Techniques of TIPS in the treatment of liver cirrhosis ...

Liver cirrhosis is one of the important risk factors that promote the formation of PVT8. PVT may aggravate portal hypertension, which lead to variceal gastrointestinal bleeding and refractory ascites. These complications negatively influency the survival. In this study, all patients had developed PVT. All of the patients had portal hypertension, such as variceal gastrointestinal bleeding or refractory ascites or a combination of both above.

PVT was once considered to be a contraindication of TIPS. But as the improvement of the technology, more than 500 PVT patients have been reported to underwent TIPS until 20129. Han G et al.6 had reported 75 subjects, whose success rate of TIPS was 75% (43/75). Angelo Luca et al.7 had reported 70 subjects, whose success rate of TIPS was 100%. As some studies reported, the degree of the PVT is an essential factor for the technical success6,7, and portal vein thrombosis with cavernous transformation usually cause the failure of TIPS10. Some scholars have classified the thrombosis in the portal vein system7,10,11. For example, Angelo Luca7 has classified PVT according to the proportion of thrombosis in the portal vein: Grade 0 invisible PVT; Grade I 1–25%, Grade II 26–50%, Grade III 51–75%, and Grade IV 76–100%. Most previous studies believe that conventional TIPS procedure can be directly applied to patients with images showing clear portal veins, while percutaneous transhepatic puncture combined with TIPS was applied in patients with images showing unclear portal veins. In this study, based on Angelo Luca’s grade of thrombosis, the thrombosis location was classified comprehensively mainly on the main portal vein thrombosis (Table 2). Based on this more specific classification, appropriate TIPS procedures were performed, leading to a high technical success rate of 95.8% (183/191). The technical success rate of TIPS for PVT in previous reports varied between 71–100%6,7. According to the results above, the technical success rate in our study is better. We also reported individual technical success rates of different types of PVT, and found there was statistically significant difference between Grade II and IV of the main PVT in TIPS success rates. We think the patients with lower grade of PVT have better technical success rate of TIPS.

Full size table

Angelo et al.7 have reported that TIPS was conducted in 70 PVT patients, the preoperative lumen occupancy (%) was p 49 ± 28, m 37 ± 29, s 10 ± 21, respectively, while the postoperative lumen occupancy (%) was p 10 ± 18, m 10 ± 17, s 3 ± 11, respectively. Only 57.1% (40/70) of the patients achieved complete restoration of portal vein blood flow after TIPS, while PVT was significantly reduced in 30% (21/70) of the patients. However, nearly 12.9% (9/70) of PVT patients did not show any significant improvement after TIPS. As we all know, survival time is the most important factor for prognosis. In view of the research of Angelo above, we have a hypothesis, and that is whether vascular patency is an essential factor for cumulative survival time. After a TIPS procedure, we immediately performed angiography of PV, splenic and mesenteric veins to identify the patency of vessels. Patency of vessels conferred a better survival. Successful TIPS insertions can maintain the persistent portal vein patency, and avoid thrombus extension into the portal venous system5, which is perhaps the reason. Regarding the analysis of survival in case of mesenteric and splenic vein thrombosis, because some patients have different proportion of main portal vein thrombosis could be a competing effect for the survival analysis. Further investigation should be conducted to make the conclusion more convincing. I will make a cox regression analysis in the future.

Previous studies reported that TIPS has a mortality rate of 0–3%12. In this study, no body died because of TIPS, but two patients developed severe complications. The first one suffered from acute shunt occlusion because the distal stent was embedded in the PVT, so the shunt patency could be attained only after implanting the stent. The second patient got thoracic cavity bleeding when the percutaneous transhepatic stunt was implanted. The lateral and posterior costophrenic angle is deepened in patients with liver cirrhosis, and the puncture must have crossed the right diaphragm in this patient. Therefore, the bleeding occurred in the thoracic cavity due to negative pressure. The bleeding stopped when we performed emergency radiofrequency ablation of the puncture stunt. Although a combination of TIPS and percutaneous transhepatic portal puncture tends to improve the success rate of TIPS, the risks of thoracic and abdominal bleeding (seven cases have been described in our next article) should not be ignored.

Han et al.6 conducted a study on 57 patients, which included 35 cases of partial thrombosis (61.4%) and 22 cases of completely occlusive thrombosis (38.6%). The cumulative 1-year and 2-year shunt obstruction were 21% and 32%, respectively, while the incidences of hepatic encephalopathy were 25% and 27%, respectively. Moreover, the incidences of cumulative 1-year and 5-year variceal bleeding were 10% and 28%, respectively. Although we selected patients with incompletely occlusive thrombosis in the main portal veins, 48 cases (26.2%) suffered from Grade IV main PVT. Moreover, 31 cases were complicated as they had developed Grade IV mesenteric vein and/or splenic vein thrombosis. Therefore, the results are not only associated with thrombosis characteristics (such as degree, range, involved target vessels etc.), but they are also influenced by various parameters, such as liver function, technical procedures, postoperative treatment, symptoms relapse, thrombosis relapse, and follow-up time. We will make a cox regression analysis in the future.

In summary, we conclude that appropriate TIPS procedures and lower grade of PVT are essential for better technical success rate of TIPS. The patency of target vessels is important for survival.