What makes this perspective particularly valuable is the way it organizes a sprawling pipeline into four mechanism-based therapeutic domains: molecular-targeted therapies, immunotherapies and vaccines, oncolytic viruses and gene therapies, and agents directed at tumor metabolism and the microenvironment. This taxonomy is more than a convenient filing system. It reflects a deliberate shift away from the historical practice of treating glioblastoma as a single, uniform disease, and toward an approach in which treatments are matched to the molecular fingerprints of individual tumors. Across all four domains, the authors emphasize that patient selection is increasingly guided by biomarkers such as MGMT promoter methylation, IDH mutation status, EGFR amplification and BRAF V600E alterations. These molecular characteristics define biologically distinct patient groups and can determine who is enrolled in trials testing particular therapeutic strategies, a movement that mirrors the biomarker-driven revolution already realized in lung cancer, breast cancer and melanoma.

The biological obstacles that have frustrated glioblastoma treatment for decades remain formidable, and the review does not shy away from them. Glioblastoma tumors harbor genetically and functionally diverse populations of cancer cells, meaning that a drug that eradicates one clone may leave others to repopulate the tumor. The tumor microenvironment is profoundly immunosuppressive, populated by suppressive macrophages, myeloid-derived suppressor cells and regulatory T cells that blunt antitumor immunity. The blood-brain barrier, the brain’s highly selective vascular gatekeeper, restricts the ability of many systemically administered drugs to reach therapeutic concentrations within the tumor. Hypoxia and other features of the tumor environment further promote treatment resistance, allowing glioblastoma cells to adapt their metabolism in response to therapeutic pressure. Any honest account of the emerging pipeline must weigh each new agent against these barriers, and the Oncoscience perspective does exactly that.

Within the molecular-targeted domain, several programs stand out for their biological rationale. The review discusses approaches targeting B7-H3, CDK4/6, the RAS/MAPK signaling pathway and BRAF alterations, among other vulnerabilities. One noteworthy example is abemaciclib, a CDK4/6 inhibitor with relatively favorable brain penetration, which is being investigated in combination with the ERK inhibitor LY3214996. The logic of such a combination is elegant: blocking CDK4/6 restrains cell-cycle progression, while inhibiting ERK disrupts downstream signaling from the RAS pathway, making it harder for tumor cells to compensate through parallel routes. Regorafenib, a multi-kinase inhibitor that suppresses both angiogenic and proliferative signaling, offers a cautionary but instructive case. In the Phase 2 REGOMA trial for recurrent glioblastoma, regorafenib improved median overall survival compared with lomustine, yet it was also associated with more severe adverse events, a reminder that in a disease where the therapeutic window is narrow, efficacy and tolerability must be evaluated together.

Immunotherapy represents perhaps the most rapidly evolving arm of the pipeline. Dendritic cell vaccines, designed to expose the immune system to tumor-associated antigens and generate cytotoxic T-cell responses, continue to be refined, with DCVax-L among the most prominent examples. The review also highlights plasmid DNA vaccines INO-5401 and INO-9012 in combination with the PD-1 inhibitor cemiplimab, a strategy intended to prime immune responses against multiple tumor antigens while simultaneously releasing the brakes that glioblastoma places on T-cell activity. The overarching goal is to convert what is often called a cold tumor, immunologically silent and heavily suppressed, into one that can be recognized and destroyed by the patient’s own immune system. That conversion remains difficult, but the diversity of vaccine platforms now in testing signals a maturing understanding of how to combine antigen delivery with immune checkpoint release.

Chimeric antigen receptor T-cell therapy, which has produced dramatic results in hematologic malignancies, is being adapted to confront the molecular heterogeneity that makes solid brain tumors so elusive. Because glioblastoma cells can simply stop expressing any single target antigen, investigators are developing bivalent and multivalent CAR constructs capable of recognizing more than one tumor-associated antigen at once. The perspective describes a bivalent CAR T-cell strategy targeting both EGFR and IL-13Rα2 in recurrent glioblastoma, as well as engineered tumor-infiltrating lymphocytes designed to secrete anti-PD-1 antibodies directly within tumors, effectively building the checkpoint inhibitor into the cellular therapy itself. The challenges, however, are sobering: antigen loss, T-cell exhaustion, neurotoxicity, questions about the optimal delivery route, and the hostile immunosuppressive microenvironment all remain obstacles between laboratory promise and durable clinical benefit.

