Objectives: Transcatheter arterial embolization is used to control active hemorrhage at different anatomic locations. Because hematomas can suddenly deteriorate and become life threatening for transplant patients, they require prompt diagnosis and intervention rather than conservative management. Here, we evaluated computed tomography in treatment planning and transcatheter embolization effectiveness for hematoma management in pediatric liver transplant patients.
Materials and Methods: Between June 2012 and December 2021, 10 pediatric liver transplant patients were referred to our interventional radiology unit. Computed tomography and angiograms were reviewed for hematoma location and presence of extravasation. We analyzed correlations between computed tomography and angiography findings and technical and clinical success of the endovascular interventions.
Results: Active leak of contrast material during arterial phase was detected on 9/10 CT scans. Although there was no active bleeding on CT in 1 patient, active arterial bleeding was detected on angiography. On the contrary, in 2 patients, although active bleeding was observed on computed tomography, it was not detected on angiography. Source of bleeding was superior mesenteric artery branches in 4, hepatic artery branch in 2, superior epigastric artery in 1, and phrenic artery in 1 patient. Six of 8 patients with active bleeding were treated with endovascular procedures. The remaining 2 patients received surgery: 1 had bleeding from liver cut surface originating from a hepatic artery branch and received open surgery because the bleeding branch was too thin for catheterization, and 1 was hemodynamically unstable and selective catheterization of the internal thoracic artery would take time. Two patients received embolization procedures with N-butyl 2-cyanoacrylate (glue) diluted with iodized oil, and 1 patient had coil and glue with iodized oil. Embolization with coils was performed in 3 patients. Rate of success with transcatheter arterial embolization was 75%. No complications related to patient comorbidities or embolization procedures were shown. No deaths occurred due to progression of the hematoma.
Conclusions: Transcatheter arterial embolization is effective and safe for treatment of pediatric liver transplant patients with hematomas. Computed tomography has value in identifying the bleeding source and its anatomic relationships and may enhance our intervention abilities to become quicker, more effective, and more secured.
Key words : Arterial bleeding, Computed tomography, Transcatheter embolization
Liver transplant is an effective treatment for children with end-stage liver cirrhosis or acute liver failure. The first living donor liver transplant (LDLT) (adult-to-child) was carried out by Raia and colleagues.1 From that time to the present, despite recent progress in surgical techniques and perioperative management, complications are still a major challenge in pediatric liver transplant recipients. Unlike other types of major abdominal surgery in pediatric patients, LDLT is related to various conditions that increase the risk of arterial bleeding.2-5 As a result of their underlying liver condition, pediatric transplant recipients have some level of coagulation disorder, particularly during the perioperative period. In addition, the LDLT procedure involves liver graft with a cut surface of variable size and multiple anastomotic sites that are potential sources of bleeding after transplant.2-5 Although the incidence and mortality rates of arterial bleeding after liver transplant in children has not been definitively verified, this complication can result in death. Arterial bleeding may occur immediately after surgery or months after the transplant. There are few published studies on pediatric posttransplant bleeding, its diagnosis with computed tomography (CT), and treatment with transcatheter arterial embolization (TAE).
Bleeding after transplant is a potentially dangerous condition that requires immediate identification and treatment rather than conservative therapy. At our institution, dynamic CT is the primary diagnostic imaging modality in transplant recipients with suspected acute hemorrhage, and TAE has been the first choice for the treatment of arterial bleeding for many years. Dynamic CT is essential for detection of hematoma to obtain a positive topographic diagnosis and determine the severity. Moreover, CT can also be used to plan the interventional radiological procedure. Detection of an active arterial bleeding is an indication for the use of TAE. In this study, we evaluated the efficiency of CT in the treatment planning and reported the effectiveness of TAE for the management of hematomas in pediatric transplant patients.
Materials and Methods
Between June 2012 and December 2021, 10 pediatric liver transplant recipients (7 boys, 3 girls; mean age of 61 months; age range, 5 months to 16 years) were referred to our interventional radiology unit. We obtained retrospective medical and surgical records and radiological imaging results for all patients. The sites of arterial bleeding, the technical success rates of TAE, and the complications were all recorded retrospectively. The underlying causes of hepatic failure and liver transplant in these patients were biliary atresia (n = 3), Wilson disease (n = 1), progressive familial intrahepatic cholestasis (n = 2), cryptogenic cirrhosis (n = 1), fulminant hepatitis (n = 1), benign intrahepatic cholestasis (n = 1), and primary hyperoxaluria (n = 1). The grafts used in LDLT were left lobe or left lobe lateral segment. Table 1 presents patient characteristics, information on bleeding foci, and treatment modalities.
