EXPERIMENTAL AND CLINICAL TRANSPLANTATION

In patients with complete portal vein thrombosis, the main portal vein is obstructed, resulting in devel-opment of hepatopetal collateral vessels. In cases of complete portal vein thrombosis, interventional procedures are challenging, with the greatest difficulty in the form of passing the guide wire across the level of obstruction. A recognizable main portal vein remnant has been deemed as a mandatory criterion in previous reports. Here, we report a case of cavernous trans-formation of the portal vein in a pediatric patient after liver transplant who had no obvious detectable portal vein remnant on radiologic imaging. Using digital subtraction angiography, we were successful in passing a guide wire through the level of obstruction and placing a stent, thus causing successful reca-nalization of the occluded segment.

Key words : Pediatric liver transplantation, Portal vein thrombosis, Portal vein stenting

The development of reduced-size hepatic transplant procedures in the pediatric population has resulted in improved patient survival and decreased wait times.¹ One of the challenges of the procedure in pediatric patients is creating an adequate portal vein (PV) diameter. Postoperative portal vein thrombosis (PVT) has been reported in 7% of pediatric partial liver transplant recipients. Late PVT (> 1 mo after transplant) has been reported in 4.5% of cases.² In patients with PVT, the hepatopetal flow obstruction results in development of prehepatic portal hyper-tension and subsequent cavernous replacement of the PV. In these patients, morbidity is due to portal biliopathy, sequelae of hypersplenism, variceal bleeding, ascites, and growth retardation.^3-5 When encountered, this vascular complication is alleviated effectively with surgery or percutaneous venoplasty.^4,5 Recurrent and elastic stenoses have been treated by metallic stents.^6-8

Here, we report a case of successful percutaneous recanalization and stenting of a chronically throm-bosed PV in a pediatric patient who had undergone left lateral split liver transplant.

The institutional review board at our hospital waived the need for approval for this type of retrospective brief case study.

An 8-month-old female patient, who had been diagnosed with cirrhosis secondary to biliary atresia, received an orthotopic liver transplant with a split liver left lateral graft from a deceased donor. Graft Doppler studies on day 1 and day 2 posttransplant were normal. Graft Doppler on day 7 suggested a lack of flow in the main PV. A triphasic computed tomography was done that revealed extrahepatic PVT. However, no active intervention was planned, as the patient was asymptomatic and liver function tests were normal.

Injectable anticoagulation was started for 7 days. The patient had a short stay in the intensive care unit for management of sepsis due to suspected cytomegalovirus infection. The patient was dis-charged at 4 weeks posttransplant, once recovery of function in the transplanted organ, normal liver function tests, and no ascites were shown.

Five years after transplant, the patient presented with melena and low hemoglobin levels. Liver function tests and upper and lower gastrointestinal scope investigations were normal. However, hemo-globin levels continued to persistently decrease despite packed red cell transfusions and injection of terlipressin. A computed tomography angiography did not show an extrahepatic PV. A large collateral vessel was seen arising from the superior mesenteric vein (SMV) and leading to cavernoma formation along the hepaticojejunostomy site, replacing the portal flow (Figure 1). The collateral vessels tra-versing the jejunum were the likely cause of melena. An interventional radiology procedure was planned to attempt percutaneous recanalization of the main PV.

The procedure was performed under general anesthesia in an angiographic suite, with a flat-panel detector-based system (Allura Xpert FD20, Philips, Andover, MA, USA). A trans-splenic puncture of the splenic vein was performed under real-time ultrasonographic guidance, and a 6F vascular sheath (Cordis, Miami Lakes, FL, USA) was inserted in the splenic vein lumen using the Seldinger technique. Contrast injection showed opacification of the splenic vein and the splenic-mesenteric confluence and a collateral pathway arising from the SMV forming the cavernoma around jejunal loops, and delayed acquisitions showed filling of PV branches in the transplanted liver (Figure 2). No recognizable remnant of the main PV was seen above the splenic-mesenteric confluence.

