Our study demonstrates that TIPS is effective in achieving immediate clinical success and preventing future variceal bleeding in children

The risk of medically refractory HE is minimal

Our study shows the need for high-quality comparative studies of TIPS versus surgical bypass in the pediatric population.

Complications of pHTN can range from variceal bleeding, ascites to hepatorenal syndrome and thrombocytopenia.[2] There are two main categories behind the etiology of pHTN, intrahepatic and extrahepatic, with the former being the most common type in children.[3] The present meta-analysis covered a variety of primary diagnoses, such as biliary atresia, cystic fibrosis, congenital hepatic fibrosis, and Budd-Chiari syndrome [Table 1]. TIPS is a well-validated percutaneous intervention that can alleviate pHTN-related complications through the creation of a decompressive shunt between the portal venous and systemic venous systems.[2] Despite TIPS being standard-of-care in selected adult patients, there is limited evidence supporting its use in pediatric patients. Based on 11 retrospective cohorts, the present meta-analysis showed that TIPS is technically feasible and effective for managing pHTN in children.

Similar to the adult population, the most commonly encountered indications in the present meta-analysis were variceal bleeding and refractory ascites [Table 1]. All included patients with ongoing or recurrent variceal bleeding had already failed conservative medical and endoscopic management. In the acute setting, TIPS achieved nearly 100% immediate hemostatic success rate [Figure 2c]. Among the two patients who had continued variceal bleeding and hemoperitoneum post-TIPS, one patient was a 6-year-old female with biliary atresia who received a liver transplant on post-TIPS day 3, while the other patient was a 15-year-old male with Child-Pugh C cryptogenic cirrhosis with concurrent pulmonary and renal disease who succumbed to multisystem organ failure.[6,10] Recurrent or de novo GIB occurred in 14% of patients, most secondary to shunt stenosis or thrombosis [Figure 2d]. The cumulative TIPS dysfunction rate was 27%. The primary patency rate was 71–83% and 60–64% at 1 and 2 years, respectively, consistent with reported rates in both pediatric and adult demographic.[18,19] The adopted follow-up and surveillance algorithms mirrored those of the adult population: Doppler ultrasound at 1 week and every 3–6 months within the 1st year post-TIPS. All reported dysfunctions were successfully managed through standard shunt maintenance without reported complications. While the use of covered versus non-covered stents can affect shunt patency, study-level subgroup analysis of such matter was not possible, as most individual studies included both types of stents.

According to Trebicka et al., the effectiveness of an 8 mm stent surpasses that of a fully or under dilated 10 mm stent, with no difference in shunt dysfunction rates and significantly lower HE complication rates.[20,21] Nonetheless, in younger children, dilating to 8 mm may still be too wide for successful treatment. Several retrospective studies have revealed positive outcomes resulting from under dilation of PTFE stents measuring up to 8 mm diameter with no accompanying elevations in complication risks.[22,23] While self-expandable stents can expand spontaneously and progressively which is beneficial especially for pediatric patients due to their rapid growth rate.[21,23,24] Schepis et al. indicated that this expansion effect does not occur when an 8 mm PTFE stent is under dilated. Hence, the optimal approach toward mitigating complications associated with these interventions would entail using “controlled-expansion” adjustable-diameter devices that lack inherent post-deployment expansion ability.[25,26] It is important to note that selection of stent depends on various other factors and data is not robust concerning the most appropriate diameter for a TIPS stent in a pediatric population.

Refractory ascites was the second most frequent indication among patients of the present study. Whereas 96% of ascites improved after TIPS placement, two patients had persistent mild symptoms controlled with diuretics. Less commonly, TIPS was performed in four cases to treat hypersplenism, a sequelae of pHTN associated with thrombocytopenia [Table 2]. An increase in platelet count and decreased splenic size was noted in several studies, though it was not consistently statistically significant [Table 3]. Other indications of TIPS, such as hydrothorax and hepatorenal syndrome, were not observed in the present study.

TIPS placement was technically successful in 94% of cases (Figure 2a, 95% CI: 86–99%) which are similar to that reported in adult population.[27,28] From a technical perspective, the creation of TIPS in children can be more technically challenging. On the one hand, these pediatric patients have a higher prevalence of hepatic vascular anomalies such as portal venous cavernous transformation and Budd-Chiari occlusive venous sequelae.[9,12] On the other hand, tools such as the standard TIPS kit were initially designed for adults. The size of the needle used in shunt creation can be rather large for younger and smaller children, leading to a high risk of iatrogenic trauma (i.e., hemoperitoneum). Tableside modifications of available tools and intravascular ultrasound have been used to increase technical success and minimize risks.[7,17] With careful manipulation, TIPS can be successfully deployed in patients weighing as low as 11 kg.[7] An additional consideration is a growth, although this patient population is notorious for predicted growth delay secondary to underlying liver disease. For patients with a predicted TIPS for longer than a year, an under-dilated endoprosthesis can be placed initially (i.e., using an 8 mm balloon to dilate a 10 mm endoprosthesis); based on physiological surveillance, future percutaneous adjustment can then be performed if needed.[12,15] Furthermore, because venous thrombosis can be more common in the pediatric population, preprocedural recanalization of the portal vein demands additional technical consideration before TIPS creation.[7,13] A variety of techniques may be needed for successful TIPS placement: secondary percutaneous access such as trans-splenic and transhepatic routes may be necessary and direct intrahepatic portosystemic shunt.[13] Nevertheless, operators should be cognizant of underlying pre-procedural hepatic vascular anatomy.

Despite the aforementioned technical challenges, TIPS is a safe procedure in the pediatric population with bleeding-related complications occurring in 2.7% (5/187) of pooled data, which is consistent with reported literature in the adult population.[29] In terms of long-term side effects, pediatric patients with pHTN are less likely to develop HE as compared to their adult counterparts (Incidence varies from 15% to 48%), this can be explained by the greater proportion of non-cirrhotic etiologies in children.[28] Although the blood ammonia level increased after TIPS placement in multiple studies, only 10.6% (21/198) developed HE. The majority (85.7%, 18/21) resolved with medical management alone.

The present study should be interpreted with caution. First, 10/11 studies are single-arm cohort and case series, which is considered level IV evidence and subject to selection bias. Comparative analyses between TIPS and other options, such as surgery or transplantation, were not performed, whereas its comparison with no TIPS could be difficult to perform for refractory GIB, as precluding a patient from receiving a life-saving intervention would be ethically challenging. Second, the patient population was rather heterogeneous, with a significant variety of primary diagnoses included in the study. Diseases such as Budd-Chiari syndrome are associated with a higher risk of thrombotic events, leading to an increased risk of future shunt dysfunction.[30] Yet, subgroup analysis based on each etiology was not possible due to the low quality of available patient-level data. Further, clinical outcomes were rarely reported based on time intervals. For example, TIPS patency and risk of HE likely increases as the follow-up interval increases, but only a few studies analyzed outcomes using Kaplan–Meier curves. In addition, the survival was poorly reported because many studies had a wide range of follow-up duration and did not include the specifics for individual patients. Noteworthy is the fact that the relationship between final PSG and clinical outcomes was not identified. In adults, a post-TIPS final PSG <12 mmHg or <10 mmHg is the recommended endpoint in the treatment of variceal bleeding and an even lower PSG may be needed for ascites management. By contrast, post-TIPS final PSG values remain generally undefined in children. Slowik et al. have shown that pediatric patients are still at risk of developing variceal bleeding, using adult PSG standard, while Bertino et al. demonstrated that a PSG <12 mmHg may be associated with promising clinical outcomes in the pediatric population.[6,17]