Oncolytic and gene-therapy platforms are pushing the field toward treatments that exploit the very biology of tumor cells. Retroviral replicating vectors such as DB107-RRV and DB107-FC are engineered to spread selectively through dividing tumor cells and convert an otherwise innocuous prodrug into a cytotoxic compound precisely where it is needed, within the tumor itself. Other experimental gene-therapy approaches aim to couple this tumor-selective activity with immune stimulation, pairing direct cell killing with the activation of antitumor immunity. These strategies remain strictly investigational, and the review is candid about their challenges, which include achieving adequate delivery across the blood-brain barrier, evading antiviral immune responses, and assembling the specialized infrastructure required to manufacture and administer such sophisticated treatments.

A fourth and increasingly active frontier targets the altered metabolism of glioblastoma cells. The review describes BPM31510, an agent that targets mitochondrial bioenergetics and redox homeostasis, and ASC40, an inhibitor of fatty acid synthase, both of which are being evaluated in clinical trials. The rationale is compelling: glioblastoma cells display remarkable metabolic flexibility, rewiring their energy production and lipid synthesis to survive under the hypoxic and nutrient-limited conditions that characterize the tumor core. By disrupting these metabolic adaptations, investigators hope to starve tumor cells of the biochemical options they use to endure treatment pressure. Because metabolic vulnerabilities can differ between tumors and even between regions of the same tumor, this domain may ultimately depend heavily on the same biomarker-guided selection strategies that are reshaping targeted therapy.

Despite the expanding arsenal, the authors are unambiguous that none of these emerging approaches has yet displaced the temozolomide-based standard of care. The evolving strategy is instead centered on biomarker-directed combinations, in which targeted, immune-based or metabolic treatments are layered onto existing therapy for molecularly defined groups of patients. As the authors put it, the combination of precision molecular targeting with new immune activation methods, CNS-based drug delivery and metabolic interference establishes a foundation for developing combination therapies that should use proven predictive biomarkers for maximum effectiveness. Major barriers persist, from intratumoral heterogeneity that allows resistant clones to escape single-target treatments, to the practical problem that differences among clinical trials in eligibility criteria, endpoints, imaging assessments and corticosteroid use make it difficult to compare results across studies. The authors argue that future progress may hinge on adaptive clinical trial platforms capable of evaluating multiple therapies and molecular subgroups simultaneously, rational combinations that attack several facets of glioblastoma biology at once, personalized neoantigen vaccines, improved methods for delivering drugs across the blood-brain barrier, and emerging liquid-biopsy and imaging biomarkers that could allow treatment response to be monitored in real time.

The broader message of the Oncoscience perspective is one of cautious, methodical optimism. The glioblastoma treatment landscape is unmistakably moving away from broadly applied cytotoxic approaches and toward mechanism-based, biomarker-guided and combination strategies, with molecular targeting, immune activation, gene and oncolytic therapies, metabolic intervention and improved CNS drug delivery all being pursued in parallel. The results of ongoing Phase II and Phase III studies will be decisive in determining which of these approaches ultimately advances toward regulatory consideration and clinical adoption. For a disease that has seen repeated waves of promising science fail at the bedside, the current pipeline’s emphasis on biological sophistication, rigorous patient selection and combination logic may prove to be exactly the kind of strategic reset that glioblastoma research has been waiting for, provided the field can translate mechanistic breadth into measurable survival gains for patients.

