Objectives: Early hepatic artery thrombosis is rare but devastating in living-donor liver transplant recipients. In this study, our aim was to evaluate the results of all patients with early hepatic artery thrombosis at our center.
Materials and Methods: Between October 2004 and November 2015, 841 patients underwent liver transplant for end-stage liver disease at our center. All recipients with early hepatic artery thrombosis were identified and retrospectively analyzed. Early hepatic artery thrombosis was defined in our study as its occurrence within the first postoperative week.
Results: Early hepatic artery thrombosis was seen in 12 patients (1.8%). Seven of these 12 patients developed hepatic artery thrombosis on postoperative day 1 after intraoperative dissection of the hepatic artery with necessity of repeated reconstruction. However, the primary liver function of these patients was excellent, with nearly normal serum liver panel results. The other 5 patients were diagnosed after sudden significant increases of liver values followed by graft dysfunction within the first postoperative week. All patients were listed as high urgent and underwent retransplant.
Two patients died perioperatively due to primary nonfunction, and 2 other patients died due to liver cirrhosis based on recurrent autoimmune hepatitis 29 and 106 months after retransplant. The remaining patients were alive with good liver function after a median time of 18 months (interquartile range, 13-45 mo).
Conclusions: Early hepatic artery thrombosis is not always associated with graft dysfunction. Retransplant is still necessary due to ischemic cholangiopathy in the long-term follow-up. However, an elective retransplant, which could mean better outcomes, should be preferred instead of an emergency retransplant.
Key words : Arterial complications, Early hepatic artery thrombosis, End-stage liver disease, Retransplant
Hepatic artery reconstruction is the Achilles heel of liver transplant procedures. Complications after hepatic artery anastomosis are rare but dreaded. These complications can lead to significant increases in morbidity and mortality rates.1 The incidence of hepatic artery thrombosis (HAT) is reported to be between 2.7% and 8%.2,3 Several strategies, such as improved microsurgical techniques and use of intraoperative Doppler-ultrasonography assessments, have been developed to decrease the rate of arterial complications. Factors that can play important roles for sufficient hepatic artery reconstruction include length and diameter of the arterial ends, no intimal injury, and no arteriosclerosis. For living-donor liver transplant (LDLT) procedures, hepatic artery reconstruction can be complicated because the artery of the graft is smaller, narrower, and shorter than for deceased-donor liver transplant procedures.
It is generally accepted that early HAT without urgent revascularization or retransplant almost always leads to early graft failure and mortality.4-6 Therefore, early HAT is an accepted indication of a high urgent listing for retransplant. However, we had observed that early HAT in several patients was not associated with any graft dysfunction or increased blood liver values. In this study, our aim was to evaluate all patients with early HAT at our center.
Materials and Methods
Between October 2004 and November 2015, there were 841 liver transplant procedures performed at our center. Of these patients, 192 received deceased-donor liver transplant procedures and 649 received LDLT procedures. All patients who had intra- or postoperative HAT were retrospectively identified. Patient demographics, indication for transplant, clinical presentation, Model for End-Stage Liver Disease (MELD) score, operative data, postoperative course, primary immunosuppressive therapy, donor criteria, ischemic time, follow-up of the donor, and graft and patient survival data were recorded.
All recipients and donors were evaluated for LDLT in accordance with our standard evaluation program, and the LDLT was performed in all cases per our standard technique as described elsewhere by the same transplant team.7 All hepatic artery reconstructions were performed under a microscope (Carl Zeiss, Jena, Germany) with a microvascular surgical technique. Hepatic artery reconstruction was done between the graft’s hepatic artery and the recipient’s left/right or common hepatic artery with interrupted suture using the end-to-end technique. If these arteries were not suitable, the next option was the reconstruction with the gastroduodenal artery. The arterial reconstruction was assessed intraoperatively by Doppler ultrasonography. Flow and resistance were measured. There was no daily postoperative routine assessment of the arterial flow by Doppler ultraso-nography. The vascular flow was assessed by Doppler ultrasonography or computed tomography only in patients with intraoperative dissection of the artery and repeated arterial reconstruction, postoperative complications, or increased serum liver values.
