Original Clinical ScienceçLiver
Pretransplant Portal Vein Recanalization— Transjugular Intrahepatic Portosystemic Shunt in Patients With Complete Obliterative Portal Vein Thrombosis Riad Salem, MD MBA,1,2 Michael Vouche, MD,1 Talia Baker, MD,2 Jose Ignacio Herrero, MD,3,4 Juan Carlos Caicedo, MD,2 Jonathan Fryer, MD,2 Ryan Hickey, MD,1 Ali Habib, MD,1 Michael Abecassis, MD, MBA,2 Felicitas Koller, MD,2 Robert Vogelzang, MD,1 Kush Desai, MD,1 Bartley Thornburg, MD,1 Elias Hohlastos, MD,1 Scott Resnick, MD,1 Robert J. Lewandowski, MD,1 Kent Sato, MD,1 Robert K Ryu, MD,1 Daniel Ganger, MD,5 and Laura Kulik, MD5
Background. Chronic, obliterative portal vein (PV) thrombosis (PVT) represents a relative contraindication to liver transplantation
(LT) in some centers. When PV thromboembolectomy is not feasible, alternative techniques (portacaval hemitransposition, portal arterialization, multivisceral transplantation) are associated with suboptimal outcomes. In cases where a chronically thrombosed PV has become obliterated, we developed PV recanalization (PVR)-transjugular intrahepatic portosystemic shunt (TIPS) to potentiate LT. We evaluated the impact of PVR-TIPS on liver function, transplant eligibility, and long-term outcomes after LT. Methods. Forty-four patients with chronic obliterative main PVT were identified during our institutional LT selection committee. After joint imaging review by transplant surgery/radiology, these patients underwent PVR-TIPS to potentiate transplant eligibility. Patients were followed by hepatology/transplant until LT, and ultimately in posttransplant clinic. The TIPS venography and serial ultrasound/MRI were used subsequently to document PV patency. Results. The main PV (MPV) was completely thrombosed in 17 of 44 (39%) patients; near complete (>95%) occlusion was noted in 27 of 44 (61%) patients. Direct transhepatic and transsplenic punctures were required in 11 of 43 (26%) and 3 of 43 (7%) cases, respectively. Technical success was 43 of 44 (98%) cases. At PVRTIPS completion, persistence of MPV thrombus was noted in 33 of 43 (77%) cases. One-month TIPS venography demonstrated complete resolution of MPV thrombosis in 22 of 29 (76%) without anticoagulation. Thirty-six patients were listed for transplantation; 18 (50%) have been transplanted. Eighty-nine percent MPV patency rate and 82% survival were achieved at 5 years. Conclusions. The PVR-TIPS may be considered for patients with obliterative PVTwho are otherwise appropriate candidates for LT. The high rate of MPV patency post-TIPS placement suggests flow reestablishment as the dominant mechanism of thrombus resolution. (Transplantation 2015;99: 2347–2355)
P
ortal vein (PV) thrombosis (PVT) is a complicating factor of cirrhotic patients that leads to increased intrahepatic resistance, decreased portal velocities, splanchnic vasodilatation, and stagnant flow.1-5 Chronic PVT may present with decompensation of liver disease/variceal bleeding.6 In addition, portal hypertension may result in encephalopathy along with hepatopulmonary or renal syndromes.7 The diagnosis of PVT is made frequently on surveillance scans for hepatocellular carcinoma or upon investigation of an acute decompensation; its discovery may compromise the patient's candidacy for liver transplantation (LT).
Although a patent PV is preferred and typically simplifies transplantation, chronic PVT may in some cases be a contraindication. Although established surgical intraoperative methods for the management of PVT have been used by transplant surgeons, these options are associated with significant morbidity: (a) arterialization of the PV: fibrosis/ cirrhosis/right heart failure; (b) caval transposition: lower extremity venous insufficiency, ascites/variceal bleeding from unrelieved portal hypertension; (c) multivisceral transplantation: inferior survival rates and morbidity; (d) native renoportal anastomosis: limited to patients with
Received 28 December 2014. Revision received 16 February 2015.
J.I.H. is supported by a grant of the Colegio Oficial de Medicos de Navarra.
Accepted 19 February 2015.
The authors declare conflicts of interest.
1
Department of Radiology, Section of Interventional Radiology, Northwestern University, Chicago IL.
Correspondence: Riad Salem, MD, MBA, Department of Radiology, Northwestern University, 676N. St. Clair, Suite 800, Chicago, IL 60611. ([email protected]).
2
Supplemental digital content (SDC) is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (www.transplantjournal.com).
Department of Transplantation, Northwestern University, Chicago, IL.
3
Liver Unit, Clinica Universidad de Navarra, Pamplona, Navarra, Spain.
4
Centro de Investigacion Biomedica en Red de Enfermedades Hepaticas y Digestivas (CIBERehd), Barcelona, Spain.
5
Division of Hepatology, Department of Medicine, Northwestern University, Chicago, IL.
Transplantation
■
November 2015
■
Volume 99
■
Number 11
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0041-1337/15/9911-2347 DOI: 10.1097/TP.0000000000000729
www.transplantjournal.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2347
2348
Transplantation
■
November 2015
■
Volume 99
■
Number 11
a constructed or spontaneous large splenorenal shunts. Notwithstanding these surgical strategies, the extent of chronic PVT (cavernous transformation) and propagation of thrombosis into the splenic vein (SV)/superior mesenteric vein (SMV) may jeopardize candidacy for LT. In cases of completely obliterative chronic PVT, these methods to reestablish portal flow may not even be feasible. Furthermore, the University of Miami group recently published a series confirming that LT involving nonphysiologic portal flow (cavoportal hemitranspositions, renoportal anastomoses) was associated with higher incidence of PV rethrombosis, gastrointestinal (GI) bleeding and worsened survival.8 Hence, the ability to reestablish physiologic native portal flow enabling end-to-end anastomosis at LT represents an attractive scenario that would obviate compromising LT based solely on PVT. Toward this end, our group has taken an aggressive approach to restore main PV flow in patients with chronic PVT (permitting endto-end LT anastomosis) with the intention of improving long-term outcomes. Portal vein recanalization (PVR)-transjugular intrahepatic portosystemic shunt (TIPS) is a concept aimed at restoring main PV (MPV) flow in chronic PVT. The routine application of PVR-TIPS in potential LT patients with chronic obliterative PVT is novel, and in this study, we review the safety/ efficacy of this approach. We also report on the posttransplant outcomes and highlight the unique features of PVR-TIPS, which to our knowledge, have only been reported in limited series.9 METHODS Patient Cohort
This study was approved by Institutional Review Board and was Health Insurance Portability and Accountability Act compliant. Between 2009 and 2014, 44 cirrhotic patients with advanced occlusive PVT exhibiting clinical sequelae of end-stage liver disease were presented at our multidisciplinary liver conference. All patients had been considered high risk for transplant listing because of occlusive main PVT. Patients were subsequently presented at a multidisciplinary conference (radiologists, hepatologists, transplant surgeons) and the determination to proceed with PVR-TIPS was made by consensus. Inclusion criteria were: (1) chronic advanced thrombosis of the PV with or without cavernomatous transformation, (2) patent splenic and SMVs (clot from PVextending to SMV permissible), (3) medically in need of LT (ascites, variceal bleeding, encephalopathy). The aim was to restore main PV flow. Alternatives were considered, including continued observation and anticoagulation. Anticoagulation was not prescribed given its controversial role and risks (bleeding) in this high-risk population. This study represents data from a large comprehensive transplant center with significant expertise in interventional radiology techniques and LT (>1500 liver transplants, >1500 TIPS over 2 decades). Patients who underwent PVR-TIPS were followed and if transplanted, technical surgical details were reviewed. Survival data were closed on June 1, 2014. All patients had decompensated liver disease, and all were considered potentially suitable for transplantation by hepatology/ transplant surgery because of complications of cirrhosis. Technical considerations related to the extent of PVT were the major surgical barriers to listing for LT.
