P01.156 Verbal fluency as a brief cognitive screening tool in newly diagnosed brain tumour patients A potential red flag

Abstract

BACKGROUND: Early diagnosis of brain tumours is one of the top 10 research priorities. Early symptoms are often non-specific such as headache or cognitive/personality changes which will not cause high enough suspicion of a brain tumour for a General Practitioner (GP). As a result, there are often a delay in patient presentation to primary care, referral for investigation and ultimately in diagnosis. We assessed the utility of a simple 1-minute cognitive screening tool, verbal fuency test, as a potential risk assessment tool for GPs for a suspected brain tumour. MATERIAL AND METHODS: This is a prospective cohort study of pre-surgical 102 adult brain tumour patients with a history of headache and 90 controls referred for direct-access head imaging for investigation of headache who completed semantic (animal) and phonemic (letter P) verbal fuency tasks. Recruitment period was between August 2016 and February 2018. Fluency performance was reported as a total score, mean cluster and switching score. RESULTS: Mean age was similar across both groups. There were more females in the control group. Tumour patients comprised of a heterogenous histology population: 33.3% had glioblastoma multiforme (GBM), followed by meningioma (21.6%), cerebral metastases (17.6%), glial series (12.7%), others (pituitary, schwannoma, haemangioblastoma) (11.8%) and CNS lymphoma (2.9%). Tumour patients obtained significantly lower scores on all fuency test measures, however the largest effect size difference was observed for semantic total and phonemic total scores (Cohen's d = -0.89 and -0.44, respectively). Brain tumour patients named on average 4.6 animals less (95%CI-6.7, -3.13) (p<0.001) than controls. On a letter P task, there was a 2 word-mean difference (95%CI-3.4, -0.73) (p=0.003) for the groups. Fourty-eight (47.1%) tumour patients and 22(23.4%) controls had subjective memory disturbance. Tumour patients performed equally on both fuency tasks regardless of memory complaints (p>0.5). There were however significant differences on both animal and letter P fuency total scores for controls with subjective memory disturbance compared to those without (p=0.04 and p=0.001, respectively). CONCLUSION: of both fuency tasks, semantic test is more reliable to discriminate patients with a brain tumour. In contrast to patients without a brain tumour, subjective memory impairment is not associated with a reduced performance on verbal fuency tasks, perhaps suggesting that the test should be performed in all patients suspected of a brain tumour in primary care.