Subject of Research: Emerging pharmacological agents and biomarker-guided combination strategies in glioblastoma clinical trials

Article Title: Emerging glioblastoma therapies move toward biomarker-guided and combination strategies

Article References: Emerging glioblastoma therapies move toward biomarker-guided and combination strategies. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: glioblastoma, clinical trials, molecular targeted therapy, immunotherapy, CAR T-cell therapy, dendritic cell vaccines, oncolytic virus, gene therapy, tumor metabolism, blood-brain barrier, biomarkers, combination therapy

The logic of therapeutic vaccination differs fundamentally from preventive immunization. Rather than blocking infection, these vaccines are designed to treat existing cancers by activating tumor-specific immunity and strengthening immune surveillance. Yet the review is candid about the field’s central frustration: even though endogenous T-cell responses against tumors exist in most patients, they typically fail to control disease, and immune checkpoint blockade (ICB) therapy—drugs that release the molecular brakes on T cells—often cannot rescue those responses on its own. Vaccines are meant to supply what the tumor microenvironment suppresses: a large, specific, and activated pool of anti-tumor T cells. Combining vaccines with checkpoint inhibitors has produced encouraging outcomes, including improved survival and tumor reduction across multiple cancer types, but low immunogenicity and tumor heterogeneity remain stubborn obstacles.

Nucleic acid vaccines are the fastest-moving segment of the field. DNA vaccines deliver bacterial plasmids encoding tumor-associated antigens (TAAs) or neoantigens into the cell nucleus, where they are transcribed into mRNA and translated into target proteins. They are stable, cheap to manufacture, easy to store, and can encode multiple antigens to hedge against tumor heterogeneity. The trade-offs are significant: DNA must cross the nuclear envelope, an extra step that lowers antigen production efficiency, and theoretical genomic integration remains a safety consideration unique to DNA. Because their intrinsic immunogenicity is low, DNA vaccines generally require adjuvants, electroporation, or nanoparticle delivery, and few clinical successes have been reported. One notable exception is GNOS-PV02, a DNA plasmid encoding up to 40 personalized neoantigens, which is being tested with an IL-12-encoding plasmid and the PD-1 antibody pembrolizumab in a phase 1/2 trial in hepatocellular carcinoma with promising early results.

mRNA vaccines have captured most of the attention, and the review argues the hype is at least partly earned. mRNA only needs to reach the cytoplasm to begin producing antigen, avoiding the nuclear barrier, and it cannot integrate into the genome. Advances in lipid nanoparticle (LNP) delivery have dramatically improved mRNA stability and cellular uptake, while self-amplifying mRNA (samRNA) promises prolonged antigen expression at lower doses, though cancer applications remain at an early preclinical and clinical stage. A double-edged sword is innate immune sensing: mRNA triggers pattern recognition receptors such as RIG-I and Toll-like receptors, which can act as a built-in adjuvant but can also induce type-I interferon responses that inhibit translation of the vaccine itself. The flagship candidate is mRNA-4157, which encodes up to 35 neoantigen epitopes and is delivered in LNPs. Combined with pembrolizumab, it improved recurrence-free survival versus pembrolizumab alone in a phase 2 trial of patients with completely resected, high-risk cutaneous melanoma, and phase 3 trials are now underway in melanoma, non-small cell lung cancer, and cutaneous squamous cell carcinoma.

Peptide vaccines take a more direct approach, administering synthetic polypeptides that mimic TAAs or neoantigens, which are captured by antigen-presenting cells and used to prime CD8-positive and CD4-positive T cells. Their Achilles’ heel is size: short peptides can load directly onto MHC molecules on any cell, including non-professional antigen-presenting cells that induce tolerance rather than immunity. Synthetic long peptides containing both MHC class I and class II epitopes avoid this trap and reliably induce both T-cell compartments. To overcome weak immunogenicity, researchers pair peptides with Toll-like receptor agonists, checkpoint inhibitors, or nanoparticle carriers such as virus-like particles and liposomes. Clinical results are accumulating: EVX-01, a personalized vaccine of up to 15 neoantigen epitopes with the CAF09b adjuvant, is in phase 1/2 testing alongside checkpoint inhibitors, and SurVaxM, targeting survivin in newly diagnosed glioblastoma, significantly prolonged patient survival in a phase 2 trial combined with pembrolizumab. A new phase 1 trial is evaluating a nine-peptide vaccine with the TLR1/2 ligand XS15 in acute myeloid leukemia patients in first remission.

