SURG-17. Impact of introduction of sagittal dimensions to predictive model of post-operative cerebellar mutism (CMS): A report of the Nottingham CMS Workshop

Abstract

Abstract BACKGROUND: The Nottingham-Liverpool risk stratification tool for predicting post-operative cerebellar mutism syndrome (CMS) requires further refinement. Previous smaller studies identified tumours arising in cerebellar hemispheres (CH+ve) as very low risk and proposed that larger sagittal dimensions may be associated with a higher CMS risk. METHODS: An international multi-centre workshop assembled clinical, imaging and surgical data from 242 posterior fossa tumour patients from 9 centres (age 2-23.8, mean 8.23 years; histology 122 medulloblastoma, 75 pilocytic astrocytoma, 25 ependymoma, 18 other tumour types, 2 not known). Post-operative CMS developed in 0/66 CH+ve and 40/176 CH-ve patients (23%, range 13-33%) and. Tumour measurements in the sagittal plane were incorporated into the predictive model. RESULTS: In CH-ve tumours, CMS was associated higher D-Sag (CMS+ve 3.91±1.08 vs CMS-ve 3.01±1.56 cm p=0.001), d-Sag (CMS+ve 0.63±0.27 vs CMS-ve 0.44±0.31 cm p=0.001) and D-Sag*d-Sag (CMS+ve 2.60±1.36 vs CMS-ve 1.65±1.45 cm2 p=0.001). The Rotterdam model identified CMS rates for CH-ve tumours for low, intermediate and high risk groups as: 4% (2/45), 20% (16/79) and 42% (22/52), respectively. CONCLUSIONS: Incorporating D-sag and d-Sag estimates of tumour size, proposed, and adopted by the Beijing and Rotterdam groups, respectively, enhanced the objectivity and supports the model’s CMS predictive capacity for the purposes of consent and future trials. Higher CMS risk, associated with larger sagittal dimensions, may account for variability in CMS rates between cohorts and, as a new finding, may have broader clinical implications around the timing of diagnosis.