www.transplantjournal.com
Liver Function and Baseline Characteristics
Abdominal imaging and laboratory tests were used to assess baseline characteristics (Child-Turcotte-Pugh, model for end-stage liver disease [MELD] score). Transplant Eligibility and PV Thrombosis Diagnosis/Staging
All patients were eligible for LT as determined by our selections committee, but their candidacy was put into question by our surgeons because of the presence of advanced, obliterative PVT. The diagnosis and extent of the PVT was assessed by computed tomography/MRI and confirmed at venography during PVR-TIPS. Anatomical PVT staging was performed by Jamieson classification (stage 1: PVT, limited to the MPV; stage 2: MPV, thrombus involving the proximal SMV; stage 3: MPV and diffuse thrombosis of the splanchnic venous system with large collaterals; stage 4: same but with fine collaterals).10 Degree of MPV was categorized as complete (100%) or near occlusive (>95%). Cavernous transformation, when present, was also recorded. PVR-TIPS Technical Description
All procedures were performed under general anesthesia. A 10-French vascular sheath was advanced, and right atrial pressure measurements were obtained. Hepatic wedge venography was performed, looking for small, peripheral intrahepatic portal radicals. When catheterization of peripheral portal radicals was not possible from a transjugular route, percutaneous methods were used. In such cases, a long, 21-gauge needle was advanced percutaneously into peripheral radicals. Once a small portal radical was catheterized, access to the MPV was achieved; this was successful using glidewire/catheter in all cases. Through the 10-French sheath, a 10-mm snare was placed in the parenchyma and used as a target. Once the snare was punctured (transjugular route), exchange length glidewire wire access was advanced to the mesenteric system. Splenoportography was performed, assessing the extent of the thrombosis and the presence of varices. Portal vein thrombectomy/angioplasty was performed; lytic agents or anticoagulation were never used. ATIPS (Viator, Gore, Arizona; range 10 4 2 to 10 6 2) stent was used, covering the parenchymal tract and minimal main PV to facilitate LT and sufficient nonstented PV access for endto-end anastomosis. Varices and/or splenorenal shunts were not embolized during the initial PVR-TIPS (concern for lack of mesenteric outflow should the newly recanalized PVocclude). Later in our experience, the transsplenic approach was used for accessing the PV and subsequent targeting for TIPS placement. Under ultrasound guidance, an intrasplenic venous branch was punctured and a guidewire/catheter advanced into the main SV and across the PVT (recanalization). A 10-mm snare was used as a target for needle puncture. The MPV was dilated and thrombectomized (occasionally); a TIPS was placed (2 cm of peripheral PV stented). Once complete, transsplenic tract was coil embolized. Post–PVR-TIPS systemic anticoagulation was never used. Persistent thrombus on completion splenoportogram at the end of PVR-TIPS was purposefully not stented (to preserve long unstented PV for LT). After PVR-TIPS, patients were re-reviewed, and once portal flow was confirmed, listed for LT. In summary, the specific definition of PVR-TIPS includes: short-length TIPS stent (not covering the MPV),
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
2349
TABLE 1.
Statistical Analysis
Baseline characteristics
Continuous variables were presented as median/range; categorical variables as number/percentages. Kaplan-Meier curves of survival and patency of the MPV post-PVR-TIPS were obtained. Wilcoxon tests comparing MELD scores and liver function pre–/post–PVR-TIPS were performed. A P value less than 0.05 was considered significant.
Total (N = 44)
Age, y Sex Etiology of liver disease
PVT stage
Cavernoma Manifestations of cirrhosis
Status main portal vein on imaging (CT/MRI) Creatinine Platelets INR Child-Turcotte-Pugh score
Median Range Male Female NASH HCV HBV Autoimmune Alcohol Regenerative nodular hyperplasia Alpha-1 antitrypsin deficiency Hemochromatosis Primary sclerosing cholangitis Cryptogenic 1 2 3 4 Yes No Ascites Hydrothorax Spontaneous bacterial peritonitis Hepatorenal syndrome Gastro-intestinal bleeding Hepatic encephalopathy Hepatocellular carcinoma Hepatopulmonary syndrome complete thrombosis near complete (>95%) thrombosis
MELD score
10 ≤10 11-14 15-18 >18
57 22-75 25 19 13 9 1 7 4 2 1 2 2 3 22 16 6 0 13 31 32 4 2 4 20 20 3 1 17 27 0.95 (0.41-5.9) 62.5 (34-284) 1.4 (1.0-2.4) 10 13 13 8 15 19 4 6
CT indicates computed tomography; INR, international normalized ratio; NASH, nonalcoholic steatohepatitis.
mechanical thrombectomy (occasionally) and angioplasty, listing by the transplant team and ultimately LT. Figures S1-S12, SDC, http://links.lww.com/TP/B148, demonstrate a case example of the PVR-TIPS technique using the transsplenic approach. Clinical Follow-Up and LT
PVR-TIPS adverse events and post-transplant outcomes were assessed. PV patency was assessed by routine ultrasonography and confirmed by direct TIPS interrogation.11 Untransplanted patients underwent TIPS venography at 1, 3, and 6 months to embolize varices and/or splenorenal shunts. Abnormal findings on ultrasound prompted a TIPS venogram with or without intervention.
RESULTS Baseline Characteristics
Table 1 lists baseline characteristics. Forty-four patients were treated by PVR-TIPS; technical success was 43 (98%). Portal vein thrombosis was limited to the MPV in 22 of 44 (50%) patients, extended to the SMV in 16 of 44 (36%) patients, and involved the complete portal system but with large collateral vessels in 6 of 44 (14%) patients. The MPV thrombosis was complete (100%) in 17 of 44 (39%) patients, near complete (>95%) in 27 (61%) patients. Cavernomas were noted in 13/44 (30%) patients. While MELD score was ≤18 in 38/44 (86%), most patients were ChildTurcotte-Pugh B or C (N = 34; 77%), justifying this approach before LT (Table 1). Mean pre– and post–PVR-TIPS pressure gradients were 19 (confidence interval [CI]: 17-21) and 9 (CI: 7-11), respectively (P < 0.0001). Treatment Characteristics
Although the transjugular approach was used in the majority of cases, direct transhepatic puncture of an intrahepatic PV was required in 11 of 43 (26%). The transsplenic approach (a later technical development) was used in 3 cases. Follow-up TIPS venogram demonstrated 1 case of arterioportal fistula and 9 cases of moderate TIPS narrowing corrected by embolization (fistula) or dilatation (stenosis). Three cases of recurrent PV or TIPS thrombosis were noted. Two were TABLE 2.
Treatment characteristics and adverse events Total (n = 43a)
Required transhepatic puncture 11/43 (26%) Required transplenic puncture 3/43 (7%) Persistence of thrombus at 33/43 (77%) the end of the procedure Status of PV thrombus Complete resolution of thrombus 22/29 (76%) at FU venogramb Persistent partial thrombosis 7/29 (24%) Adverse events at anytime Right heart failure 1 during follow-up Minor hemoperitoneum 1 Minor contrast extravasation 4 Pulmonary embolus 0 Skin radiation burn 1 Right upper quadrant pain 3 TIPS/MPV rethrombosis 3c TIPS stenosis 9 Arterioportal fistula 1 Transient encephalopathy 7 Recurrent encephalopathy 2 Fever 2 Transient cardiac hypokinesis 1 a b c
1 TIPS placement failure. 14 1-month venograms not available (eg, patient underwent transplant). 2 successfully recanalyzed, 1 partially rethrombosed.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2350
Transplantation
■
November 2015
■
Volume 99
■
Number 11
www.transplantjournal.com
recanalized, and one showed partial thrombosis not treated until LT (80% persistent thrombus, end-to-end anastomosis was possible). Seven patients presented with transient hepatic encephalopathy; one of these patients required a TIPS narrowing (Table 2) without improvement (noncompliant with lactulose). None of the patients with a history of variceal hemorrhage rebled; patients with ascites were well controlled with PVR-TIPS. PVR-TIPS Outcomes
One patient died after 11 days because of right heart failure. Of the initial 44 patient cohort initially ineligible for transplant at our center because of PVT, 18 of 36 ultimately listed patients (50%) were transplanted (at data closure) after PVR-TIPS with median PVR-TIPS to transplant of 75 days (3-1346). Eighteen (41%) other patients were eligible for transplantation; 2 of these patients died before transplantation (right heart failure 5 month after TIPS, GI/peritoneal bleeding from coagulopathy 1.5 month after TIPS) (Figure 1). The 6 remaining patients were ultimately deemed not to be candidates by hepatology/transplant surgery and denied listing for multifactorial causes. By Kaplan-Meier analyses, PV patency was 89% and survival 82% at 5 years (Figures 2 and 3). Despite being present in 77% (33/43) of the cases at completion portography of PVR-TIPS, the 1-month TIPS venogram demonstrated complete PVT resolution in 76% (23/29) of the cases without intervention/anticoagulation. Technical Adverse Events
Periprocedural adverse events included 1 minor hemoperitoneum not requiring transfusion. No adverse events required
FIGURE 2. Kaplan-Meier patency of main portal vein.
surgical intervention. No pulmonary embolus or stroke was reported. There were no episodes of splenic hematoma or bleeding requiring embolization or splenectomy in the transsplenic cases. MELD score/liver function
The MELD scores (median [range]) were not altered 1 (12 [3-46]), 3 (12 [6-30]) , and 6 months (11 [4-17]) after the TIPS placement (P > 0.05). Three patients acutely decompensated (MELD increase 13 to 41, 23 to 47, and 14 to 25) and were successfully transplanted within 1 month. Another patient exhibited MELD decrease from 36 to 25 in the first month. Total bilirubin level (median [range]) slightly increased after 1 month (from 1.7 [0.3-37.2] to 2.4 [0.6-53.5],
FIGURE 1. Study flow chart.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
2351
secure suture placement and significant elasticity allowing expected accommodation of growth factor. Posttransplant Outcomes
Table 3 provides detailed information about transplanted patients. Eighteen of the 36 (50%) listed patients have been transplanted: 7 within 2 months; 4 at 2 to 3 months; and 7 after 4, 4.2, 7, 29, 30, 34, and 45 months, respectively. There were 4 live and 14 cadaveric donors. At transplantation, no significant PVT was noted except in 1 case; surgical thromboembolectomy was possible permitting uneventful end-to-end portal anastomoses. One interposition vascular graft for a living donor was necessary for a short PV. Three transplanted patients died at day 30 (anoxic encephalopathy), day 35 (small for size syndrome), and 34 months (unknown etiology). Five-year survival rate was 82% (Figure 3). FIGURE 3. Kaplan-Meier overall survival.
