ABSTRACT
BACKGROUND: Complications of cirrhosis may persist after liver transplantation. When indicated, partial splenic embolization (PSE) is an alternative to splenectomy but can cause severe infection. The identification of modifiable risk factors when performing PSE in immunocompromised liver transplant recipients may help reduce the risk of severe infection. METHODS: Data were collected retrospectively for all PSE performed after liver transplantation at a single institution and included demographics, etiology of liver disease, indication for PSE, vaccination status, laboratory findings, procedural details, extent and pattern of splenic infarction, hospital length-of-stay, readmissions, procedural complications, and mortality. Statistical analysis included 2-tailed t test, Fisher exact test, and Kaplan-Meier survival curves, with significance defined as P < .05. RESULTS: Sixteen patients received 22 embolizations, with 11 patients undergoing a single session and 5 patients undergoing multiple sessions. Indications included hypersplenism, gastrointestinal hemorrhage, ascites, and autoimmune hemolytic anemia. PSE produced significant and sustained cell count increases, improved ascites, and controlled hemorrhage. Splenic abscess, septic shock, need for splenectomy, and PSE-related mortality were seen in the group with large confluent splenic infarction but not in peripheral/wedge-shaped infarction. Multiple-session PSE exclusively using particles for embolization correlated with the pattern of peripheral/wedge-shaped infarction and avoided severe infection and PSE-related mortality. CONCLUSIONS: PSE in the immunosuppressed liver transplant recipient is an effective alternative to splenectomy, but carries substantial infectious risk. The risk is decreased when PSE performed with polyvinyl alcohol particles results in a pattern of peripheral/wedge-shaped infarction, which correlates with smaller infarction volumes, favorable length-of-stay, and minimal risk of abscess, sepsis, and mortality.
Subject(s)
Embolization, Therapeutic/mortality , Liver Cirrhosis/complications , Liver Transplantation , Postoperative Complications/therapy , Splenic Diseases/therapy , Abscess , Adult , Aged , Ascites/etiology , Ascites/therapy , Embolization, Therapeutic/methods , Female , Gastrointestinal Hemorrhage/etiology , Gastrointestinal Hemorrhage/therapy , Humans , Hypersplenism/etiology , Hypersplenism/therapy , Infarction , Kaplan-Meier Estimate , Liver Cirrhosis/surgery , Liver Transplantation/adverse effects , Male , Middle Aged , Polyvinyl Alcohol , Postoperative Complications/etiology , Postoperative Complications/mortality , Retrospective Studies , Risk Factors , Sepsis/etiology , Sepsis/prevention & control , Splenectomy/statistics & numerical data , Splenic Diseases/etiology , Splenic Diseases/mortality , Young AdultABSTRACT
Hepatocellular carcinoma (HCC) recurrence in patients undergoing liver transplantation (LT) with donation after brain death (DBD) and donation after cardiac death (DCD) allografts has not previously been investigated. Rates and patterns of HCC recurrences were investigated in patients undergoing DBD (N = 1633) and DCD (N = 243) LT between 2003 and 2012. LT for HCC was identified in 397 patients (340 DBD and 57 DCD). No difference in tumor number (p = 0.26), tumor volume (p = 0.34) and serum alphafetoprotein (AFP) (p = 0.47) was seen between the groups. HCC recurrence was identified in 41 (12.1%) patients in the DBD group and 7 (12.3%) patients in the DCD group. There was no difference in recurrence-free survival (p = 0.29) or cumulative incidence of HCC recurrence (p = 0.91) between the groups. Liver allograft was the first site of recurrence in 22 (65%) patients in the DBD group and two (37%) patients in the DCD group (p = 0.39). LT for HCC with DBD and DCD allografts demonstrate no difference in the rate of HCC recurrence. Previously published differences in survival demonstrated between recipients with HCC receiving DBD and DCD allografts despite statistical adjustment can likely be explained by practice patterns not captured by variables contained in the SRTR database.
