Abstract
INTRODUCTION: Vestibular schwannoma (VS) often shows transient post-radiosurgery swelling ("pseudoprogression," PP) that mimics true progression (TP). We validated volumetric rules to separate PP from TP and linked growth patterns to clinical outcome. METHOD: We retrospectively reviewed 259 sporadic VS cases treated with single-fraction stereotactic radiosurgery (2012-2024). SRS was delivered with a median margin dose of 12 Gy, prescribed to the institutional peripheral isodose line, using a median of 5 isocentres. Tumours were manually segmented on serial MRIs ≥ 24 months. PP was defined as > 20% volume rise ≤ 12 months followed by ≥ 10% fall ≤ 24 months; TP as > 15% rise > 36 months with persistent growth. Two alternative thresholds (> 25% rise any time; >10% rise > 24 months without regression) were tested. Longitudinal trajectories were clustered with Gaussian-mixture models. Outcomes included hearing, facial and trigeminal function, and Dizziness Handicap Index (DHI). RESULTS: Four growth clusters emerged: early PP (35.5%), late PP (11.2%), stable (41.3%) and TP (12%). The PP rule yielded 86% sensitivity and 93% specificity (AUC 0.92); the TP rule 77%/95% (AUC 0.90). Alternative thresholds performed worse (AUC 0.81-0.85). Serviceable hearing was preserved in ≥ 86% of PP or stable tumours but only 61% in TP (p < 0.01). CONCLUSION: Time-anchored volumetric rules accurately distinguish transient post-SRS swelling from genuine progression. Incorporating these criteria into routine surveillance can prevent premature salvage therapy while ensuring timely intervention for the minority of tumours that truly grow.