P = 0.011); this change was not clinically relevant (Figure 4). Albumin, international normalized ratio, and creatinine were also not changed (Figure 5). Surgical Findings at Transplantation
Post–PVR-TIPS hepatectomy was approached in the standard fashion. The porta hepatis was skeletonized and PV isolated for clamping. In this series, no excessive fibrotic reaction was encountered at the level of the extrahepatic PV; the dissections were uncomplicated. Normal appearing, normal walled PVs were encountered in all cases after clamping/transecting the PVat the bifurcation. In most cases, no residual intraluminal clot was found. In 1 case, very mild residual intraluminal clot was encountered; an uncomplicated thromboembolectomy was performed using standard eversion technique. The new liver was sewn in using cavoplasty technique, PV continuity was restored after an end-to-end anastomosis (running 5-0 prolene/large growth factor) was performed. The quality of the PV post–PVR-TIPS was found to be excellent with easy,
DISCUSSION Portal vein thrombosis is a challenging problem in patients with cirrhosis. The presence of extensive, obliterative PVT makes for challenging LT and hampers/prevents placement of an effective/durable TIPS. Our institutional multidisciplinary group believed the guarded prognosis warranted an aggressive attempt to recanalize the PV and potentially permit standard end-to-end portal anastomosis at LT. Based on our extensive experience with endovenous recanalization methods, we devised and implemented the techniques which currently constitutes PVR-TIPS. Safety, Efficacy, and Transplantability
Our results demonstrate acceptable safety/efficacy of PVRTIPS. Of the 44 patients considered technically challenging for LT because of chronic obliterative PVT, including 17 of 44 with complete (100%) thrombosis of the MPV, there was 1 technical failure and 1 periprocedural death from right heart failure. Extensive cardiac work-up was within normal limits in this latter patient.12 The MPV patency and/or successful LT were obtained in most patients. These results are
FIGURE 4. Change in bilirubin after PVR-TIPS (10 for enhanced display). Max value for baseline, 1 month and 3 months 372, 535, and 121.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2352
Transplantation
■
November 2015
■
Volume 99
■
Number 11
www.transplantjournal.com
FIGURE 5. Change in creatinine, MELD, albumin, and international normalized ratio (INR) after PVR-TIPS.
consistent with other series reported in the literature.13-17 The use of short TIPS stent was favored, leaving a sufficiently long length (4-5 cm) uncovered MPV permitting end-to-end anastomosis. Most notably for LT, PVR-TIPS frequently
transformed the PV from a small, virtually nonexistent sclerotic channel/fibrotic cord, to a vein of normal diameter without thickening, induration or scarring which was readily sewn to the donor PV without difficulty. This
TABLE 3.
Characteristics of transplanted patients post–PVR-TIPS
Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Cadaveric versus living-related
Time to transplant, mo
MELD at TIPS
MELD at transplantation
Cadaveric Cadaveric Cadaveric Living Cadaveric Cadaveric Cadaveric Cadaveric Cadaveric Cadaveric Cadaveric Living Cadaveric Living Cadaveric Cadaveric Cadaveric Living
0.10 0.40 0.43 0.70 0.93 1.00 1.30 2.33 2.33 2.67 2.80 4.00 4.23 7.03 29.3 29.9 33.8 44.9
11 32 24 10 14 14 18 8 39 9 15 18 12 14 14 12 16 12
25 >40 >40 8 >40 31 28 29 >40 30 12 22 28 10 21 22 16 17
Calculated MELD if upgraded
Comments
Donation after cardiac death Liver/kidney Split liver Acute Decompensation 25 14
Hepatocellular carcinoma Hepatocellular carcinoma Repeat Liver/Kidney Stable MELD for 2 mo, then drastic increase to 30 Liver declined by other center
14
Hepatocellular carcinoma
16
Donation after cardiac death Hepatocellular carcinoma, blood group B Blood group B
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
operative finding was confirmed in 18 transplanted patients after PVR-TIPS. To ensure that any hepatic decompensation could be rescued with transplantation, all 44 patients had their LT workup completed before PVR-TIPS. This was important for the 3 patients that decompensated acutely and subsequently transplanted. Otherwise, there were no significant adverse events despite percutaneous and transsplenic access in this patient cohort with thrombocytopenia. Technical Considerations
The TIPS placement is usually indicated in cases of cirrhosis with complications of portal hypertension. Guidelines recommend TIPS be avoided in patients with MELD score greater than 18.18 For this reason, we performed PVR-TIPS only in patients who had been determined to otherwise be LT candidates. Whereas our initial experience with PVRTIPS included only patients with extensive main PVT, our success with these patients led us to include patients with extensive PVT not yet completely occlusive (>95%). Our approach has emphasized that patients with a broad range of liver function may be candidates for this technique. Six of 44 patients had a MELD score greater than 18. Of these, 4 underwent PVR-TIPS following a specific request by surgery to restore PV flow before listing (1 decompensated and underwent LT weeks after PVR-TIPS), 1 patient had a significant MELD score improvement but died 1.5 months later from bleeding and coagulopathy, and 1 remains on the transplant list. We recognize some data suggest anticoagulation results in PVT recanalization in 40% to 60%; we are not sure these findings apply to complete (100%) thrombosis, not to mention the inherent risks of anticoagulation.19,20 Our study shows that liver function was preserved in the majority of patients after PVR-TIPS. However, some exhibited hepatic decompensation. Hence, it appears important to consider all patients as potential LT candidates before PVR-TIPS. Adverse events were successfully managed. Conceptually, the primary aim of PVR-TIPS is to facilitate surgery during LT rather than improve liver function. The 1-month venograms yielded 1 significant counterintuitive finding. Although 77% of PVR-TIPS patients exhibited persistent thrombus at the end of the procedure, 76% of these had complete resolution at 1 month despite lack of anticoagulation. This is counter to the classical teaching of high rethrombosis rates in unstented recanalized veins. This observation raises a question regarding the need for aggressive thrombectomy versus the benefit of reestablishing flow through creation of the TIPS. Thus, whereas we pursued an aggressive angioplasty/thrombectomy approach at the time of PVR-TIPS, the continued resolution of PVT suggests that once PV venous flow is reestablished, high portal/mesenteric pressures generate sufficient flow resulting in clot dissolution and PV patency. The small group of patients in whom we used the transsplenic approach to the occluded PV is deserving of further discussion. This technique was considered after favorable experiences using a combination of transjugular and transhepatic approach to the PV. We had observed that these methods often required long procedure times. Using the transsplenic approach, overall procedure (not fluoroscopic) times were 45, 90, and 120 minutes. These cases were remarkably straightforward and simplified our earlier approaches;
2353
the SV was readily catheterized through the spleen. Crossing the chronically occluded PV required minimal effort (glidewire, catheter). The PVR-TIPS was then carried out by targeting a snare placed in a PV branch. Given the technical simplicity of the transsplenic approach, this is now our initial method of choice, with routine splenic tract embolization. Comparison With Other Series
Our study differs from others in many respects. Luca et al17 treated patients with partial PVT; in our series, patients had either complete (100%) or near-complete (>95%) luminal occlusion. They also excluded patients with cavernomas; it was present in 30% of our patients. We did however confirm their finding of significant resolution of PVT after TIPS placement. We also do agree with their recommendation not to cover the MPV. Han et al16 included patients with acute PVT; in our series, we only treated patients with chronic PVT in the transplant setting. They exhibited 75% success. They too used a transsplenic approach when necessary. Finally, in contradistinction to our approach, they anticoagulated patients after TIPS; our findings suggest this is not needed. Senzolo et al14 reported successfully using TIPS in a heterogeneous cohort of 28 patients, only one of which was LT the primary indication. In our series, preparation for LT was the primary objective. Comparison With Anticoagulation
Delgado et al20 published a series studying anticoagulation for acute/chronic mesenteric thrombosis including the SMV; they excluded patients with cavernomas. This differs from our series where we focused specifically chronic MPV thrombosis and included cavernomas. Our series also focused on the intent of rendering the PV anatomy favorable for LT. Hence, comparing anticoagulation in the acute setting with PVR-TIPS for chronic MPV thrombosis appears spurious. Anticoagulation in this patient population also carries an increased risk of bleeding and rethrombosis.20 The role of anticoagulation in the management of PVT remains controversial.21,22 Comparison to University of Miami and Other
Hibi et al8 recently published a series extremely relevant to our study. The authors compared the clinical outcomes of PVT patients able to achieve physiologic restoration of flow through end-to-end PV anastomoses to those with severe PVT not permitting restoration of native portal flow. Patients in the latter group were those with complete/near complete obliterative main PVT, just as in our series presented herein. They reported that nonphysiologic PV inflow (caval hemitransposition, renoportal anastomosis, PV arterialization) at transplantation exhibited higher rates of rethrombosis, GI bleeding, intraoperative blood loss, highest in-hospital mortality rates, as well as lower survival compared with physiologic PV flow restoration. In our opinion, their findings support the importance of reestablishing native PV flow before LT; our posttransplant survival supports this contention. The PVRTIPS simplified LT by reestablishing main portal flow; this was confirmed at surgery in 18 transplanted cases to date, where end-to-end anastomoses were achieved. In our series, PVR-TIPS resulted in high PV patency and excellent survival. Alves et al23 also used surgical techniques (shunts, grafts) to deal with PVT in 29 patients, achieving 75% actuarial
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2354
Transplantation
■
November 2015
■
Volume 99
■
Number 11
survival. Robles et al24 achieved 69% 5-year survival in 32 patients treated with thromboendovenectomy. Lessons Learned
Several lessons or clinical “pearls” deserve highlighting. First, the presence of cavernoma did not impede our success rate, especially when the transsplenic route was used. Second, identifying a peripheral intra-SV in direct line with the main SV was essential for straight access to the thrombosed PV. Third, establishing portal flow despite the presence of mild residual thrombus in the PV at PVR-TIPS termination was satisfactory; complete removal was not necessary. Fourth, a stent extending only 1 to 2 cm into the PV is recommended, maximizing unstented PV length for LT. Fifth, the unexpected finding of normal-walled PVs at transplantation suggests stenting the main PV is not necessary; this finding challenges the previously long-held belief that chronically thrombosed veins require stenting. Finally, the gastroesophageal varices and splenorenal shunts did not require embolization during the first procedure. Rather, these were reserved for the 1-month TIPS venogram, where PV flow status and embolization was performed.
