New treatment approaches for glioblastoma: reused drugs, resistance challenges, personalized medicine, and new technologies

This review includes studies on repurposed drugs with anticancer effects on glioblastoma (GBM), focusing on mechanisms such as apoptosis, autophagy, resistance pathways, and TME interactions. Studies employing human GBM cell lines, animal models, or clinical data were considered, particularly those addressing critical pathways such as the blood–brain barrier (BBB), glioblastoma stem cells (GSCs), and drug resistance. Research on advanced diagnostics (e.g., liquid biopsy, next-generation sequencing) and novel therapies, including targeted treatments, combination regimens, and immunotherapies, was also included. Only peer-reviewed, English-language studies published from 2008 to 2024 were considered.

Studies were excluded if they were irrelevant to GBM, drug repurposing, or advanced therapies; did not address molecular mechanisms or clinical applications; lacked experimental evidence or clear conclusions on GBM treatment efficacy; or were purely theoretical without empirical validation. Additionally, studies with significant biases, poor controls, reproducibility issues, or those not focused on repurposed drugs for GBM were excluded.

To minimize bias and ensure methodological rigor, a systematic approach was employed to prioritize experimental and clinical studies, while excluding unsupported or irrelevant research. Bias was minimized by using a diverse set of data sources and avoiding a preference for positive findings. Methodological rigor was ensured through critical evaluation of study controls, statistical analyses, and reproducibility. Funding bias was mitigated by excluding studies with potential conflicts of interest. However, the focus on English-language, peer-reviewed articles introduces potential language and information biases. Variability in study methodologies, particularly reliance on preclinical data for some repurposed drugs, may limit conclusions regarding clinical efficacy. Despite efforts to reduce bias, these limitations should be considered. Future reviews should explore ongoing or unpublished clinical trials to enhance comprehensiveness.

Angiogenesis plays a crucial role in the progression of GBM, primarily driven by the release of pro-angiogenic factors such as vascular endothelial growth factor (VEGF), fibroblast growth factor, and hepatocyte growth factor [18]. However, the newly formed blood vessels are often disorganized, which impairs blood flow and complicates the delivery of therapeutic drugs [19]. Hypoxia, resulting from high cell density and blocked blood vessels, leads to the creation of a necrotic core within the tumor. This oxygen-deprived environment contributes to therapy resistance, particularly because conventional chemotherapy relies on oxygen to generate reactive oxygen species (ROS), which are crucial for cancer cell destruction. In response to hypoxia, GBM cells increase the production of hypoxia-inducible factor-1 (HIF-1), which boosts angiogenesis through VEGF upregulation and enhances tumor invasion [20] (Fig. 4).

The BBB adds another layer of complexity to effective GBM treatment. This barrier regulates the exchange of substances between the bloodstream and the central nervous system (CNS), playing a key role in maintaining CNS homeostasis. The BBB consists of endothelial cells, basement membranes, pericytes, and astrocytes, and its selective permeability is controlled by tight junctions between endothelial cells and high transendothelial electrical resistance (TEER). These mechanisms restrict the passage of water-soluble and polar molecules, while allowing small, lipophilic molecules like oxygen and carbon dioxide to pass freely. Nutrient transporters also regulate the influx of essential compounds such as glucose and amino acids, as well as the efflux of harmful substances, including many drugs [21].

This selective permeability presents major challenges for GBM treatment, as most chemotherapy drugs are large, hydrophilic molecules that struggle to cross the BBB. GBM cells themselves worsen this issue in multiple ways. For instance, they overproduce VEGF, leading to the formation of immature, leaky blood vessels that weaken the tight junctions in the BBB, increasing its permeability and hindering drug delivery. Additionally, GBM cells invade surrounding brain tissue, triggering inflammation that further disrupts the BBB. The release of matrix metalloproteinases by GBM cells degrades the extracellular matrix and vascular basement membrane, further compromising the BBB’s integrity [22].

