XGBoost Improves Classification of MGMT Promoter Methylation Status in IDH1 Wildtype Glioblastoma.

Nguyen Quoc Khanh Le,Edward Kien Yee Yapp,Fang-Ying Chiu,Hui-Yuan Yeh,Cheng-Yu Chen,Duyen Thi Do

doi:10.3390/jpm10030128

Abstract

Approximately 96% of patients with glioblastomas (GBM) have IDH1 wildtype GBMs, characterized by extremely poor prognosis, partly due to resistance to standard temozolomide treatment. O6-Methylguanine-DNA methyltransferase (MGMT) promoter methylation status is a crucial prognostic biomarker for alkylating chemotherapy resistance in patients with GBM. However, MGMT methylation status identification methods, where the tumor tissue is often undersampled, are time consuming and expensive. Currently, presurgical noninvasive imaging methods are used to identify biomarkers to predict MGMT methylation status. We evaluated a novel radiomics-based eXtreme Gradient Boosting (XGBoost) model to identify MGMT promoter methylation status in patients with IDH1 wildtype GBM. This retrospective study enrolled 53 patients with pathologically proven GBM and tested MGMT methylation and IDH1 status. Radiomics features were extracted from multimodality MRI and tested by F-score analysis to identify important features to improve our model. We identified nine radiomics features that reached an area under the curve of 0.896, which outperformed other classifiers reported previously. These features could be important biomarkers for identifying MGMT methylation status in IDH1 wildtype GBM. The combination of radiomics feature extraction and F-core feature selection significantly improved the performance of the XGBoost model, which may have implications for patient stratification and therapeutic strategy in GBM.

Highlights

Glioblastomas (GBMs), the most aggressive and exceptionally invasive brain tumors, are characterized by their frequent resistance to chemotherapy and always recurrence following surgical treatment [1]
As the availability of public medical imaging datasets (e.g., TCGA and The Cancer Imaging Archive (TCIA)) by which patient information is extracted from different angles increases [21], it becomes possible to considerably improve the predictive efficiency of molecular subtype, in general, and methylguanine-DNA methyltransferase (MGMT) genotype, in particular, by fusing information originating from the correlation of high-throughput data with radiomics
We investigated the role of MRI features and radiomics-based XGBoost model in predicting MGMT genotype in patients with IDH1 wildtype GBM

Summary

Introduction

Glioblastomas (GBMs), the most aggressive and exceptionally invasive brain tumors, are characterized by their frequent resistance to chemotherapy and always recurrence following surgical treatment [1]. GBM treatment is considered most challenging in clinical oncology. 96% of patients with GBM have IDH1 wildtype mutations, and the treatment success rate for these patients (i.e., concomitant adjuvant temozolomide (TMZ) therapy) can be predicted via the O6-methylguanine-DNA methyltransferase (MGMT) gene promoter. It is well known that GBM with MGMT promotor methylation responds to temozolomide better than the unmethylated counterpart. Because GBM with MGMT promotor methylation responds to temozolomide better than the unmethylated counterpart, MGMT methylation status is considered a critical factor of temozolomide resistance and poor progression-free survival. A noninvasive imaging biomarker for determining the MGMT promoter status in IDH1-wildtype GBM could lead to improved GBM treatment, with accurate treatment guidance

Methods

Results

Discussion

Conclusion