文献速递妇产影像泌尿生殖US

超声分类O-RADS US v2022、O-RADS US v2020与IOTA SR鉴别附件良恶性肿块的诊断准确性:O-RADS US v2022联合肿瘤标志物HE4的提升

Diagnostic accuracy of ultrasound classifications - O-RADS US v2022, O-RADS US v2020, and IOTA SR - in distinguishing benign and malignant adnexal masses: Enhanced by combining O-RADS US v2022 with tumor marker HE4.
PubMed 文献 · Eur J Radiol · 2024-11-09

回顾性纳入2018年1月至2023年12月连续613例因附件肿块拟手术女性,以病理为参考标准。O-RADS US v2022与v2020的AUC分别为0.940和0.937,均优于IOTA SR的0.862;O-RADS US v2022联合HE4的AUC达0.98,优于两者单独使用。

要点

  • O-RADS US v2022与v2020诊断效能相当(AUC 0.940 vs 0.937),均优于IOTA SR(AUC 0.862)。
  • O-RADS US v2022联合HE4的AUC为0.98,优于O-RADS US v2022单独(0.94)或HE4单独(0.92)。
  • O-RADS US v2022、v2020与IOTA SR的Kappa值分别为0.933、0.891、0.923,阅片者间一致性高。

核心结论

  • O-RADS US v2022预测卵巢恶性病变的效能与O-RADS US v2020相当,且优于IOTA SR;O-RADS US v2022联合HE4较两者单独使用诊断效能更优。

研究设计与样本

  • 回顾性研究,2018年1月至2023年12月连续纳入613例因附件肿块拟手术的女性,以组织病理学结果为参考标准。

诊断标准与关键数值

  • O-RADS US v2022的AUC为0.940,O-RADS US v2020的AUC为0.937,两者比较P=0.02,超过校正显著性水平0.0167。
  • IOTA SR的AUC为0.862,O-RADS US v2022与v2020分别与之比较均P<0.0001。
  • 多因素分析显示O-RADS US v2022与恶性病变相关(OR 9.148,95%CI 4.912-17.039,P<0.001),HE4亦相关(OR 1.023,95%CI 1.010-1.036,P=0.001)。
  • O-RADS US v2022联合HE4的AUC为0.98,优于O-RADS US v2022单独(AUC=0.94)或HE4单独(AUC=0.92)。
  • O-RADS US v2022、O-RADS US v2020与IOTA SR的Kappa值分别为0.933、0.891、0.923。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

PURPOSE: To assess the diagnostic accuracy of O-RADS Ultrasound (O-RADS US) v2022, O-RADS US v2020, and IOTA SR, and to evaluate whether combining imaging findings with tumor markers enhances the diagnosis of adnexal masses.

METHODS: This retrospective study, conducted between January 2018 and December 2023, included consecutive women with adnexal masses scheduled for surgery. Histopathologic results served as the reference standard. Risk factors for malignancy were identified using univariate and multivariate logistic regression analyses. ROC analysis was employed to assess diagnostic test performances, while Kappa statistics evaluated inter-reviewer agreement.

RESULTS: A total of 613 women (mean age, 49.39 ± 12.81 years; range, 16-87 years) with pelvic masses were included. O-RADS US v2022 exhibited comparable performance to O-RADS US v2020, with areas under the curve (AUC) values of 0.940 and 0.937, respectively (p = 0.02, exceeding the adjusted significance level of 0.0167). Both O-RADS models outperformed the IOTA SR, which had an AUC of 0.862 (p < 0.0001 for both comparisons). Multivariate analysis revealed that O-RADS US v2022 [OR 9.148, 95 %CI (4.912-17.039), p < 0.001] and HE4 [OR 1.023, 95 %CI (1.010-1.036), p = 0.001] were significant factors associated with malignant lesions. Furthermore, the combination of O-RADS US v2022 and HE4 demonstrated an AUC of 0.98, significantly outperforming either O-RADS US v2022 alone (AUC = 0.94) or HE4 alone (AUC = 0.92). The Kappa values for O-RADS US v2022, O-RADS US v2020 and IOTA SR were 0.933, 0.891 and 0.923, respectively, indicating substantial inter-reader agreement.

CONCLUSIONS: The O-RADS US v2022 demonstrates comparable performance in predicting ovarian malignant lesions when compared to O-RADS US v2020, while surpassing the performance of IOTA SR. Additionally, the combination of O-RADS US v2022 and HE4 provides improved diagnostic effectiveness over using either O-RADS US v2022 or HE4 alone.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Liu Y, Cao L, Chen S, et al. Diagnostic accuracy of ultrasound classifications - O-RADS US v2022, O-RADS US v2020, and IOTA SR - in distinguishing benign and malignant adnexal masses: Enhanced by combining O-RADS US v2022 with tumor marker HE4. Eur J Radiol. 2024. PMID: 39541614. doi:10.1016/j.ejrad.2024.111824

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。