Moreover, GBM cells alter the function of pericytes, which are vital for BBB stability and maintaining vascular tone, further weakening the barrier and promoting tumor growth and invasion. The BBB is not uniform throughout the tumor—some areas may exhibit a compromised barrier that allows limited drug penetration, while other regions may remain intact, leading to incomplete treatment and eventual tumor recurrence (Fig. 5). Adding to the difficulty, GBM cells often overexpress drug efflux pumps like P-glycoprotein (P-gp), which limits the uptake of therapeutic agents and reduces the effectiveness of treatment. P-gp is overexpressed in the endothelial cells of the BBB, making it even more challenging for drugs to penetrate the brain. In GBM cells, P-gp overexpression plays a significant role in resistance to drugs such as temozolomide (TMZ) [23].

Tumor acidosis plays a vital role in shaping the TME and has a significant impact on the effectiveness of therapies, particularly chemotherapy and immunotherapy. The acidic conditions in the TME contribute to treatment resistance by enhancing markers associated with glioma stem cells (GSCs), promoting tumor growth through the upregulation of HIF-2α-regulated angiogenic factors, and increasing autophagic activity, which supports the maintenance and aggressiveness of GSCs. Moreover, acidosis activates cathepsin L, an enzyme that converts plasminogen into plasmin, which in turn degrades extracellular matrix proteins and activates matrix metalloproteinases, facilitating tumor invasion [24].

From a therapeutic standpoint, tumor acidosis hinders the effectiveness of chemotherapy, especially for weak base drugs like doxorubicin and vincristine. The acidic environment induces ion trapping, which prevents these drugs from being absorbed properly and enhances the activity of P-glycoprotein (P-gp), a drug efflux pump, further diminishing their therapeutic efficacy. Acidosis also impairs immune responses, particularly by inhibiting the function of CD8 + T-cells, reducing cytokine production, receptor expression, and overall immune signaling, all of which hinder anti-tumor immunity. Furthermore, acidosis suppresses the effector functions of T-cells and monocytes, increases the recruitment of myeloid-derived suppressor cells (MDSCs), and reduces the cytotoxic activity of natural killer (NK) and NKT cells, all of which contribute to resistance against immunotherapy [25].

Elevated levels of glutathione in GBM play a critical role in reducing oxidative stress, thereby supporting disease progression and promoting resistance to reactive oxygen species (ROS)-dependent chemotherapeutics. Resistant GBM cells exhibit higher glutathione levels and lower ROS levels compared to TMZ-sensitive cells, further underscoring its role in mediating therapy resistance. By neutralizing ROS, glutathione contributes to a cellular environment that enhances tumor survival and limits the effectiveness of treatments that rely on ROS to induce cancer cell death. Targeting the glutathione pathway presents a promising avenue for future GBM therapies, particularly within the context of drug repurposing. Repurposed drugs that modulate glutathione levels or its associated pathways could potentially reverse or reduce the resistance GBM exhibits to conventional therapies, offering new therapeutic strategies [26].

The early symptoms of GBM are often vague and non-specific, making timely diagnosis and intervention challenging. As a result, the disease is frequently not detected until it reaches more advanced stages. Common early signs include persistent or recurrent headaches, which may intensify with activities such as coughing or changes in posture. Neurological symptoms, such as focal weakness, cognitive decline, and personality changes, are also frequently observed [27]. The onset of new seizures, particularly in individuals over 20 years old, is another concerning indicator of GBM [28]. Aphasia, which refers to difficulties with speech and language, typically occurs when the tumor affects language-processing regions of the brain, such as Broca’s or Wernicke’s areas [29] (Fig. 6). The delayed appearance of these clinical symptoms underscores the difficulty of early detection and highlights the critical need for more effective diagnostic tools and timely treatments to improve patient outcomes and quality of life.

Sanger sequencing, the first commercially available DNA sequencing method [30], utilized 2′-deoxynucleotides (dNTPs) for DNA synthesis and 2′,3′-dideoxynucleotides (ddNTPs) to terminate the synthesis, generating DNA fragments of various lengths. These fragments were analyzed using high-resolution gel electrophoresis. The automation of Sanger sequencing with fluorescently labeled primers or ddNTPs facilitated major milestones, including the completion of the Human Genome Project. Despite its groundbreaking success, Sanger sequencing faced limitations such as high costs, long processing times, low throughput, and inefficiency, which led to the development of next-generation sequencing (NGS) [31].

