We retrospectively reviewed the clinical records of all patients with
cancer and brachial plexopathy who had an MRI of the brachial plexus b
etween 1984 and 1993 (71 patients). The MRIs were reevaluated in a bli
nded fashion. The presence of a mass adjacent to the brachial plexus o
n MRI was highly predictive of tumor infiltration as determined by cli
nicopathologic criteria and was the most useful feature in distinguish
ing radiation plexopathy from tumor infiltration. Increased T-2 signal
in or near the brachial plexus was commonly seen in both groups and w
as not useful in this distinction. MRI was very sensitive for brachial
plexus abnormalities in this condition, and limited comparison with C
T suggested that MRI is superior to CT as an imaging modality. CT perf
ormed sufficiently well, however, to suggest that a prospective compar
ison study of the cost effectiveness and clinical utility of the two i
maging modalities in this clinical setting is warranted.