Authors: Deborah Boyett; Zachary Englander, MD; George Zanazzi, MD, PhD; Tamara Marie, BA; Guy McKhann, MD; Michael Sisti, MD; Jack Grinband, PhD; Peter Canoll, MD, PhD; Jeffrey Bruce, MD (New York, NY)

Introduction: Evaluating for tumor recurrence in post-treatment glioma is a challenge as radiographically appearing contrast-enhancing (CE) regions are a mixture of tumor cells and treatment effect. This study characterizes intratumoral heterogeneity using quantitative digital pathology to correlate intraoperative MRI-localized biopsies with histopathology in the post-treatment setting. These findings are being used to inform a multiparametric radiographic model of intratumoral heterogeneity. Methods: A retrospective review was performed on adult patients with MRI-localized biopsies obtained during resection for post-treatment recurrent high grade glioma. 46 patients and 133 MRI-localized samples were analyzed (median 2 samples/patient). A histopathological classification was developed to assess each sample for relative abundances of treatment effect and recurrent tumor. Immunohistochemistry (IHC) with SOX2, CD68, and Ki67 was used to validate the classification and further characterize the samples. Slides were digitized and quantified using an automated cell-counting algorithm. IHC quantification was compared across histological groups using ANOVA and paired t- tests. Results: Of 34 patients with multiple biopsies, 25 (74%) demonstrated heterogeneity (as assessed by histopathological classification). 71/133 (53%) biopsies showed predominantly treatment effect and of those, 40/71 (56%) specimens were CE. 19% (12/62) of non-enhancing specimens contained predominantly recurrent tumor. Cell density was correlated with histopathological classification ( p =2.2e-16). SOX2 and Ki67 staining were higher in specimens containing predominantly tumor ( p =2.2e-16, p =1.269e-14). CD68 staining was higher in specimens containing substantial treatment effect ( p =1.733e-07). Conclusion: This study shows that contrast enhancement is not a reliable predictor of tumor presence in post-treatment recurrent malignant gliomas. A majority of patients demonstrated marked intratumoral heterogeneity, highlighting the difficulty of adequate tumor sampling and diagnostic accuracy in the post-treatment setting. In addition to validating the histological classification, IHC revealed that SOX2 is a useful marker for quantifying tumor burden. These results are being used in the development of radiographic models to map intratumoral heterogeneity.