Building on the principles of Sanger sequencing, NGS uses polymerases, modified nucleotides, and fluorescent detection to sequence DNA more efficiently [32]. Platforms like Illumina, Ion Torrent, and PacBio provide much higher throughput at lower costs. Illumina's "sequencing by synthesis," Ion Torrent's proton-release detection, and PacBio's single-molecule real-time (SMRT) sequencing each contribute to the advancements in NGS technology [33]. For example, NGS platforms can sequence around 5,000 Mb per day at a cost of $0.50/Mb, while Sanger sequencing is limited to approximately 6 Mb per day at $500/Mb [34].

Since its introduction in 2005, NGS has revolutionized genetic research, enabling extensive studies on genetic variations, RNA profiles, methylation patterns, and chromatin immunoprecipitation. It has significantly advanced genomic medicine, facilitating personalized treatments, improving diagnostics, and enhancing our understanding of genetic factors in health. In the context of GBM, NGS plays a crucial role in identifying key molecular markers, such as mutations and gene expression patterns, which may be targeted by existing drugs. This makes NGS particularly valuable for drug repurposing, as it allows for the identification of genetic profiles that match available therapies to the unique molecular characteristics of a patient's GBM. However, widespread adoption of NGS requires significant infrastructure, skilled personnel, and raises ethical concerns regarding data ownership and the secondary use of genetic information [35].

Third-generation sequencing improves on traditional NGS by allowing for single-molecule long reads, which addresses many of the limitations of earlier technologies [36]. Techniques like PacBio’s SMRT sequencing and nanopore sequencing offer real-time sequencing with less need for extensive sample preparation [37]. In particular, nanopore sequencing uses an electric field to pull single-stranded DNA or RNA through tiny pores, converting the resulting electrical signals into sequence data [21]. This approach makes it possible to read long DNA strands, often spanning several kilobases, with ongoing improvements in nanopore technology helping to address challenges like translocation speed and the size of recognition regions [22].

These advances in sequencing are helping move personalized medicine forward by enabling more detailed gene analysis, full exome and genome sequencing, and RNA profiling. Nanopore sequencing, for example, has shown it can sequence an entire human genome using ultra-long reads (up to 882 kb), all while being faster and more affordable, even in areas with fewer resources. This technology is also being tested for cancer diagnostics, allowing for the rapid identification of mutations in genes like IDH1, IDH2, TP53, H3F3A, and TERT promoter, as well as the 1p/19q codeletions. Notably, nanopore sequencing can provide 0.1X genome coverage in just six hours, and it can even detect DNA methylation profiles without needing special DNA conversion methods like bisulfite treatment [22, 38].

For GBM, which is often difficult to diagnose early due to vague symptoms, third-generation sequencing could significantly improve diagnostic accuracy. With its ability to deliver real-time sequencing and quickly identify mutations, nanopore sequencing could help doctors spot key genetic changes, such as those in IDH1 or TP53, which are crucial in GBM. This fast, high-resolution sequencing could help reduce diagnostic delays, allowing for quicker and more personalized treatment decisions that could ultimately improve patient outcomes. However, despite its impressive speed, portability, and affordability, nanopore sequencing still faces challenges, such as the need for large amounts of DNA (~ 700 μg) and the optimization required when dealing with fragmented nucleic acids. Ongoing improvements in both the technology itself and the software used to analyze it are necessary to unlock its full potential for widespread clinical use [38].

The Cancer Genome Atlas (TCGA) has provided valuable insights into the key genetic changes present in GBM, highlighting three major pathways: receptor tyrosine kinase (RTK)/rat sarcoma (RAS)/PI3K (found in 88% of cases), p53 (78%), and retinoblastoma (RB) (87%). Some of the most common mutations include changes in the EGFR (45%), PTEN (36%), cyclin-dependent kinase inhibitors 2A (CDKN2A) and 2B (CDKN2B) (52% each), TP53 (35%), and RB deletions (11%) [39]. While mutations in IDH are rare in primary GBM that show alterations in EGFR and PTEN, these mutations are more frequently observed in lower-grade gliomas and in those that progress to higher-grade tumors [40].

