Transjugular intrahepatic portosystemic shunt (TIPS) dysfunction: quantitative assessment of flow and perfusion changes using 2D-perfusion angiography following shunt revision.

PURPOSE: To analyze the feasibility of 2D-perfusion angiography (2D-PA) to quantify flow and perfusion changes pre- and post-transjugular intrahepatic portosystemic shunt (TIPS) revision.

MATERIALS AND METHODS: Fifteen consecutive patients (54 ± 14 years, seven men and eight women) scheduled for TIPS revision were included in this study. To quantify flow and perfusion changes caused by TIPS revision, digital subtraction angiography (DSA) series acquired during the revision were post-processed using a dedicated software. Reference region-of-interest (ROI) in the main portal vein (input function) and target ROIs in the TIPS lumen, the liver parenchyma and in the right atrium were placed in corresponding areas on DSA pre- and post-TIPS revision. 2D-PA evaluation included time to peak (TTP), peak density (PD), and the area under the curve (AUC) assessment. The ratios of reference ROI to target ROIs pre- and post-TIPS revision were calculated (TTPparenchyma /TTPinflow , PDparenchyma /PDinflow , AUCparenchyma /AUCinflow , TTPTIPS /TTPinflow , PDTIPS /PDinflow , AUCTIPS /AUCinflow , TTPatrium /TTPinflow , PDatrium /PDinflow , and AUCatrium /AUCinflow ). Pressure measurements pre- and post-TIPS revision were performed and correlated to the 2D-PA parameters. Reproducibility of 2D-PA was assessed by the intra-class correlation coefficient (ICC).

RESULTS: The portosystemic pressure gradient was significantly reduced following TIPS revision (17.1 ± 6.3 vs. 8.9 ± 4.3 mmHg; p < 0.0001). PDTIPS /PDinflow (0.22 vs. 0.35; p = 0.0014) and AUCTIPS /AUCinflow (0.24 vs. 0.39; p = 0.0012) increased significantly. Likewise, PDatrium /PDinflow (0.32 vs. 0.78; p = 0.0004) and AUCatrium /AUCinflow (0.3 vs. 0.79; p < 0.0001) increased, whereas PDparenchyma /PDinflow decreased significantly (0.14 vs. 0.1; p = 0.0084). Pressure gradient changes correlated significantly with the increase in PDatrium /PDinflow (r = - 0.77, p = 0.0012) and AUCatrium /AUCinflow (r = - 0.76, p = 0.0018). ICC of the 2D-PA parameters was in the range of 0.88-0.99.

CONCLUSION: 2D-PA offers a feasible approach to quantify flow and perfusion changes during TIPS revision. Therefore, 2D-PA may be a valuable amendment to mere pressure measurements.