Table 3. Cox proportional hazards regression results for a total of 9 models: Uniparametric using rCBVmax at infiltrated peripheral edema (IPE) and Verhaak subtypes and multiparametric using their combination.
| rCBVmax at IPE | Verhaak Subtype | ||||
|---|---|---|---|---|---|
| HR (CI95) | p-value | HR (CI95) | p-value | AUC | |
| rCBVmax at IPE | 1.81 (1.2, 2.7) | 0.0045* | - | - | 0.5815 |
| Classical | - | - | 0.99 (0.5, 2.0) | 0.9748 | 0.5037 |
| Mesenchymal | - | - | 0.83 (0.4, 1.6) | 0.5904 | 0.5407 |
| Neural | - | - | 0.61 (0.3, 1.3) | 0.1958 | 0.5403 |
| Proneural | - | - | 2.58 (1.2, 5.4) | 0.0113* | 0.5847 |
| rCBVmax+ Classical | 1.81 (1.2, 2.7) | 0.0045* | 0.97 (0.5, 2.0) | 0.9469 | 0.5806 |
| rCBVmax+ Mesenchymal | 1.87 (1.2, 2.9) | 0.0058* | 1.15 (0.6, 2.4) | 0.7066 | 0.5676 |
| rCBVmax+ Neural | 1.77 (1.2, 2.7) | 0.0064* | 0.65 (0.3, 1.4) | 0.2741 | 0.5898 |
| rCBVmax+ Proneural | 1.55 (1.0, 2.5) | 0.0759 | 1.67 (0.7, 4.1) | 0.2577 | 0.6102 |
* for statistical significance.