Glioblastomas are molecularly diverse, which makes them particularly challenging to treat. These tumors can be divided into four subtypes: classical, proneural, neural, and mesenchymal [41]. Thanks to advances in single-cell RNA sequencing (scRNA-seq), we now have a much clearer understanding of the tumor cells and the surrounding microenvironment, including important profiles of GSCs, macrophages, microglia, and gene expressions linked to tumor development. Additionally, next-generation sequencing (NGS) panels have improved the accuracy of diagnosing GBM, offering a more precise molecular profile [42].

However, the complexity within GBM, including their genetic diversity, complicates treatment options. Tumors are often made up of different subclones, meaning that a single biopsy might not fully represent the genetic variability of the tumor. Treatments like TMZ can induce mutations that may drive tumor progression further. The genetic complexity of GBM, including mutations in key pathways like EGFR, PTEN, and TP53, plays a crucial role in therapy resistance. These findings point to the potential of repurposing existing drugs to target these specific mutations, offering a promising strategy for overcoming resistance. Transcriptome analyses, which study the RNA profiles of tumors, can refine GBM signatures, improve diagnostic accuracy, predict how tumors will respond to treatment, and help assess patient outcomes [43].

Epigenetic modifications, such as DNA methylation, histone modifications, and chromatin remodeling, play a crucial role in regulating gene expression without changing the DNA sequence itself [44]. Chromatin, which consists of DNA wrapped around histone proteins (H1, H2A, H2B, H3, and H4) [45], forms nucleosomes that are essential for controlling gene activity [46]. When histones are acetylated by enzymes called histone acetyltransferases (HATs), the DNA–histone interactions are loosened, promoting gene expression. Conversely, deacetylation by histone deacetylases (HDACs) silences gene activity [47]. In pediatric gliomas, mutations in the H3.3 histone, such as K27M and G34R/V, are common. In contrast, adult GBMs, as defined by the current WHO classification, are typically IDH-wildtype. IDH1-mutant gliomas are now classified separately as astrocytomas (Grades II-IV). IDH1 mutations lead to the accumulation of 2-hydroxyglutarate (2-HG), an oncometabolite that inhibits histone demethylases, thereby altering H3K27 and H3K36 methylation patterns and affecting gene expression [47]. Methylation of lysine residues on histones can mark chromatin regions as either active (K4, K36, K39) or inactive (K9, K27), thus influencing gene transcription [48]. Chromatin remodeling mechanisms, such as the SWI/SNF core complex and the BRG1 protein, also contribute to GBM progression by affecting glioma stem cell tumorigenicity [49].

DNA methylation, where methyl groups are added to cytosine at CpG sites, is crucial for regulating gene expression. This modification helps silence transposable elements, inactivate X-chromosome genes, and maintain genome stability. While adenine methylation is less studied in mammals, it plays important roles in other organisms. In GBM, global hypomethylation is found in about 80% of cases and promotes the activation of oncogenes, aiding tumor growth [50]. On the other hand, hypermethylation at specific gene loci silences tumor suppressor genes. Methylation of CpG islands typically correlates with reduced gene expression. Studies have identified specific methylation patterns with prognostic value. For instance, the 1p/19q co-deletion (Codel) is associated with better prognosis in oligodendrogliomas, while glioma-CpG island methylator phenotype (G-CIMP) classification further refines prognosis, with G-CIMP-high tumors generally linked to better outcomes and G-CIMP-low tumors associated with worse prognosis. One key epigenetic biomarker is the methylation of the MGMT promoter, which is the only known predictor of response to TMZ. When the MGMT promoter is methylated, the tumor is more sensitive to chemotherapy, significantly improving patient survival. For methylated cases, the median survival increases from 16 to 43 months, compared to 11–36 months for unmethylated cases [51].

These epigenetic changes present valuable opportunities for drug repurposing. Targeting HDACs or using DNA demethylating agents could provide new ways to overcome treatment resistance [52]. By understanding how existing drugs might influence these epigenetic pathways, we could improve the effectiveness of treatments and better tailor them to the specific epigenetic profile of each GBM tumor.

