We evaluated 14 consecutive patients with leptomeningeal metastasis prospectively, using both T1-weighted (T1W) gadolinium-DTPA-enhanced MR (Gd-MR) and contrast-enhanced CT (CE-CT). Thirteen had positive CSF cytology; the remaining patient had an atypical CSF lymphocytosis and primary CNS lymphoma. The patients (8M/6F) ranged in age from 8 to 70 years (median, 42 years). Tumor histology included 3 systemic and 2 primary CNS lymphomas, 3 breast carcinomas, 2 leukemias, 1 malignant schwannoma, 1 small cell lung cancer, 1 prostate cancer, and 1 melanoma. Both imaging methods demonstrated parenchymal volume loss equally well in all patients. Gd-MR revealed abnormal enhancement of meninges or parenchyma in 10 patients, including all 5 patients with positive CE-CT. Neither technique revealed any foci of abnormal enhancement in 4 patients. Gd-MR was superior to CE-CT in demonstrating and quantifying enhancing subarachnoid and parenchymal nodules in 6 patients and in demonstrating sulcal, dural, cisternal, tentorial, and ependymal enhancement. Our findings indicate that T1W Gd-MR is the preferred imaging modality in leptomeningeal metastasis and suggest that CE-CT is unnecessary.

To assess the utility of gadolinium-DTPA (Gd) and of MR imaging in the evaluation of spinal cord tumors, ten consecutive patients were prospectively evaluated. T1-proton density-, and T2-weighted images were obtained in sagittal or axial planes. T1-weighted images were obtained before and after intravenous administration. Five tumors were within the cervical spinal cord; 3 neoplasms were within the thoracic cord; 1 neoplasm extended from the cervical to the thoracic cord and 1 neoplasm extended from the cervical cord to the conus medullaris. Four tumors were ependymomas; 3 were astrocytomas; 1 was an hemangioblastoma, and 1 was a metastatic malignant peripheral nerve sheath tumor. The remaining patient died prior to spinal surgery and no autopsy was obtained. Of the precontrast sequences, tumors were best evaluated using T1-weighted images. Abnormal findings included cord widening, presence of a tumor mass, intratumoral or other associated cyst(s), and hemorrhage. Nevertheless, T1-weighted images obtained following the administration of GD were superior relative to all other pre- and post-contrast sequences for defining tumor margins, characterizing cyst(s) and delineating tumor masses. Based primarily on their appearance on post-contrast T1-weighted images, tumor-associated cysts could be subcategorized into 3 types: intratumoral cysts (found within the contrast-enhancing soft tissue mass); nonenhancing extratumoral cysts (found eitherrostral or caudal to the enhancing tumor mass); and enhancing extratumoral cysts (having an enhancing wall or containing an enhancing nodule).(ABSTRACT TRUNCATED AT 250 WORDS)