Bidimensional shear wave ultrasound elastography with supersonic imaging to predict presence of oesophageal varices in cirrhosis
Bidimensional shear wave ultrasound elastography with supersonic imaging to predict presence of oesophageal varices in cirrhosis
复制标题
二维剪切波超声弹性成像与超声成像预测肝硬化食管静脉曲张的存在
DOI：
10.1111/liv.13418
复制
发表时间：
2017
影响因子：
6.7
通讯作者：
F. Piscaglia
中科院分区：
文献类型：
--
作者：
H. Stefanescu;G. Allegretti;V. Salvatore;F. Piscaglia
我们怀着极大的兴趣阅读了詹森等人的手稿，1最近发表在国际肝脏。为了与他们的发现保持一致，我们还将肝脏和脾脏弹性成像（L/S2 DSWE）作为非侵入性评估胰腺炎患者的重要工具。在他们的工作中，Jansen 1没有评估CSPH并发症的预测，如食管静脉曲张（EV）的存在。Baveno VI共识指出，当肝脏硬度（通过瞬时弹性成像测量）<20 kPa且同时血小板（PLT）计数>150 × 103/mL时，EV可能被排除。2嵌入传统超声扫描仪的新弹性成像技术，如2DSWE with Aixplorer® Supersonic Imagine（2D SWE.SSI）很有吸引力，但尚未根据Baveno VI建议进行测试。因此，我们前瞻性地测试了L/S2 DSWE的作用。SSI排除了73例代偿性心脏病患者队列中的EV，并获得了我院伦理委员会的批准（93/2013/U/ Sper）。73例患者中44例（60.27%）有EV。L2 DSWE、S2 DSWE和PLT单独检测时预测EV存在的准确度适中，AUC为0.753在最佳临界值19 kPa、38 kPa和100 × 103/mL时，分别为0.6230.883、0.747（95% CI：0.6170.876）和0.773（95% CI：0.6480.898）。在Baveno VI建议的设置中，L2DSWE.SSI（<20 kPa）和PLT（>150 × 103）排除了EV，准确性为68.50%（PPV为80%;特异性为95.45%）。以L2DSWE.SSI<19 kPa、PLT>100 × 103为界值，排除EV的总准确率为76.71%，PPV为87.50%，特异性为95.45%。考虑到逐步方法（L2 D SWE.SSI <19 and PLT>100 × 103=无EV; L2 DSWE. SSI>19且PLT<100 × 103=EV可能），34例患者仍处于灰色区域（L2 DSWE. SSI <19且PLT<100 × 103或L2 DSWE. SSI>19且PLT>100 × 103）。S2DSWE。SSI（</≥38 kPa）用于对这些落入灰色区域的患者进行进一步分类。使用这种改进的算法，EV被排除，准确性为83.07%（PPV为77.8%;特异性为84.6%; NPV为86.8%;灵敏度为80.8%），73例内窥镜检查中有54例（74%）可以避免。这些数据证实了Jansen等人的结论，1指出L/S2 DSWE是一个强大的临床工具，用于治疗糖尿病患者。虽然这两组结果仍需要外部验证，但L/S2DSWE.SSI似乎是一种有前途的无创技术，可用于评估代偿性晚期慢性肝病患者的CSPH和EV。Baveno VI算法中使用的最佳截止值仍有待商定，并在较新的剪切波弹性成像机器的情况下进行确认。
We read with great interest the manuscript by Jansen et al.,1 recently published in Liver International. In keeping with their findings, we also hold liver and spleen elastography (L/S2DSWE) as important tools for the noninvasive assessment of cirrhotic patients. In their work, Jansen1 did not assess the prediction of complications of CSPH, like presence of oesophageal varices (EV). The Baveno VI consensus states that EV may be ruledout when liver stiffness (measured by transient elastography) is <20 kPa and concurrently platelet (PLT) count >150 × 103/mL.2 New elastography technologies, embedded in conventional ultrasound scanners, such as 2DSWE with Aixplorer® Supersonic Imagine (2D SWE.SSI) are appealing, but have not been tested in connection with the Baveno VI recommendations. For this reason, we prospectively tested the role of L/S2DSWE. SSI in ruling out EVs in a cohort of 73 compensated cirrhotic patients and approved by the Ethical Committee of our hospital (93/2013/U/ Sper). Fortyfour out of 73 patients (60.27%) had EV. L2DSWE, S2DSWE and PLT had a modest accuracy to predict the presence of EV when tested individually, showing AUCs of 0.753 (95% CI: 0.6230.883), 0.747 (95% CI: 0.6170.876) and 0.773 (95% CI: 0.6480.898), at best cutoff values of 19 kPa, 38 kPa and 100 × 103/mL, respectively. In the settings of Baveno VI recommendations, L2DSWE.SSI (<20 kPa) and PLT (>150 × 103) ruledout EV with 68.50% accuracy (80% PPV; 95.45% specificity). Adopting instead our own cutoff values (L2DSWE.SSI<19 kPa; PLT>100 × 103), EV were ruledout with 76.71% overall accuracy (87.50% PPV and 95.45% specificity). Considering a stepwise approach (L2D SWE.SSI<19 and PLT>100 × 103=no EV; L2DSWE.SSI>19 and PLT<100 × 103=EV probable), 34 patients remained in the grey zone (L2DSWE.SSI<19 and PLT<100 × 103 or L2DSWE.SSI>19 and PLT>100 × 103). S2DSWE. SSI (</≥38 kPa) was used to further classify these patients falling in the grey zone. With this refined algorithm, EV were ruledout with 83.07% accuracy (77.8% PPV; 84.6% specificity; 86.8% NPV; 80.8% sensitivity) and 54/73 (74%) endoscopies could have been spared overall. These data confirm the conclusion of Jansen et al.,1 pointing out L/S2DSWE as a robust clinical tool for the management of cirrhotic patients. Although both sets of results still need external validations, L/S2DSWE.SSI appears to be a promising noninvasive technique for the assessment of CSPH and EVs in patients with compensated advanced chronic liver diseases. The best cutoff values to be used in the Baveno VI algorithm are still to be agreed and validated in the instance of newer shearwave elastography machines.
影响因子：
13.5
作者：
Castera, Laurent;Foucher, Juliette;de Ledinghen, Victor
通讯作者：
de Ledinghen, Victor