GBM is a highly heterogeneous tumor, meaning it varies greatly in terms of its genetic and phenotypic traits. It has several molecular subtypes, including proneural, classical, and mesenchymal types, each influenced by distinct genetic and phenotypic factors. According to the WHO 2021 classification, GBM is strictly defined as a Grade IV astrocytoma that is IDH-wildtype, reinforcing its aggressive nature, invasive properties, and treatment resistance. Astrocytoma, IDH-mutant (Grades II-IV), previously classified as “secondary GBM,” is now recognized as a separate tumor entity due to distinct molecular and clinical differences [53]. Moreover, GBM cells have the ability to switch between different phenotypes in response to changes in their surrounding environment, adding further complexity to treatment approaches [54] (Fig. 7).

Molecular classification systems, like the one developed by The Cancer Genome Atlas (TCGA), identify four main subtypes of GBM. Classical GBM is marked by the overexpression of the EGFR and mutations in the PTEN gene. Mesenchymal GBM, on the other hand, is characterized by the activation of the NF-κB pathway, increased necrosis, and inflammation, as well as the expression of mesenchymal markers such as CHI3L1 and MET, all of which contribute to a poorer prognosis. Proneural GBM is associated with the expression of genes like PDGFRA and IDH1, while neural GBM expresses proteins related to neural activity, such as NEFL and SLC1A3, although its clinical significance is still not well understood [55]. However, subsequent studies revealed that the neural subtype largely represented contamination from normal neuronal cells rather than a distinct tumor intrinsic classification. As a result, the neural subtype was removed, and the classification was refined into three main subtypes: classical, mesenchymal, and proneural (Fig. 8).

Beyond these molecular subtypes, GBM can also exhibit distinct histological variations. Some glioblastomas may present with histological features that resemble oligodendrogliomas; however, the defining 1p/19q co-deletion is exclusively observed in IDH-mutant, Grade II or III oligodendrogliomas and is not a feature of GBM (IDH-wildtype), as per the current WHO classification. Similarly, certain GBMs may display characteristics reminiscent of pilocytic astrocytoma, further complicating differential diagnosis. A rare histological variant of GBM, known as gliosarcoma, contains a sarcomatous component and is associated with a particularly poor prognosis. Importantly, IDH mutations are no longer considered part of GBM classification under the updated WHO guidelines. Instead, IDH-mutant Grade IV gliomas are now classified separately as astrocytoma, IDH-mutant (Grade II-IV), reflecting their distinct molecular and clinical behavior. This distinction is critical, as IDH-mutant astrocytomas generally have a better prognosis compared to IDH-wildtype GBM, reinforcing the importance of IDH status as a key biomarker for accurate diagnosis and personalized treatment planning [55] (Fig. 8).

By using these molecular and histological classifications, doctors can develop personalized treatment plans, including the possibility of repurposing existing drugs. Tailoring therapies to fit the specific genetic and phenotypic profile of each GBM subtype holds great potential for improving patient outcomes and overcoming the challenge of treatment resistance.

Beyond transcriptomic subtyping, a growing area of research has focused on the TME and its role in GBM progression and therapeutic response. A recent classification proposed by White et al. introduces a TME-based model for IDH-wildtype GBM, categorizing tumors into TME-high, TME-med, and TME-low groups. This classification is particularly relevant for immunotherapy stratification, as TME-high tumors, which exhibit increased immune cell infiltration, may respond differently to immune checkpoint inhibitors compared to TME-low tumors. As GBM treatment strategies evolve, incorporating both transcriptomic and TME-based classifications may aid in precision medicine approaches and personalized treatment strategies.

Combination therapy, which involves using two or more treatments with different mechanisms, has shown great promise in improving outcomes for patients with GBM. A key example is the combination of TMZ with radiotherapy, which is currently the standard treatment for GBM. This combination works well because TMZ, an alkylating agent that damages DNA, complements the effects of radiotherapy, which targets fast-growing tumor cells [65].

Efforts to enhance the effectiveness of TMZ continue, with several promising combinations. One approach pairs TMZ with MGMT inhibitors, which target the DNA repair enzyme MGMT and may help overcome TMZ-resistance. This combination is particularly relevant in patients with unmethylated MGMT promoters, where MGMT is actively expressed and counteracts TMZ-induced DNA damage. In contrast, in methylated MGMT tumors, where MGMT expression is silenced, TMZ remains more effective without the need for MGMT inhibition [66]. Other strategies include combining TMZ with inhibitors that target the PI3K/mTOR pathway, such as XL765, which has shown to increase cancer cell death in animal models of GBM [67]. Similarly, combining TMZ with bromo-domain inhibitors like JQ1 can promote DNA damage and help trigger cancer cell death [68]. Additionally, pairing TMZ with the anti-EGFR antibody nimotuzumab has shown potential for improving treatment effectiveness by targeting the EGFR mutation common in many GBM cases [69].

