PDF(Pubmed)
Abstract：
Managing complications of liver cirrhosis such as varices needing treatment (VNT) and clinically significant portal hypertension (CSPH) demands precise and non-invasive diagnostic methods. This study assesses the efficacy of spleen stiffness measurement (SSM) using a 100-Hz probe for predicting VNT and CSPH, aiming to refine diagnostic thresholds. A retrospective analysis was conducted on 257 cirrhotic patients, comparing the diagnostic performance of SSM against traditional criteria, including Baveno VII, for predicting VNT and CSPH. The DeLong test was used for statistical comparisons among predictive models. The success rate of SSM@100 Hz was 94.60%, and factors related to SSM failure were high body mass index and small spleen volume or length. In our cohort, the identified SSM cut-off of 38.9 kPa, which achieved a sensitivity of 92% and a negative predictive value (NPV) of 98% for detecting VNT, is clinically nearly identical to the established Baveno threshold of 40 kPa. The predictive capability of the SSM-based model for VNT was superior to the LSM ± PLT model (p = 0.017). For CSPH prediction, the SSM model notably outperformed existing non-invasive tests (NITs), with an AUC improvement and significant correlations with HVPG measurements (obtained from 49 patients), highlighting a correlation coefficient of 0.486 (p < 0.001) between SSM and HVPG. Therefore, incorporating SSM into clinical practice significantly enhances the prediction accuracy for both VNT and CSPH in cirrhosis patients, mainly due to the high correlation between SSM and HVPG. SSM@100 Hz can offer valuable clinical assistance in avoiding unnecessary endoscopy in these patients.
摘要：
管理肝硬化的并发症，例如需要治疗的静脉曲张（VNT）和临床上显着的门脉高压（CSPH），需要精确且无创的诊断方法。这项研究评估了使用100Hz探头预测VNT和CSPH的脾硬度测量（SSM）的功效。旨在完善诊断阈值。对257例肝硬化患者进行了回顾性分析，将SSM的诊断性能与传统标准进行比较，包括BavenoVII,用于预测VNT和CSPH。DeLong检验用于预测模型之间的统计比较。SSM@100Hz的成功率为94.60%,与SSM失败相关的因素是高体重指数和小脾脏体积或长度。在我们的队列中，确定的SSM截止值为38.9kPa，检测VNT的灵敏度为92%，阴性预测值为98%，在临床上与建立的40kPa的Baveno阈值几乎相同。基于SSM的VNT模型的预测能力优于LSM±PLT模型(p=0.017)。对于CSPH预测，SSM模型的性能明显优于现有的非侵入性测试(NIT)，AUC改善且与HVPG测量值显着相关（从49名患者获得），强调SSM和HVPG之间的相关系数为0.486(p<0.001)。因此,将SSM纳入临床实践显着提高了肝硬化患者VNT和CSPH的预测准确性，主要是由于SSM和HVPG之间的高度相关性。SSM@100Hz可以为这些患者提供有价值的临床帮助，避免不必要的内窥镜检查。