www.transplantjournal.com
may be an attractive approach for live donors when vessels may not be available for venous conduits. We also believe PVR-TIPS provides transplant surgeons with an added tool (along with surgical techniques) in the management of chronic PVT before LT. Finally, although the generalizability of this technique may be open for debate, we have recently updated a new (and readily transferrable) method using the transsplenic route; this is now our default approach for PVR-TIPS.25 CONCLUSIONS Pretransplant PVR in patients with chronic, obliterative PVT provides surgeons an additional option to address PVT and obviate the need for complex surgical grafts. Our study demonstrates PVR-TIPS to be an effective intervention with a favorable safety profile. This study explores a procedure that allows an entirely new subpopulation (patients deemed untransplantable by some centers) access to LT without the risks of anticoagulation. Once these data are confirmed, we anticipate that the management of cirrhotic patients with chronic PVT will be positively impacted.
Strengths and Limitations
REFERENCES
We recognize this represents single-center experience with high level technical expertise. It does represent the largest series of an aggressive approach to patients in the transplant setting. They were not compared to a control population (such as anticoagulation); these patients had cavernomas and GI bleeds, conditions where anticoagulation is not routinely used. However, patency rates and outcomes in our series exceed those described in studies of anticoagulation. The surgical perspective is also important because, whereas most transplant surgeons would posit they can technically overcome PVT, most would agree that a patent PV during LT is reassuring. Other than a simple eversion thrombectomy, surgical procedures (graft transposition, cavoportal hemitransposition, multivisceral transplantation) are generally associated with poor outcomes.8 It cannot be overemphasized that the Miami group data are the most appropriate comparison cohort for our study group—these were not cases of routine advanced PVT but rather PVT with obliteration of the main portal venous channel. Because this series has no comparator, survival outcomes should be interpreted with caution. We acknowledge that this is a unique subset of patients with low baseline MELD scores and compensated liver disease with chronic obliterative PVT. Finally, we recognize the technical limitations of PVRTIPS for patients with thrombosis of the entire portal/ mesenteric system, where this technique may not be effective given the absence of PV inflow from the SMV or SV, significant contributing factors for PVR-TIPS patency. There are strengths. This is the first study focusing on PVR-TIPS with the specific intent of physiologic portal flow restoration supported by surgical confirmation at LT in 50% (N = 18/36) of the patients. It is also the first moderately sized cohort study to focus on this major clinical challenge; complete/near-complete obliterative MPV thrombosis in cirrhotics. We have mature follow-up permitting assessment of PV patency with ultrasound/cross-sectional imaging/venography. The patients exhibited clinical manifestations of cirrhosis, justifying the need for TIPS and ultimate LT. The PVR-TIPS
1. Ponziani FR, Zocco MA, Campanale C, et al. Portal vein thrombosis: insight into physiopathology, diagnosis, and treatment. World J Gastroenterol. 2010;16:143–155. 2. Amitrano L, Guardascione MA, Brancaccio V, et al. Risk factors and clinical presentation of portal vein thrombosis in patients with liver cirrhosis. J Hepatol. 2004;40:736–741. 3. Luca A, Caruso S, Milazzo M, et al. Natural course of extrahepatic nonmalignant partial portal vein thrombosis in patients with cirrhosis. Radiology. 2012;265:124–132. 4. Maruyama H, Okugawa H, Takahashi M, et al. De novo portal vein thrombosis in virus-related cirrhosis: predictive factors and long-term outcomes. Am J Gastroenterol. 2013;108:568–574. 5. Zocco MA, Di Stasio E, De Cristofaro R, et al. Thrombotic risk factors in patients with liver cirrhosis: correlation with MELD scoring system and portal vein thrombosis development. J Hepatol. 2009;51:682–689. 6. Francoz C, Valla D, Durand F. Portal vein thrombosis, cirrhosis, and liver transplantation. J Hepatol. 2012;57:203–212. 7. Garcia-Pagan JC, Valla DC. Portal vein thrombosis: a predictable milestone in cirrhosis? J Hepatol. 2009;51:632–634. 8. Hibi T, Nishida S, Levi DM, et al. When and why portal vein thrombosis matters in liver transplantation: a critical audit of 174 cases. Ann Surg. 2014;259:760–766. 9. Bauer J, Johnson S, Durham J, et al. The role of TIPS for portal vein patency in liver transplant patients with portal vein thrombosis. Liver Transpl. 2006;12:1544–1551. 10. Jamieson NV. Changing perspectives in portal vein thrombosis and liver transplantation. Transplantation. 2000;69:1772–1774. 11. Rossle M. TIPS: 25 years later. J Hepatol. 2013;59:1081–1093. 12. Bilbao JI, Quiroga J, Herrero JI, Benito A. Transjugular intrahepatic portosystemic shunt (TIPS): current status and future possibilities. Cardiovasc Intervent Radiol. 2002;25:251–269. 13. Van Ha TG, Hodge J, Funaki B, et al. Transjugular intrahepatic portosystemic shunt placement in patients with cirrhosis and concomitant portal vein thrombosis. Cardiovasc Intervent Radiol. 2006;29:785–790. 14. Senzolo M, Tibbals J, Cholongitas E, et al. Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with and without cavernous transformation. Aliment Pharmacol Ther. 2006;23:767–775. 15. Perarnau JM, Baju A, D'Alteroche L, et al. Feasibility and long-term evolution of TIPS in cirrhotic patients with portal thrombosis. Eur J Gastroenterol Hepatol. 2010;22:1093–1098. 16. Han G, Qi X, He C, et al. Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with symptomatic portal hypertension in liver cirrhosis. J Hepatol. 2011;54:78–88. 17. Luca A, Miraglia R, Caruso S, et al. Short- and long-term effects of the transjugular intrahepatic portosystemic shunt on portal vein thrombosis in patients with cirrhosis. Gut. 2011;60:846–852.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
18. Boyer TD, Haskal ZJ. The role of transjugular intrahepatic portosystemic shunt (TIPS) in the management of portal hypertension: update 2009. Hepatology. 2010;51:306. 19. D'Avola D, Bilbao JI, Zozaya G, et al. Efficacy of transjugular intrahepatic portosystemic shunt to prevent total portal vein thrombosis in cirrhotic patients awaiting for liver transplantation. Transplant Proc. 2012;44:2603–2605. 20. Delgado MG, Seijo S, Yepes I, et al. Efficacy and safety of anticoagulation on patients with cirrhosis and portal vein thrombosis. Clin Gastroenterol Hepatol. 2012;10:776–783. 21. Huard G, Bilodeau M. Management of anticoagulation for portal vein thrombosis in individuals with cirrhosis: a systematic review. Int J Hepatol. 2012;2012:672986.
