Twenty-seven cutting needle biopsies were performed on 25 children wit
h suspected malignancy using computed tomographic (CT, 22) or ultrasou
nd (US, 5) guidance. Anatomical sites were: retroperitoneum 6, liver 4
, kidney 4, abdomen/pelvis 4, thorax 4, bowel 2, neck 1. Sixteen patie
nts (64%) underwent subsequent open biopsy (5), marrow biopsy (2) or r
esection (9). There was complete concordance between the histological
findings from the open or marrow biopsy and the previous needle biopsy
in 12 of these 16 patients; in two patients the needle biopsy was mis
leading, causing inappropriate initial treatment in one. In two other
patients needle biopsy was correct but lacked specific diagnostic feat
ures. Needle biopsies were performed under general, local or Ketamine
anaesthesia. There were no apparent complications related to these pro
cedures. We believe that radiologically-guided cutting needle biopsy s
hould replace open biopsy in most children with solid malignant lesion
s. It can easily be performed during a single anaesthetic episode whic
h allows radiological evaluation, biopsy, bone marrow and cerebrospina
l fluid sampling. However, the potential for sampling error and histol
ogical variation within these tumours needs to be borne in mind.