Fourteen retrospective studies (n greater than 14,417) were included: 13 retrospective cohorts and one cohort with a nested case-control study (n=2,710). The number of patients in one study was unclear.
Methodological problems with the studies included poorly described study group, incomplete follow-up, no blinding of the outcome assessors, inadequate definition of risk estimation and lack of adjustment for other risk factors. Further details were given in the report.
The cumulative incidence of clinical CVE ranged from 0.3% to 22.8% (9 studies). Standardised incidence ratios could not be estimated from the data provided.
The cumulative incidence of CVM ranged from 0% to 3.5% (11 studies).
SMRs (adjusted for duration of follow-up, age and gender) ranged from 0 to 68, based on 6 studies of paediatric Hodgkin's lymphoma. Four studies showed significantly increased SMRs after radiotherapy in comparison with the general population. The pooled SMR of 28.4 was based on 4 studies.
The absolute excess risk ranged from -0.4 to +17.7 excess cardiac deaths per 10,000 person-years adjusted for age, gender, race and calendar period (4 studies). The actuarial risk for CVM increased from 0.3% at 10 years to 10.2% after 25 years (2 studies).
The RR of CVD (both CVE and CVM) following radiation was reported in 3 studies. Two of the 3 studies showed a significantly increased risk with radiation compared with no radiation. In one study the risk of CVM was significantly greater in those given radiation compared with no radiation (RR 2.2, 95% CI: 1.2, 4.4, P<0.05) after adjusting for gender, age at diagnosis and follow-up time. One study found that the risks of clinical CVE were increased with increasing dose of radiation, for female patients and with anthracycline dose. In one study there was no difference in risk when adjusting for follow-up (RR 1.75, 95% CI: 0.56, 5.43, P=0.34).