Detection of brain metastases: comparison of contrast-enhanced MR with unenhanced MR and enhanced CT.

Contrast-enhanced MR studies were compared with noncontrast MR and contrast-enhanced CT scans in the evaluation of intraparenchymal brain metastases. Fifty consecutive inpatients were studied with short and long repetition time (TR) sequences before and after the administration of gadopentetate dimeglumine. In addition, a delayed short TR sequence was performed. The contrast CT, noncontrast MR, immediate postcontrast short TR sequence, postcontrast long TR sequence, and delayed postcontrast short TR sequence were each read blindly and independently by two neuroradiologists. These results were then compared with a final interpretation, reached by all the neuroradiologists in the study, using all the clinical information and imaging findings. Postcontrast short TR scans proved to be superior to other sequences. They were particularly useful in the detection of metastases in the posterior fossa and cortex. The delayed postcontrast short TR scan held no definite advantage over the immediate postcontrast short TR scan, although metastases were sometimes seen slightly better after the delay. While long TR sequences were not always sensitive or specific, they often did provide ancillary information and were particularly useful in cases of hemorrhagic metastases. Because of these findings, we recommend that the evaluation of intraparenchymal metastases consist of a single postcontrast long TR scan followed by a single postcontrast short TR scan. While these sequences should be very accurate in the detection of metastases, we also generally perform a single precontrast short TR scan as well, since the question of hemorrhage or bone lesion may be clinically relevant.