Table 2 MRI imaging biomarkers for assessment of the immune and tumor environment of glioblastoma
| Hallmark | Threshold | Advantages | Limitations |
|---|---|---|---|
| Cellular proliferation |
MRS: ↑Chomax, ↑Chomean, ↓NAA, ↓Cr, ↑mI, ↓NAA/Cr, ↑Cho/Cr, ↑Cho/NAA ratio MRI: ↓ADC | Specificity |
Tumor heterogeneity Low sensitivity (mM) No specific patterns Acquisition time (MRS) Peripheral lesions (MRS: pitfalls with bone and skin) No absolute reference value (ADC) |
| Membrane proliferation | MRS: ↑Chomax, ↑Chomean, ↑Cho/Cr, ↑Cho/NAA ratio | Low sensitivity (mM) | |
| Structural complexity |
↑Diffusion kurtosis imaging ↓ Fractional anisotropy (brain fibers) |
Specificity Sensitivity |
Availability Still experimental |
| Aminoacid metabolism | – | – | – |
| Glucose metabolism | MRS: ↑free lipids | – | Pitfalls: lymphoma, lactates (TE = 35 ms) |
| Angiogenesis |
↑Ktrans on DCE-MR (permeability) ↓BOLD fMRI signal | Level of evidence |
Software Steroids Non specific |
| Perfusion |
DSC MRI: ↑CBV↑CBF↑rCBV > 4 ↑Relative CBV on DCE-MR | Robust software |
Normalization is required Operator dependent Nonspecific of gliomas |
| Invasiveness |
↑FLAIR ↓ADC (except in the edema) ↓FA (experimental) | Sensitivity |
Non-standardized Operator dependent Normalization is required |
| Hypoxia |
MRS: lactate > 0 (no lactate accumulation in healthy tissue) ↓19F-MRI | Specificity | Non-standardized |
| Necrosis |
↓DWI ↓ADC MRS: lipids > 0 (TE = 35 ms), no lipid accumulation in healthy tissue, Lactate > 0 | Specificity | Non-standardized |
| Edema |
↑ADC ↑T2FLAIR | Sensitivity | Specificity |
| Infiltration of cytotoxic T cells | – | – | – |
| Anergy of T cells | – | – | – |
| Activated microglia | – | – | – |
