Table 3.
OTFG (n = 32) | OGCT (n = 21) | Univariate analysis (p) | LASSO estimate | Κ-value | ||
---|---|---|---|---|---|---|
Lesion size (cm), mean ± SD | 12.1 ± 5.3 | 11.8 ± 7.5 | 0.331 a | -0.071 | 0.927 d | |
Morphology | 0.472 a | -0.034 | 0.758 | |||
Multiple nodular | 10 (31.3%) | 8 (38.1%) | ||||
Oval | 13 (40.6%) | 9 (42.9%) | ||||
Shallow lobulated | 9 (28.1%) | 4 (19.0%) | ||||
Capsule | 0.004 b | 0.033 | 0.903 | |||
Absent | 13 (40.6%) | 1 (4.8%) | ||||
Present | 19 (59.4%) | 20 (95.2%) | ||||
Solid and cystic components | 0.013 a | 0.000 | 0.942 | |||
Solid | 13 (40.6%) | 1 (4.8%) | ||||
Partly cystic | 7 (21.9%) | 8 (38.1%) | ||||
Solid and cystic | 8 (25.0%) | 5 (23.8%) | ||||
Partly solid | 4 (12.5%) | 7 (33.3%) | ||||
Peripheral cystic areas | < 0.001 b | 0.097 | 0.762 | |||
Absent | 16 (50.0%) | 1 (4.8%) | ||||
Present | 16 (50.0%) | 20 (95.2%) | ||||
Cyst size (cm), mean ± SD | 4.2 ± 5.9 | 5.0 ± 4.5 | 0.135 a | 0.000 | 0.985 d | |
Cystic pattern | 0.014 a | 0.000 | 0.903 | |||
None | 13 (40.6%) | 1 (4.8%) | ||||
Mainly large | 12 (37.5%) | 11 (52.4%) | ||||
Mainly small | 3 (9.4%) | 6 (28.6%) | ||||
Mixed | 4 (12.5%) | 3 (14.3%) | ||||
Hemorrhagic component | 0.004 c | 0.000 | 0.958 | |||
Absent | 26 (81.2%) | 8 (38.1%) | ||||
Present | 6 (18.8%) | 13 (61.9%) | ||||
Endometrial hyperplasia | 0.671 b | 0.002 | 0.899 | |||
Absent | 29 (90.6%) | 3 (9.4%) | ||||
Present | 18 (85.7%) | 3 (14.3%) | ||||
Uterine fibroid or adenomyosis | 0.739 b | 0.000 | 0.801 | |||
Absent | 11 (34.4%) | 21 (42.9%) | ||||
Present | 9 (65.6%) | 12 (57.1%) | ||||
Pelvic fluid | 0.013 a | 0.000 | 0.828 | |||
None | 2 (6.3%) | 0 (0.0%) | ||||
Small | 8 (25.0%) | 15 (71.4%) | ||||
Large | 22 (68.8%) | 6 (28.6%) |
Except where otherwise indicated, data are the number (%) of patients. LASSO least absolute shrinkage and selection operator, OTFG ovarian thecoma-fibroma group, OGCT ovarian granulosa cell tumor, SD standard deviation
aData were tested using the Mann–Whitney U test
bData were tested using Fisher’s exact test
cData were tested using the chi-squared test
dThe intra-class correlation (ICC) was calculated instead because of continuous variables