Spatial patterns of progression in reported glioblastoma cohorts after upfront chemoradiation and salvage therapy: a systematic review and meta-analysis

PURPOSE: Spatial patterns of glioblastoma (GBM) progression inform focal salvage eligibility, but interpretation across studies is limited by heterogeneous spatial definitions, high between-study heterogeneity, and incomplete reporting of non-enhancing/FLAIR-dominant progression. We synthesized reported progression-location involvement after upfront radiotherapy/temozolomide (RT/TMZ) and salvage therapy. METHODS: We systematically searched PubMed and Scopus from January 1, 2000, to February 1, 2026, and performed arm-level random-effects meta-analysis of logit-transformed proportions. Enhancing progression was harmonized as non-mutually exclusive local/in-field, marginal/field-edge, and distant/out-of-field involvement among evaluable cases, using each study's spatial framework. The supplementary broad escape-pattern construct was defined as reported progression beyond purely local/in-field enhancing failure. RESULTS: We included 107 studies (122 arms; 10,529 patients). At first progression after upfront RT/TMZ ± TTF, reported local/in-field enhancing involvement among evaluable cases was 79.2% (95% CI 75.7-82.3; k = 94; n = 6,989; I²=84.8%), and distant/out-of-field involvement was 16.9% (95% CI 14.3-19.9; k = 89; n = 6,493; I²=85.2%). After salvage therapy, local/in-field involvement remained the majority pattern reported among evaluable cases (59.8%, 95% CI 52.8-66.4; k = 27; n = 1,572; I²=80.5%), with distant/out-of-field involvement of 17.5% (95% CI 12.9-23.3; k = 21; n = 1,307; I²=75.5%) and a supplementary broad escape-pattern estimate of 38.6% (95% CI 31.9-45.9; k = 27; n = 1,572; I²=80.1%). Non-enhancing/FLAIR-dominant progression was sparsely and likely selectively reported; only six recurrent/salvage arms contributed data, yielding an exploratory pooled estimate of 27.8% (95% CI 12.8-50.3; n = 216; I²=75.7%). CONCLUSION: Published GBM cohorts show predominantly local/in-field enhancing progression after upfront therapy (~ 80%) and persistent majority local/in-field involvement after salvage therapy (~ 60%). Standardized reporting of enhancing, non-enhancing/FLAIR-dominant, disseminated, posterior fossa, and brainstem progression is needed.