Endovascular thromboaspiration with neurointerventional devices for early hepatic artery thrombosis after split liver transplant

Abstract

We report the case of a 38-year-old woman who underwent orthotopic ‘split’ liver transplant, complicated by hepatic artery thrombosis on the first postoperative day. The patient was successfully treated with an endovascular approach by mechanical thromboaspiration of the hepatic artery, using neurovascular devices, angioplasty and stenting at the site of the surgical anastomosis.

Background

Despite advances in surgical techniques, hepatic artery thrombosis (HAT) remains a dreaded complication after orthotopic liver transplant (OLT), jeopardising transplant success and patient survival.¹ Surgeons still claim that early-HAT, occurring within 30 days after OLT, requires immediate re-transplantation for survival, but today, endovascular intervention represents a valid therapeutic alternative, granting equivalent results with a significantly lower invasiveness. Endovascular treatments for HAT traditionally include intra-arterial fibrinolysis, percutaneous transluminal angioplasty (PTA) and stent placement.² Instead, hepatic artery (HA) thrombectomy remains a less explored strategy. We report a case of post-OLT hyperacute HAT, treated by endovascular mechanical thromboaspiration, combined with PTA and stenting.

Case presentation

A 38-year-old woman affected by polycystic liver disease underwent to OLT with ‘split liver’ technique. The graft comprised a right lobe from a cadaveric donor. The arterial reconstruction was performed by ‘end-to-end’ anastomosing the donor’s right HA to the recipient’s common HA. Surgery was uncomplicated.

On the first postoperative day, the patient was clinically stable, presented normal blood chemistry (Haemoglobin 90 g/L, platelets (PLT) 105 000/mm³), altered hepatic tests (alanine amino-tranferase (ALT),60 000 U/L ca., aspartate amino-transferase (AST), 50 000 U/L ca.), International normalised Ratio (INR), 1.18, slightly altered D-dimer (2000 ng/mL) and normal blood lactates.

Investigations

Routine Doppler sonographic evaluation always performed at 12 hours after transplantation showed absence of flow in the HA both in perihilar localization and in intrahepatic branches. Regular flow was observable in portal and hepatic veins. A small amount of fluid collection was noticed in the perihepatic space.

The diagnosis of HAT was clear and no other imaging investigation was performed even to avoid waste of time for potential life-saving treatment.

Differential diagnosis

Hyperacute reject must always be considered as a potential cause for graft failure.

In post-transplant care, Doppler imaging can rule out hyperacute causes of graft failure and, in particular, early HAT.

In this specific case, the split liver transplant technique suggested the diagnosis of HAT, due to the potential difference in calibre between donor and receiver arteries.

Treatment

After a multidisciplinary discussion involving hepatobiliary surgeons, interventional radiologists and anesthesiologists, the patient was taken to the angiographic suite.

An 8 Fr introducer was positioned through the right femoral access and the coeliac axis was catheterized by a 6 Fr KMP (Cook Medical, Bloomington, Indiana), slid over a 0.035″ J-standard guidewire (Radifocus Guidewire, Terumo, Shibuya, Tokyo). The Digital Subtraction Angiography (DSA), showing no opacification of the HA, confirmed the diagnosis of HAT (figure 1A). The guidewire was exchanged to a 0.014″ 300 cm (Transend, Stryker, Kalamazoo, Michigan) and manipulated beyond the occlusion reaching the intrahepatic arterial district. At first, a 3 mm balloon-catheter (Armada, Abbott Vascular) was dilated across the anastomotic site (figure 1B). The subsequent DSA showed partial recanalisation of common and proper HA, with a dim visualisation of the intrahepatic arterial tree; a stenotic, irregular aspect persisted in the site of the surgical anastomosis (figure 1C). Moreover, after few minutes, both extrahepatic and intrahepatic segments of the artery were newly occluded (figure 1D). Thus, the KMP catheter was exchanged with a long introducer (6 Fr, 80 cm-long NeuronMax 0.88, Penumbra), the tip of which was positioned in the proximal tract of the coeliac trunk, and a Catalyst-5 catheter (5 Fr, Stryker) was inserted to perform mechanical thromboaspiration.

Figure 1.

(A) Angiography from the coeliac trunk shows no HA opacification, confirming the diagnosis of HAT. (B) A 0.014″ guidewire is delivered in the intrahepatic arterial district, beyond the occlusion, and a 3 mm balloon-catheter is dilated across the anastomotic site. (C) The common and right HA are partially recanalised, and the intrahepatic arterial tree is dimly visible, but the anastomotic site appears stenotic (black arrow). (D) After few minutes, the HA is newly occluded. HA, hepatic artery; HAT, hepatic artery thrombosis.

The Catalyst-5 was pushed against the edge of the thrombosed segment, then slowly withdrawn while keeping a 60 mL syringe in ‘under vacuum’ aspiration (with no back flow through the syringe) and following evidence of aspirated clots (figure 2A). Nevertheless, after 1 min, angiograms showed worsening of the stenosis, with diffuse narrowing of the anastomotic site, and tendency to re-thrombosis in the downstream territory.

Figure 2.

