Management of Venous Outflow Complications After Liver Transplantation

Management of Venous Outflow Complications After Liver Transplantation Michael D. Darcy, MD Liver transplantation can be complicated by stenosis of the hepatic venous or inferior vena cava outflow. Venous outflow stenosis occurs at rates of 1 to 6% depending on the type of anastomosis. Stenoses can develop acutely as a result of technical problems or can present much later after the transplant due to intimal hyperplasia or perianastomotic fibrosis. Common clinical presentations include hepatic dysfunction, liver engorgement, ascites, abdominal pain, and occasionally variceal bleeding. Treatment can generally be accomplished via a transjugular approach, but percutaneous transhepatic access may be needed when the anastomosis cannot be catheterized from the jugular access. Angioplasty can achieve technical success in restoring anastomotic patency in close to 100% of cases, but restenosis is frequent. Repeat angioplasties may be needed. In adults and pediatric patients with adult sized hepatic veins, stenting may be a better option. Resolution of clinical signs and symptoms is seen in 73 to 100% of cases. Major complications are uncommon, with stent migration being one of the more difficult complications to manage. Tech Vasc Interventional Rad 10:240-245 © 2007 Elsevier Inc. All rights reserved. KEYWORDS hepatic transplantation, venous anastomotic stenosis, venous angioplasty, venous stenting

L

iver transplantation has become a common operation and although results are often very good, the various vascular anastomoses can become compromised. Fortunately these complications can usually be managed by interventional techniques. The key role of interventional techniques was nicely illustrated in the study by Cavallari and coworkers in which mortality after interventional graft salvage procedures was 11.1% compared with 41.6% mortality for those patients managed by retransplantation.1 This current article will focus on compromise of the venous outflow (due to either hepatic vein or inferior vena cava stenosis), which can lead to significant morbidity and even death.

Clinical Presentation Overall venous outflow complications are relatively uncommon. Stenosis of the inferior vena cava (IVC) is reported to occur in less than 2% of liver transplants.2 Stenosis of the hepatic vein (HV) outflow is slightly more common, but the incidence varies with the type of transplant. In a review of 600 pediatric liver transplants, the rates of HV stenosis was

Interventional Radiology Section, Mallinckrodt Institute of Radiology, Washington University, St Louis, MO. Address reprint requests to: Michael D. Darcy, MD, Professor of Radiology (Section Chief), Interventional Radiology Section, Mallinckrodt Institute of Radiology, Washington University, 510 South Kingshighway Blvd, St Louis, MO 63110. E-mail: [email protected]

240

1089-2516/07/$-see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1053/j.tvir.2007.09.018

1% for whole liver grafts and 2% for living related grafts, but 4% for reduced size or split grafts.3 The piggyback type of anastomosis is said to reduce problems with anastomotic stenosis. In one review of 264 piggyback transplants only 2 cases (0.8%) of delayed HV obstruction occurred4; however, in another series of piggyback transplants the HV stenosis rate was 3.4%.5 On the other hand, HV stenosis is more likely to occur after living related transplants because of the different ways these are anastomosed. In living related transplants a patch of the donor IVC is usually not removed as can be done in cadaveric donors. Egawa and coworkers6 reported HV stenosis in 3 of 48 (6%) children post living related donor transplants. There are different causes of outflow obstruction, which will vary with both the type of transplant and the length of time elapsed since the operation. If the obstruction occurs acutely, it is often due to a technical problem such as too tight of an anastomosis, donor-recipient size mismatch, twisting of the veins, or an abnormal intimal flap (Fig. 1). Anastomotic stenoses presenting in delayed fashion are more likely to be secondary to perivascular fibrosis or intimal hyperplasia. Obstruction does not necessarily need to occur right at the anastomosis. Graft edema may cause hepatic enlargement and extrinsic impression on the IVC thus obstructing caval and/or HV outflow. Normal morphologic change of the liver may also cause problems. Partial liver grafts show considerable growth after implantation and this morphologic change may lead to twisting of the veins.

