Jp. Krischer et al., RISK-FACTORS FOR EARLY ANTHRACYCLINE CLINICAL CARDIOTOXICITY IN CHILDREN - THE PEDIATRIC-ONCOLOGY-GROUP EXPERIENCE, Progress in pediatric cardiology, 8(2), 1997, pp. 83-90
Purpose: To evaluate risk factors for clinical cardiotoxicity from ant
hracycline chemotherapy in children with cancer and to estimate the re
lative risk associated with each factor singly and with different comb
inations of risk factors. Patients and Methods: The study population c
onsisted of 6493 children with cancer who had received anthracycline c
hemotherapy on Pediatric Oncology Group protocols during the period fr
om 1974 to 1990. Cardiotoxicity, defined as congestive heart failure n
ot due to other causes, abnormal measurements of cardiac function that
prompted the discontinuation of therapy, or sudden death from presume
d cardiac causes, was determined by a review of protocol records. Resu
lts: Cardiotoxicity was confirmed in 106 patients (1.6%): 58 had conge
stive heart failure, 43 had changes in measures of cardiac function th
at prompted the discontinuation of therapy and five died suddenly from
presumed cardiac causes. In a multivariate analysis, factors contribu
ting to the relative risk (RR) of toxicity were a cumulative dose of a
nthracycline greater than or equal to 550 mg/m(2) of body-surface area
(RR = 5.2), a maximal dose of 50 mg/m(2) (RR = 2.8), female sex (RR =
1.9), black race (RR = 1.7), the presence of trisomy 21 (RR = 3.4) an
d exposure to amsacrine (RR = 2.6). The relative risk of early clinica
l cardiotoxicity increased with increasing numbers of risk factors and
was projected to exceed 405 when all six statistically significant ri
sk factors were present. Conclusion: Early clinical cardiotoxicity in
children treated with anthracycline is rare. A high maximal dose, or c
umulative dose of anthracycline, female sex, black race, the presence
of trisomy 21 and treatment with amsacrine increase the risk for anthr
acycline-associated cardiotoxicity. The cumulative effect of multiple
risk factors can be estimated as the product of the relative risks ass
ociated with each. (C) 1998 Elsevier Science Ireland Ltd.