Quantitative Evaluation of Gastrorenal Shunt Morphology Using Three-Dimensional Computed Tomography Portography: Comparison With Intraoperative Venography

TL;DR

This study compares 3DCT portography with intraoperative venography for evaluating gastrorenal shunt anatomy, finding that 3DCT is reliable for preprocedural planning in gastric variceal treatment.

Contribution

The study provides quantitative validation of 3DCT portography as a reliable alternative to intraoperative venography for preprocedural planning in BRTO.

Findings

3DCT portography showed high agreement with venography for vessel diameters and stenosis rates.

The vertebral level of stenosis was more caudal on 3DCT portography compared to venography.

Key anatomical features like angulation and confluence with the left adrenal vein were consistently identified by 3DCT.

Abstract

Background: Gastric varices are a serious complication of portal hypertension and may cause life-threatening bleeding when ruptured. Balloon-occluded retrograde transvenous obliteration (BRTO) is a standard endovascular treatment performed via the gastrorenal shunt (GRS). The present study examined concordance between three-dimensional computed tomography (3DCT) portography and intraoperative venography at the GRS outflow and evaluated the utility of 3DCT portography for preprocedural planning.

Methods: Nineteen patients who underwent BRTO between 2017 and 2024 were retrospectively analyzed. Preoperative 3DCT portography and intraoperative venography were used to assess the following parameters: the diameter of the stenosis, the diameter of the common trunk, the stenosis rate, the distance from the left renal vein to the stenosis, and the vertebral level of the stenosis. Paired t-tests, a Bland-Altman analysis, and correlation analysis were performed.

Results: No significant differences were observed in vessel diameters, the stenosis rate, or the distance from the left renal vein between the two modalities. The vertebral level of the stenosis was significantly more caudal on 3DCT portography (p = 0.0022). Bland-Altman plots showed good agreement, and all parameters showed moderate to strong correlations. In all cases, the stenosis was located just cranial to the confluence with the left adrenal vein, followed by a dorsal angulation averaging 62.9°, located approximately 20 mm cranial to the stenosis.

Conclusion: 3DCT portography showed high concordance with venography and consistently identified key anatomical features. It is a reliable tool for a preprocedural assessment in BRTO, aiding in risk stratification and individualized planning.

Figures