BIOS-04. THE ROLE OF MACHINE LEARNING, PREDICTIVE MODELING, AND DEEP LEARNING IN ASSESSING METABOLIC BIOMARKERS TO IMPROVE PROGNOSTICATION IN GLIOBLASTOMA MULTIFORME

Abstract

Abstract Glioblastoma multiforme (GBM) is the most common primary malignant brain tumor in the United States, accounting for approximately 56.6% of all gliomas and 47.7% of all primary malignant CNS tumors. The prognosis of GBM is notably grim, with a 1-year relative survival rate of 41.4% and a 5-year survival rate of 5.8% following diagnosis. Recent efforts to identify potential therapeutic targets have utilized tumor omics data integrated with clinical information that leverages machine learning (ML) algorithms. However, there remains a paucity of studies assessing the value of these ML models as prognostic tools in GBM. A systematic search adhering to PRISMA guidelines was conducted to identify all studies describing the use of a ML algorithm involving GBM metabolic biomarkers and each algorithm's accuracy. Ten studies were included for final analysis. They were diagnostic (n = 3, 30%), prognostic (n = 6, 60%), or both (n = 1, 10%), respectively. Most studies analyzed data from multiple databases, while 50% (n = 5) included additional original samples. At least 2,536 data samples were run through a ML algorithm. 27 ML algorithms were recorded with a mean 2.8 algorithms per study. Algorithms were supervised (n = 22, 79%) or unsupervised (n = 6, 21%), and continuous (n = 21, 75%) or categorical (n = 7, 25%). The mean reported accuracy and AUC of ROC was 95.63% and 0.779, respectively. 106 metabolic markers were identified, but only EMP3 was reported in multiple studies. Many studies have identified potential biomarkers for GBM diagnosis and prognostication. These algorithms show promise; although, a consensus on even a handful of biomarkers has not been made. An integration of ML algorithms for biomarker detection combined with radiomics-based tumor imaging will be necessary to ascertain the greatest level of accuracy and precision.