Table 3.
ROC analysis using CA125, HE4, IL-6, IL-8, RMI score, ROMA alone and predicted probability in combination for distinguishing patients with malignant HGSOC from those with benign ovarian mass, malignant HGSOC from patients with normal ovaries, and benign ovarian mass from patients with normal ovaries in the training phase.
| Markers | Malignant vs Benign | Malignant vs Normal | Benign vs Normal | ||||||
|---|---|---|---|---|---|---|---|---|---|
| AUC | 95% CI | p value | AUC | 95% CI | p value | AUC | 95% CI | p value | |
| CA125 | 0.96 | (0.91–1.00) | 3.0 × 10–6 | 0.99 | (0.98–1.00) | 1.2 × 10−9 | 0.73 | (0.53–0.95) | 0.027 |
| RMI score | 0.97 | (0.92–1.00) | 2.0 × 10−6 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.95 | (0.88–1.00) | 2.3 × 10−6 |
| HE4 | 0.99 | (0.99–1.00) | 4.0 × 10−7 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.84 | (0.69–0.99) | 0.0013 |
| ROMA | 0.99 | (0.98–1.00) | 4.9 × 10−7 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.84 | (0.69–0.99) | 0.0013 |
| IL6 | 0.92 | (0.85–1.00) | 1.4 × 10−6 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.91 | (0.76–1.00) | 1.4 × 10−6 |
| IL8 | 0.64 | (0.43–0.84) | 0.17 | 0.72 | (0.59–0.86) | 0.006 | 0.72 | (0.51–0.93) | 0.036 |
| IL6 + CA125 | 0.99 | (0.98–1.00) | 4.3 × 10−7 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.99 | (0.96–1.00) | 4.0 × 10−6 |
| IL6 + RMI score | 0.99 | (0.99–1.00) | 6.4 × 10−7 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.97 | (0.91–1.00) | 1.0 × 10−5 |
| IL6 + HE4 | 1.00 | (1.00–1.00) | 3.7 × 10−7 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 0.97 | (0.91–1.00) | 1.0 × 10−5 |
| IL6 + ROMA | 1.00 | (1.00–1.00) | 3.7 × 10−7 | 1.00 | (1.00–1.00) | 7.8 × 10−10 | 1.00 | (1.00–1.00) | 8.5 × 10−6 |
ROC-receiver operating characteristic, AUC- area under curve, 95% CI- 95% Confidence Interval.
PPV- positive predictive value, NPV-negative predictive value.
ROC-receiver operating characteristic, AUC- area under curve, 95% CI- 95% Confidence Interval.