Cell-based vaccines exploit whole cells rather than defined antigens. Dendritic cell (DC) vaccines harvest a patient’s own antigen-presenting cells, load them ex vivo with tumor lysates, peptides, or antigen-encoding RNA, and reinfuse them to activate cytotoxic T cells. Whole tumor cell vaccines, using irradiated or genetically modified tumor cells, present a broad antigen repertoire that hedges against heterogeneity, and are often supercharged with adjuvants such as GM-CSF. The category’s proof of principle is Provenge (sipuleucel-T), the FDA-approved autologous DC vaccine for prostate cancer. GVAX, an irradiated GM-CSF-secreting tumor cell vaccine, improved disease-free and overall survival in pancreatic cancer patients when combined with the antibodies urelumab and nivolumab in a phase 2 trial. Ongoing phase 2 studies are testing the neoantigen-loaded DC vaccine NeoDC-Vac after chemoradiotherapy in locally advanced esophageal squamous cell carcinoma, and an autologous tumor lysate-loaded cDC1 vaccine in ovarian cancer. The review notes that high production costs, labor-intensive manufacturing, and variable patient responses still constrain this platform.

Virus-based strategies come in two flavors. Oncolytic viruses preferentially infect and replicate inside tumor cells, exploiting abnormal surface receptor expression and the weakened antiviral signaling—blunted interferon and PKR pathways—characteristic of cancer. As infected cells lyse, they release reactive oxygen species, cytokines, and damage-associated molecular patterns that ignite both innate and adaptive immunity, and epitope spreading helps the immune system recognize tumor antigens beyond those the virus carries. The landmark agent is T-VEC, the first FDA-approved oncolytic therapy: an attenuated herpes simplex virus deleted for ICP34.5 to enhance tumor-selective replication and reduce neurovirulence, and engineered to carry the GM-CSF gene to recruit dendritic cells and macrophages into the tumor. Viral vector vaccines, by contrast, use engineered adenoviruses, lentiviruses, or vaccinia viruses to deliver antigen genes into host cells, mimicking natural infection so effectively that adjuvants are often unnecessary. Their vulnerability is pre-existing vector immunity, which can neutralize the vaccine—a problem managed by heterologous prime-boost strategies that switch vectors between doses.

Two technological currents are reshaping the entire field. The first is artificial intelligence: machine learning tools such as NetMHC and DeepHLA mine genomic and transcriptomic data to predict which tumor mutations will yield immunogenic neoantigens that bind a patient’s HLA molecules, bypassing central immune tolerance and reducing off-target risk. GNOS-PV02 already uses such algorithms to select neoantigens by predicted binding affinity. But the review cautions that limited training datasets and ethnic bias in HLA haplotype data can undermine the generalizability of predictions. The second is delivery engineering: LNP surface charge can be tuned to redirect biodistribution to different organs, cholesterol analogues such as C-24 alkyl phytosterols enhance in vivo mRNA delivery, hydrogels act as depots for sustained antigen release—with one study finding that D- versus L-glutamate chirality in polypeptide hydrogels profoundly altered immune cell infiltration and T-cell exhaustion—and lentiviral particles pseudotyped with Sindbis virus glycoprotein can target mRNA directly to dendritic cells, boosting both humoral and T-cell responses.

The review’s most sobering conclusion is that no vaccine platform, however elegant, is likely to deliver durable clinical benefit alone. Combination is the operative word: pairing vaccines with anti-PD-1 and anti-CTLA-4 antibodies, cytokines such as IL-2 and IL-15 superagonist ALT-803, adoptive cell transfers like tumor-infiltrating lymphocyte therapy, or even radiotherapy and chemotherapy that can convert immunologically cold tumors into inflamed ones by breaking down physical barriers of tumor-associated fibroblasts and poor vascularization. Underlying biology still constrains outcomes: HLA diversity between patients complicates neoantigen prediction and blocks some patients from accessing promising TAA epitopes, while individual T-cell receptor repertoires and pMHC densities produce highly variable T-cell avidity—and recent work suggests low-avidity T cells actually drive endogenous tumor immunity. Toxicity, patient variability, and treatment resistance complicate combination regimens. Still, with 15,000 accesses and growing citations since its May 2025 publication, the review captures a field at an inflection point: after decades of disappointment, therapeutic cancer vaccines—personalized, AI-designed, and rationally combined—are finally delivering on their immunological promise.

