Path-06. histopathological characteristics of treatment-induced effects versus progression in 180 re-resection samples from irradiated gliomas grade 2-4

Abstract Histopathological studies of re-resected glioma tissue are considered the gold standard to distinguish actual tumor progression (TP) and treatment-induced effects (TIE) after radiotherapy. The objectives of this mono-center, retrospective cohort study were to provide detailed histopathological characterization of re-resected glioma tissue, and to correlate pathological findings with survival (data driven). Consecutive adults with a diffuse glioma (WHO grade 2-4), previously treated with (chemo-)radiation and progressive lesion were reviewed. New histological slides from the recurrent-resected tissue were prepared and stained with routine and exploratory immunohistochemical (IHC) stains. Two neuropathologists used a scoring sheet to independently assess 33 items on standard H&E and IHC stains. We included 180 samples, from 165 patients. Of these, 159 samples (88.3%) originated from a second resection (first re-resection). TIE were diagnosed in 34 (19%), a mixed TIE and TP lesion in 91 (51%), and TP in 55 (31%). Median survival after first re-resection for the groups TIE, mixed and TP were respectively 14.4 (interquartile range (IQR)=8.7-42.4), 7.6 (IQR=4-12.9) and 11.5 months (IQR=7.2-25.4). A diagnosis of TIE was associated with histological findings of hyalinization, endothelial thickening, cortical infiltration, (radio)necrosis, thrombosis, lymphocytic infiltration (CD3, CD8 positive), and positivity for CD34. Percentage of viable tumor and PTBP1 stain correlated negatively with TIE. Survival in the group TIE was significantly longer than in the groups with any viable tumor (p-value of log-rank test 0.008). In multivariable Cox regression analysis, hemorrhage and higher PTBP1 positivity were associated with a poor overall survival after re-resection. Dystrophic calcifications, IDH mutations, and a higher KPS correlated with improved survival. In conclusion, we confirm the previously described characteristics of treatment-induced effects in a large consecutive cohort of re-resected diffuse gliomas. However, none of these characteristics correlated with survival. PTBP1 could be a promising marker to differentiate TIE from TP in treated glioma samples.