简化三平面T2加权盆腔MRI用于宫颈癌局部分期
回顾性单中心研究纳入47例宫颈癌患者,评估仅用轴位、矢状位、冠状位T2加权快速自旋回波序列的简化盆腔MRI方案。两名放射科医生独立评估肿瘤大小、宫旁浸润、阴道受累及局部T分期,以麻醉下检查、腹腔镜分期及组织病理为复合参考标准。结果显示该方案诊断准确性高、观察者间一致性良好,采集时间约13分钟。
要点
- 简化T2三平面方案对无膀胱或直肠受累的宫颈癌局部分期具有较高诊断准确性
- 宫旁浸润平均敏感度94.0%、特异度88.6%,阴道受累平均敏感度95.5%、特异度96.0%
- 采集时间约13分钟,适用于时间紧迫或资源受限场景
核心结论
- 非成角三平面T2加权盆腔MRI方案在无膀胱或直肠受累的早期至局部晚期宫颈癌中可提供高诊断准确性及观察者间一致性
- 该简化方案采集时间约13分钟,可作为时间受限或资源受限环境下标准多参数MRI的可行替代方案
研究设计与样本
- 采用1.5T系统,图像判读仅限轴位、矢状位和冠状位T2加权快速自旋回波序列
- 两名取得执业资格的放射科医生独立评估,其中一名为妇科影像亚专科医生,一名为普通放射科医生
- 复合参考标准由麻醉下检查、腹腔镜分期和组织病理学构成
关键影像表现/征象
- 肿瘤大小测量在MRI与参考标准之间一致性良好,平均差异为3.1至4.8 mm
- 宫旁浸润和阴道受累在T2加权图像上可由两名阅片者以高度一致的方式进行判定
- 观察者间一致性方面,肿瘤大小测量ICC为0.946至0.992,宫旁浸润κ=0.828,阴道受累κ=0.915
诊断标准与关键数值
- 宫旁浸润检测的平均敏感度为94.0%,平均特异度为88.6%
- 阴道受累检测的平均敏感度为95.5%,平均特异度为96.0%
- 肿瘤大小测量观察者间ICC为0.946至0.992
- 宫旁浸润观察者间κ=0.828,阴道受累观察者间κ=0.915
- 简化方案采集时间约13分钟,简化机构常规方案约21分钟,标准多参数MRI为30至45分钟
临床应用与阅片要点
- 阅片时以轴位、矢状位、冠状位T2加权序列为唯一依据即可完成宫颈癌关键分期参数的评估
- 适用于临床已确认无膀胱或直肠受累的早期至局部晚期宫颈癌患者
- 在资源紧张或检查时间受限的临床场景中可优先考虑该简化T2三平面方案
局限性
- 回顾性单中心研究设计
- 样本量仅47例
- 结论仅适用于无膀胱或直肠受累的宫颈癌患者
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 evaluate the diagnostic performance and inter-reader reliability of an abbreviated non-angulated three-plane T2-weighted pelvic MRI protocol for assessing local staging key parameters in cervical cancer.
METHODS: In this retrospective single-center study, 47 patients with cervical cancer who underwent pelvic MRI between January 2023 and December 2024 were included. MRI was performed on a 1.5-T system using a simplified protocol; image interpretation was restricted to axial, sagittal, and coronal T2-weighted turbo spin-echo sequences. Two board-certified radiologists (one subspecialized in gynecological imaging, one general radiologist) independently assessed tumor size, parametrial invasion, vaginal involvement, and local T stage. A composite reference standard consisting of examination under anaesthesia, laparoscopic staging, and histopathology was used. Diagnostic performance metrics and inter-reader agreement (Cohen's κ, ICC) were calculated.
RESULTS: Parametrial invasion was detected with high diagnostic accuracy, with a mean sensitivity of 94.0% and specificity of 88.6% across both readers. For vaginal invasion, mean sensitivity and specificity were 95.5% and 96.0%, respectively. Inter-reader agreement was excellent for tumor size measurements (ICC 0.946-0.992) and high for categorical variables categorical variables (κ = 0.828 for parametrial and 0.915 for vaginal invasion). Agreement between MRI and the reference standard for tumor size was good, with mean differences of 3.1-4.8 mm. The abbreviated protocol corresponded to an acquisition time of approximately 13 min compared to 21 min for the simplified institutional protocol and 30-45 min for standard multiparametric MRI.
CONCLUSION: A non-angulated three-plane T2-weighted pelvic MRI protocol provides high diagnostic accuracy and inter-reader agreement for assessing local staging key parameters in early-stage to locally advanced cervical cancer without bladder or rectal involvement and may represent a feasible alternative in time-constrained or resource-limited settings.
引用
- Winkelmann, Bareiss, Winkelmann, et al. Abbreviated T2w-only pelvic MRI for local tumor staging of cervical cancer: a pragmatic approach for time-critical and resource-limited settings. Abdom Radiol (NY). 2026. PMID: 42474717. doi:10.1007/s00261-026-05691-y
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。