Gastric variceal pressure gradient in patients with large spontaneous shunts: A novel approach to local hemodynamic assessment

Figure 1 Endoscopic ultrasound-guided portal pressure measurement process. A: Endoscopic ultrasound (EUS)-guided portal pressure measurement process; B: Zero calibration of the pressure transducer at the right atrial level; C: Gastric fundal varix on coronal computed tomography (CT) imaging; D: Gastric fundal varix on sagittal CT imaging; E: Endoscopic views of gastric fundal varices; F: Endoscopic views of gastric fundal varices; G: Visualization of the main portal vein under EUS, the arrow indicates the main portal vein; H: Visualization of the inferior vena cava under EUS, the arrow indicates the inferior vena cava; I: Visualization of gastric fundal varices under EUS, the arrow indicates gastric fundal varices; J: EUS-guided needle puncture of the main portal vein, the arrow indicates the main portal vein; K: EUS-guided needle puncture of the inferior vena cava, the arrow indicates the inferior vena cava; L: EUS-guided needle puncture of the gastric varix, the arrow indicates gastric fundal varices.

Figure 2 Hemodynamic comparisons and diagnostic performance of different portal pressure parameters. A: Paired comparison of hepatic venous pressure gradient (HVPG), portal pressure gradient (PPG), and gastric varices pressure gradient (GVPG). GVPG consistently exceeded both PPG and HVPG (P < 0.001); B: Correlation between portal vein pressure (PVP) and gastric variceal pressure (GVP) (Spearman’ r = 0.856, P < 0.001); C: Bland-Altman analysis of GVP and PVP [mean bias, 7.20 mmHg; 95% confidence interval (CI): 5.77-8.63]; D: Correlation between HVPG and PPG (Spearman’ r = 0.721, P < 0.001); E: Correlation between PPG and GVPG (Spearman’ r = 0.796, P < 0.001); F: Correlation between GVPG and HVPG (Spearman’ r = 0.537, P = 0.006); G: Bland-Altman analysis of HVPF and PPG (mean bias, -2.88 mmHg; 95%CI: -4.68 to -1.08); H: Bland-Altman analysis of PPG and GVPG (mean bias, -7.28 mmHg; 95%CI: -8.70 to -5.86); I: Bland-Altman analysis of HVPG and GVPG (mean bias, -10.16 mmHg; 95%CI: -12.32 to -8.00); J: Comparison of GVPG between patients with and without recent variceal bleeding. GVPG was significantly higher in patients with recent bleeding (18.40 ± 6.07 mmHg vs 13.20 ± 5.13 mmHg, P = 0.030); K: Receiver operating characteristic (ROC) curves of PVP, GVP, and GVPG. The area under the ROC curve (AUC) for discriminating patients with and without recent variceal bleeding were 0597, 0.670, and 0.760, respectively; L: ROC curves of HVPG, PPG, and GVPG. The AUCs for discriminating patients with and without recent variceal bleeding were 0730, 0.713, and 0.760, respectively. HVPG: Hepatic venous pressure gradient; PPG: Portal pressure gradient; GVPG: Gastric varices pressure gradient; PVP: Portal vein pressure; GVP: Gastric variceal pressure; CI: Confidence interval; VB: Variceal bleeding; TPR: True positive rate; FPR: False positive rate; AUC: Area under the curve.