ROC curve analysis results for the prediction of response to induction chemotherapy based on the pretreatment tumor volume and the perfusion-associated parameters*
| Response | Cutoff Value | Sensitivity (%) | Specificity (%) | AUC | P Value |
|---|---|---|---|---|---|
| Pretreatment tumor volume (mL) | ≤15.3 | 100 (99.1–100) | 33.3 (26.5–40.7) | 0.50 | .84 |
| Blood flow (mL/100 g/min) | ≤66.2 | 84.2 (79.8–89.7) | 66.7 (59.3–73.5) | 0.63 | .0001 |
| Blood volume (mL/100 g) | >7.4 | 100 (98.9–100) | 30.6 (23.9–37.8) | 0.59 | .0005 |
| Mean transit time (s) | <8.5 | 43.7 (38.4–49.1) | 100 (98–100) | 0.62 | .0001 |
| Permeability (mL/100 g/min) | >8.6 | 65.4 (60.1–70.4) | 100 (98–100) | 0.78 | .0001 |
Note:—AUC indicates area under the curve.
* The values (95% CI) represent thresholds above or below of which the patients will show a response.