Serial cerebrospinal fluid examinations to diagnose hematological malignancy causing neurological disease

Abstract

To determine the diagnostic utility of serial cerebrospinal fluid (CSF) examinations for hematological malignancy causing neurological disease. All CSF cytology reports at Mayo Clinic Rochester from 2005 to 2014 (n = 20,018) were reviewed. Study inclusion criteria were: repeated CSF examinations within 1 year in patients without known hematological malignancy performed to determine if hematological malignancy was the cause of neurological disease. Exclusion criteria were: preexisting hematological malignancy; >1 year between CSF examinations, serial CSF examinations for infection, tumor surveillance or intrathecal therapies, and for assessment or treatment of CSF dynamics (e.g. idiopathic intracranial hypertension, CSF shunt or persistent CSF leak). The initial study population included patients undergoing three or more serial CSF examinations; subsequently those undergoing two serial CSF examinations were investigated. A total of 613 patients met the study criteria with 477 having two CSF examinations and 136 having three or greater CSF examinations. Of those with three or greater serial CSF examinations none were found to have hematological malignancy exclusively on the third or subsequent CSF examinations. Of those with two CSF examinations 0.4 % (2/477) were found to have hematological malignancy (large B cell lymphomas) exclusively on the second CSF. Ten patients (1.6 %) had suspicious hematological abnormalities on initial CSF examinations confirmed on subsequent CSF examinations. Serial CSF examinations are of low yield to diagnose hematological malignancy as a cause of neurological disease but may confirm atypical features observed in an initial CSF examination.