In addition to enhancing TMZ treatments, researchers are also focusing on targeting survival pathways in GBM cells to make treatments more effective. One example is the use of cyclopamine, a drug that blocks the Shh pathway, in combination with TMZ. This combination helps to prevent tumor recurrence [70]. Another promising approach uses γ-secretase inhibitors (GSIs) alongside radiation therapy, sensitizing GBM cells to radiation and improving the response to treatment [71]. Additionally, pairing TMZ with morphine has shown potential in overcoming drug resistance, specifically by blocking a mechanism (P-gp1) that prevents TMZ from entering tumor cells. This combination helps reduce tumor growth and improve the long-term success of treatment [72].

Beyond traditional treatments, combination therapy is also being explored in immunotherapy. One exciting strategy involves combining anti-PD-1 antibodies with other immunotherapies to enhance immune response in GBM. While early studies in animal models suggested potential benefits, clinical trials evaluating anti-PD-1 monotherapy, including Phase III trials and its upfront use with the Stupp protocol, have not demonstrated significant survival benefits. As a result, current research is focusing on combining anti-PD-1 therapy with other immunotherapeutic approaches, such as CTLA-4, tumor vaccines, or adoptive T-cell therapies, in an effort to improve treatment efficacy in GBM [73]. Another promising combination is using both anti-PD-1 and anti-CTLA-4 antibodies, which together enhance the immune response by reducing the number of immunosuppressive cells (MDSCs) in the TME, helping the immune system to target the tumor more effectively [74] (Fig. 10).

The molecular complexity of GBM underscores the need for reliable biomarkers that can improve diagnosis, prognosis, and the personalization of treatment. Established biomarkers, including IDH1, MGMT, and EGFR, provide crucial insights into the tumor's biology and are essential for guiding therapy decisions.

IDH1 mutations are predominantly found in lower-grade gliomas and IDH-mutant astrocytomas (Grades II-IV), which were previously referred to as secondary glioblastomas. However, under the current WHO classification, GBM is strictly defined as IDH-wildtype, and IDH-mutant high-grade gliomas are now classified separately as astrocytoma, IDH-mutant, Grades II-IV. These mutations are associated with a better prognosis and improved response to therapy, making IDH status a key molecular marker in the WHO classification system for gliomas [75, 76].

MGMT, an enzyme involved in repairing DNA damage, plays a key role in treatment effectiveness. When the MGMT gene promoter is methylated, it improves the tumor's response to chemotherapy with TMZ [77]. EGFR, frequently overexpressed or mutated in GBM, particularly in the EGFRvIII variant, is associated with aggressive tumor behavior and poor survival outcomes, although therapies targeting EGFR have had limited success [56]. Additionally, emerging biomarkers like ATRX and TP53 mutations are explored to refine prognosis and treatment strategies further [78].

Another set of important biomarkers includes mutations in the TERT promoter, which activate telomerase and promote telomere elongation, enabling sustained tumor cell growth. This makes TERT a crucial marker for prognosis and potential therapeutic targets [79]. VEGF, which is overexpressed in GBM, stimulates angiogenesis and helps form dense blood vessels that supply the tumor with nutrients and oxygen, contributing to rapid tumor growth and resistance to therapy. Similarly, the concurrent loss of CDKN2A and CDKN2B, which encode critical cell cycle regulators, contributes to uncontrolled cell division and tumor progression. This deletion leads to the loss of p16 (INK4a), which regulates Rb signaling; p14 (ARF), which modulates the p53 tumor suppressor pathway; and p15 (INK4b), which inhibits CDK4 activity, collectively promoting GBM proliferation and resistance to cell cycle control mechanisms [80]. NF1, another tumor suppressor gene, inhibits the Ras signaling pathway. Mutations in NF1 are linked with the mesenchymal subtype of GBM, which tends to be more invasive and associated with poorer survival [81].

