Recording neurological impairment in clinical trials of glioma.

The criteria for clinical response to treatment in cerebral glioma remain poorly defined, but could be made more objective if simple measures of neurological impairments were included in the definitions. We assessed the utility of simple fast previously validated tests of limb impairment (Timed nine hole peg test and 10 meter walk), memory (Williams delayed recall test) and language (Boston Aphasia Severity Rating Scale) in fifty patients with primary brain tumours to see if they could act as a surrogate for neurological impairment. The tests were compared with established measures of physical disability (Barthel Disability Index [BDI]) and handicap. Timed tests of hand function and gait were sensitive to minor impairments and were abnormal in patients with physical disability on BDI. Timed tests correlated well with handicap (rank correlation 0.734). Short term memory was impaired more commonly with tumours involving the left hemisphere (p < 0.01). Dysphasia limited testing of memory in 8%. Depression was associated with problems in limb function (p < 0.01), memory (p < 0.001), language (p < 0.001), BDI (p < 0.001) and handicap (p < 0.001). The number of abnormal fast tests also correlated with the severity of handicap (rank correlation 0.786) indicating that memory impairment and aphasia contribute to handicap and should be assessed. Median time to complete all assessments was 7 minutes 20 seconds. Utilization of these simple tests will add sensitivity and objectivity to evaluation of neurological response in clinical trials and can be performed quickly by non medical staff.