Tixier Florent, Um Hyemin, Bermudez Dalton, Iyer Aditi, Apte Aditya, Graham Maya S, Nevel Kathryn S, Deasy Joseph O, Young Robert J, Veeraraghavan Harini
Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, New York, USA.
Department of Neurology, Memorial Sloan Kettering Cancer Center, New York, New York, USA.
Oncotarget. 2019 Jan 18;10(6):660-672. doi: 10.18632/oncotarget.26578.
Glioblastoma (GBM) is the most common malignant central nervous system tumor, and promoter hypermethylation in this tumor has been shown to be associated with better prognosis. We evaluated the capacity of radiomics features to add complementary information to status, to improve the ability to predict prognosis.
159 patients with untreated GBM were included in this study and divided into training and independent test sets. 286 radiomics features were extracted from the magnetic resonance images acquired prior to any treatments. A least absolute shrinkage selection operator (LASSO) selection followed by Kaplan-Meier analysis was used to determine the prognostic value of radiomics features to predict overall survival (OS). The combination of status with radiomics was also investigated and all results were validated on the independent test set.
LASSO analysis identified 8 out of the 286 radiomic features to be relevant which were then used for determining association to OS. One feature (edge descriptor) remained significant on the external validation cohort after multiple testing (p=0.04) and the combination with identified a group of patients with the best prognosis with a survival probability of 0.61 after 43 months (p=0.0005).
Our results suggest that combining radiomics with is more accurate in stratifying patients into groups of different survival risks when compared to with using these predictors in isolation. We identified two subgroups within patients who have methylated : one with a similar survival to unmethylated patients and the other with a significantly longer OS.
胶质母细胞瘤（GBM）是最常见的恶性中枢神经系统肿瘤，该肿瘤中的启动子高甲基化已被证明与较好的预后相关。我们评估了放射组学特征为[具体内容缺失]状态添加补充信息的能力，以提高预测预后的能力。
本研究纳入了159例未经治疗的GBM患者，并将其分为训练集和独立测试集。从任何治疗前获取的磁共振图像中提取了286个放射组学特征。采用最小绝对收缩选择算子（LASSO）选择，随后进行Kaplan-Meier分析，以确定放射组学特征对预测总生存期（OS）的预后价值。还研究了[具体内容缺失]状态与放射组学的组合，并在独立测试集上验证了所有结果。
LASSO分析确定286个放射组学特征中有8个相关，然后用于确定与OS的关联。经过多次检验后，一个特征（边缘描述符）在外部验证队列中仍然显著（p = 0.04），并且与[具体内容缺失]的组合确定了一组预后最佳的患者，43个月后的生存概率为0.61（p = 0.0005）。
我们的结果表明，与单独使用这些预测指标相比，将放射组学与[具体内容缺失]相结合在将患者分层为不同生存风险组时更准确。我们在甲基化的患者中确定了两个亚组：一个与未甲基化的患者生存期相似，另一个OS明显更长。