Venous outflow complications

241 The signs and symptoms of hepatic vein stenosis may be similar to those of portal hypertension. Ascites, increasing abdominal girth, and weight gain is one of the commonest presentations. The patient may also have shortness of breath from either hydrothorax or massive ascites. Variceal bleeding may also be the initial presentation. Abdominal pain is another common symptom and may relate to either the ascites or the engorgement and distension of the liver. Rarely fatigue and weight loss associated with protein losing enteropathy can be the primary presentation of hepatic vein obstruction.7 On physical examination the liver will feel enlarged and firm. Abdominal distension from ascites may be significant. When IVC stenoses occur they usually occur above the HV anastomosis. Thus IVC stenoses often present with hepatic dysfunction or symptoms similar to the HV stenoses. Isolated IVC stenosis below the hepatic veins may present with lower extremity edema and ascites but without hepatic enlargement or dysfunction. A severe stenosis of the IVC can lead to IVC thrombosis, which may cause more severe lower extremity edema.8 Laboratory findings of hepatic dysfunction can help confirm the clinical suspicion or may actually precede symptoms or physical findings. There is one situation in particular in which the liver function tests may not parallel the clinical symptoms. Isolated stenosis of the right hepatic vein after piggyback transplantation has been reported to cause development of ascites despite relatively preserved liver function.9 Another laboratory manifestation of HV outflow obstruction is decreased clearance of transplant immunosuppressive drugs like tacrolimus, resulting in higher trough levels. In one study,10 50% of 28 cases with HV stricture had increased trough levels of tacrolimus. Elevated serum creatinine may occur in cases of IVC obstruction.

Imaging Diagnosis Confirmatory imaging is done when venous obstruction is suspected. Although anastomotic stenoses can certainly be identified on computed tomographic (CT) or magnetic resonance (MR) scans, Doppler ultrasound examinations are the most commonly used noninvasive examination because of advantages of cost, easy availability, and the ability to do the examination portably. Doppler examinations in a patient with HV stenosis will show decreased mean velocities in both the hepatic veins and portal vein. The hepatic vein wave form will be very dampened when there is an outflow obstruction.6 Typically the Doppler waveform in a normal hepatic vein is triphasic, but after transplantation the waveform is often biphasic even

Figure 1 (A) IVC venogram done intraoperatively on a patient who developed hepatic congestion when the donor liver was placed back into the abdominal cavity. The venogram demonstrated an IVC flap (arrow) that had a 15 mm Hg gradient across it. Contrast filled the HV trunk (arrowhead), which has an unusual appearance because the anastomosis was redone before this venogram. (B) Venogram after stent deployment, which was necessitated by lack of response to PTA. The obstruction was resolved, the hepatic congestion resolved, and the transplant operation was able to be completed.

M.D. Darcy

242

Figure 2 (A) Hepatic venogram done in a pediatric patient with a living related left lobe transplant. The catheter obstructs the stenosis completely so no contrast flows into the IVC. (B) Simultaneous injection through a sheath and the HV catheter defines that the obstruction is very focal. (C) Balloon angioplasty of the HV anastomotic stenosis. (D) Post angioplasty venogram showing a widely patent HV anastomosis.

without any other signs or symptoms of flow obstruction.11 However, when significant stenosis develops, the waveform usually degrades to a monophasic pattern. Other findings may include reversal of HV flow, accelerated flow with aliasing just beyond the stenosis, and visualization of the stenosis on gray scale imaging. Although Doppler ultrasound is useful, venography and pressure measurements are still considered to be the gold standard. Venography should be done whenever the diagnosis of venous outflow compromise is suspected. This is partially because the symptoms of outflow obstruction may be nonspecific. In one study12 of 26 patients referred for suspected HV stenosis, only 20 (77%) actually had an abnormal gradient across the hepatic venous anastomosis. Thus venography and pressure measurements are necessary to prove the diagnosis. However, there is considerable question about what constitutes an abnormal gradient. Anywhere from 3 to 20 mm Hg has been considered to be the threshold of abnormality. A gradient greater than 10 mm Hg is one commonly used threshold.13,14 However, in a series by Weeks and coworkers,15 the initial gradients ranged from 3 to 14 mm Hg.

