Abstract:

Objective

To assess the clinical value of ultrasound features combined with the risk of ovarian malignancy algorithm （ROMA） index in predicting type I and type II epithelial ovarian cancer（EOC）.

Methods

A total of 195 patients with EOC confirmed by surgical pathology at Tongji University Affiliated Obstetrics and Gynecology Hospital （Shanghai First Maternity and Infant Hospital） from January 2021 to January 2024 were retrospectively included， and all patients underwent transvaginal ultrasound examination before treatment.Among them， 64 patients had type I EOC and 131 had type II.The differences in ultrasound features between the two groups were compared， and multivariate Logistic regression analysis was used to screen the independent predictors of type II EOC.The efficacy of ROMA index， ultrasound features， and their combination in predicting EOC subtypes was assessed by receiver operating characteristic（ROC） curve analysis.

Results

Type II EOC was mostly solid lesions （67/131， 51.14%）， while type I EOC was mostly single-chamber cystic-solid lesions （29/64， 45.3%）.Type I EOC was mostly unilateral lesions（58/64， 90.6%）， while type II EOC was mostly bilateral lesions （85/131， 64.9%）， and often accompanied by ascites （69/131， 52.67%）.The cystic part of type II EOC usually had anechoic echoes （43/64， 67.19%）， and papillary protrusions were more common （80/131， 61.07%）， while the cystic part of type I EOC usually had non-anechoic echoes （38/54， 70.4%）， and papillary protrusions were relatively rare （28/64， 43.7%）.All the above differences between the two groups were statistically significant （P＜0.05）.When the critical value of ROMA index was set at 58.48%， it had the highest predictive efficacy for type I and type II EOC， with an area under the ROC curve （AUC） of 0.805， and its sensitivity， specificity， positive predictive value， and negative predictive value were 74.8%， 75.0%， 57.8%， and 82.4%， respectively.The AUC of ultrasound feature prediction was 0.919， with a sensitivity， specificity， positive predictive value， and negative predictive value of 82.4%， 89.1%， 73.4%， and 91.6%， respectively.The combination of ultrasound features and ROMA index showed the highest diagnostic performance， with an AUC of 0.963， and a sensitivity， specificity， positive predictive value， and negative predictive value of 90.8%， 90.6%， 84.4%， and 92.4%， respectively.

Conclusion

ROMA index and ultrasound features have appreciated predictive value for type I and type II EOC， and the combination of ultrasound features and ROMA index has the best efficacy.

Key words: Epithelial ovarian carcinoma, Ultrasonography, Risk of ovarian malignancy algorithm