2355
22. Congly S, Lee S. Portal vein thrombosis: should anticoagulation be used? Curr Gastroenterol Rep. 2013;15:1–7. 23. Alves J, Boin I, Cholongitas E, et al. Surgical technique used for portal vein thrombosis when thrombectomy is not possible during liver transplantation. Surg Sci. 2011;2:767–775. 24. Robles R, Fernandez JA, Hernandez Q, et al. Eversion thromboendovenectomy in organized portal vein thrombosis during liver transplantation. Clin Transplant. 2004;18:79–84. 25. Habib A, Desai K, Hickey R, et al. Portal vein recanalizationtransjugularintrahepatic portosystemic shunt using the transsplenic approach to achieve transplant candidacy in patients with chronic portal vein thrombosis. J Vasc Interv Radiol. 2015;S1051-0443:01183.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Pretransplant Portal Vein Recanalization— Transjugular Intrahepatic Portosystemic Shunt in Patients With Complete Obliterative Portal Vein Thrombosis Riad Salem, MD MBA,1,2 Michael Vouche, MD,1 Talia Baker, MD,2 Jose Ignacio Herrero, MD,3,4 Juan Carlos Caicedo, MD,2 Jonathan Fryer, MD,2 Ryan Hickey, MD,1 Ali Habib, MD,1 Michael Abecassis, MD, MBA,2 Felicitas Koller, MD,2 Robert Vogelzang, MD,1 Kush Desai, MD,1 Bartley Thornburg, MD,1 Elias Hohlastos, MD,1 Scott Resnick, MD,1 Robert J. Lewandowski, MD,1 Kent Sato, MD,1 Robert K Ryu, MD,1 Daniel Ganger, MD,5 and Laura Kulik, MD5
Background. Chronic, obliterative portal vein (PV) thrombosis (PVT) represents a relative contraindication to liver transplantation
(LT) in some centers. When PV thromboembolectomy is not feasible, alternative techniques (portacaval hemitransposition, portal arterialization, multivisceral transplantation) are associated with suboptimal outcomes. In cases where a chronically thrombosed PV has become obliterated, we developed PV recanalization (PVR)-transjugular intrahepatic portosystemic shunt (TIPS) to potentiate LT. We evaluated the impact of PVR-TIPS on liver function, transplant eligibility, and long-term outcomes after LT. Methods. Forty-four patients with chronic obliterative main PVT were identified during our institutional LT selection committee. After joint imaging review by transplant surgery/radiology, these patients underwent PVR-TIPS to potentiate transplant eligibility. Patients were followed by hepatology/transplant until LT, and ultimately in posttransplant clinic. The TIPS venography and serial ultrasound/MRI were used subsequently to document PV patency. Results. The main PV (MPV) was completely thrombosed in 17 of 44 (39%) patients; near complete (>95%) occlusion was noted in 27 of 44 (61%) patients. Direct transhepatic and transsplenic punctures were required in 11 of 43 (26%) and 3 of 43 (7%) cases, respectively. Technical success was 43 of 44 (98%) cases. At PVRTIPS completion, persistence of MPV thrombus was noted in 33 of 43 (77%) cases. One-month TIPS venography demonstrated complete resolution of MPV thrombosis in 22 of 29 (76%) without anticoagulation. Thirty-six patients were listed for transplantation; 18 (50%) have been transplanted. Eighty-nine percent MPV patency rate and 82% survival were achieved at 5 years. Conclusions. The PVR-TIPS may be considered for patients with obliterative PVTwho are otherwise appropriate candidates for LT. The high rate of MPV patency post-TIPS placement suggests flow reestablishment as the dominant mechanism of thrombus resolution. (Transplantation 2015;99: 2347–2355)
P
ortal vein (PV) thrombosis (PVT) is a complicating factor of cirrhotic patients that leads to increased intrahepatic resistance, decreased portal velocities, splanchnic vasodilatation, and stagnant flow.1-5 Chronic PVT may present with decompensation of liver disease/variceal bleeding.6 In addition, portal hypertension may result in encephalopathy along with hepatopulmonary or renal syndromes.7 The diagnosis of PVT is made frequently on surveillance scans for hepatocellular carcinoma or upon investigation of an acute decompensation; its discovery may compromise the patient's candidacy for liver transplantation (LT).
Although a patent PV is preferred and typically simplifies transplantation, chronic PVT may in some cases be a contraindication. Although established surgical intraoperative methods for the management of PVT have been used by transplant surgeons, these options are associated with significant morbidity: (a) arterialization of the PV: fibrosis/ cirrhosis/right heart failure; (b) caval transposition: lower extremity venous insufficiency, ascites/variceal bleeding from unrelieved portal hypertension; (c) multivisceral transplantation: inferior survival rates and morbidity; (d) native renoportal anastomosis: limited to patients with
Received 28 December 2014. Revision received 16 February 2015.
J.I.H. is supported by a grant of the Colegio Oficial de Medicos de Navarra.
Accepted 19 February 2015.
The authors declare conflicts of interest.
1
Department of Radiology, Section of Interventional Radiology, Northwestern University, Chicago IL.
Correspondence: Riad Salem, MD, MBA, Department of Radiology, Northwestern University, 676N. St. Clair, Suite 800, Chicago, IL 60611. ([email protected]).
2
Supplemental digital content (SDC) is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (www.transplantjournal.com).
Department of Transplantation, Northwestern University, Chicago, IL.
3
Liver Unit, Clinica Universidad de Navarra, Pamplona, Navarra, Spain.
4
Centro de Investigacion Biomedica en Red de Enfermedades Hepaticas y Digestivas (CIBERehd), Barcelona, Spain.
5
Division of Hepatology, Department of Medicine, Northwestern University, Chicago, IL.
Transplantation
■
November 2015
■
Volume 99
■
Number 11
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0041-1337/15/9911-2347 DOI: 10.1097/TP.0000000000000729
www.transplantjournal.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2347
2348
Transplantation
■
November 2015
■
Volume 99
■
Number 11
a constructed or spontaneous large splenorenal shunts. Notwithstanding these surgical strategies, the extent of chronic PVT (cavernous transformation) and propagation of thrombosis into the splenic vein (SV)/superior mesenteric vein (SMV) may jeopardize candidacy for LT. In cases of completely obliterative chronic PVT, these methods to reestablish portal flow may not even be feasible. Furthermore, the University of Miami group recently published a series confirming that LT involving nonphysiologic portal flow (cavoportal hemitranspositions, renoportal anastomoses) was associated with higher incidence of PV rethrombosis, gastrointestinal (GI) bleeding and worsened survival.8 Hence, the ability to reestablish physiologic native portal flow enabling end-to-end anastomosis at LT represents an attractive scenario that would obviate compromising LT based solely on PVT. Toward this end, our group has taken an aggressive approach to restore main PV flow in patients with chronic PVT (permitting endto-end LT anastomosis) with the intention of improving long-term outcomes. Portal vein recanalization (PVR)-transjugular intrahepatic portosystemic shunt (TIPS) is a concept aimed at restoring main PV (MPV) flow in chronic PVT. The routine application of PVR-TIPS in potential LT patients with chronic obliterative PVT is novel, and in this study, we review the safety/ efficacy of this approach. We also report on the posttransplant outcomes and highlight the unique features of PVR-TIPS, which to our knowledge, have only been reported in limited series.9 METHODS Patient Cohort
This study was approved by Institutional Review Board and was Health Insurance Portability and Accountability Act compliant. Between 2009 and 2014, 44 cirrhotic patients with advanced occlusive PVT exhibiting clinical sequelae of end-stage liver disease were presented at our multidisciplinary liver conference. All patients had been considered high risk for transplant listing because of occlusive main PVT. Patients were subsequently presented at a multidisciplinary conference (radiologists, hepatologists, transplant surgeons) and the determination to proceed with PVR-TIPS was made by consensus. Inclusion criteria were: (1) chronic advanced thrombosis of the PV with or without cavernomatous transformation, (2) patent splenic and SMVs (clot from PVextending to SMV permissible), (3) medically in need of LT (ascites, variceal bleeding, encephalopathy). The aim was to restore main PV flow. Alternatives were considered, including continued observation and anticoagulation. Anticoagulation was not prescribed given its controversial role and risks (bleeding) in this high-risk population. This study represents data from a large comprehensive transplant center with significant expertise in interventional radiology techniques and LT (>1500 liver transplants, >1500 TIPS over 2 decades). Patients who underwent PVR-TIPS were followed and if transplanted, technical surgical details were reviewed. Survival data were closed on June 1, 2014. All patients had decompensated liver disease, and all were considered potentially suitable for transplantation by hepatology/ transplant surgery because of complications of cirrhosis. Technical considerations related to the extent of PVT were the major surgical barriers to listing for LT.