The significance of these stenoses was validated both by presenting symptoms and the response to treatment.

Techniques An internal jugular access will provide the best overall approach to study either HV or IVC anastomoses since it provides a good trajectory to cannulate the target veins and there is very little associated morbidity with jugular access. Where to search for the HVs will depend on the type of anastomosis; hence it is important to review the transplant operative summary. Once the HV has been catheterized, the stenosis can usually be adequately defined by contrast injection through the catheter used to select the HV. However, the stenosis may be tight enough that the catheter completely obstructs flow preventing opacification of the caval side of the obstruction. In that case, also injecting through a sheath positioned just above the stenosis will allow better definition of the stenosis (Fig. 2). If the HV obstruction prevents HV catheterization from the jugular access, an alternative approach is to do a percutaneous transhepatic puncture into the peripheral as-

Venous outflow complications pects of the obstructed hepatic vein, which can usually be done using ultrasound guidance. Similarly caval stenoses can usually be catheterized from the jugular access, but if the IVC is completely obstructed, a catheter may need to be passed from a femoral venous access to define the caudal extent of the obstruction. Attempts to cross the obstruction may be more successful from the femoral access since the cava provides support for the catheter and prevents some of the wire buckling that occurs in the right atrium when trying to cross the obstruction from above. If unable to cross the IVC stenosis, sharp recanalization may be necessary.8 This involves passing a needle across the obstruction while aiming at a balloon or snare placed on the other side of the obstruction. This technique should be restricted to short obstructions and cases where the needle can be passed in a straight line toward the target without significant angulation or curvature. Rarely thrombus may be present below the stenotic anastomosis and thrombolysis may be necessary to reduce the clot burden before restoring patency across the anastomosis. Thrombolytic infusion for posttransplant IVC anastomotic stenosis with superimposed thrombosis has been described.14 Although no complications were reported, one should be cautious about thrombolysis in the early posttransplant period for fear of stimulating anastomotic bleeding. Mechanical thrombolytic devices may also be used to avoid bleeding. The next step is to open the stenosis and balloon angioplasty is usually the first approach. The balloon needs to be slightly oversized compared with the vessel diameter and often a high pressure balloon will be needed to adequately open the anastomosis. Prolonged inflation (1-2 minutes) may help stretch and tear the tissue causing the stenosis and reduce immediate recoil. However, close monitoring is required since prolonged balloon inflation will reduce blood return to the heart and can lead to temporary hypotension. For tough strictures that fail to dilate with standard percutaneous transluminal angioplasty (PTA) balloons, cutting balloons can be useful. We have seen cases where a stricture resistant to high pressure balloons opened up after just a single inflation of a cutting balloon. Because of the high incidence of restenosis, stenting is an attractive option. While stenting may be preferable in adults, there is some debate about their use in pediatric patients. Venous stents can develop intimal hyperplasia and thus are not totally protective against restenosis. In addition there is concern that stents will not grow in size with the child and could end up creating a relative narrowing of the HV. However, it should be recognized that the HVs may be adult size if the liver is derived from an adult donor. Rerksuppaphol and coworkers16 suggest that stenting is superior even in pediatric cases. They reported 3 infants that were successfully stented; however, their follow-up was short (2-20 months). Thus how stents will fare once the child has grown is unclear. If an adult size stent (10 mm or greater diameter) can be deployed, the concern about growth of the child is probably a moot point. If the hepatic veins are small (⬍8 mm in size), then stenting may not be the best option and instead repeated PTA is probably a better strategy. For both HV and IVC stenting, using a stent with large interstices is commonly recommend to avoid blocking HV flow if the stent ends up partially or completely over the HV or HV branches. Despite this theoretic risk, some authors14 have placed Wallstents across the HV orifice to treat IVC stenoses and

243 not encountered hepatic vein thrombosis despite that Wallstents have the smallest interstices of commercially available stents. Another theoretic problem with jailing the HV with an IVC stent is that future interventions would be made more difficult if a separate HV stenosis should develop. Use of balloon expandable stents is sometimes recommended for HV stenoses because they have higher radial strength compared with self-expanding stents and generally can be placed more precisely. This assertion is supported by comparing the 92% clinical success of the Stanford group5 using Palmaz stents to the 72% clinical success reported by Ko and coworkers17 using self-expanding stents. Stent diameter will vary according to the vein treated but will usually be at least 8 to 12 mm for hepatic vein stenoses and 14 to 20 mm for most IVC stenoses.