Immunosuppressant therapy after transplant included calcineurin inhibitors and prednisone, sup-plemented with mycophenolate mofetil. Modifi-cations in doses or compounds were done on an individual basis, dependent on the clinical course. After trans-plant, patients were followed at least monthly for the first year, every third month from the second to the fourth year, and then annually. Follow-up examina-tions included history, physical examination, Doppler ultrasonography of the liver, and routine blood tests.
The study was conducted in accordance with criteria of the Declaration of Helsinki (2000) and the Declaration of Istanbul (2008).
Results are reported as medians and interquartile range (IQR). Continuous variables were compared using the Mann-Whitney test. Comparisons of proportions were performed using the Fisher exact test. The Kaplan-Meier method was used for survival calculations, and comparisons between groups were performed using the log-rank test.
There were 12 patients (1.8%), including 6 males (50%) and 6 females (50%), who underwent LDLT and developed early HAT. The median age of the 10 adult patients was 41 years (IQR, 26-47 y). The other 2 patients were children (age of 4 and 5 y). Cause of liver cirrhosis in the 2 pediatric patients (17%) was metabolic disease. Causes in the remaining 10 patients included chronic hepatitis B virus infection in 3 patients (25%), cryptogenic liver cirrhosis in 2 patients (17%), Budd-Chiari in 2 patients (17%), primary biliary cirrhosis in 1 patient (8%), auto-immune hepatitis in 1 patient (8%), and Wilson disease in 1 patient (8%). Of these patients, 1 patient also had hepatocellular carcinoma (HCC) without interventional treatment preoperatively for the HCC. The median MELD score was 14 (range, 8-17). All patients with a MELD score below 15 had either HCC in the liver or at least 1 complication of their liver cirrhosis, such as hepatic encephalopathy, recurrent variceal bleeding, therapy-resistant ascites, or complications of their metabolic disease.
Right lobe LDLT was performed in 10 patients; the 2 pediatric patients underwent LDLT with left lateral lobe graft. All liver grafts were with single right or left artery. There were no reconstructions with arterial conduit. In 7 of the 12 patients (58%), which included both pediatric patients, an intra-operative dissection of the graft’s hepatic artery was seen after the arterial reconstruction, and the arterial reconstruction was repeated several times until arterial flow could be assessed by intraoperative Doppler ultrasonography. There were no other intraoperative complications. The reconstruction of the hepatic artery in the remaining 5 patients was performed without any problems.
Intraoperative Doppler ultrasonography showed excellent flow on all vascular reconstruction, inclu-ding hepatic artery reconstruction. The anastomosis of the bile duct was duct to duct in 10 patients (83%) with trans-choledochal and trans-anastomotic placement of a 5F feeding catheter. Hepatico-jejunostomy with Roux-en-Y reconstruction was necessary in 2 patients (17%).
The median number of intraoperative transfused red blood cells was 3 units (IQR, 2-7 units). The median time of surgery was 530 minutes (IQR, 495-665 min). There were no other intraoperative complications. All patients were followed post-operatively in the intensive care unit (ICU) for a median of 4 days (IQR, 2-5 d). All patients received tacrolimus-based immunosuppressant therapy in combination with prednisone, which was sup-plemented with mycophenolate mofetil. Daily routine postoperative Doppler ultrasonography was only performed in patients with intraoperative arterial difficulties. All other patients received postoperative Doppler ultrasonography on the 1st and 5th postoperative day because the rate of HAT in patients with sufficient intraoperative Doppler ultrasonography was low at our center (0.7%). From our experience, retransplant in these particular patients is the better treatment option versus revascularization because the thrombosis often continues toward the intra-hepatic artery.
The median donor age was 39 years (IQR, 22-48 y). There were 8 female (67%) and 4 male donors (33%) with a median body mass index of 25 kg/m2 (IQR, 21-28 kg/m2). All donors were at least 4th-grade relatives of patients. None of the donors had intra- or postoperative complications. All donors were followed for 1 night in the ICU and then transferred to the normal surgical ward on postoperative day 1. The median hospital stay was 6 days (IQR, 6-10 d).