Subject of Research: Therapeutic cancer vaccine platforms and their combination with immunotherapy

Article Title: Therapeutic vaccine strategies in cancer: advances, challenges, and future directions

Article References: Wang, C., & Hao, P. (2025). Therapeutic vaccine strategies in cancer: advances, challenges, and future directions. Clinical Cancer Bulletin, 4(1), Article 10. https://doi.org/10.1007/s44272-025-00039-x

Image Credits: AI Generated

DOI: 10.1007/s44272-025-00039-x

Keywords: therapeutic cancer vaccines, cancer immunotherapy, mRNA vaccines, DNA vaccines, neoantigens, peptide vaccines, dendritic cell vaccines, oncolytic viruses, viral vectors, immune checkpoint inhibitors, lipid nanoparticles, artificial intelligence

The study, conducted by researchers at The Affiliated Hospital of Qingdao University and published in the Journal of Ovarian Research, systematically mined the ClinicalTrials.gov registry for ovarian cancer vaccine trials registered between 1996 and June 2026. The team extracted and analyzed data on study phase, recruitment status, enrollment size, geographic distribution, funding source, vaccine platform, combination strategy, clinical context, and endpoint selection. Their findings paint a detailed portrait of a field that has sustained momentum since the mid-1990s but has yet to translate that activity into approved therapies.

The numbers tell a striking story. The 136 unique studies collectively involved 4,180 participants. Nearly half of all trials, 48.53 percent, were Phase I studies, while 23.53 percent were Phase II and 21.32 percent were combined Phase I/II designs. Only two Phase III trials appeared in the entire registry. Completion rates were moderate, with 55.88 percent of studies finished, but transparency lagged far behind: just 22.79 percent of trials had publicly posted results. Geographically, the United States dominated the landscape, participating in 105 of the registered studies, reflecting the concentration of immunotherapy infrastructure and funding in American academic and industry centers.

On the technical side, the analysis identified five major vaccine platforms competing for dominance. Peptide and antigen vaccines, which deliver synthetic fragments of tumor-associated proteins to prime immune responses, formed the largest category. Dendritic cell vaccines, in which a patient’s immune cells are harvested, loaded with tumor antigens in the laboratory, and reinfused to act as powerful antigen-presenting scouts, represented another major approach. Whole tumor cell or lysate vaccines offer the advantage of presenting the full antigenic repertoire of a patient’s tumor, while DNA and RNA vaccines instruct the patient’s own cells to produce tumor antigens in situ. Viral and vector vaccines use engineered viruses to deliver tumor antigens and provoke strong cellular immunity. Each platform carries distinct advantages in manufacturing complexity, antigen breadth, and the type of immune response it preferentially elicits.

The mechanistic logic underlying all these platforms is the same: break immune tolerance to tumor antigens and generate durable populations of cytotoxic T lymphocytes capable of recognizing and killing ovarian cancer cells. Common antigenic targets across trials have included Wilms tumor 1, p53, and antigens associated with BRCA-mutated tumors, alongside the widely monitored biomarker CA-125. Adjuvants such as granulocyte-macrophage colony-stimulating factor have frequently been deployed to supercharge dendritic cell activation, and human leukocyte antigen typing has shaped patient eligibility in many peptide-based studies, since vaccine peptides must be presented on specific MHC molecules to be seen by T cells.

Perhaps the most important trend the analysis uncovered is a decisive shift away from single-agent vaccination toward combination strategies. Modern trials increasingly pair therapeutic vaccines with immune checkpoint inhibitors such as antibodies targeting PD-1, PD-L1, or CTLA-4, which release the molecular brakes that tumors place on T cells. Others combine vaccines with chemotherapy, exploiting the ability of certain cytotoxic drugs to promote immunogenic tumor cell death and antigen release, or with immune adjuvants that amplify the vaccine-primed response. This combinatorial logic mirrors the broader evolution of cancer immunotherapy, where vaccines are increasingly viewed as one component of a multi-pronged attack rather than a standalone cure.