www.transplantjournal.com
Liver Function and Baseline Characteristics
Abdominal imaging and laboratory tests were used to assess baseline characteristics (Child-Turcotte-Pugh, model for end-stage liver disease [MELD] score). Transplant Eligibility and PV Thrombosis Diagnosis/Staging
All patients were eligible for LT as determined by our selections committee, but their candidacy was put into question by our surgeons because of the presence of advanced, obliterative PVT. The diagnosis and extent of the PVT was assessed by computed tomography/MRI and confirmed at venography during PVR-TIPS. Anatomical PVT staging was performed by Jamieson classification (stage 1: PVT, limited to the MPV; stage 2: MPV, thrombus involving the proximal SMV; stage 3: MPV and diffuse thrombosis of the splanchnic venous system with large collaterals; stage 4: same but with fine collaterals).10 Degree of MPV was categorized as complete (100%) or near occlusive (>95%). Cavernous transformation, when present, was also recorded. PVR-TIPS Technical Description
All procedures were performed under general anesthesia. A 10-French vascular sheath was advanced, and right atrial pressure measurements were obtained. Hepatic wedge venography was performed, looking for small, peripheral intrahepatic portal radicals. When catheterization of peripheral portal radicals was not possible from a transjugular route, percutaneous methods were used. In such cases, a long, 21-gauge needle was advanced percutaneously into peripheral radicals. Once a small portal radical was catheterized, access to the MPV was achieved; this was successful using glidewire/catheter in all cases. Through the 10-French sheath, a 10-mm snare was placed in the parenchyma and used as a target. Once the snare was punctured (transjugular route), exchange length glidewire wire access was advanced to the mesenteric system. Splenoportography was performed, assessing the extent of the thrombosis and the presence of varices. Portal vein thrombectomy/angioplasty was performed; lytic agents or anticoagulation were never used. ATIPS (Viator, Gore, Arizona; range 10 4 2 to 10 6 2) stent was used, covering the parenchymal tract and minimal main PV to facilitate LT and sufficient nonstented PV access for endto-end anastomosis. Varices and/or splenorenal shunts were not embolized during the initial PVR-TIPS (concern for lack of mesenteric outflow should the newly recanalized PVocclude). Later in our experience, the transsplenic approach was used for accessing the PV and subsequent targeting for TIPS placement. Under ultrasound guidance, an intrasplenic venous branch was punctured and a guidewire/catheter advanced into the main SV and across the PVT (recanalization). A 10-mm snare was used as a target for needle puncture. The MPV was dilated and thrombectomized (occasionally); a TIPS was placed (2 cm of peripheral PV stented). Once complete, transsplenic tract was coil embolized. Post–PVR-TIPS systemic anticoagulation was never used. Persistent thrombus on completion splenoportogram at the end of PVR-TIPS was purposefully not stented (to preserve long unstented PV for LT). After PVR-TIPS, patients were re-reviewed, and once portal flow was confirmed, listed for LT. In summary, the specific definition of PVR-TIPS includes: short-length TIPS stent (not covering the MPV),
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
2349
TABLE 1.
Statistical Analysis
Baseline characteristics
Continuous variables were presented as median/range; categorical variables as number/percentages. Kaplan-Meier curves of survival and patency of the MPV post-PVR-TIPS were obtained. Wilcoxon tests comparing MELD scores and liver function pre–/post–PVR-TIPS were performed. A P value less than 0.05 was considered significant.
Total (N = 44)
Age, y Sex Etiology of liver disease
PVT stage
Cavernoma Manifestations of cirrhosis
Status main portal vein on imaging (CT/MRI) Creatinine Platelets INR Child-Turcotte-Pugh score
Median Range Male Female NASH HCV HBV Autoimmune Alcohol Regenerative nodular hyperplasia Alpha-1 antitrypsin deficiency Hemochromatosis Primary sclerosing cholangitis Cryptogenic 1 2 3 4 Yes No Ascites Hydrothorax Spontaneous bacterial peritonitis Hepatorenal syndrome Gastro-intestinal bleeding Hepatic encephalopathy Hepatocellular carcinoma Hepatopulmonary syndrome complete thrombosis near complete (>95%) thrombosis
MELD score
10 ≤10 11-14 15-18 >18
57 22-75 25 19 13 9 1 7 4 2 1 2 2 3 22 16 6 0 13 31 32 4 2 4 20 20 3 1 17 27 0.95 (0.41-5.9) 62.5 (34-284) 1.4 (1.0-2.4) 10 13 13 8 15 19 4 6
CT indicates computed tomography; INR, international normalized ratio; NASH, nonalcoholic steatohepatitis.
mechanical thrombectomy (occasionally) and angioplasty, listing by the transplant team and ultimately LT. Figures S1-S12, SDC, http://links.lww.com/TP/B148, demonstrate a case example of the PVR-TIPS technique using the transsplenic approach. Clinical Follow-Up and LT
PVR-TIPS adverse events and post-transplant outcomes were assessed. PV patency was assessed by routine ultrasonography and confirmed by direct TIPS interrogation.11 Untransplanted patients underwent TIPS venography at 1, 3, and 6 months to embolize varices and/or splenorenal shunts. Abnormal findings on ultrasound prompted a TIPS venogram with or without intervention.
RESULTS Baseline Characteristics
Table 1 lists baseline characteristics. Forty-four patients were treated by PVR-TIPS; technical success was 43 (98%). Portal vein thrombosis was limited to the MPV in 22 of 44 (50%) patients, extended to the SMV in 16 of 44 (36%) patients, and involved the complete portal system but with large collateral vessels in 6 of 44 (14%) patients. The MPV thrombosis was complete (100%) in 17 of 44 (39%) patients, near complete (>95%) in 27 (61%) patients. Cavernomas were noted in 13/44 (30%) patients. While MELD score was ≤18 in 38/44 (86%), most patients were ChildTurcotte-Pugh B or C (N = 34; 77%), justifying this approach before LT (Table 1). Mean pre– and post–PVR-TIPS pressure gradients were 19 (confidence interval [CI]: 17-21) and 9 (CI: 7-11), respectively (P < 0.0001). Treatment Characteristics
Although the transjugular approach was used in the majority of cases, direct transhepatic puncture of an intrahepatic PV was required in 11 of 43 (26%). The transsplenic approach (a later technical development) was used in 3 cases. Follow-up TIPS venogram demonstrated 1 case of arterioportal fistula and 9 cases of moderate TIPS narrowing corrected by embolization (fistula) or dilatation (stenosis). Three cases of recurrent PV or TIPS thrombosis were noted. Two were TABLE 2.
Treatment characteristics and adverse events Total (n = 43a)
Required transhepatic puncture 11/43 (26%) Required transplenic puncture 3/43 (7%) Persistence of thrombus at 33/43 (77%) the end of the procedure Status of PV thrombus Complete resolution of thrombus 22/29 (76%) at FU venogramb Persistent partial thrombosis 7/29 (24%) Adverse events at anytime Right heart failure 1 during follow-up Minor hemoperitoneum 1 Minor contrast extravasation 4 Pulmonary embolus 0 Skin radiation burn 1 Right upper quadrant pain 3 TIPS/MPV rethrombosis 3c TIPS stenosis 9 Arterioportal fistula 1 Transient encephalopathy 7 Recurrent encephalopathy 2 Fever 2 Transient cardiac hypokinesis 1 a b c
1 TIPS placement failure. 14 1-month venograms not available (eg, patient underwent transplant). 2 successfully recanalyzed, 1 partially rethrombosed.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2350
Transplantation
■
November 2015
■
Volume 99
■
Number 11
www.transplantjournal.com
recanalized, and one showed partial thrombosis not treated until LT (80% persistent thrombus, end-to-end anastomosis was possible). Seven patients presented with transient hepatic encephalopathy; one of these patients required a TIPS narrowing (Table 2) without improvement (noncompliant with lactulose). None of the patients with a history of variceal hemorrhage rebled; patients with ascites were well controlled with PVR-TIPS. PVR-TIPS Outcomes
One patient died after 11 days because of right heart failure. Of the initial 44 patient cohort initially ineligible for transplant at our center because of PVT, 18 of 36 ultimately listed patients (50%) were transplanted (at data closure) after PVR-TIPS with median PVR-TIPS to transplant of 75 days (3-1346). Eighteen (41%) other patients were eligible for transplantation; 2 of these patients died before transplantation (right heart failure 5 month after TIPS, GI/peritoneal bleeding from coagulopathy 1.5 month after TIPS) (Figure 1). The 6 remaining patients were ultimately deemed not to be candidates by hepatology/transplant surgery and denied listing for multifactorial causes. By Kaplan-Meier analyses, PV patency was 89% and survival 82% at 5 years (Figures 2 and 3). Despite being present in 77% (33/43) of the cases at completion portography of PVR-TIPS, the 1-month TIPS venogram demonstrated complete PVT resolution in 76% (23/29) of the cases without intervention/anticoagulation. Technical Adverse Events
Periprocedural adverse events included 1 minor hemoperitoneum not requiring transfusion. No adverse events required
FIGURE 2. Kaplan-Meier patency of main portal vein.
surgical intervention. No pulmonary embolus or stroke was reported. There were no episodes of splenic hematoma or bleeding requiring embolization or splenectomy in the transsplenic cases. MELD score/liver function
The MELD scores (median [range]) were not altered 1 (12 [3-46]), 3 (12 [6-30]) , and 6 months (11 [4-17]) after the TIPS placement (P > 0.05). Three patients acutely decompensated (MELD increase 13 to 41, 23 to 47, and 14 to 25) and were successfully transplanted within 1 month. Another patient exhibited MELD decrease from 36 to 25 in the first month. Total bilirubin level (median [range]) slightly increased after 1 month (from 1.7 [0.3-37.2] to 2.4 [0.6-53.5],
FIGURE 1. Study flow chart.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
2351
secure suture placement and significant elasticity allowing expected accommodation of growth factor. Posttransplant Outcomes
Table 3 provides detailed information about transplanted patients. Eighteen of the 36 (50%) listed patients have been transplanted: 7 within 2 months; 4 at 2 to 3 months; and 7 after 4, 4.2, 7, 29, 30, 34, and 45 months, respectively. There were 4 live and 14 cadaveric donors. At transplantation, no significant PVT was noted except in 1 case; surgical thromboembolectomy was possible permitting uneventful end-to-end portal anastomoses. One interposition vascular graft for a living donor was necessary for a short PV. Three transplanted patients died at day 30 (anoxic encephalopathy), day 35 (small for size syndrome), and 34 months (unknown etiology). Five-year survival rate was 82% (Figure 3). FIGURE 3. Kaplan-Meier overall survival.
