Artificial intelligence for early prediction of treatment response in glioblastoma

Abstract

AIMS: Gliomas are the most frequent brain tumours, representing 75% of all primary malignant brain tumours in adults. IDH1 (and IDH2) driver mutations occur in >80% of low grade gliomas and secondary GBMs, in <10% of primary GBMs and other cancers. How IDH1/2 mutations contribute to tumorigenesis is mostly unknown. IDH1/2 convert isocitrate to α-ketoglutarate, but when mutated possess a novel enzymatic function that reduces α-ketoglutarate to D2-hydroxyglutarate (2HG). Indeed 2HG accumulates in IDH1/2-mutant tumours, and this discovery suggested that 2HG may have a role in IDH1/2-mutant tumours onset and progression, possibly by causing dysregulations of various enzymes in the cells. Studies are undergoing to clarify the causative role of 2HG in IDH1/2-mutant tumours , but it is still not clear whether 2HG is the driver/oncometabolite. Our aim is to understand the role of 2HG in developing and adult mouse tissues and whether its accumulation might cause features of gliomagenesis. METHOD: A constitutive D2hgdh Knockout mouse (D2hgdh KO) was generated and the relative molecular and cellular analysis were performed. RESULTS: Brains dissected from D2hgdh KO mice appeared to be histologically normal. No differences were found in the proliferation and labelling retaining capacity of neural stem and progenitors cells (NSC/NPC) of the D2hgdh KO mice compared to controls. A comprehensive metabolites analysis showed that D2hgdh KO mouse accumulated 2HG in various organs and tissues, included total brains and in the NSC/NPC microdissected from the subventricular zone, the site of origin of many human gliomas. The DNA amount of 5mC and 5hmC extracted from brains of D2hgdh KO mice was similar to controls. A normal number of haematopoietic progenitors was also found. CONCLUSION: Although D2hgdh KO mice accumulated 2HG in all tissues analysed, they did not develop any abnormalities and remained completely asymptomatic. This suggests that a mere increment of 2HG in developing and adult tissues may be not sufficient to cause tumori-genesis (and gliomagenesis), leading some doubts on the oncogenic roles of the 2HG in IDH1/2-mutant tumours. AIMS: Treatment response assessment in glioblastoma is challenging. Patients routinely undergo conventional magnetic resonance imaging (MRI), but it has a low diagnostic accuracy for distinguishing between true progression (tPD) and pseudoprogression (psPD) in the early post-chemoradiotherapy time period due to similar imaging appearances. The aim of this study was to use artificial intelligence (AI) on imaging data, clinical characteristics and molecular information within machine learning models, to distinguish between and predict early tPD from psPD in patients with glioblastoma. METHOD: The study involved retrospective analysis of patients with newly-diagnosed glioblastoma over a 3.5 year period (n=340), undergoing surgery and standard chemoradiotherapy treatment, with an increase in contrast-enhancing disease on the baseline MRI study 4-6 weeks post-chemoradiotherapy. Studies had contrast-enhanced T1-weighted imaging (CE-T1WI), T2-weighted imaging (T2WI) and apparent diffusion coefficient (ADC) sequences, acquired at 1.5 Tesla with 6-months follow-up to determine the reference standard outcome. 76 patients (mean age 55 years, range 18-76 years, 39% female, 46 tPD, 30 psPD) were included. Machine learning models utilised information from clinical characteristics (age, gender, resection extent, performance status), O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status and 307 quantitative imaging features; extracted from baseline study CE-T1WI/ADC and T2WI sequences using semi-automatically segmented enhancing disease and perilesional oedema masks respectively. Feature selection was performed within bootstrapped cross-validated recursive feature elimination with a random forest algorithm and Naïve Bayes five-fold cross-validation to validate the final model. RESULTS: Treatment response assessment based on the standard-of-care reports by clinical neuroradiologists showed an accuracy of 33% (sensitivity/specificity 52%/3%) to distinguish between tPD and psPD from the early post-treatment MRI study at 4-6 weeks. Machine learning-based models based on clinical and molecular features alone demonstrated an AUC of 0.66 and models using radiomic features alone from the early post-treatment MRI demonstrated an AUC of 0.46-0.69 depending on the feature and mask subset. A combined clinico-radiomic model util-ising top common features demonstrated an AUC of 0.80 and an accuracy of 74% (sensitivity/specificity 78%/67%). The features in the final model were age, MGMT promoter methylation status, two shape-based features from the enhancing disease mask (elongation and sphericity), three radiomic features from the enhancing disease mask on ADC (kurtosis, correlation, contrast) and one radiomic feature from the perilesional oedema mask on T2WI (dependence entropy). CONCLUSION: Current standard-of-care glioblastoma treatment response assessment imaging has limitations. In this study, the use of AI through a machine learning-based approach incorporating clinical characteristics and MGMT promoter methylation status with quantitative radiomic features from standard MRI sequences at early 4-6 weeks post-treatment imaging showed the best model performance and a higher accuracy to distinguish between tPD and psPD for early prediction of glioblastoma treatment response.