Supplementary Table 3.
Stage distribution of the cancers detected and missed by each of the models
| Primary invasive epithelial ovarian cancer |
Borderline epithelial cancers |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Models |
Overall, n |
Early stage |
Late stage |
Overall, n |
Early stage |
Late stage |
Nonepithelial |
||||
| Total | 48 | 19 | 29 | 24 | 21 | 3 | 6 | ||||
| Detected cancers | No | No | % | No | % | No | No | % | No | % | No |
| Ultrasound plus CA125 plus HE4a | 43 | 15 | 78.9 | 28 | 96.6 | 14 | 11 | 52.4 | 3 | 100.0 | 2 |
| Ultrasound plus CA125a | 43 | 16 | 84.2 | 27 | 93.1 | 13 | 10 | 47.6 | 3 | 100.0 | 2 |
| ROMA (CA125 plus HE4) | 42 | 14 | 73.7 | 28 | 96.6 | 11 | 8 | 38.1 | 3 | 100.0 | 1 |
| Ultrasound plus HE4a | 37 | 14 | 73.7 | 23 | 79.3 | 14 | 11 | 52.4 | 3 | 100.0 | 2 |
| CA125a | 39 | 14 | 73.7 | 25 | 86.2 | 12 | 9 | 42.9 | 3 | 100.0 | 1 |
| RMI-mod | 39 | 14 | 73.7 | 25 | 86.2 | 10 | 7 | 33.3 | 3 | 100.0 | 1 |
| Ultrasounda | 30 | 14 | 73.7 | 16 | 55.2 | 11 | 8 | 38.1 | 3 | 100.0 | 2 |
| HE4a | 35 | 10 | 52.6 | 25 | 86.2 | 6 | 5 | 23.8 | 1 | 33.3 | 2 |
| Missed cancers | |||||||||||
| Ultrasound plus CA125 plus HE4a | 5 | 4 | 21.1 | 1 | 3.4 | 10 | 10 | 47.6 | 0 | 0.0 | 4 |
| Ultrasound plus CA125a | 5 | 3 | 15.8 | 2 | 6.9 | 11 | 11 | 52.4 | 0 | 0.0 | 4 |
| ROMA (CA125 plus HE4) | 6 | 5 | 26.3 | 1 | 3.4 | 13 | 13 | 61.9 | 0 | 0.0 | 5 |
| Ultrasound plus HE4a | 11 | 5 | 26.3 | 6 | 20.7 | 10 | 10 | 47.6 | 0 | 0.0 | 4 |
| CA125a | 9 | 5 | 26.3 | 4 | 13.8 | 12 | 12 | 57.1 | 0 | 0.0 | 5 |
| RMI-mod | 9 | 5 | 26.3 | 4 | 13.8 | 14 | 14 | 66.7 | 0 | 0.0 | 5 |
| Ultrasounda | 18 | 5 | 26.3 | 13 | 44.8 | 13 | 13 | 61.9 | 0 | 0.0 | 4 |
| HE4a | 13 | 9 | 47.4 | 4 | 13.8 | 18 | 16 | 76.2 | 2 | 66.7 | 5 |
Early stage includes stages I and II. HE4, human epididymis 4; RMI, Risk of Malignancy Index; ROMA, Risk of Ovarian Malignancy Algorithm.
Gentry-Maharaj et al. HE4 in diagnosis of ovarian cancer. Am J Obstet Gynecol 2020.
a
All models incorporate age.