P = 0.011); this change was not clinically relevant (Figure 4). Albumin, international normalized ratio, and creatinine were also not changed (Figure 5). Surgical Findings at Transplantation
Post–PVR-TIPS hepatectomy was approached in the standard fashion. The porta hepatis was skeletonized and PV isolated for clamping. In this series, no excessive fibrotic reaction was encountered at the level of the extrahepatic PV; the dissections were uncomplicated. Normal appearing, normal walled PVs were encountered in all cases after clamping/transecting the PVat the bifurcation. In most cases, no residual intraluminal clot was found. In 1 case, very mild residual intraluminal clot was encountered; an uncomplicated thromboembolectomy was performed using standard eversion technique. The new liver was sewn in using cavoplasty technique, PV continuity was restored after an end-to-end anastomosis (running 5-0 prolene/large growth factor) was performed. The quality of the PV post–PVR-TIPS was found to be excellent with easy,
DISCUSSION Portal vein thrombosis is a challenging problem in patients with cirrhosis. The presence of extensive, obliterative PVT makes for challenging LT and hampers/prevents placement of an effective/durable TIPS. Our institutional multidisciplinary group believed the guarded prognosis warranted an aggressive attempt to recanalize the PV and potentially permit standard end-to-end portal anastomosis at LT. Based on our extensive experience with endovenous recanalization methods, we devised and implemented the techniques which currently constitutes PVR-TIPS. Safety, Efficacy, and Transplantability
Our results demonstrate acceptable safety/efficacy of PVRTIPS. Of the 44 patients considered technically challenging for LT because of chronic obliterative PVT, including 17 of 44 with complete (100%) thrombosis of the MPV, there was 1 technical failure and 1 periprocedural death from right heart failure. Extensive cardiac work-up was within normal limits in this latter patient.12 The MPV patency and/or successful LT were obtained in most patients. These results are
FIGURE 4. Change in bilirubin after PVR-TIPS (10 for enhanced display). Max value for baseline, 1 month and 3 months 372, 535, and 121.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2352
Transplantation
■
November 2015
■
Volume 99
■
Number 11
www.transplantjournal.com
FIGURE 5. Change in creatinine, MELD, albumin, and international normalized ratio (INR) after PVR-TIPS.
consistent with other series reported in the literature.13-17 The use of short TIPS stent was favored, leaving a sufficiently long length (4-5 cm) uncovered MPV permitting end-to-end anastomosis. Most notably for LT, PVR-TIPS frequently
transformed the PV from a small, virtually nonexistent sclerotic channel/fibrotic cord, to a vein of normal diameter without thickening, induration or scarring which was readily sewn to the donor PV without difficulty. This
TABLE 3.
Characteristics of transplanted patients post–PVR-TIPS
Number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Cadaveric versus living-related
Time to transplant, mo
MELD at TIPS
MELD at transplantation
Cadaveric Cadaveric Cadaveric Living Cadaveric Cadaveric Cadaveric Cadaveric Cadaveric Cadaveric Cadaveric Living Cadaveric Living Cadaveric Cadaveric Cadaveric Living
0.10 0.40 0.43 0.70 0.93 1.00 1.30 2.33 2.33 2.67 2.80 4.00 4.23 7.03 29.3 29.9 33.8 44.9
11 32 24 10 14 14 18 8 39 9 15 18 12 14 14 12 16 12
25 >40 >40 8 >40 31 28 29 >40 30 12 22 28 10 21 22 16 17
Calculated MELD if upgraded
Comments
Donation after cardiac death Liver/kidney Split liver Acute Decompensation 25 14
Hepatocellular carcinoma Hepatocellular carcinoma Repeat Liver/Kidney Stable MELD for 2 mo, then drastic increase to 30 Liver declined by other center
14
Hepatocellular carcinoma
16
Donation after cardiac death Hepatocellular carcinoma, blood group B Blood group B
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
operative finding was confirmed in 18 transplanted patients after PVR-TIPS. To ensure that any hepatic decompensation could be rescued with transplantation, all 44 patients had their LT workup completed before PVR-TIPS. This was important for the 3 patients that decompensated acutely and subsequently transplanted. Otherwise, there were no significant adverse events despite percutaneous and transsplenic access in this patient cohort with thrombocytopenia. Technical Considerations
The TIPS placement is usually indicated in cases of cirrhosis with complications of portal hypertension. Guidelines recommend TIPS be avoided in patients with MELD score greater than 18.18 For this reason, we performed PVR-TIPS only in patients who had been determined to otherwise be LT candidates. Whereas our initial experience with PVRTIPS included only patients with extensive main PVT, our success with these patients led us to include patients with extensive PVT not yet completely occlusive (>95%). Our approach has emphasized that patients with a broad range of liver function may be candidates for this technique. Six of 44 patients had a MELD score greater than 18. Of these, 4 underwent PVR-TIPS following a specific request by surgery to restore PV flow before listing (1 decompensated and underwent LT weeks after PVR-TIPS), 1 patient had a significant MELD score improvement but died 1.5 months later from bleeding and coagulopathy, and 1 remains on the transplant list. We recognize some data suggest anticoagulation results in PVT recanalization in 40% to 60%; we are not sure these findings apply to complete (100%) thrombosis, not to mention the inherent risks of anticoagulation.19,20 Our study shows that liver function was preserved in the majority of patients after PVR-TIPS. However, some exhibited hepatic decompensation. Hence, it appears important to consider all patients as potential LT candidates before PVR-TIPS. Adverse events were successfully managed. Conceptually, the primary aim of PVR-TIPS is to facilitate surgery during LT rather than improve liver function. The 1-month venograms yielded 1 significant counterintuitive finding. Although 77% of PVR-TIPS patients exhibited persistent thrombus at the end of the procedure, 76% of these had complete resolution at 1 month despite lack of anticoagulation. This is counter to the classical teaching of high rethrombosis rates in unstented recanalized veins. This observation raises a question regarding the need for aggressive thrombectomy versus the benefit of reestablishing flow through creation of the TIPS. Thus, whereas we pursued an aggressive angioplasty/thrombectomy approach at the time of PVR-TIPS, the continued resolution of PVT suggests that once PV venous flow is reestablished, high portal/mesenteric pressures generate sufficient flow resulting in clot dissolution and PV patency. The small group of patients in whom we used the transsplenic approach to the occluded PV is deserving of further discussion. This technique was considered after favorable experiences using a combination of transjugular and transhepatic approach to the PV. We had observed that these methods often required long procedure times. Using the transsplenic approach, overall procedure (not fluoroscopic) times were 45, 90, and 120 minutes. These cases were remarkably straightforward and simplified our earlier approaches;
2353
the SV was readily catheterized through the spleen. Crossing the chronically occluded PV required minimal effort (glidewire, catheter). The PVR-TIPS was then carried out by targeting a snare placed in a PV branch. Given the technical simplicity of the transsplenic approach, this is now our initial method of choice, with routine splenic tract embolization. Comparison With Other Series
Our study differs from others in many respects. Luca et al17 treated patients with partial PVT; in our series, patients had either complete (100%) or near-complete (>95%) luminal occlusion. They also excluded patients with cavernomas; it was present in 30% of our patients. We did however confirm their finding of significant resolution of PVT after TIPS placement. We also do agree with their recommendation not to cover the MPV. Han et al16 included patients with acute PVT; in our series, we only treated patients with chronic PVT in the transplant setting. They exhibited 75% success. They too used a transsplenic approach when necessary. Finally, in contradistinction to our approach, they anticoagulated patients after TIPS; our findings suggest this is not needed. Senzolo et al14 reported successfully using TIPS in a heterogeneous cohort of 28 patients, only one of which was LT the primary indication. In our series, preparation for LT was the primary objective. Comparison With Anticoagulation
Delgado et al20 published a series studying anticoagulation for acute/chronic mesenteric thrombosis including the SMV; they excluded patients with cavernomas. This differs from our series where we focused specifically chronic MPV thrombosis and included cavernomas. Our series also focused on the intent of rendering the PV anatomy favorable for LT. Hence, comparing anticoagulation in the acute setting with PVR-TIPS for chronic MPV thrombosis appears spurious. Anticoagulation in this patient population also carries an increased risk of bleeding and rethrombosis.20 The role of anticoagulation in the management of PVT remains controversial.21,22 Comparison to University of Miami and Other
Hibi et al8 recently published a series extremely relevant to our study. The authors compared the clinical outcomes of PVT patients able to achieve physiologic restoration of flow through end-to-end PV anastomoses to those with severe PVT not permitting restoration of native portal flow. Patients in the latter group were those with complete/near complete obliterative main PVT, just as in our series presented herein. They reported that nonphysiologic PV inflow (caval hemitransposition, renoportal anastomosis, PV arterialization) at transplantation exhibited higher rates of rethrombosis, GI bleeding, intraoperative blood loss, highest in-hospital mortality rates, as well as lower survival compared with physiologic PV flow restoration. In our opinion, their findings support the importance of reestablishing native PV flow before LT; our posttransplant survival supports this contention. The PVRTIPS simplified LT by reestablishing main portal flow; this was confirmed at surgery in 18 transplanted cases to date, where end-to-end anastomoses were achieved. In our series, PVR-TIPS resulted in high PV patency and excellent survival. Alves et al23 also used surgical techniques (shunts, grafts) to deal with PVT in 29 patients, achieving 75% actuarial
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
2354
Transplantation
■
November 2015
■
Volume 99
■
Number 11
survival. Robles et al24 achieved 69% 5-year survival in 32 patients treated with thromboendovenectomy. Lessons Learned
Several lessons or clinical “pearls” deserve highlighting. First, the presence of cavernoma did not impede our success rate, especially when the transsplenic route was used. Second, identifying a peripheral intra-SV in direct line with the main SV was essential for straight access to the thrombosed PV. Third, establishing portal flow despite the presence of mild residual thrombus in the PV at PVR-TIPS termination was satisfactory; complete removal was not necessary. Fourth, a stent extending only 1 to 2 cm into the PV is recommended, maximizing unstented PV length for LT. Fifth, the unexpected finding of normal-walled PVs at transplantation suggests stenting the main PV is not necessary; this finding challenges the previously long-held belief that chronically thrombosed veins require stenting. Finally, the gastroesophageal varices and splenorenal shunts did not require embolization during the first procedure. Rather, these were reserved for the 1-month TIPS venogram, where PV flow status and embolization was performed.