The clinical context of these trials is also revealing. Most studies enrolled patients with advanced, metastatic, or recurrent ovarian cancer, populations with urgent unmet needs but also with tumors that have already evolved sophisticated immune evasion mechanisms. This creates a fundamental tension: the patients most likely to benefit from vaccination in principle, those with minimal residual disease after cytoreductive surgery and platinum-sensitive relapse, are underrepresented relative to heavily pretreated cohorts whose immune systems and tumor microenvironments are profoundly immunosuppressed. The authors argue that biomarker-guided patient selection will be essential to move the field forward, matching vaccine platforms to patients whose tumors express the right antigens and whose immune milieus remain permissive.

Endpoint selection emerged as another critical weakness. Safety and toxicity measures and immunogenicity readouts dominated the trial landscape, while progression-free survival, overall survival, quality of life, and biomarker-based outcomes were comparatively underrepresented. This skew reflects the exploratory nature of most studies but also highlights why so few vaccines have progressed: regulators and clinicians ultimately need evidence that vaccine-induced immune responses translate into longer survival and better lives, not merely stronger laboratory measurements. The immune monitoring data were particularly inconsistent. Although 92 trials, or 67.65 percent, reported immune response outcomes, only 58 trials, or 42.65 percent, specified the assay method used, and 34 trials described an immune domain without naming any assay at all. Techniques as varied as chromium-51 release assays, intracellular cytokine staining, fluorescence-activated cell sorting, immunohistochemistry, delayed-type hypersensitivity testing, and ELISPOT-based measurements appeared across the literature, making cross-trial comparisons nearly impossible.

This heterogeneity in immune-response reporting is more than a bureaucratic nuisance; it is a scientific bottleneck. Without prespecified, harmonized, vaccine-specific immune monitoring frameworks, the field cannot reliably determine which platforms, antigens, adjuvants, and combinations are genuinely superior. A promising response measured in one trial with one assay cannot be meaningfully compared to a disappointing result in another using a different technique. The study’s authors call for standardized monitoring protocols to be built into trial design from the outset, alongside clinically meaningful endpoints and immune-related response criteria such as irRC and iRECIST that account for the atypical response patterns seen with immunotherapies.

Looking ahead, the researchers outline a roadmap for the next decade of ovarian cancer vaccine development. Personalized vaccines, tailored to the unique mutation profile of each patient’s tumor, hold particular promise in a disease like ovarian cancer where homologous recombination deficiency and BRCA mutations generate abundant neoantigens. Large-scale, multicenter, randomized Phase II and III trials are urgently needed to convert three decades of immunological proof-of-principle into confirmatory clinical evidence. If the field can align biomarker-driven enrollment, combination immunotherapy, standardized immune monitoring, and survival-focused endpoints, the long-pursued dream of a therapeutic ovarian cancer vaccine may finally move from the registry pages of ClinicalTrials.gov into routine clinical practice.

Subject of Research: Clinical trial landscape of therapeutic vaccines for ovarian cancer

Article Title: Clinical landscape of therapeutic vaccines for ovarian cancer: current status, challenges, and future directions

Article References: Shan, Y., Wang, Q., Wan, W., Zhang, Y., & Chu, Y. (2026). Clinical landscape of therapeutic vaccines for ovarian cancer: current status, challenges, and future directions. Journal of Ovarian Research. https://doi.org/10.1186/s13048-026-02253-0

Image Credits: AI Generated

DOI: 10.1186/s13048-026-02253-0

Keywords: ovarian cancer, therapeutic vaccines, immunotherapy, clinical trials, dendritic cell vaccines, peptide vaccines, immune checkpoint inhibitors, combination immunotherapy, immune monitoring, personalized vaccines, ClinicalTrials.gov, translational research