The ability of positron emission tomography (PET) to detect spinal cord tum
ors was studied prospectively in 14 patients presenting over a 5-year perio
d. Abnormal uptake by [F-18]-fluorodeoxyglucose (FDG) or C-11-methionine wa
s detected in all except one. These data were assessed in relation to magne
tic resonance imaging (MRI) findings with regard to tumor type and extent p
reoperatively, findings at operation, and subsequent clinical course. The g
roup consisted of six astrocytomas, five ependymomas, one mixed ependymoma
and astsocytoma, one schwannoma, and one ganglioglioma, all confirmed histo
logically. This is the largest study comparing spinal PET to MRI. Accurate
preoperative correlation between PET and MRI was found in all eight patient
s scanned at first presentation. The PET uptake was in keeping with the low
-grade histology of the tumors. Postoperatively, PET and MRI findings were
in agreement in nine patients. In eight of these the findings were in keepi
ng with the subsequent clinical course. In three patients, however, the PET
findings were at variance with the clinical course and MRI findings. In on
e, persistent FDG uptake after radiotherapy was seen where there was subseq
uent tumor resolution. In two patients with low-grade astrocytomas, scanned
with FDG and C-11-methionine, respectively, tracer was not taken up by res
idual tumor. In this small group of patients, PET did not provide additiona
l useful information. This could be because all tumors studied were low gra
de and the limited spatial resolution of PET does not lend itself to imagin
g small spinal cord tumors. The prospective study of larger numbers of pati
ents with a wider range of tumor types is required, but this might be diffi
cult to achieve given the rarity of spinal cord tumors.