www.transplantjournal.com
may be an attractive approach for live donors when vessels may not be available for venous conduits. We also believe PVR-TIPS provides transplant surgeons with an added tool (along with surgical techniques) in the management of chronic PVT before LT. Finally, although the generalizability of this technique may be open for debate, we have recently updated a new (and readily transferrable) method using the transsplenic route; this is now our default approach for PVR-TIPS.25 CONCLUSIONS Pretransplant PVR in patients with chronic, obliterative PVT provides surgeons an additional option to address PVT and obviate the need for complex surgical grafts. Our study demonstrates PVR-TIPS to be an effective intervention with a favorable safety profile. This study explores a procedure that allows an entirely new subpopulation (patients deemed untransplantable by some centers) access to LT without the risks of anticoagulation. Once these data are confirmed, we anticipate that the management of cirrhotic patients with chronic PVT will be positively impacted.
Strengths and Limitations
REFERENCES
We recognize this represents single-center experience with high level technical expertise. It does represent the largest series of an aggressive approach to patients in the transplant setting. They were not compared to a control population (such as anticoagulation); these patients had cavernomas and GI bleeds, conditions where anticoagulation is not routinely used. However, patency rates and outcomes in our series exceed those described in studies of anticoagulation. The surgical perspective is also important because, whereas most transplant surgeons would posit they can technically overcome PVT, most would agree that a patent PV during LT is reassuring. Other than a simple eversion thrombectomy, surgical procedures (graft transposition, cavoportal hemitransposition, multivisceral transplantation) are generally associated with poor outcomes.8 It cannot be overemphasized that the Miami group data are the most appropriate comparison cohort for our study group—these were not cases of routine advanced PVT but rather PVT with obliteration of the main portal venous channel. Because this series has no comparator, survival outcomes should be interpreted with caution. We acknowledge that this is a unique subset of patients with low baseline MELD scores and compensated liver disease with chronic obliterative PVT. Finally, we recognize the technical limitations of PVRTIPS for patients with thrombosis of the entire portal/ mesenteric system, where this technique may not be effective given the absence of PV inflow from the SMV or SV, significant contributing factors for PVR-TIPS patency. There are strengths. This is the first study focusing on PVR-TIPS with the specific intent of physiologic portal flow restoration supported by surgical confirmation at LT in 50% (N = 18/36) of the patients. It is also the first moderately sized cohort study to focus on this major clinical challenge; complete/near-complete obliterative MPV thrombosis in cirrhotics. We have mature follow-up permitting assessment of PV patency with ultrasound/cross-sectional imaging/venography. The patients exhibited clinical manifestations of cirrhosis, justifying the need for TIPS and ultimate LT. The PVR-TIPS
1. Ponziani FR, Zocco MA, Campanale C, et al. Portal vein thrombosis: insight into physiopathology, diagnosis, and treatment. World J Gastroenterol. 2010;16:143–155. 2. Amitrano L, Guardascione MA, Brancaccio V, et al. Risk factors and clinical presentation of portal vein thrombosis in patients with liver cirrhosis. J Hepatol. 2004;40:736–741. 3. Luca A, Caruso S, Milazzo M, et al. Natural course of extrahepatic nonmalignant partial portal vein thrombosis in patients with cirrhosis. Radiology. 2012;265:124–132. 4. Maruyama H, Okugawa H, Takahashi M, et al. De novo portal vein thrombosis in virus-related cirrhosis: predictive factors and long-term outcomes. Am J Gastroenterol. 2013;108:568–574. 5. Zocco MA, Di Stasio E, De Cristofaro R, et al. Thrombotic risk factors in patients with liver cirrhosis: correlation with MELD scoring system and portal vein thrombosis development. J Hepatol. 2009;51:682–689. 6. Francoz C, Valla D, Durand F. Portal vein thrombosis, cirrhosis, and liver transplantation. J Hepatol. 2012;57:203–212. 7. Garcia-Pagan JC, Valla DC. Portal vein thrombosis: a predictable milestone in cirrhosis? J Hepatol. 2009;51:632–634. 8. Hibi T, Nishida S, Levi DM, et al. When and why portal vein thrombosis matters in liver transplantation: a critical audit of 174 cases. Ann Surg. 2014;259:760–766. 9. Bauer J, Johnson S, Durham J, et al. The role of TIPS for portal vein patency in liver transplant patients with portal vein thrombosis. Liver Transpl. 2006;12:1544–1551. 10. Jamieson NV. Changing perspectives in portal vein thrombosis and liver transplantation. Transplantation. 2000;69:1772–1774. 11. Rossle M. TIPS: 25 years later. J Hepatol. 2013;59:1081–1093. 12. Bilbao JI, Quiroga J, Herrero JI, Benito A. Transjugular intrahepatic portosystemic shunt (TIPS): current status and future possibilities. Cardiovasc Intervent Radiol. 2002;25:251–269. 13. Van Ha TG, Hodge J, Funaki B, et al. Transjugular intrahepatic portosystemic shunt placement in patients with cirrhosis and concomitant portal vein thrombosis. Cardiovasc Intervent Radiol. 2006;29:785–790. 14. Senzolo M, Tibbals J, Cholongitas E, et al. Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with and without cavernous transformation. Aliment Pharmacol Ther. 2006;23:767–775. 15. Perarnau JM, Baju A, D'Alteroche L, et al. Feasibility and long-term evolution of TIPS in cirrhotic patients with portal thrombosis. Eur J Gastroenterol Hepatol. 2010;22:1093–1098. 16. Han G, Qi X, He C, et al. Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with symptomatic portal hypertension in liver cirrhosis. J Hepatol. 2011;54:78–88. 17. Luca A, Miraglia R, Caruso S, et al. Short- and long-term effects of the transjugular intrahepatic portosystemic shunt on portal vein thrombosis in patients with cirrhosis. Gut. 2011;60:846–852.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Salem et al
© 2015 Wolters Kluwer
18. Boyer TD, Haskal ZJ. The role of transjugular intrahepatic portosystemic shunt (TIPS) in the management of portal hypertension: update 2009. Hepatology. 2010;51:306. 19. D'Avola D, Bilbao JI, Zozaya G, et al. Efficacy of transjugular intrahepatic portosystemic shunt to prevent total portal vein thrombosis in cirrhotic patients awaiting for liver transplantation. Transplant Proc. 2012;44:2603–2605. 20. Delgado MG, Seijo S, Yepes I, et al. Efficacy and safety of anticoagulation on patients with cirrhosis and portal vein thrombosis. Clin Gastroenterol Hepatol. 2012;10:776–783. 21. Huard G, Bilodeau M. Management of anticoagulation for portal vein thrombosis in individuals with cirrhosis: a systematic review. Int J Hepatol. 2012;2012:672986.
2355
22. Congly S, Lee S. Portal vein thrombosis: should anticoagulation be used? Curr Gastroenterol Rep. 2013;15:1–7. 23. Alves J, Boin I, Cholongitas E, et al. Surgical technique used for portal vein thrombosis when thrombectomy is not possible during liver transplantation. Surg Sci. 2011;2:767–775. 24. Robles R, Fernandez JA, Hernandez Q, et al. Eversion thromboendovenectomy in organized portal vein thrombosis during liver transplantation. Clin Transplant. 2004;18:79–84. 25. Habib A, Desai K, Hickey R, et al. Portal vein recanalizationtransjugularintrahepatic portosystemic shunt using the transsplenic approach to achieve transplant candidacy in patients with chronic portal vein thrombosis. J Vasc Interv Radiol. 2015;S1051-0443:01183.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