Other markers like matrix metalloproteinases (MMPs), especially MMP-2 and MMP-9, play a critical role in GBM invasion by degrading the extracellular matrix, making them potential targets for limiting tumor progression [82]. The c-MET receptor, which is overexpressed in GBM, enhances cell proliferation and migration, positioning it as a valuable prognostic biomarker [83]. S100 proteins, such as S100B, are elevated in GBM and influence cell growth, offering potential for both diagnostic and therapeutic applications [84].

Liquid biopsy has emerged as a non-invasive method for monitoring GBM using biomarkers like circulating tumor cells (CTCs), extracellular vesicles (EVs), and cell-free nucleic acids. Liquid biopsy enables real-time treatment monitoring, early detection of recurrence, and differentiation between pseudoprogression and actual tumor progression. There are two primary strategies for liquid biopsy: one that detects tumor-specific materials like circulating tumor DNA (ctDNA), and another that analyzes biofluid components altered by GBM, such as immune cells. Both strategies significantly enhance diagnostic accuracy and enable more personalized treatment options [85].

CTCs, released from tumors through apoptosis or necrosis, provide valuable information about tumor size, mutations, and recurrence. Despite challenges posed by the BBB, ctDNA can help detect mutations in critical genes like TERT, MGMT, and EGFR, complementing invasive biopsy results. Cerebrospinal fluid (CSF) ctDNA has shown higher diagnostic accuracy than blood ctDNA, with advanced techniques like methylation-based PCR and next-generation sequencing (NGS) improving sensitivity [86].

EVs, including exosomes and microvesicles, carry nucleic acids, proteins, and metabolites that reflect the tumor's biology. EV markers such as EGFRvIII and PD-L1 show promise as diagnostic and prognostic tools. Elevated plasma EV levels correlate with tumor progression and recurrence, and techniques like Raman spectroscopy and flow cytometry are advancing the detection of EVs for clinical use [87].

Tumor-associated RNAs, including microRNAs (miRNAs) and circular RNAs (circRNAs), are stable and abundant biomarkers in GBM. Oncogenic miRNAs, like miR-10b, are elevated in GBM, while tumor suppressor miRNAs, such as miR-29a, decrease during tumor progression. Multi-RNA signatures, particularly combinations of miRNAs like miR-21 and miR-15b, have demonstrated superior diagnostic accuracy. Advances in RNA sequencing further support the development of reliable liquid biopsy assays for GBM [88].

Finally, cell-based approaches that analyze circulating immune cells and CTCs provide additional insights. Although isolating CTCs remains difficult due to their low frequency, global blood analysis can overcome this limitation. GBM induces systemic immunosuppressive effects via monocytes, which differentiate into suppressive cells such as myeloid-derived suppressor cells (MDSCs) and M2 macrophages. Elevated levels of CD163 + monocytes and tumor-educated platelets, which express VEGFR1/2 and vWF, can serve as biomarkers for GBM progression and therapy resistance [89] (Fig. 11).

In cancer diagnostics, traditional methods such as microscopy often depend on expert interpretation, which can be subjective and prone to variability [95]. AI enhances diagnostic accuracy and efficiency by automating image analysis, such as using augmented reality microscopes integrated with real-time AI [96]. Machine learning (ML) utilizes both supervised (labeled) and unsupervised (unlabeled) algorithms to identify diagnostic and therapeutic patterns. Additionally, ML is increasingly used in drug discovery, accelerating the identification of potential therapeutics and reducing costs by enabling predictive modeling and drug repurposing strategies. ML algorithms can analyze vast datasets from platforms like DrugBank, which provides comprehensive molecular drug data, to identify novel drug–target interactions, predict drug efficacy, and optimize lead compounds for further investigation. Predictive models, often referred to as “drug scoring,” allow for preemptive (a priori) or retrospective (a posteriori) assessments of treatment efficacy. Despite its potential, challenges remain, such as data privacy concerns, ensuring equitable access to AI technologies and ensuring algorithm reliability across diverse populations [97].