Postoperative Doppler ultrasonography and computed tomography of grafts were performed in the 7 patients with intraoperative dissection of the hepatic artery; patients were diagnosed with HAT on postoperative day 1. All 7 patients were listed as high urgent for retransplant. However, none of the 7 patients developed graft dysfunction. The initially increased liver function tests were decreased during follow-up to nearly normal serum liver values (Table 1). None of these patients had postoperative complications. The 2 pediatric patients were dis-charged during postoperative week 3 and readmitted for retransplant after 33 and 36 days. The median time between the first and second transplant was 9 days (IQR, 7-29 d). The other 5 patients underwent retransplant without any complications after a median time of 12 days (IQR, 6-20 d). Perioperative death due to primary nonfunction after retransplant occurred in 1 patient. Another patient died 29 months after retransplant due to liver cirrhosis based on recurrent autoimmune hepatitis. The remaining 5 patients are still alive with good liver function.
The other 5 patients with normal intraoperative arterial flow had initially normal postoperative liver values with good primary liver function. Within the first postoperative week, they developed sudden significant increases in liver values. All were in good general condition. Doppler ultrasonography and computed tomography showed thrombosis of the hepatic artery, which continued most likely into the intrahepatic area. Consequently, these patients were immediately listed as high urgent for liver retrans-plant since revascularization was not an effective option. Liver values continued to increase over further follow-up with graft dysfunction. Retransplant procedures were performed a median of 7 days (IQR, 6-9 d) after the first transplant without any complications. At the time of retransplant, all patients still presented with acceptable general condition without necessity of ICU treatment but with graft dysfunction, which was reflected as increased liver enzymes and increased international normalized ratio values (Table 2). One patient required addi-tional surgical procedures after retransplant due to postoperative bleeding. Another patient developed partial thrombosis of the vena cava after retransplant, which was treated successfully with conservative anticoagulation therapy. Perioperative death was seen in 1 patient due to graft failure and sepsis. There were no other postoperative com-plications. One patient died after 106 months due to graft failure based on recurrent autoimmune hepatitis.
A comparison of the 2 groups showed no significant differences in time between first and second transplant; however, liver values after the first transplant were significantly different. Of the 12 patients with early HAT and retransplant, 8 patients were still alive after a median follow-up of 18 months (IQR, 13-45 mo) (Table 3).
Liver transplant is the accepted standard therapy for end-stage liver disease, with long-term survival rates of over 85%.8-10 One serious complication of liver transplant is early HAT, which is one of the leading causes of early graft failure and patient death.4 Although the cause of HAT is not absolutely clear, there are some accepted surgical and nonsurgical risk factors described in the literature.3,11 With increasing experience and improved surgical techniques, these surgical factors are not a major risk factor for HAT in high-volume centers.12,13 The prolonged hepatic artery anastomosis time (> 80 min) is described as an independent risk factor for early HAT, which is often associated with complex artery reconstruction, such as arterial conduits or more than 1 hepatic artery of the graft.13-15 Further nonsurgical risk factors are donor age older than 50 years, lack of ABO compatibility, intraoperative use of 7 or more units of transfused blood and 6 or more units of fresh frozen plasma, previous upper abdominal surgery, and retransplant procedures.16-19 Another important factor, which is discussed as a causative aspect for HAT, is the dysfunctional coagulation system of cirrhotic patients with hypercoagulability. This may persist over variable time periods after transplant, causing vascular thrombosis.20 Some studies reported that prophylactic anticoagulant treatment could reduce the incidence of HAT.21-24 At our center, we do not perform prophylactic anticoagulation treatment during the early postoperative time. However, the rate of HAT at our center is 1.8%, which is lower than values reported in the literature of between 2.7% and 9%.23
Early HAT is defined as thrombosis of the artery within the first postoperative month. Thrombosis of the hepatic artery after the first month is considered as late HAT and is often associated with a mild clinical course without any significant increase of liver values. Retransplant is only necessary in those patients who have graft dysfunction.24,25 However, patients with early HAT almost always show signs of graft dysfunction with significantly increased liver values followed by fulminant hepatic necrosis. It is generally accepted that early HAT without urgent revascularization or retransplant almost always leads to mortality.4-6
We defined early HAT in our patients as HAT within the first postoperative week. Some of our patients with early HAT had excellent graft function and normal liver values without arterial flow on Doppler ultrasonography or on computed tomo-graphy. Our patients with intraoperative dissection of the graft hepatic artery and repeated arterial reconstruction were listed as high urgent for retransplant. However, all of these patients in our retrospective study recovered completely with good graft function. Patients were transferred to the normal surgical ward with almost normal liver values, with 2 even discharged and then readmitted for retransplant.
