This prospective randomized cooperative group study was conducted in patients with neoplastic meningitis treated with intrathecal methotrexate or thiotepa to assess response rates and survival, prognostic factors, and the toxicity of these regimens.

Fifty-nine adults with nonleukemic malignancies, performance status of 0 to 3, and positive CSF cytologies were assigned to receive intrathecal methotrexate (10 mg) or thiotepa (10 mg) twice weekly. Radiation, was administered to mass lesions and/or symptomatic sites and appropriate systemic therapy was given concomitantly.

Fifty-two patients were assessable. Most were female (79%), nonambulatory (77%), had been pretreated with radiation (52%) and chemotherapy (77%), and had evidence of systemic disease (65%). Most primary cancers were of the breast (48%), lung (23%), or lymphatics (19%). Treatment arms were well balanced, except that more patients randomized to methotrexate had breast cancer (61% v 33%) and were without evidence of systemic cancer (21% v 4%). No patient had important neurologic improvement with therapy, and 75% deteriorated neurologically within 8 weeks of initiating therapy. Survival ranged from 4 days to 110.5+ weeks. Median survival for patients receiving methotrexate was 15.9 weeks and 14.1 weeks for patients treated with thiotepa. Factors predictive of shorter survival included progressive systemic disease (P = .0005), poor performance status (P = .03), and significant cranial nerve palsies (P = .02). Although serious toxicities were similar, mucositis (P = .04) and neurologic complications (P = .008) were more common in patients who received methotrexate.

The efficacy and overall toxicities of intraventricular methotrexate and thiotepa seem similar and neither reverses fixed neurologic deficits. Early diagnosis and treatment and new therapeutic approaches are needed to improve the outcome for patients with neoplastic meningitis.