Embolization of spontaneous portosystemic shunts for management of severe persistent hepatic encephalopathy


  • Potential conflict of interest: Nothing to report.

To the Editor:

We read with great interest the recent article by Laleman et al.[1] on the efficacy and safety of embolization of large portosystemic shunts for management of refractory hepatic encephalopathy (HE).

In our experience at the Mayo Clinic, Rochester, Minnesota (2003-2012), 23 patients with severe persistent HE were found to have large spontaneous portosystemic shunts on cross-sectional imaging and were considered for embolization of the shunt. Of these, eight patients did not undergo the procedure (technically not feasible [four patients], shunts too large to occlude [two patients], very little flow through shunts [two patients]; contraindicated due to advanced liver disease [MELD >25] in two patients; perceived risk of postembolization variceal bleeding in a patient who was a Jehovah's witness [one patient]; and acute-on-chronic kidney injury [one patient]). A total of 15 patients (mean age ± SD, 64 ± 12 years; eight males) underwent successful embolization (using coils [13 patients], Amplatzer Vascular Plug [two patients]). Ten patients (67%) had cirrhosis (Model for Endstage Liver Disease [MELD] score, median [range]: 13 [9-18]). Eight patients had a large splenorenal shunt, whereas five patients had an inferior or superior mesenteric vein to gonadal or ovarian vein shunt. Eight patients were on lactulose monotherapy (with or without dietary modification); seven patients were on concomitant antibiotics (rifaximin or neomycin), with or without zinc supplementation.

Median follow-up was 6 months (interquartile range, 3-11 months). Overall, 14/15 patients reported significant improvement in HE symptoms within 1 week of the procedure (93.3% immediate effectiveness). Of the 14 patients who responded initially, 13 patients had sustained improvement in HE at 2 months (overall short-term effectiveness, 86.7%) (Table 1). Among 10 patients with cirrhosis (mean age ± SD, 59 ± 10 years; six males), all reported significant improvement in HE symptoms within 1 week of embolization (100% immediate effectiveness), and nine had sustained improvement in HE at 2 months (overall short-term effectiveness, 90%). On long-term follow-up, three patients had mild relapse of HE symptoms (2-18 months, postprocedure) due to progressive hepatic dysfunction with no recurrence of large spontaneous portosystemic shunting. Three patients did not require any long-term HE medications, six were managed with lactulose or rifaximin monotherapy, and one patient required dual therapy. On the other hand, of the five patients (mean age ± SD, 74 ± 8 years; two males) with spontaneous portosystemic shunting without cirrhosis, four patients (80%) had immediate improvement in HE, with no recurrence of encephalopathy and required no HE medications. One patient, a 78 year-old female with unexplained confusion and hyperammonemia, underwent successful embolization with resolution of hyperammonemia, but did not have improvement in her mental status.

Table 1. Baseline Characteristics and Clinical Outcomes After Embolization of Portosystemic Shunt
Patient IDAge at EmbolizationSexEtiology of Underlying Liver DiseasePreprocedureFollow-up, Postembolization (Months)Clinical Improvement In Mental Status, at 2 MonthsPostprocedure (at 2 months)Adverse Events
MELDaMedications for HEMELDMedications for HE
  1. a

    For patients with underlying cirrhosis.

  2. b

    Hyperammonemia resolved, but underlying altered mental status of unclear etiology, persisted.

  3. c

    Improved immediately postprocedure for 6 weeks, then gradual relapse due to progression of underlying liver disease with no signs of recurrence of large portosystemic shunting.

  4. Abbreviations: EtOH, alcoholic liver disease; HCV, hepatitis C virus-associated; HE, hepatic encephalopathy; L, lactulose; MELD, Model of Endstage Liver Disease; Neo, neomycin; NASH, nonalcoholic steatohepatitis; NR, not reported; R, rifaximin; Zn, zinc.

Patients With Cirrhosis
163MNASH14L, Neo, Zn4.7Yes10L, as neededNone
257MNASH10L, R13.4YesNRL as neededNone
365MNASH13L, R6.7Yes13RNew ascites, 6 months
453MEtOH15L, R2.2YesNRRNone
547MEtOH12L, R, Zn2.9Yes10L, RNone
673FCryptogenic9L, Diet6.0Noc17LNone
854FNASH11L, Diet19Yes17NoneNone
949MHCV15L8.4Yes15L, as neededNone
1078FNASH18L, R, Zn3.2Yes12NoneNone
Patients Without Cirrhosis
1471FNoL, R10.9YesNoneNone
1582MNoL, Diet2.4YesNoneHepatic abscess

Overall, the procedure was well tolerated. One patient developed a hepatic abscess within 2 weeks postprocedure, and one patient with cirrhosis developed new onset of ascites 6 months after embolization, which was managed with diuretics (overall, 13.3% complication rate). No variceal bleeding was noted during follow-up. Our findings are comparable to that observed by Laleman et al., who noted 75% improvement in HE-related symptoms after the procedure in 37 patients with cirrhosis and refractory HR with underlying large portosystemic shunts.[1] Of note, we have two additional patients with cirrhosis and refractory HE who underwent surgical ligation of a large patent umbilical vein with excellent results.

In conclusion, in carefully selected patients (MELD scores ≤15), with severe persistent HE and large spontaneous portosystemic shunts (large single portosystemic shunt, absence of advanced hepatic dysfunction, no significant portal hypertensive complications), embolization of shunts may be a safe and effective treatment option.

  • Siddharth Singh, M.D. Patrick S. Kamath, M.D. James C. Andrews, M.D. Michael D. Leise, M.D.

  • Division of Gastroenterology and Hepatology Mayo Clinic Rochester, MN