A 5F Kumpe catheter (Cook, Bjæverskov, Denmark) was taken up to the SMV and exchanged with a glide catheter, which was advanced into the collateral forming the cavernoma. Active bleeding was identified through one of the collateral vessels, with contrast opacifying the jejunal lumen. The main portal remnant could not be cannulated successfully. Therefore, a transhepatic puncture of PV was performed, and a needle was placed into one of the intrahepatic portal radicles. Manual portography showed patent intrahepatic portal radicles. The gradient across the thrombosed vein was 22 mm Hg.

A Terumo guide wire was used to probe the distal end of the occluded segment (Figure 3a), and the guide wire was successfully passed across the segment and advanced into the SMV. A 5F Kumpe catheter was inserted over it and replaced with a 5F sizing catheter (Centimeter sizing catheter, Cook, Bjæverskov, Denmark) for accurate measurement of the distance between the splenic-mesenteric con-fluence and the Rex segment of PV. Unfortunately, there was no recognizable normal-caliber PV beyond the splenic-mesenteric confluence.

After the stenosed segment was predilated with an EverCross balloon (8 mm × 40 mm; eV3, Plymouth, MN, USA), a 10-mm-expanded-diameter, 60-mm-length uncovered self-expanding E-Luminexx stent (Bard, Tempe, AZ, USA) was finally deployed into the PV with the proximal end in the SMV and the distal end in the Rex segment of the PV (Figure 3b). The gradient across the stent decreased to 8 mm Hg. Portography showed good flow into the stent, with no filling of the collateral vessels. The transhepatic and trans-splenic tracks were embolized with 18-14-3 Nester coils (Cook).

The patient’s recovery was uneventful, and she was discharged 3 days later. The patient remains symptom free at 8 months after the procedure; a Doppler study during each follow-up visit has shown a patent PV stent.

Portal vein thrombosis in patients after liver transplant is amenable to both surgical and interventional treatment. The aim of each procedure is to restore the physiologic hepatopetal flow. The only surgical procedure with the potential to restore physiologic flow is the meso-Rex shunt. Studies have shown good clinical results and patency with this shunt.^9,10 However, it is not free of complications, which can include shunt thrombosis or stenosis. In a recent study, 6 of 14 meso-Rex shunts performed in patients with PVT after liver transplant required a surgical revision.¹¹ On the other hand, good clinical success has been shown with percutaneous minimally invasive radiologic treatments in patients with main PVT.¹² However, in cases of complete PVT, interventional procedures are challenging, with the greatest difficulty in the form of passing the guide wire across the level of obstruction.

Recent cases of challenging bidirectional approaches (ie, transhepatic and transmesenteric through a mini-laparotomy for successful stent placement) in pediatric recipients with PVT have been described.^13,14 In this scenario, it is useful to be aware of the possibility of catheterizing the PV remnant. Miraglia and associates reported that clear detectability of the PV remnant by axial imaging, patency of major splanchnic vessels, and temporal evidence of recent cavernous replacement of PV should be used as criteria to select pediatric trans-plant recipients with PVT who could benefit from minimally invasive procedures as a first therapeutic approach.¹⁵ However, there are some lacunae in the criteria, including a paucity of data in the literature, rendering it difficult to define a recent cavernous replacement of PV. Our patient, who had chronic documented evidence of PVT and collateral development, had no detectable PV remnant on imaging. However, using fluoroscopic angiography, we were successful in recanalization of the occluded PV segment. Preservation of the patency of the splenic-mesenteric confluence and the Rex segment of the left PV for possible future surgical procedures was deemed mandatory by Miraglia and associates in selection of candidates for endovascular intervention.¹⁵ However, in our patient, because obstruction and occlusion were to the level of the splenic-mesenteric confluence, an uncovered stent was placed with its proximal part within the SMV. With medium patency already previously documented,¹² our presented solution can be a long-term cure with repeat dilatation or additional stent placement if stent occlusion occurs. The long-term patency results are still not clear.

In conclusion, percutaneous recanalization of the PV remnant can be considered as a first therapeutic approach in pediatric transplant recipients with chronic anastomotic PVT, especially in clinically challenging circumstances. Further studies are needed to study the long-term patency rates of PV stents.

Volume: 18 Issue: 2 April 2020

FULL TEXT