We do not have any data to clarify the exact pathophysiology behind this phenomenon. One possible hypothesis is that there was no reperfusion injury since the graft may be never had sufficient arterial flow, which could lead to recovery of liver function with portal flow. In a report describing 3 cases with similar phenomenon, all 3 patients developed early HAT, with 2 having intraoperative HAT and probably never having sufficient arterial flow. The patients underwent endovascular or surgical revascularization, which failed to restore arterial flow. All patients recovered completely without arterial flow and showed good liver function and normal liver values without retransplant after follow-up of 32 months, 11 months, and 4 months.26
In a rat experimental study, no differences in survival rates were shown between liver transplant rats with hepatic artery reconstruction versus liver transplant rats without hepatic artery reconstruction.27 The detailed mechanisms of hepatic arterial collateral formation are not clear. There are studies indicating that arterioportal communication occurs through the peribiliary plexus, which represents a collateral source of arterial blood to the liver when the hepatic artery is occluded.27-29 One possible explanation could be that arterioportal communication may develop better in grafts without sufficient arterial flow and any reperfusion injury. However, these are only possible explanations, and we do not have any data to prove these hypothesis. The most important consideration is whether this knowledge changes anything in the clinical course and treatment of these patients.
Incidence of HAT within 7 days after liver transplant is an indication for emergency transplant status (United Network for Organ Sharing status 1).30 Liver retransplant, especially emergency retrans-plant, is associated with a significantly higher mortality rate versus elective retransplant.31-34 Another accepted treatment option is early endovascular or surgical revascularization, which has been shown to achieve the same long-term survival rates as shown in patients without HAT.35,36
From our experience and other published data, another problem is that, even with early and suc-cessful revascularization, most patients develop ischemic cholangiopathy with necessity of repeated radiologic interventions, several periods of hospi-talization, and emotional stress.37 One more possible treatment option is interventional revascularization.38 However, interventional treatment is very much dependent on the expertise of the interventional radiologist, and even then it is not proven to be effective at reversing HAT.39 When all of these factors are considered, we believe that retransplant is the optimal treatment choice, even in patients with HAT and good liver function, since ischemic cholangiopathy is likely to occur over the long-term follow-up. However, a high-risk emergency liver retransplant may not be necessary. From our experience, it can be justified that these patients could wait as long as their biochemical and hemodynamic status remains stable and then undergo an elected retransplant, which can result in significantly better graft and patient survival. Although the number of patients in our study was too small to draw conclusions based on strong statistically proven facts, to the best of our knowledge, our study is the first to describe this phenomenon on early HAT, which adds significant value in management of early HAT and treatment of these patients to achieve significantly improved outcomes.
Early HAT is not always associated with graft dysfunction and elevated liver function tests. Especially in patients with repeated arterial recon-struction due to dissection, the liver graft is able to recover completely with normal liver values. Retransplant is in our opinion still necessary due to ischemic cholangiopathy in long-term follow-up. However, urgent retransplant in an emergency setting may be not required in these patients. An elective retransplant, which has significantly better outcomes, can be preferred.
DOI : 10.6002/ect.2017.0363
From the Department of General and Transplantation Surgery, University Hospital
of the Istanbul Bilim University, Istanbul, Turkey
Acknowledgements: This work was previously presented at the ILTS 21st Annual International Congress in Chicago, Illinois, USA (abstract confirmation number 672). The authors received no funding for this study and have no conflicts of interest to declare.
Corresponding author: Arzu Oezcelik, Department of General and Transplantation Surgery, University Hospital of the Istanbul Bilim University, Abide-i Hurriyet Cad. No: 164 Sisli, Istanbul, Turkey
Phone: +90 212 224 4950
Table 1. Liver Values of the 7 Patients With Intraoperative Hepatic Artery Complications and Repeated Arterial Reconstruction
Table 2. Median Liver Values of the 5 Patients With Postoperative Hepatic Artery Thrombosis
Table 3. Blood Values of Patients Alive After Median Follow-Up of 18 Months