(A) After the first attempt of HA thromboaspiration using the CAT5 catheter, the removal of some clots allows visualisation of both extrahepatic and intrahepatic HA. A stenotic narrowing persists at the anastomotic site (black arrow). (B) The NeuronMax introducer is slid over the CAT5 and used to aspirate inside the HA, closely to the anastomotic site. (C) A balloon-expandable drug-eluting coronary stent is deployed across the anastomotic site. (D) Final DSA shows revascularisation of the intrahepatic and extrahepatic arterial tree and restoration of an adequate HA calibre in the site of the surgical anastomosis. HA, hepatic artery.

Thromboaspiration was then repeated, this time sliding the NeuronMax over the Catalyst-5 and using it to aspirate inside the HA, pushing the introducer close to the anastomotic site, with aspiration of a greater quantity of clots (figure 2B). The DSA demonstrated thrombosis resolution, with persisting stenosis at the anastomotic site; therefore, a 2.5 mm×8 mm balloon-expandable drug-eluting coronary stent (STENTYS, Paris, France) was deployed across the anastomotic site (figure 2C).

Final DSA showed stenosis resolution, regular representation of the intraparenchymal arterial tree and no signs of distal embolization (figure 2D).

Outcome and follow-up

The patient remained clinically stable and afebrile in the following days, and routine Doppler sonograms confirmed HA patency, with good Resistive Index (RI≈0.7). Liver transaminases decreased and normalised on day 3; blood and coagulation tests remained normal (PLT 120 000/mm³, INR 1.09). Double antiplatelet therapy was inserted in the therapy scheme, for prophylaxis of in-stent thrombosis: aspirin 100 mg/die starting the day after the procedure, and clopidogrel 75 mg/die, starting one more day later, and kept for the next 6 months, according to ESC guidelines.³ Prophylactic anticoagulation was administered during the first week at reduced dose to balance the risk of bleeding (subcutaneous enoxaparin 2000 UI). Broad range antibiotic therapy was also held for 1 week.

On 16th postoperative day, the patient was discharged. At the 6-month follow-up visit, the patient is in good clinical condition, Doppler US shows a patent HA, with good flow indices, blood and hepatic tests are normal.

Discussion

Among the possible causes underlying to post-LT HAT, discrepancy in size between graft and recipient arteries bears a major impact, particularly in the context of ‘split’ OLT, since the smaller right HA may not match with the recipient’s common HA when they are sutured together.⁴ The clinical presentation of HAT is variable. The most prominent feature on early-HAT is a dramatic transaminases increase, possibly accompanied by a general septic state, worsened by pharmacological immunosuppression. If untreated, this condition carries a 34% of overall mortality and, among survivors, a 75% rate of allograft failure or initial dysfunction.⁵ Our patient developed an early increase of liver enzymes with normal lactates, as occurs in a minority of cases. Indeed, even asymptomatic patients are at risk for delayed complications, such as abscess formation or biliary strictures or leakage, owing to the ischaemic damage on the bile duct epithelium.

Once detected HAT, a collegial discussion took place, involving the main specialists in charge: the less invasive impact of the endovascular approach makes it a rational choice, especially 12 hours after major surgery, such as liver transplantation. Moreover, the recent advances made in endovascular therapy can assure the expertise and proper materials adaptable to many settings, in particular arterial thrombosis, even of small or delicate vessels. In this case, we considered the difference in size between receiver–donor artery as the main cause of the onset of acute thrombosis: therefore and endovascular ‘correction’ was considered more appropriate to gently remodelling the arterial size.

Several techniques have been described for treating HAT: intra-arterial infusion of thrombolytic drugs (urokinase, rtPA, streptokinase), angioplasty with or without the use of drug-coated balloons, stenting, and even thrombectomy with peripheral devices.^{6 7} Results of these techniques reported HA patency rates at 1 year, 2 years, 3 years and 5 years, respectively, of 96%, 93%, 83% and 83%, and overall survival rate at 1 year, 2 years, 3 years and 5 years, respectively, of 82%, 73%, 57% and 57%.² To our knowledge, this is the first case of post-LT HAT treated by employing a neurovascular introducer coupled with a neurovascular aspiration catheter to perform thromboaspiration. We decided to manage the surgical complication as a neurovascular emergency, applying materials and techniques to the hepatic ambit. The advantages of this approach were a less traumatic manipulation of a recent surgical site, despite using a device with a big internal lumen, thus resulting in a complete and safer thromboaspiration without administration of thrombolytic intra-arterial drugs. The underlying anastomotic stenosis repeatedly triggered HAT, as proved by our many attempts before ultimate revascularisation. Thus, stent placement was necessary to adjust the arterial calibre mismatch at the anastomotic site. Balloon-expandable coronary stents are usually the first choice for HA, because of their dimensions, increased radial force and conformability. Moreover, we employed a drug-coated stent to reduce the risk of late in-stent intimal proliferation, stenosis and occlusion, even considering the young age of the Patient. Still, these stent are designed for coronary arteries and their implantation in the HA remains off-label.⁸

Some authors support splenic artery embolization to treat post-LT splenic artery steal syndrome.⁹ Though angiographies showed intense blood flow in the splenic artery even in our case, a consultation with the surgeons convinced us to spare the splenic artery, to ensure its availability in case of future re-transplantation with spleno-hepatic arterial anastomoses.

Learning points.

• Endovascular approach should always be considered in case of post-LT hepatic artery thrombosis (HAT).

• Mechanical thromboaspiration with neurovascular devices and techniques can be efficiently applied to HAT treatment, achieving complete HA revascularisation.

• Stent placement helpfully resolves underlying anastomotic stenosis.