Abstract
Background This study developed a nomogram using quantitative indices of this subregion before chemoradiotherapy (CRT) to predict early GBM recurrence. Methods Adult patients with GBM diagnosed between October 2018 and October 2022 were retrospectively analyzed and randomly divided into training and validation groups. Using T1-weighted imaging enhancement and FLAIR fusion maps, the CRT extra-residual FLAIR high-signal area was segmented into categories and the signal intensity of each subzone was measured.The study compared clinical, pathological, and imaging indexes between recurrent and non-recurrent groups, identified independent prognostic risk factors, and developed a prediction model using univariate Cox analysis and LASSO Cox regression analysis. The discriminatory ability of the model was assessed using the C-index, and its performance was evaluated through calibration curves and decision curves. Results A study found that 53.4% of 129 patients with GBM experienced postoperative recurrence. Factors such as the subventricular zone involvement, enhanced regional outside the residual cavity (ER) median, enhanced + unenhanced regional outside the residual cavity (ER + UR) rFLAIR, and corpus callosum involvement were identified as independent predictors of recurrence. The model had a C-index of 0.733 in the training group and 0.746 in the validation group for predicting recurrence at 1 year post-surgery. Patients were also stratified based on these factors.Patients were divided into high and low-risk groups based on their nomogram score, showing a significant difference in progression-free survival between the two groups. Conclusions Quantitative assessment of FLAIR high signal areas in GBM after segmenting subregions shows promise for predicting survival prognosis. Emphasizing specific subregions may improve predictive accuracy.