All 10 patients with suspicion of bleeding received dynamic CT angiography, which included precontrast, arterial, portal, and late phases. The CT examinations were performed with third-generation dual-energy Siemens Somatom Force (Siemens Healthcare AG) or 16-detector CT device (Siemens, Sensation). A bleeding focus was defined as a contrast agent extravasation that was possibly the cause of bleeding. After dynamic CT, all patients were evaluated with angiography. If bleeding was detected on the angiography, TAE or surgery were used to stop the bleeding, according to the mutual agreement between the surgeons and interventional radiologists. The first treatment option was TAE; however, if the bleeding points were too challenging or risky for endovascular embolization, open surgery was the next preferred treatment. Six patients with active bleeding were treated with endovascular techniques. Bleeding sources detected on angiography were superior mesenteric artery branches (n = 4), hepatic artery branches (n = 2), superior epigastric artery (n = 1), and phrenic artery (n = 1).
In all patients, informed consent was obtained from the patient’s family before the angiography. A 4F vascular sheath was placed into the right or left common femoral artery under the guidance of ultrasonography and fluoroscopy. Catheters were used based on the attending interventional radiologist’s preference. All angiograms were obtained using digital subtraction angiography. When the bleeding focus was identified, superselective catheterization was performed with a microcatheter (Fast Tracker, Target Therapeutics; or Direxion HI-FLO, Boston Scientific), which was inserted coaxially through the guiding angiographic catheter. The microcatheter was then advanced as close to the bleeding source as possible, allowing for superselective arteriography and embolization. Embolization procedures were performed with coils (Azur cx 18, Terumo) in 3 patients, with N-butyl 2-cyanoacrylate (NBCA; Histoacryl Braun) diluted with iodized oil (Lipiodol, Andre Guerbet) in 2 patients, and with NBCA and coils in 1 patient (Figure 1 and Figure 2).
Computed tomography and angiograms were reviewed for the location of the hematoma and presence of extravasation. We analyzed correlations between CT and angiography findings and technical and clinical success of the endovascular interventions. The technical success of TAE was defined as the complete disappearance of arterial bleeding following TAE. Technically impossible or incomplete embolization was regarded as technical failure.
Active leak of contrast material during arterial phase was detected in 9 of 10 CT scans (90%). Although there was no active bleeding on CT in 1 patient, angiography revealed active arterial bleeding. In 2 patients, although active bleeding was observed on CT, the bleeding point could not be detected on angiography. The distribution of the bleeding foci is shown in Table 1. The bleeding foci originated most frequently from the superior mesenteric artery branches (4/10, 40%). Six of 8 patients with active bleeding on angiography were treated with endovascular procedures (Table 1). In these patients, control selective angiograms after embolization did not demonstrate recurrent bleeding at the same or any other bleeding foci. Two of 8 patients with active bleeding on angiography were treated surgically. One of these patients (Table 1, patient 7) required surgery because the patient was hemodynamically unstable and selective catheterization of the internal thoracic artery would take time. The other patient, who showed bleeding to the liver cut surface originating from a hepatic artery branch (Table 1, patient 10), was treated by open surgery because the bleeding branch was too thin for catheterization.
Two patients did not show active bleeding focus in angiography. These cases were managed with conservative therapy, including fluid resuscitation, transfusion, treatment of predisposing factors, and close clinical observation. Signs of active bleeding ameliorated in these 2 patients who were treated conservatively.
Except for 1 patient, CT detected bleeding source correctly in all treated patients, and the success rate of TAE was found to be 75%. No complications related to patient comorbidities or the embolization procedure (such as mistarget embolization, reflux of the glue to another innocent vessel, catheter adherence) were shown. None of the patients died due to a progression of the hematoma.
Pediatric patients undergoing liver transplant are at risk for significant bleeding.5 In both pediatric and adult transplant recipients, bleeding estimates are difficult to quantify given variability in the definition of major bleeding; however, the rate has been reported to range from approximately 5% to 9%.5-8 The risk of bleeding is high during the early postoperative period.4 In our study, all arterial bleeding occurred during the early postoperative period, ranging between 1 and 62 days after transplant.
Hemorrhage after liver transplant might be arterial or venous in origin. If arterial bleeding is not detected and treated promptly, it can be lethal. Computed tomography angiography is the main imaging modality in our institution for transplant patients with suspicion of acute hemorrhage. Computed tomography angiography is a correct and reasonable imaging modality to determine active bleeding and to plan TAE or surgery by visualizing anatomic location of active bleeding.9 In our study, prior CT examinations allowed a faster and easier embolization because selective catheterizations were started with the target artery defined with the CT. Other possible arteries were not catheterized unless there was no active bleeding on the CT. Except for 1 patient, CT detected active bleeding source correctly in all of our treated patients. In this patient, angiography revealed active arterial bleeding, which was treated with TAE. In our opinion, the absence of contrast extravasation on CT should not exclude intervention in patients with persistent bleeding or hemodynamic compromise. According to our findings, a prior CT examination can aid in correct embolization and improvement of clinical success. In pediatric transplant recipients, we suggest that patients should have a CT angiography prior to endovascular intervention whenever possible because it is beneficial for diagnosing the hematoma and planning the TAE.
