Diagnosis and outcome measures in trials for neoplastic meningitis: a review of the literature and clinical experience.

Innovative approaches to the treatment of neoplastic meningitis are being widely tested. Unfortunately, research on diagnostic strategies and outcome measures on which any advances in treatment ultimately depend, has not been avidly pursued. A critical review of the literature on neoplastic meningitis published since 1978 was undertaken by using MEDLINE and other English language databases. All articles addressing the issues of diagnostic or response criteria were included. Randomized clinical trials (RCTs) were emphasized. Prospectively collected data from the authors' institution correlating the results of cerebrospinal fluid (CSF) cytological examinations with Karnofsky Performance Scale (KPS) score are also discussed. Twenty-six studies (representing 1208 patients) fulfilled search criteria. Only three were RCTs. Cerebrospinal fluid cytology was the sole diagnostic criterion in two-thirds of studies. The results of CSF cytological examination alone or in combination with other clinical or laboratory endpoints constituted the primary outcome measure in 85%. Few studies attempted to address known deficiencies in the reliability and validity of these measures, and correlation between measures was poor. Quality of life was never used as a primary outcome measure. All currently available measurements, including CSF cytology, biochemistry, immunological, and molecular markers, neuroimaging studies, clinical examination, and survival, suffer from poor sensitivity and/or specificity, and often correlate poorly with each other. Although CSF cytological examination, performed according to a rigorous, research-supported protocol, may be the optimum diagnostic and outcome measure at this time, additional research is a prerequisite for any further advances in the clinical care of patients with neoplastic meningitis.