Latest News

• Researchers Uncover How an Aggressive Brain Tumor Evades Treatment        

  This is an interesting new paper from researchers at Yale, Massachusetts General Hospital, the Weizmann Institute of Science, and the University of Miami.

  Using one of the largest longitudinal single-cell datasets in IDH-mutant glioma, the researchers showed that progression can occur through both genetic evolution and cell-state changes. In some tumors, acquired genetic alterations expand proliferative, stem-like cell populations. In others, tumors adopt a mesenchymal-like state associated with immune suppression and poorer outcomes, even without major new genetic changes.
  
  Thus, treatment resistance is not necessarily driven solely by accumulating mutations. Instead, tumors under treatment pressure can evolve toward distinct biological trajectories, including stem-like or more glioblastoma-like programs. Two IDH-mutant gliomas could have similar DNA yet behave very differently because of the cell types and cellular programs active within the tumor. Single-cell sequencing is a powerful tool for capturing these differences, and as our understanding of these evolutionary trajectories improves, it may help identify early warning signs of resistance, predict treatment response, and potentially guide strategies to keep tumors on more treatment-sensitive paths.

  
  
• Tumor transcriptional state predicts survival in immune-checkpoint-blockade-treated glioblastoma        

  In this study, another group of researchers reports a similar idea - that response to therapy in glioblastoma (GBM) may depend just as much on tumor cell state and the surrounding immune environment as on genetic mutations. The study analyzed 181 immune checkpoint inhibitor (ICI)-treated glioblastoma (GBM) samples using bulk DNA and RNA sequencing along with single-nucleus RNA sequencing to understand what predicts response in IDH-wildtype GBM.

  The main finding was that the tumor’s baseline cell state, rather than its mutational burden, best predicted outcomes. Tumors with a mesenchymal (MES) subtype had better survival with ICIs compared to non-MES tumors, even though this same subtype did not benefit as much from standard chemoradiation. This MES state was linked to higher immune signaling (HLA class I expression) and more T cell infiltration, consistent with a more inflamed tumor environment. In contrast, overall mutation burden was not predictive, while specific genetic changes such as PDGFRA and CDKN2A were associated with worse outcomes after ICI but not standard therapy.

  Long-term analyses suggested that ICI treatment may shift how tumors evolve over time, with selection of resistant subclones and changes away from MES-like states, suggesting that immunotherapy and chemoradiation may push tumors along different evolutionary paths.

  
  
• MRI-guided adaptive radiotherapy for high grade glioma (UNITED): a single-centre, single-arm, non-inferiority, phase 2 trial        

  The UNITED trial was a single-center phase 2 study testing MRI-guided adaptive radiotherapy for high-grade glioma using a 1.5T MR-Linac. Ninety-eight patients received standard chemoradiation, but with weekly MRI scans used to adjust radiation plans in real time. This allowed doctors to tighten treatment margins around the tumor while still accounting for microscopic spread.

  The key concern with this approach had always been whether smaller margins would lead to more “missed” tumor at the edges. In this study, that did not happen. The rate of marginal failure was low (4%), meeting the study’s non-inferiority goal compared with historical outcomes. Treatment was generally well tolerated, with lymphopenia as the most common significant side effect. Overall, these results provide reassuring evidence that MRI-guided adaptive radiotherapy can safely reduce radiation margins in glioma, and support further testing in randomized trials.

  
  
• Personalized machine learning-guided radiation dose escalation in newly diagnosed glioblastoma: prospective pilot study        

  A small, prospective pilot study from researchers at the University of Pennsylvania, including the Musella Foundation's Chief Scientific Advisor Dr. Steven Brem, evaluated machine learning-guided personalized precision radiation therapy (PPRT) in newly diagnosed IDH-wildtype glioblastoma (GBM). Twenty patients were enrolled after gross total resection, with efficacy analyses reported for 17 patients. The PPRT used an AI-based model to estimate patterns of microscopic tumor infiltration beyond what was visible on MRI to personalize radiation targeting, followed by standard temozolomide.

  Compared with a matched historical control group, PPRT was associated with improved median progression-free survival (24.4 vs 11.6 months) and overall survival (35.4 vs 17.7 months). Treatment was generally well tolerated, with no grade ≥3 acute toxicities, but radiation necrosis was more frequent in the PPRT group. These findings will need further validation in larger studies, but we hope to see the survival gains replicated.