Background: Glioblastoma multiforme (GBM) is the most common and aggressive primary brain tumour, associated with poor survival outcomes and significant clinical challenges. Conventional diagnostic methods, including MRI, CT, and histopathological analysis of tissue biopsies, are limited by their inability to reliably distinguish treatment effects from true tumour progression, often resulting in misdiagnosis and delayed intervention. Repeated tissue biopsies are also invasive and unsuitable for longitudinal monitoring. Liquid biopsy, a minimally invasive approach analysing tumour-derived material in biofluids such as blood and cerebrospinal fluid (CSF), offers a promising alternative. This review aims to evaluate current evidence on circulating biomarkers including circulating tumour cells (CTCs), circulating tumour DNA (ctDNA), microRNAs (miRNAs), extracellular vesicles (EVs), and proteins in GBM diagnosis and monitoring, and to assess the potential role of artificial intelligence (AI) in enhancing their clinical application.Methods: A narrative synthesis of the literature was undertaken, focusing on studies that have investigated blood- and CSF-derived biomarkers in GBM patients. Key aspects evaluated included biomarker biology, detection techniques, diagnostic and prognostic value, current technical challenges, and progress towards clinical translation. Studies exploring AI and machine learning (ML) approaches for biomarker integration and analysis were also reviewed.Results: Liquid biopsy enables repeated and minimally invasive sampling of tumour-derived material, reflecting the genetic, epigenetic, proteomic, and metabolomic landscape of GBM. Although promising, its translation into routine clinical practice is hindered by the low abundance of circulating biomarkers and lack of standardised collection and analysis protocols. Evidence suggests that combining multiple biomarkers improves sensitivity and specificity compared with single-marker approaches. Emerging AI and ML tools show significant potential for improving biomarker discovery, integrating multi-omic datasets, and enhancing diagnostic and prognostic accuracy.Conclusions:Liquid biopsy represents a transformative tool for GBM management, with the capacity to overcome limitations of conventional diagnostics and provide real-time insights into tumour biology. By integrating multiple circulating biomarkers and leveraging AI-driven approaches, liquid biopsy could enhance diagnostic precision, enable dynamic disease monitoring, and improve clinical decision-making. However, large-scale validation and standardisation are required before routine clinical adoption can be achieved.
背景：多形性胶质母细胞瘤（GBM）是最常见且最具侵袭性的原发性脑肿瘤，其生存预后差，临床治疗面临重大挑战。传统的诊断方法，包括磁共振成像（MRI）、计算机断层扫描（CT）以及组织活检的病理学分析，均存在局限性，无法可靠地区分治疗效果与真正的肿瘤进展，常导致误诊和干预延迟。重复的组织活检具有侵入性，也不适用于纵向监测。液体活检作为一种微创方法，通过分析血液、脑脊液等生物体液中的肿瘤源性物质，提供了一种前景广阔的替代方案。本综述旨在评估当前关于循环肿瘤细胞（CTCs）、循环肿瘤DNA（ctDNA）、微小RNA（miRNAs）、细胞外囊泡（EVs）和蛋白质等循环生物标志物在GBM诊断与监测中的证据，并评估人工智能（AI）在增强其临床应用方面的潜在作用。 方法：本研究对文献进行了叙述性综合，重点关注了在GBM患者中研究血液和脑脊液来源生物标志物的相关研究。评估的关键方面包括生物标志物的生物学特性、检测技术、诊断与预后价值、当前面临的技术挑战以及向临床转化的进展。同时，也回顾了探索利用人工智能和机器学习方法进行生物标志物整合与分析的研究。 结果：液体活检能够对肿瘤源性物质进行重复、微创的采样，反映GBM的遗传、表观遗传、蛋白质组和代谢组学特征。尽管前景广阔，但其向常规临床实践的转化受到循环生物标志物丰度低以及缺乏标准化的采集与分析方案的阻碍。证据表明，与单一标志物方法相比，联合多种生物标志物可提高检测的敏感性和特异性。新兴的人工智能和机器学习工具在改进生物标志物发现、整合多组学数据集以及提高诊断和预后准确性方面显示出巨大潜力。 结论：液体活检代表了GBM管理的一种变革性工具，有能力克服传统诊断方法的局限性，并提供对肿瘤生物学的实时洞察。通过整合多种循环生物标志物并利用人工智能驱动的方法，液体活检有望提高诊断精度，实现动态疾病监测，并改善临床决策。然而，在实现常规临床应用之前，仍需进行大规模验证和标准化工作。
The Quest for Non-Invasive Diagnosis: A Review of Liquid Biopsy in Glioblastoma
肿瘤(癌症)患者之家