Deep learning (DL), a more advanced form of ML, processes raw, unstructured data to reduce subjectivity in analysis. By modeling complex relationships, such as gene expression patterns and GBM survival, DL identifies biologically significant features that may be missed by traditional methods. DL techniques have successfully classified GBM into subtypes linked to survival outcomes, identified key progression-related genes, and uncovered hidden cellular pathways that influence disease progression. This advancement takes GBM diagnosis from traditional histological analysis to molecular profiling, leveraging modern, personalized algorithms that continuously self-learn and improve. This progression, facilitated by next-generation sequencing, holds promise for improving patient outcomes and advancing precision care [98].

Resistance to GBM therapies, particularly TMZ, remains a major obstacle. MGMT overexpression counteracts TMZ efficacy by repairing DNA damage, suggesting that MGMT inhibitors could help restore drug sensitivity. P-glycoprotein (P-gp), a membrane transporter, actively expels chemotherapeutic agents, reducing drug effectiveness; thus, targeting P-gp inhibitors could improve chemotherapy uptake. Additionally, GBM stem cells (GSCs) contribute to tumor recurrence and therapy resistance. Investigating key molecular pathways, such as Wnt/β-catenin and Notch signaling, may provide novel strategies to target GSCs and enhance GBM treatment.

The GBM TME is highly immunosuppressive, facilitating immune evasion and therapy resistance. PD-1/PD-L1 and immunosuppressive cytokines (e.g., TGF-β) play critical roles in this process, making immune checkpoint inhibitors a promising avenue to restore immune surveillance. Additionally, hypoxia and exosome secretion contribute to tumor progression and resistance by promoting intercellular communication and metastasis. Strategies such as hypoxia-targeted therapies and exosome inhibition could enhance current immunotherapies.

Recent innovations have improved GBM diagnosis, treatment monitoring, and personalized care. Circulating tumor DNA (ctDNA) liquid biopsies enable real-time tumor monitoring, mutation detection, and treatment response assessment. Extracellular vesicles (EVs), which influence GBM progression, are also emerging as potential biomarkers. Furthermore, single-cell RNA sequencing (scRNA-seq) helps characterize tumor heterogeneity, facilitating targeted therapy development. Artificial intelligence (AI)-driven multi-omics analysis offers new opportunities for optimizing diagnostics, predicting therapy responses, and identifying novel drug targets.

Immunotherapy remains a promising avenue for GBM treatment. Chimeric antigen receptor (CAR) T-cell therapy, which engineers T-cells to target tumor-specific antigens (e.g., EGFRvIII), is under clinical investigation. However, overcoming GBM’s immunosuppressive TME remains a challenge. Oncolytic viruses, genetically engineered to selectively infect and kill tumor cells, are another emerging approach, particularly in combination with immune checkpoint inhibitors. While immune-based therapies have transformed treatment in other cancers, their application in GBM requires further refinement to enhance efficacy.

Epigenetic alterations play a crucial role in GBM progression and therapy resistance. Histone deacetylase inhibitors (HDACis) have shown potential in resensitizing GBM cells to chemotherapy and promoting apoptosis. Combining epigenetic therapies with TMZ or radiation could enhance treatment effectiveness. Additionally, combination therapies integrating chemotherapy, targeted agents, and immunotherapies may address resistance mechanisms and improve survival outcomes.

The BBB remains a major challenge in GBM treatment, limiting drug penetration into the tumor. Nanocarriers, including liposomes and nanoparticles, offer promising strategies for targeted drug delivery, bypassing the BBB and improving therapeutic specificity. Temporary BBB disruption techniques allow drug passage but require careful modulation to avoid damage to healthy brain tissue. Advancements in nanoparticle engineering to target GBM-specific biomarkers could enhance drug delivery precision while minimizing systemic toxicity.

Personalized medicine is transforming GBM treatment by tailoring therapies to individual tumor characteristics. Biomarkers such as MGMT methylation and EGFR variants guide treatment selection, optimizing patient outcomes. The integration of multi-omics data (genomic, proteomic, and transcriptomic) further enhances precision therapy design. Clinical trials exploring drug repurposing, combination therapies, and adaptive trial designs are crucial for translating promising preclinical findings into clinical applications.

As research and technology continue to evolve, a multidisciplinary approach combining molecular insights, technological innovations, and novel therapies will be essential for advancing GBM treatment and improving patient survival.

The authors declare no competing interests.

Not applicable.

Not applicable.

No datasets were generated or analyzed during the current study.