Results The definitions of technical success for therapy of HV/IVC stenoses are not uniform in the literature. Sometimes PTA or stent deployment that leads to morphologic improvement is considered a technical success whereas others define success as elimination or significant reduction of the trans-stenotic pressure gradient. However it is defined, the technical success rates for treating these entities is often reported to be 100%.5,17 Lorenz and coworkers18 reported 94% technical success and their one technical failure was due to inability to cross the obstructed segment. These optimistic numbers need to be tempered by the realization that although HV stenoses can be resolved with PTA alone they usually require several sessions to achieve a durable result.10 Ko and coworkers17 noted that 3 of 5 patients with HV stenoses treated by PTA developed recurrent stenoses as soon as 1 to 5 weeks after the first PTA and these patients required multiple PTAs. Once the venous obstruction has been successfully treated, clinical success will often be apparent fairly quickly. Signs and symptoms such as edema or ascites often resolve within a few days after a successful endovascular procedure. Laboratory abnormalities may also return to the normal range within days. If prolonged hepatic clearance of immunosuppressive drugs was the initial presentation, the drug trough levels also quickly improve after PTA.19 Lack of response suggests either rapid recoil/restenosis or that there is additional underlying hepatic pathology. Short-term clinical success is generally excellent with some authors reporting clinical improvement in 100% of their cases.12 Others have reported lower clinical success rates of 73 to 89% despite having higher technical success rates.15,17 It is important to realize that clinical improvement may not occur despite restoration of venous patency. This may be due to other coexisting conditions such as acute rejection or permanent hepatic damage caused by HV congestion. In one series of 13 patients who underwent successful stenting, 1 patient did not improve and was found to have hepatic necrosis believed to be due to prolonged hepatic venous congestion.5 Assessing patency during follow-up can be done through a combination of clinical observation and use of Doppler ultrasound. If restenosis occurs, the initial presenting symptoms, physical signs, or laboratory abnormalities will usually return.

M.D. Darcy

244

Figure 3 (A) Initial hepatic venogram in a pediatric patient with stenosis (arrow) of a common HV trunk at the anastomosis to the recipient HV trunk. (B) Postangioplasty venogram showing only a mild residual stenosis, which has no pressure gradient across it. (C) Hepatic venogram done only 2 months after the PTA because of recurrent ascites. The venogram shows high grade restenosis.

Doppler ultrasound can also be used to screen for recurrent stenoses. After successful venous therapy, the HV velocities will increase and hepatic waveform will again be more pulsatile.6 However, when restenosis occurs, HV velocities will again slow and the waveforms will become monophasic. For HV stenoses it is rarely possible to achieve long-term patency with a single angioplasty (Fig. 3). Cheng and coworkers reported 1 patient with 9.5 years of documented patency after single PTA, but several others developed recurrence by 1 month.20 In one of the larger series,12 primary patency at 3, 6, 12, and 60 months was 80%, 65%, 60% and 60%, respectively; but in each case the recurrent stenoses were successfully treated yielding assisted patency of 100% throughout follow-up. In a report of 2 patients, use of cutting balloon seemed to lead to more durable results, but obviously larger numbers of patients are needed to confirm if this technology really is beneficial.21 Of the posttransplant venous stenoses, Zajko and coworkers22 noted that suprahepatic IVC stenoses are more resistant to PTA than portal vein anastomotic stenoses. While portal steno-

ses usually did not recur, 50% of the patients with IVC stenoses required repeated angioplasty. Stenting for IVC stenoses seems to lead to more durable results. In one study, 91% (20 of 22) of stented patients needed no further intervention, whereas only 40% (2 of 5) of angioplasty patients avoided further intervention.17 Given the large diameter of stents used in the IVC, intimal hyperplasia should not have much impact on patency. Weeks and coworkers15 reported 100% objective patency with an average of 491 days of follow-up.

