Suppr超能文献

CT 对 HCC 洗脱的定量评估是 TACE 早期完全缓解的预测因子。

Quantitative assessment of HCC wash-out on CT is a predictor of early complete response to TACE.

机构信息

Department of Surgical Sciences - Radiology Unit, A.O.U. Città della Salute e della Scienza di Torino, Via Genova 3, 10126, Turin, Italy.

Department of Surgical Sciences - Radiology Unit, University of Torino - A.O.U. Città della Salute e della Scienza di Torino, Via Genova 3, 10126, Turin, Italy.

出版信息

Eur Radiol. 2021 Sep;31(9):6578-6588. doi: 10.1007/s00330-021-07792-2. Epub 2021 Mar 18.

Abstract

OBJECTIVES

To investigate the predictive value of four-phase contrast-enhanced CT (CECT) for early complete response (CR) to drug-eluting-bead transarterial chemoembolization (DEB-TACE), with a particular focus on the quantitatively assessed wash-in and wash-out.

METHODS

A retrospective analysis of preprocedural CECTs was performed for 129 HCC nodules consecutively subjected to DEB-TACE as first-line therapy. Lesion size, location, and margins were recorded. For the quantitative analysis, the following parameters were computed: contrast enhancement ratio (CER) and lesion-to-liver contrast ratio (LLC) as estimates of wash-in; absolute and relative wash-out (WO and WO) and delayed percentage attenuation ratio (DPAR) as estimates of wash-out. The early radiological response of each lesion was assessed by the mRECIST criteria and dichotomized in CR versus others (partial response, stable disease, and progressive disease).

RESULTS

All quantitatively assessed wash-out variables had significantly higher rates for CR lesions (WO p = 0.01, WO p = 0.01, and DPAR p = 0.00002). However, only DPAR demonstrated an acceptable discriminating ability, quantified by AUC = 0.80 (95% CI0.73-0.88). In particular, nodules with DPAR ≥ 120 showed an odds ratio of 3.3(1.5-7.2) for CR (p = 0.0026). When accompanied by smooth lesion margins, DPAR ≥ 120 lesions showed a 78% CR rate at first follow-up imaging. No significative association with CR was found for quantitative wash-in estimates (CER and LLC).

CONCLUSIONS

Based on preprocedural CECT, the quantitative assessment of HCC wash-out is useful in predicting early CR after DEB-TACE. Among the different formulas for wash-out quantification, DPAR has the best discriminating ability. When associated, DPAR ≥ 120 and smooth lesion margins are related to relatively high CR rates.

KEY POINTS

• A high wash-out rate, quantitatively assessed during preprocedural four-phase contrast-enhanced CT (CECT), is a favorable predictor for early radiological complete response of HCC to drug-eluting-bead chemoembolization (DEB-TACE). • The arterial phase of CECT shows great dispersion of attenuation values among different lesions, even when a standardized protocol is used, limiting its usefulness for quantitative analyses. • Among the different formulas used to quantify the wash-out rate (absolute wash-out, relative wash-out, and delayed percentage attenuation ratio), the latter (DPAR), based only on the delayed phase, is the most predictive (AUC = 0.80), showing a significant association with complete response for values above 120.

摘要

目的

探讨四期对比增强 CT(CECT)对载药微球动脉化疗栓塞术(DEB-TACE)早期完全缓解(CR)的预测价值,尤其关注定量评估的强化方式和廓清方式。

方法

回顾性分析了 129 个连续接受 DEB-TACE 作为一线治疗的 HCC 结节的术前 CECT。记录病变大小、位置和边界。进行定量分析,计算以下参数:对比增强比(CER)和病变与肝脏对比比(LLC),作为强化方式的估计值;绝对廓清(WO)和相对廓清(WO)以及延迟百分比衰减比(DPAR),作为廓清方式的估计值。根据 mRECIST 标准评估每个病变的早期影像学反应,并将其分为 CR 与其他反应(部分缓解、稳定疾病和进展性疾病)。

结果

所有定量评估廓清变量在 CR 病变中的发生率均显著更高(WO,p=0.01;WO,p=0.01;DPAR,p=0.00002)。然而,只有 DPAR 具有可接受的鉴别能力,AUC=0.80(95%CI 0.73-0.88)。特别是,DPAR≥120 的结节发生 CR 的优势比为 3.3(1.5-7.2)(p=0.0026)。当病变边界光滑时,首次随访时 DPAR≥120 的病变 CR 率为 78%。定量强化估计值(CER 和 LLC)与 CR 无显著相关性。

结论

基于术前 CECT,HCC 廓清的定量评估对预测 DEB-TACE 后早期 CR 有帮助。在不同的廓清定量公式中,DPAR 具有最佳的鉴别能力。当与 DPAR≥120 和病变边界光滑同时存在时,与相对较高的 CR 率相关。

关键要点

  1. 在术前四期 CECT 中,强化廓清率高是 HCC 对 DEB-TACE 早期影像学完全缓解的有利预测因素。

  2. 即使使用标准化方案,CECT 的动脉期也显示出不同病变之间衰减值的很大差异,限制了其在定量分析中的应用。

  3. 在用于量化廓清率的不同公式中(绝对廓清、相对廓清和延迟百分比衰减比),仅基于延迟期的 DPAR 是最具预测性的(AUC=0.80),对于超过 120 的值与完全缓解有显著关联。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/2a6e/8379130/bc670677785f/330_2021_7792_Fig1_HTML.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验