Table 3.
Performance of different factors to predict neoadjuvant chemotherapy effects.
| Variable | AUC | p Value | Sensitivity % | Specificity % | +PV % | −PV % |
|---|---|---|---|---|---|---|
| Diffuse Type | 0.81 (0.7–0.9) | <0.001 | 73 (56–87) | 80 (59–93) | 83 (65–94) | 69 (49–85) |
| Poor Diff. | 0.74 (0.6–0.8) | 0.001 | 88 (72–97) | 60 (39–79) | 75 (59–87) | 79 (54–94) |
| NPS Score 3/4 | 0.79 (0.6–0.8) | <0.001 | 79 (62–91) | 68 (46–85) | 77 (60–89) | 71 (49–87) |
| T3 or T4 | 0.70 (0.6–0.8) | 0.0035 | 94 (80–99) | 20 (7–40) | 61 (47–75) | 71 (29–96) |
| Size > 3cm | 0.68 (0.5–0.8) | 0.0087 | 73 (56–87) | 64 (42–82) | 73 (55–87) | 64 (42–82) |
The performance was computed by using the Receiver Operating Characteristics (ROC) curve analysis; AUC means Area under the ROC curve; +PV means positive predictive value; −PV means negative predictive value; values in parentheses indicate 95% confidence interval; The value for the area under the ROC curve can be interpreted as follows: an area of 0.79 means that the probability of neoadjuvant chemotherapy failure (i.e., TRG > 3) in each patient with a NPS score equal or more than 3 is 79% of the time than that in patient with a NPS score less than 3.