Concordance between RTOG and EORTC prognostic criteria in low-grade gliomas

Interactive Effects of Molecular, Therapeutic, and Patient Factors on Outcome of Diffuse Low-Grade Glioma

PURPOSE

In patients with diffuse low-grade glioma (LGG), the extent of surgical tumor resection (EOR) has a controversial role, in part because a randomized clinical trial with different levels of EOR is not feasible.

METHODS

In a 20-year retrospective cohort of 392 patients with IDH-mutant grade 2 glioma, we analyzed the combined effects of volumetric EOR and molecular and clinical factors on overall survival (OS) and progression-free survival by recursive partitioning analysis. The OS results were validated in two external cohorts (n = 365). Propensity score analysis of the combined cohorts (n = 757) was used to mimic a randomized clinical trial with varying levels of EOR.

RESULTS

Recursive partitioning analysis identified three survival risk groups. Median OS was shortest in two subsets of patients with astrocytoma: those with postoperative tumor volume (TV) > 4.6 mL and those with preoperative TV > 43.1 mL and postoperative TV ≤ 4.6 mL. Intermediate OS was seen in patients with astrocytoma who had chemotherapy with preoperative TV ≤ 43.1 mL and postoperative TV ≤ 4.6 mL in addition to oligodendroglioma patients with either preoperative TV > 43.1 mL and residual TV ≤ 4.6 mL or postoperative residual volume > 4.6 mL. Longest OS was seen in astrocytoma patients with preoperative TV ≤ 43.1 mL and postoperative TV ≤ 4.6 mL who received no chemotherapy and oligodendroglioma patients with preoperative TV ≤ 43.1 mL and postoperative TV ≤ 4.6 mL. EOR ≥ 75% improved survival outcomes, as shown by propensity score analysis.

CONCLUSION

Across both subtypes of LGG, EOR beginning at 75% improves OS while beginning at 80% improves progression-free survival. Nonetheless, maximal resection with preservation of neurological function remains the treatment goal. Our findings have implications for surgical strategies for LGGs, particularly oligodendroglioma.

Machine learning based survival prediction in Glioma using large-scale registry data

Gliomas are the most common central nervous system tumors exhibiting poor clinical outcomes. The ability to estimate prognosis is crucial for both patients and providers in order to select the most appropriate treatment. Machine learning (ML) allows for sophisticated approaches to survival prediction using real world clinical parameters needed to achieve superior predictive accuracy. We employed Cox Proportional hazards (CPH) model, Support Vector Machine (SVM) model, Random Forest (RF) model in a large glioma dataset (3462 patients, diagnosed 2000-2018) to explore the most optimal approach to survival prediction. Features employed were age, sex, surgical resection status, tumor histology and tumor site, administration of radiation therapy (RT) and chemotherapy status. Concordance index (c-index) was employed to assess the accuracy of survival time prediction. All three models performed well with prediction accuracy (CI 0.767, 0.771, 0.57 for CPH, SVM, RF models respectively) with the best performance achieved when incorporating RT and chemotherapy administration status which emerged as key predictive features. Within the subset of glioblastoma patients, similar prediction accuracy was achieved. These findings should prompt stricter clinician oversight over registry data accuracy through quality assurance as we move towards meaningful predictive ability using ML approaches in glioma.

An analysis of prognostic factors in a cohort of low-grade gliomas and degree of consistency between RTOG and EORTC scores

Due to their rarity and heterogeneity and despite the introduction of molecular features in the current WHO classification, clinical criteria such as those from the European Organization for Research and Treatment of Cancer (EORTC) and the Radiation Therapy Oncology Group (RTOG) are still being used to make treatment decisions in low-grade gliomas (LGG). Patients with diffuse low-grade glioma treated at our institution between 2002 and 2018 were analyzed, retrieving and assessing the degree of consistency between the EORTC and RTOG criteria, as well as the isocitrate dehydrogenase 1 and 2 (IDH) gene mutational status. Likewise, multivariate analyses were performed to ascertain the superiority of any of the factors over the others. One hundred and two patients were included. The degree of consistency between the RTOG and EORTC criteria was 71.6% (K = 0.426; p = 0.0001). Notably, 51.7% of those assigned to low risk by the EORTC were classified as high risk according to the RTOG classification. In multivariate analysis, only complete resection, age > 40 years, size and IDH mutation status were independently correlated with OS. When the RTOG and EORTC scores were entered into the model, only the EORTC model was independently associated with mortality. The degree of consistency between the EORT and RTOG criteria is low. Therefore, there is a need to integrate clinical-molecular scores to improve treatment decisions in LGG.

