Scholarlink

Purpose <p>Distinguishing glioblastoma (GBM) from large solitary brain metastases (LBM &gt; 1.5&#xa0;cm) is crucial for tailoring therapy and optimizing outcomes. We used MRI-derived neovascularization and oxygen metabolism tumor microenvironment (TME) mapping to differentiate GBM from non-small cell lung cancer (NSCLC) origin—LBM.</p> Material and methods <p>This observational study included 21 GBM and 27 LBM patients who underwent physio-metabolic MRI. Imaging parameters related to oxygen metabolism (oxygen extraction fraction [OEF], cerebral metabolic rate of oxygen [CMRO<sub>2</sub>], tissue oxygen tension [PO<sub>2</sub>]) and neovascularization (microvessel density [MVD], vessel size index [VSI], microvessel-type indicator [MTI]) were calculated to classify tumors into five TME types: necrosis, hypoxia (with or without neovascularization), oxidative phosphorylation (OxPhos), and aerobic glycolysis. Volume fractions of different TMEs and MRI parameters in enhanced, necrotic, and edematous areas were compared using two-sample<i> t</i>-tests and FDR correction adjusting. Diagnostic performance was evaluated using receiver operating characteristic analysis.</p> Results <p>Compared to LBM, GBM showed lower hypoxia without neovascularization and total hypoxia percentage (<i>q</i> = 0.007 and 0.0035, respectively), higher aerobic glycolysis and vital tumor percentage (<i>q</i> = 0.0023 and &lt; 0.007, respectively). In subregions, GBM displayed higher PO<sub>2</sub> and MVD value in enhancing area (<i>q</i> = 0.048 and 0,032, respectively). Additionally, GBM exhibited lower OEF in edematous area (<i>q</i> = 0.048) than LBM. Vital tumor percentage and PO<sub>2</sub> in enhancing region were the best differentiation markers (AUC = 0.906 and 0.911).</p> Conclusions <p>Physio-metabolic MRI provides noninvasive, quantitative insights into the heterogeneity of tumor microenvironment characteristics, show potential for preoperative discrimination between GBM and LBM. However, larger multicenter validation studies with diverse metastasis types are needed before widely clinical implementation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Physio-metabolic MRI of oxygen metabolism and neovascularization for differentiating glioblastomas from solitary NSCLC brain metastases

  • 期刊论文

  • 发表日期：

    2026年5月6日

查看全文
  • Qun-Hui Ou-Yang,
  • Ru-tong Pan,
  • Ri-Hui Yang,
  • Ke-Lei Hua,
  • Run Cui,
  • Wu-Ming Li,
  • Si Li,
  • Wei-Xiong Fan,
  • Gui-Hua Jiang,
  • Ping Liu

摘要

Purpose

Distinguishing glioblastoma (GBM) from large solitary brain metastases (LBM > 1.5 cm) is crucial for tailoring therapy and optimizing outcomes. We used MRI-derived neovascularization and oxygen metabolism tumor microenvironment (TME) mapping to differentiate GBM from non-small cell lung cancer (NSCLC) origin—LBM.

Material and methods

This observational study included 21 GBM and 27 LBM patients who underwent physio-metabolic MRI. Imaging parameters related to oxygen metabolism (oxygen extraction fraction [OEF], cerebral metabolic rate of oxygen [CMRO2], tissue oxygen tension [PO2]) and neovascularization (microvessel density [MVD], vessel size index [VSI], microvessel-type indicator [MTI]) were calculated to classify tumors into five TME types: necrosis, hypoxia (with or without neovascularization), oxidative phosphorylation (OxPhos), and aerobic glycolysis. Volume fractions of different TMEs and MRI parameters in enhanced, necrotic, and edematous areas were compared using two-sample t-tests and FDR correction adjusting. Diagnostic performance was evaluated using receiver operating characteristic analysis.

Results

Compared to LBM, GBM showed lower hypoxia without neovascularization and total hypoxia percentage (q = 0.007 and 0.0035, respectively), higher aerobic glycolysis and vital tumor percentage (q = 0.0023 and < 0.007, respectively). In subregions, GBM displayed higher PO2 and MVD value in enhancing area (q = 0.048 and 0,032, respectively). Additionally, GBM exhibited lower OEF in edematous area (q = 0.048) than LBM. Vital tumor percentage and PO2 in enhancing region were the best differentiation markers (AUC = 0.906 and 0.911).

Conclusions

Physio-metabolic MRI provides noninvasive, quantitative insights into the heterogeneity of tumor microenvironment characteristics, show potential for preoperative discrimination between GBM and LBM. However, larger multicenter validation studies with diverse metastasis types are needed before widely clinical implementation.