Complications Complications fortunately are fairly uncommon. The rate of major complications has been reported to be 0% in several series.12,20,22 Minor complication occur in about 10% of cases and include transient hypotension and arrhythmias.12 Hypotension can occur because the large PTA balloons can occlude the IVC and decrease venous return to the heart, particularly in young pediatric patients. Guidewire manipulations in the right atrium can certainly cause arrhythmias, but just having

Venous outflow complications a large inflated balloon extending into the right atrium can also stimulate arrhythmias. Zajko and coworkers22 noted in one of their cases that leaving the balloon inflated for more than 10 seconds triggered arrhythmias. Probably the worst complication is stent migration, but it is fortunately infrequent.15 Most often stent migration is immediately obvious during the procedure; however, it can also occur in a delayed fashion. This can relate to changes in the size of the IVC. For example, hepatic congestion may cause caval compression and once the congestion is relieved the IVC diameter may increase. While not strictly migration, improper deployment causing the stent to extend too far into the right atrium can also have significant consequences. In one case a large organized thrombus formed on the stent extending into the right atrium and necessitated open heart surgery for removal.23 Although stent migration may certainly be the result of poor technique or poor judgment (picking the wrong size stent), this is a very difficult anatomic region in which to deploy stents due to the significant cardiac and respiratory motion. Other causes of migration include snagging the stent during removal of the balloon used to expand the stent or trying to deploy the stent from an angled approach such as from a femoral vein access.5 This can cause the stent to jump during final deployment because of the tension that the angled approach puts on the stent. Stent migration may also relate to the type of stent used. Simo and coworkers24 reported an usually high rate of stent dislocation in 2 of 3 cases of IVC anastomotic stenoses treated with Gianturco stents. These stents have a more open architecture and hence fewer points of contact, which may make them more prone to migration. If partial stent migration occurs during the procedure and if it has not extended too far into the IVC, the stent can be stabilized by deploying another stent in better position to lock the first stent in place.15 If excessive migration occurs, attempts should be made to snare and retrieve the stent. Leaving a migrated stent in the right atrium can have disastrous consequences. There is 1 case report25 where a migrated Wallstent could not be retrieved and it ultimately caused an aorto-right atrial fistula, which required open heart surgery to fix. One theoretic complication is bleeding due to disruption of the venous anastomosis by the angioplasty. However, according to Weeks and coworkers15 there are no reported cases of rupture after IVC venoplasty. This is remarkable considering that many series include some patients with acute stenosis who underwent PTA within days after transplantation. There have been only a couple cases of remote bleeding in the form of groin or retroperitoneal hematomas that were felt to be related to the anticoagulation, not the angioplasty.14

Summary Stenosis of the venous outflow of liver transplants is uncommon, but when it does occur it can cause significant symptoms and hepatic dysfunction. Fortunately, endovascular techniques allow these to be treated with little risk and a high degree of efficacy.

References 1. Cavallari A, Vivarelli M, Bellusci R, et al: Treatment of vascular complications following liver transplantation: multidisciplinary approach. Hepatogastroenterology 48(37):179-183, 2001

