P01.21.A Testing Semantic Verbal Fluency (SVFT) in patients with headache suspicious of cancer helps predict those with tumour on imaging

Background

Differentiating secondary headache from primary headache on history alone can sometimes be difficult. Headache “plus” other subtle cognitive features improves prediction. Patients and relatives may attribute subtle cognitive or behavioural changes to other causes (e.g. anxiety, low mood). Sematic verbal fluency - “how many animals can you think of in a minute” is commonly impaired on admission in patients with headache harbouring a tumour. We were interested whether SVFT at point of referral for scan would help predict a tumour and requested GPs report the SVFT on electronic referral for direct access scanning in cases with a history of headache suspicious of cancer. The aim of the pilot study was to establish if the SVFT was a “red flag” for secondary causes of headache. We report the results of a 20-month audit of cases.

Material and Methods

An electronic Protocol Based Referral (PBR) was developed for Headache Suspicious of Cancer to expedite scanning. The PBR sat alongside the routine e-DACI system that had been in use for some years. The GP was asked to complete the SVFT at the time of referral for scanning. Other data were gathered: PMH cancer; other symptoms/signs; co-morbid conditions and medications.

Results

GPs submitted requests for ePBR scanning in 669 cases over 20 months (62% females; Mean age 53: 60% <60 years). SVFT was completed on the request form in 381 (57%) cases. In these cases median SVFT was 17 animals. 11/381 cases were found to have cancer on scanning (2.9%): 10/188 cases with intracranial tumours had a SVFT <17 (5.32%) compared with one with a SVFT >=17. The median SVFT in cases with cancer was 10 animals. Other possible causes of SVFT <17 were - 53 psychiatric or chronic pain conditions on multiple drugs;12 were not native English speakers; 19 had co-existing dementia; 5 had small vessel disease; 4 cysts; 4 Giant Cell Arteritis ; 2 Chiari 1 malformation; PMH - encephalitis (1).

Conclusion

A SVFT result <17, at the point of referral for brain imaging, in patients with headache suspicious of cancer was associated with intracranial cancer in more than 1:20 cases, whereas a SVFT of >=17 was associated with cancer in 1:200 cases. SVFT may be an additional useful “red flag”. The most appropriate SVFT cut-off requires more research in a larger study. Low SVFT in headache patients may inform Cancer Referral Guidelines, improve the identification of secondary headache and help expedite cases.

TM1-4 Verbal fluency test in patients with a newly diagnosed brain tumour

Objectives

Utility of a simple 1 min cognitive screening tool, verbal fluency test, as a potential risk assessment tool for GPs for a suspected brain tumour.

Design

Case-control study; patients with new diagnosis of brain tumour with a history of headache and patients referred for direct-access-CT for headache without a brain tumour.

Subjects

102 brain tumour patients: 34.3% HGG, meningioma 21.6%, cerebrals metastases 17.6%, LGG 11.8%, others (pituitary, schwannoma, haemangioblastoma) 11.8%, and CNS lymphoma 2.9%.

Methods

Group differences analysed with ANCOVA with age/gender as covariates.

Results

Mean age was similar across both groups. There were more females in the control group. Tumour patients obtained significantly lower scores on fluency test measures. The largest effect size difference was observed for semantic total and phonemic total scores (Cohen’s d=−0.97 and −0.47, respectively. Brain tumour patients named on average 4.9 animals less (95% CI −6.32,–3.42) (p<0.001) than controls. On letter P task, there was a 2.2 word-mean difference (95% CI −3.6,–0.86)(p=0.001). Forty-eight (47.1%) tumour patients and 22 (23.4%) controls had subjective memory disturbance. Tumour patients performed equally on both fluency tasks regardless of memory complaints (p>0.5).

Conclusions

Semantic test is more reliable to discriminate patients with a brain tumour. In contrasts to patients without a brain tumour, subjective memory impairment is not associated with a reduced performance on verbal fluency tasks.

WP1-16 Does time to diagnosis influence surgical decision making in primary malignant brain tumours?

Objectives

To assess whether time to diagnosis influenced surgical procedure and survival in primary malignant brain tumours

Design

Retrospective cohort study of adult patients newly diagnosed with a primary intra-cerebral tumour April 2010 – March 2015 referred to neuro-oncology centre in South-East Scotland

Subjects

400 patients with primary malignant intracerebral tumours

Methods

Univariate logistic regression analysis and multivariate Cox regression survival analysis were performed

Results

Our dataset included 229 Males (57.7%), mean age 62.05 (SD14.01). Time to diagnosis did affect choice of surgical procedure. Patients undergoing biopsy had significantly longer TtD than those undergoing resection (28 vs 21 days, MWU test, p=0.035). 233 (58.3%) patients were deceased at 12 months. Having a cognitive first symptom (e.g confusion/memory changes) predicted not getting any surgical procedure (OR 2.95 [95% CI 1.8–4.8]). Where surgery was performed, first cognitive symptom was not predictive of type of surgery (biopsy vs debulking/excision) (OR 1.5 [95% CI 0.8–2.9] norhaving tumour in the frontal lobe (OR 0.9 [95% CI 0.5–1.7]. In multivariate analysis, increased hazard ratio of death at 12 month was significantly associated with a cognitive first symptom (HR 1.4 [95% CI 1.0–1.9] and KPS <70 (HR 1.69 [95% CI 1.2–2.4]).

Conclusions

Earlier diagnosis is associated with getting debulking surgery rather than biopsy, and patient’s presenting symptoms influence surgical decision making and outcome.