INTRACRANIAL BIOPSY FOR SUSPECTED TUMOURS; A COMPARISON OF SURGICAL APPROACHES

Abstract AIMS There are currently no standardised methods for brain tumour biopsy in neurosurgery. At the Walton Centre, a number of various methods are utilised and to determine the most accurate and safest method for brain tumour biopsy, we reviewed our case series at the Walton Centre with the aim of minimising repeat procedures by establishing diagnostic yield and complications per approach. METHODS Retrospective case series of over 500 patients (age range: 20-80 years, average: 58 years; female:male ratio 1:1.45) undergoing brain tumour biopsy at the Walton Centre between 01/01/2016 and 30/09/2024, excluding procedures such as abscess drainage or tumour debulking. Data including demographics, length of stay, biopsy method, histological diagnosis, complications and outcome was obtained from the Trust electronic system. RESULTS To date, we have reviewed 215 cases of which 209 had diagnosis after one procedure providing a diagnostic yield of 97%. 6 patients required repeat biopsy with an open procedure most common cause for repeat biopsy. Biopsy methods burr hole (15), CRW/Stereotactic Frame (59), EM-guided or Varioguide (Brainlab) (19), Endoscopic (2), open/mini-craniotomy (59), Stealth, Vertex or Navigus (Medtronic) (59), (13), 2 biopsies with no documented procedure type. No other significant trend observed at first stage of data collection, but post-operative surgical complications in 7 patients (3.3%). Post-operative medical complications in 15 patients (7.0%). The average length of stay: 3.9 days for elective biopsies, 20 days for emergency admission biopsies. 16.3% of patients were discharged home as a day case. CONCLUSION Permanent neurological complication rates were low, with an average biopsy length of recovery of 2.1 days. Our case series suggests a trend towards reduced diagnostic yield and greater length of stay in open biopsies, with a range of experience using navigational approaches being evaluated from the complete dataset.