Diagnosis and treatment of leptomeningeal metastases from solid tumors: experience with 90 patients.

The clinical findings and response to treatment of leptomeningeal metastases from solid tumors are analyzed in 90 patients treated at Memorial Sloan-Kettering Cancer Center during the period from January 1975 to February 1980. Patients included those who had either typical clinical findings of leptomeningeal tumor or conclusive laboratory evidence supporting the diagnosis. Carcinoma of the breast (46 patients), lung (23 patients) and melanoma (11 patients) were the common primary tumors. Symptoms of leptomeningeal metastasis occurred as the presenting sign in five patients and as late as ten years after the primary tumor was diagnosed in four other patients. Most patients had active systemic disease outside the nervous system. Signs and symptoms could be classified as involving either the brain, cranial nerves, or spinal nerves. Most patients had either symptoms or signs in more than one area at the time the diagnosis was established. The initial spinal fluid examination was abnormal in all but three patients, but only 49 had cytologic evidence of leptomeningeal metastases. Repeated spinal fluid assay yielded a positive cytology in 82 patients. Measurement of biochemical markers, including beta-glucuronidase, carcinoembryonic antigen and lactic dehydrogenase, assisted in the diagnosis. Approximately half of the patients treated by intraventricular methotrexate experienced improvement or stabilization of neurological symptoms for more than a month; median survival was 5.8 months after diagnosis, with a range of 1--29 months. In 18 patients disease was limited to the nervous system, and median survival was eight months, with four patients surviving one year and two patients for two years. Side effects of therapy were, for the most part, minor. We conclude that vigorous treatment of leptomeningeal metastases with intrathecal chemotherapeutic agents improves symptomatology in some patients, and at times prolongs survival.