurement is the main weakness, but large, dose-dependent hazard ratios are unlikely to be fully explained by reporting or surveillance bias. Single-reviewer structured judgement from the full text; a formal assessment requires two independent reviewers and a protocol.

8. Risk of bias
Domain	Judgement	Rationale
Confounding	some concerns	Sibling comparison adjusted for sex, race, income, education, smoking; within-survivor treatment models adjusted for age at diagnosis, era and other agents. Cancer type and treatment are intertwined; no blood pressure, lipids or other cardiovascular risk factors.
Measurement of exposure	low	Anthracycline doses abstracted from medical records; cardiac radiation dose estimated by standardised dosimetry. Treatment data missing for 1,111 participants.
Selection of participants	some concerns	Five-year survivors only; 81.8% participation among those contacted, 14.8% not located; non-participants had higher mortality (18% vs 12%), cardiac deaths on death certificate 0.8% vs 0.6%.
Post-exposure interventions	not applicable	Not relevant to the question of treatment-related late effects.
Missing data	some concerns	Age at onset imputed for 9%/14% of those with events; treatment data missing for 1,111; follow-up questionnaire completed by 72% of survivors.
Measurement of outcome	high	Self/proxy-reported diagnoses without validation (validation abandoned as unfeasible); possible differential surveillance of exposed survivors. Cited literature: 83% sensitivity/98% specificity for self-reported MI.
Selection of reported result	low	All four prespecified outcomes reported by diagnosis and treatment; sensitivity analyses described but 'data not shown'.

Full-text review (25 October 2026): Added absolute risks (prevalences, rates per 10,000 person-years, 30-year cumulative incidence), detailed dose-response HRs (anthracycline <250 mg/m2 also raised CHF risk, HR 2.4; cardiac radiation <1500 cGy not associated), participation and missing-data figures, abandoned outcome validation, surveillance bias and missing CV risk factors; full funding statement. ROBINS-E overall unchanged (some concerns); outcome measurement domain high.

9. Assessment of the statistical interpretation

Relative hazards versus siblings are large (5-6-fold), but absolute risks are low: 30-year cumulative incidence of CHF 4.1%, MI prevalence 0.7% at a median age of 27. Dose-response (anthracycline >=250 mg/m2, cardiac radiation >=1500 cGy) supports causality; cardiac radiation <1500 cGy was not associated with excess risk. Sibling CIs are wide because sibling events were few.

10. Relevance to the specific claim

Supports late cardiac harms in childhood-cancer survivors.

11. Independent contrary findings

A-CHM-11: later treatment eras with less anthracycline/radiation show lower late mortality; I-CHM-05: rising cure rates for childhood ALL.

12. Transferability to Germany and Europe

Moderate (historical North American protocols).

13. Overall methodological judgement

High-quality survivorship evidence for dose-dependent late cardiac effects of anthracyclines and cardiac radiation; outcomes self-reported and unvalidated. The archive's numbers are correct; absolute figures in the full text (e.g. 4.1% CHF by 30 years) show the relative hazards apply to a low baseline.

Multidimensional assessment
Multidimensional assessment
Authenticity of the source	verified
Methodological quality	moderate
Risk of bias	some concerns
Precision of results	moderate
Directness	direct
Reproducibility	replicated
Consistency with the overall evidence	consistent
Regional transferability	moderate
Adequacy of the interpretation (by the archive)	correctly reported

14. Open questions

Absolute risks under modern protocols with lower anthracycline doses, cardioprotection and reduced cardiac radiation; validated (not self-reported) outcomes; role of conventional cardiovascular risk factors.

15. Linked fact checks

Anthracycline chemotherapy (and chest radiotherapy) can cause chronic, dose-dependent heart damage that may appear or progress years after treatment. – Studies supporting the claim
Childhood-cancer survivors carry a large burden of chronic disease caused by treatment ('severe chronic conditions in 62%'). – Studies supporting the claim

Use by the Tayebat archive

Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.

What the archive says about the study: Mulrooney 2009 BMJ: 14,358 childhood-cancer survivors had hazard ratios of 5.9 for CHF, 5.0 for MI, 6.3 for pericardial disease, 4.8 for valvular abnormalities versus siblings.
A. What was actually studied?: Late cardiac outcomes in survivors vs siblings.
B. What did the authors find?: Substantially increased relative hazards, dose-related to anthracyclines and cardiac radiation.
C. What conclusions do the authors themselves draw?: Survivors need awareness and follow-up for cardiovascular disease.
D. What does the archive derive from it?: Heavy long-term cardiac price of childhood cancer therapy.
E. Is this derivation scientifically justified?: Numbers are correct. The comparison group is healthy siblings, not untreated patients; survivors exist because treatment cured them. The full text gives absolute figures the archive does not report: CHF prevalence 1.7% vs 0.2% and 30-year cumulative incidence 4.1%; risk was concentrated at anthracycline >=250 mg/m2 and cardiac radiation >=1500 cGy.
Adequacy of the interpretation (by the archive): correctly reported