Background and aims: Non-invasive methods accurately estimating hepatic venous pressure gradient (HVPG) are un unmet clinical need. Preliminary data suggested that graph analysis of dynamic contrast enhanced ultrasonography (DCE-US) of the liver using a “connectome” approach allows assessment of the liver microcirculatory derangement and mirrors the severity of portal hypertension (Amat-Roldan et al. Radiology 2015). The EC-funded prospective CLEVER study (FP7-IAPP-GA-2013-612273-CLEVER) is aimed at developing a novel automatized software based on DCE-US able to improve prognostication in cirrhosis. First extended results were developed with a Siemens Acuson Sequoia in Barcelona, showing optimal correlation with HVPG. Here we report the adaptation of this CLEVER software to DCE-US videos acquired with ESAOTE MyLab equipments in Bologna to predict HVPG in a population of patients with F≥3 hepatopathy. Method: Ten seconds long videoclip(s) of the right liver lobe were recorded in each patient producing one cycle of microbubble disruption and reperfusion during SonoVue infusion. A total of 90 videos from randomly selected 47 patients were utilized to optimize the autoselection algorithm of the computer among 5models based on platelet count and spleen diameter. Results: Applicability: the CLEVER software was technically able to provide portal pressure estimations from DCE-US in 41/90 videos corresponding to 28/47 patients (59.6%). The Spearman coefficient of correlation between CLEVER values and HVPG was r = 0.585 ( p < 0.001). The CLEVER software was then tested in a separate validation set of 17 technically successful patients, showing a correlation r = 0.701 ( p < 0.002). Conclusion: We developed and validated the DCE-US based CLEVER software which allows an automatic and quantitative non-invasive estimation of portal pressure in patients with CLD. Larger set of patients with precise subgrouping will help improving the non-invasive predictability of portal pressure by DCE-US.
Non-invasive measurement of HVPG using graph analysis based on dynamic contrast-enhanced ultrasound with ESAOTE MyLab: The CLEVER Study.
F Piscaglia;A Berzigotti;V Sansone;Stefanescu H;Allegretti G;Ravaioli F;Golfieri R;Vukotic R;
2018
Abstract
Background and aims: Non-invasive methods accurately estimating hepatic venous pressure gradient (HVPG) are un unmet clinical need. Preliminary data suggested that graph analysis of dynamic contrast enhanced ultrasonography (DCE-US) of the liver using a “connectome” approach allows assessment of the liver microcirculatory derangement and mirrors the severity of portal hypertension (Amat-Roldan et al. Radiology 2015). The EC-funded prospective CLEVER study (FP7-IAPP-GA-2013-612273-CLEVER) is aimed at developing a novel automatized software based on DCE-US able to improve prognostication in cirrhosis. First extended results were developed with a Siemens Acuson Sequoia in Barcelona, showing optimal correlation with HVPG. Here we report the adaptation of this CLEVER software to DCE-US videos acquired with ESAOTE MyLab equipments in Bologna to predict HVPG in a population of patients with F≥3 hepatopathy. Method: Ten seconds long videoclip(s) of the right liver lobe were recorded in each patient producing one cycle of microbubble disruption and reperfusion during SonoVue infusion. A total of 90 videos from randomly selected 47 patients were utilized to optimize the autoselection algorithm of the computer among 5models based on platelet count and spleen diameter. Results: Applicability: the CLEVER software was technically able to provide portal pressure estimations from DCE-US in 41/90 videos corresponding to 28/47 patients (59.6%). The Spearman coefficient of correlation between CLEVER values and HVPG was r = 0.585 ( p < 0.001). The CLEVER software was then tested in a separate validation set of 17 technically successful patients, showing a correlation r = 0.701 ( p < 0.002). Conclusion: We developed and validated the DCE-US based CLEVER software which allows an automatic and quantitative non-invasive estimation of portal pressure in patients with CLD. Larger set of patients with precise subgrouping will help improving the non-invasive predictability of portal pressure by DCE-US.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