245 2. Kraus TW, Rohren T, Manner M, et al: Successful treatment of complete inferior vena cava thrombosis after liver transplantation by thrombolytic therapy. Br J Surg 79:568-569, 1992 3. Buell JF, Funaki B, Cronin DC, et al: Long-term venous complications after full-size and segmental pediatric liver transplantation. Ann Surg 236(5):658-666, 2002 4. Sze DY, Semba CP, Razavi MK, et al: Endovascular treatment of hepatic venous outflow obstruction after piggyback technique liver transplantation. Transplantation 68:446-449, 1999 5. Wang SL, Sze DY, Busque S, et al: Treatment of hepatic venous outflow obstruction after piggyback liver transplantation. Radiology 236:352359, 2005 6. Egawa H, Tanaka K, Uemoto S, et al: Relief of hepatic vein stenosis by balloon angioplasty after living-related donor liver transplantation. Clin Transplant 7(4):306-311, 1993 7. Dousset B, Legmann P, Soubrane O, et al: Protein-losing enteropathy secondary to hepatic venous outflow obstruction after liver transplantation. J Hepatol 27:206-210, 1997 8. Mindikoglu AL, Miller JS, Borge MA, et al: Post-transplant IVC occlusion and thrombosis treated with tPA, heparin, and sharp recanalization. J Gastroenterol 40:302-305, 2005 9. Aucejo F, Winans C, Henderson JM, et al: Isolated right hepatic vein obstruction after piggyback liver transplantation. Liver Transpl 12:808812, 2006 10. Totsuka E, Hakamada K, Narumi S, et al: Hepatic vein anastomotic stricture after living donor liver transplantation. Transplant Proc 36(8): 2252-2254, 2004 11. Fujimoto M, Moriyasu F, Someda H, et al: Recovery of graft circulation following percutaneous transluminal angioplasty for stenotic venous complications in pediatric liver transplantation: assessment with Doppler ultrasound. Transpl Int 8(2):119-125, 1995 12. Kubo T, Shibata T, Itoh K, et al: Outcome of percutaneous transhepatic venoplasty for hepatic venous outflow obstruction after living donor liver transplantation. Radiology 239:285-290, 2006 13. Raby N, Karani J, Thomas S, et al: Stenoses of vascular anastomoses after hepatic transplantation: treatment with balloon angioplasty. AJR Am J Roentgenol 157(1):167-171, 1991 14. Borsa JJ, Daly CP, Fontaine AB, et al: Treatment of inferior vena cava anastomotic stenoses with the Wallstent endoprosthesis after orthotopic liver transplantation. J Vasc Interv Radiol 10(1):17-22, 1999 15. Weeks SM, Gerber DA, Jaques PF, et al: Primary Gianturco stent placement for inferior vena cava abnormalities following liver transplantation. J Vasc Interv Radiol 11(2 pt 1):177-187, 2000 16. Rerksuppaphol S, Hardikar W, Smith AL, et al: Successful stenting for Budd-Chiari syndrome after pediatric liver transplantation: a case series and review of the literature. Pediatr Surg Int 20(2):87-90, 2004 17. Ko GY, Sung KB, Yoon HK, et al: Endovascular treatment of hepatic venous outflow obstruction after living-donor liver transplantation. J Vasc Interv Radiol 13:591-599, 2002 18. Lorenz JM, Van Ha T, Funaki B, et al: Percutaneous treatment of venous outflow obstruction in pediatric liver transplants. J Vasc Interv Radiol 17:1753-1761, 2006 19. Narumi S, Hakamada K, Toyoki Y, et al: Hepatic clearance improves after angioplasty of the hepatic vein. Transplant Proc 36:3091-3092, 2004 20. Cheng YF, Chen CL, Huang TL, et al: Angioplasty treatment of hepatic vein stenosis in pediatric liver transplants: long-term results. Transpl Int 18:556-561, 2005 21. Narumi S, Hakamada K, Totsuka E, et al: Efficacy of cutting balloon for anastomotic stricture of the hepatic vein. Transplant Proc 36(10):30933095, 2004 22. Zajko AB, Sheng R, Bron K, et al: Percutaneous transluminal angioplasty of venous anastomotic stenoses complicating liver transplantation: intermediate-term results. J Vasc Interv Radiol 5(1):121-126, 1994 23. Mazariegos GV, Garrido V, Jaskowski-Phillips S, et al: Management of hepatic venous obstruction after split-liver transplantation. Pediatr Transplant 4:322-327, 2000 24. Simo G, Echenagusia A, Camunez F, et al: Stenosis of the inferior vena cava after liver transplantation: treatment with Gianturco expandable metallic stents. Cardiovasc Intervent Radiol 18:212-216, 1995 25. Guimaraes M, Uflacker R, Schonholz C, et al: Stent migration complicating treatment of inferior vena cava stenosis after orthotopic liver transplantation. J Vasc Interv Radiol 16:1247-1252, 2005