In recent years, the study of bone disorders has changed thanks to the
progressive development of imaging techniques. In our opinion, howeve
r, different from what happens in the study of pathologic conditions i
n other organs and systems, conventional radiology remains the diagnos
tic cornerstone for bone disorders and especially bone tumors. In any
case, plain radiography must always be performed first. The radiologic
approach to primary bone tumors includes a series of steps to collect
the necessary information to suggest, with the following contribution
of radiography, the most likely diagnosis. The approach is basically
statistical and many parameters should be considered. As far as bone t
umors in children are concerned, the basic role of the statistical var
iable 'patient age' is apparent and this information should be conside
red even before any others. Despite the difficulties in assessing the
frequency of bone tumors we tried to read homogeneously some of the st
atistical data reported in literature. Tumor incidence in a given age
class is a very important diagnostic tool. Lesion location is the seco
nd step in the radiologic work-up: the involved bone and bone segment
should be assessed and correlated with the preferential site of a spec
ific rumor. Regarding the bone tumors detected in the first 5 years of
life, an interesting question is whether a vertebra plana is due to a
benign or a malignant lesion. The lytic lesion will be well defined b
y CT and MRI will be very useful in depicting bone marrow and surround
ing soft tissue involvement. However. many processes have a similar im
aging pattern. We believe that before biopsy the frequent, marked infi
ltration of surrounding soft tissues in vertebra plana should suggest
a diagnosis of malignant tumor rather than of tumor-like lesion. (C) 1
998 Elsevier Science Ireland Ltd. All rights reserved.