CT in children with abdominal cancer: should we routinely include the pelvis?

Abstract   Background: It has been suggested that the pelvis should not be habitually included on abdominal CT examinations, but the potential benefit of such a practice in childhood abdominal malignancies is unknown. Objective: To estimate the yield and potential diagnostic benefit of abnormal findings on CT of the pelvis in children with malignant primary tumours in the upper abdomen. Materials and methods: From a paediatric tertiary referral hospital we retrospectively included patients having abdominal CT for primary upper abdominal tumours (1997–2004), the scan range routinely including the pelvis. We reviewed and tabulated any pelvic abnormality, and calculated group proportions with 95% confidence intervals. Results: We identified 230 children (2 days to 17 years old, median 2.9 years). Six (2.6%; 95% CI 0.5–4.7%) had abnormalities in the pelvis that would not have affected clinical management. Four (1.7%; 95% CI 0.1–3.4%) had findings that might have influenced staging, but only one was not detected by other modalities within 1 week of the CT. Conclusions: Our data suggest that diagnostically significant findings in the pelvis are rare; consequently, the habitual inclusion of the pelvis on abdominal CT for primary malignant tumours in the abdomen is not justified.