The Predictive Validity of Language Mapping for Post-operative Paediatric Epilepsy Surgery Outcome

Objectives: Estimating risk of language decline is warranted in paediatric epilepsy surgery but there is limited knowledge of the validity of language mapping techniques in this context. This paper's main aims were to (1) evaluate the state of the evidence investigating the predictive validity of any language mapping technique used in epilepsy surgery and (2) provide a synthesis of available evidence to inform future research and practice. Methods: A systematic search of MEDLINE, APA PsychNET, and Web of Science (up to October 2020) was conducted. Records of cohort studies that reported mapping techniques and surgical outcomes in paediatric epilepsy were identified. Quality appraisal was undertaken with QUADAS-2 and OCEBM evidence levels. Relationships between mapping techniques and post-operative language outcome were synthesised across studies. Results: Six studies met criteria for inclusion. These studies examined diffusion weighted imaging, electrocortical stimulation and the Wada test. The quality of five records was assessed as 'unclear' due to the lack of reporting of post-operative outcome methods and one was classified as 'low' quality. Diffusion weighted image mapping was found to have a strong relationship with language outcome across three studies. Electrocortical stimulation also demonstrated robust predictive outcome in one record but with high failure rates (48%) in another. Wada had variable success depending on reference standard thresholding. Significance: There is great need for further investigation of language mapping techniques in predicting post-operative outcome in paediatric epilepsy surgery. Preliminary evidence demonstrates promising value of diffusion weighted imaging and electrocortical stimulation techniques within paediatric practice; however, further work needs to be undertaken to support adoption of techniques into routine practice. Future research should specifically improve reporting of outcome methodologies. Other common mapping techniques should be investigated due to the successful translation of DWI and ESM into paediatric populations from adult practice.