A Risk Score Signature Consisting of Six Immune Genes Predicts Overall Survival in Patients with Lower-Grade Gliomas

Background. Lower-grade gliomas (LGGs) are less aggressive with a long overall survival (OS) time span. Because of individualized genomic features, a prognostic system incorporating molecular signatures can more accurately predict OS. Methods. Differential expression analysis between LGGs and normal tissues was performed using the Gene Expression Omnibus (GEO) datasets (GSE4290 and GSE12657). Immune-related differentially expressed genes (ImmPort-DEGs) were analyzed for functional enrichment. The least absolute shrinkage and selection operator (LASSO) analysis was performed to develop an immune risk score signature (IRSS). We extracted information from the Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA) to establish and validate the model. The relationship of model gene sets with immune infiltration was analyzed based on gene set variation analysis (GSVA) scores. Patients were divided into low- and high-risk groups based on the median score. The time-dependent receiver-operating characteristic (ROC) curve and the Kaplan-Meier curve were used to evaluate the model. Then, a precise prognostic nomogram was established, and its efficacy was verified. Results. A total of 18 related immune genes were identified, building a 6-gene IRSS (BMP2, F2R, FGF13, PCSK1, PRKCB, and PTGER3). DEGs were enriched in T cell and NK cell regulatory pathways. Immune infiltration analysis confirmed that the gene signature correlated with a decrease in innate immune cells. In terms of model evaluation, ROC curves at 1, 3, and 5 years showed moderate predictive ability of IRSS ( $\text{AUC}=0.930$ , 0.797, and 0.728). The Cox regression analysis revealed that IRSS was an independent prognostic factor, and the nomogram model had good predictive ability ( $\text{C}-\text{index}=0.828$ ). Meanwhile, the predictive power of IRSS was also confirmed in the training cohort. The Kaplan-Meier results showed that the prognosis of the high-risk group was significantly worse in all cohorts. Conclusion. IRSS may serve as a novel survival prediction tool in the classification of LGG patients.

A novel tailored immune gene pairs signature for overall survival prediction in lower-grade gliomas

•

The Immune-related gene pairs (IRGPs) pronostic signature for LGG is correlated with immune cells infiltration.

•

WGCNA presented a gene set correlating with immune cells infiltration and genes co-expression relationships were visualized.

•

The nomogram constrcted by three IRGPs and clinical factors is a novel tailored tool for individual-level prediction.

Lower-grade gliomas (LGGs) have a good prognosis with a wide range of overall survival (OS) outcomes. An accurate prognostic system can better predict survival time. An RNA-Sequencing (RNA-seq) prognostic signature showed a better predictive power than clinical predictor models. A signature constructed using gene pairs can transcend changes from biological heterogeneity, technical biases, and different measurement platforms. RNA-seq coupled with corresponding clinical information were extracted from The Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA). Immune-related gene pairs (IRGPs) were used to establish a prognostic signature through univariate and multivariate Cox proportional hazards regression. Weighted gene co-expression network analysis (WGCNA) was used to evaluate module eigengenes correlating with immune cell infiltration and to construct gene co-expression networks. Samples in the training and testing cohorts were dichotomized into high- and low-risk groups. Risk score was identified as an independent predictor, and exhibited a closed relationship with prognosis. WGCNA presented a gene set that was positively correlated with age, WHO grade, isocitrate dehydrogenase (IDH) mutation status, 1p/19 codeletion, risk score, and immune cell infiltrations (CD4 T cells, B cells, dendritic cells, and macrophages). A nomogram comprising of age, WHO grade, 1p/19q codeletion, and three gene pairs (BIRC5|SSTR2, BMP2|TNFRSF12A, and NRG3|TGFB2) was established as a tool for predicting OS. The IPGPs signature, which is associated with immune cell infiltration, is a novel tailored tool for individual-level prediction.