Gastrointestinal bleeding after liver transplant is a potentially severe situation that may require urgent diagnosis and treatment rather than conservative therapy. Endoscopy is the most common invasive procedure for identifying and treating gastrointestinal bleeding.10,11 For patients for whom endoscopic therapy has failed or is not possible, TAE offers an alternative to surgery.12,13 Transcatheter arterial embolization is becoming more popular because of advancements in catheter technology, new developments in embolization materials, and the availability of adequately trained interventional radiologists.11 Because of lower morbidity and mortality, radiologic imaging and TAE are currently the recommended approach for gastrointestinal bleeding compared with open surgery.14,15 In our study, bleeding from the jejunal wall into the lumen was observed in 4 patients on dynamic CT, which could not be treated endoscopically due to their localization. Active extravasation was detected in 3 of these 4 patients on angiography, and they were successfully treated endovascularly. Intestinal ischemia is the most common major complication in the treatment of gastrointestinal bleeding with TAE.14 However, ischemia was not observed in any of our patients.
A living donor liver transplant can involve a liver graft with a variably sized resection surface and various anastomotic sites, which could lead to posttransplant hemorrhage. The cause of bleeding in 2 patients in our study was from the intrahepatic terminal branches of the hepatic artery. For patients with bleeding from hepatic artery branches, TAE is more difficult to perform after LDLT because of the difficulty in selection of the bleeding arteries due to the presence of thin arterial feeders, a tortuous arterial course, or multifocal occurrences.4,16 In our study, 1 patient who developed bleeding from the hepatic artery end branch was successfully treated with TAE. The other patient with bleeding from liver resection margin originating from a hepatic artery branch was treated by open surgery because the bleeding branch was too thin for catheterization and thus safe embolization. Because of the difficulties highlighted, we believe that TAE should be the first therapeutic option in pediatric patients with bleeding from the liver resection margin; however, surgical options should also be strongly considered, with this option ready for patients.
In our department, NBCA (glue) with iodized oil is the primary choice of embolic agent. Compared with other embolic agents, it has some advantages. Glue with iodized oil allows fast and irreversible embolization with rapid polymerization when contacted with blood. Precise hemostasis can be obtained with a single injection with simultaneous embolization of collateral arteries related to the bleeding focus. Thus, these collateral channels may induce backbleeding or rebleeding. Moreover, NBCA can reach distal bleeding points when microcatheters cannot be advanced. It can also be used in cases with coagulopathy. Its therapeutic impact is independent of the patient’s coagulation process.17 In our study, NBCA was successfully used during embolization in 3 of 6 patients who underwent TAE.
Inferior phrenic arteries are a pair of thin vessels that supply a considerable portion of the diaphragm. The right inferior phrenic artery is crucial because it behaves as an extrahepatic collateral artery.18,19 Bleeding of the inferior phrenic artery can be caused by liver transplant. In a liver transplant recipient, hepatectomy may damage the diaphragm or perihepatic tissues during retraction of the diaphragm or separation of the liver from the surrounding tissues. The LDLT procedure tends to increase the possibility of diaphragm or perihepatic tissue injury. In adult LDLT, a smaller hepatic graft can be unstable in the large hepatectomy lodge; it requires stronger fixation with the neighboring diaphragm or perihepatic tissue to prevent twisting of the graft on the vascular pedicle, as described by Lee and colleagues and Hong and colleagues.20,21 In their study, Hong and colleagues reported 28 adult patients with inferior phrenic artery bleeding after liver transplant; in all of their patients, inferior phrenic artery bleeding was observed in the first 2 weeks.21 In our study, similar to these studies, a patient that we treated with TAE due to inferior phrenic artery bleeding was LDLT and was on the first postoperative day.
Major limitations of this study were its retrospective nature, the small population size, and absence of a control group without a prior CT scan to compare the efficacy and duration of the interventions.
In conclusion, TAE is an effective and safe treatment modality for pediatric transplant patients who have active bleedings. Contrast-enhanced CT is valuable in identifying the bleeding source and its anatomic relationships and may enhance our ability to provide interventions more quickly, more effectively, and more safely.
Volume : 20
Issue : 5
Pages : 56 - 61
DOI : 10.6002/ect.PediatricSymp2022.O19
From the 1Department of Radiology, and the 2Department of General Surgery, Baskent University Ankara Hospital, Ankara, Turkey
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Ozgur Ozen, Department of Radiology, Baskent University Ankara Hospital, Ankara, Turkey
Phone: +90 5052576404
Figure 1. Computed Tomography and Angiography of 16-Year-Old Girl With Bleeding on Day 1 Posttransplant
Figure 2. Computed Tomography and Angiography of 3-Year-Old Boy With Bleeding on Day 4 Posttransplant
Table 1. Overview of Pediatric Patient Characteristics, Bleeding Foci, and Treatment Modalities