Thrombosis of the hepatic artery following liver transplantation (LT) often results in irreversible damage to the liver and still represents one of the main causes of graft loss and liver transplant recipient mortality. Its incidence in the various series ranges between 1.6 and 8.9%; graft failure leading to retransplantation and mortality of affected patients can go beyond 50%.1 Despite the clinical relevance of this complication, the possibility of preventing it through a specific pharmacological prophylaxis using antiplatelet agents has been poorly investigated.
Liver transplant recipients often have a severely impaired coagulative function at the time of transplantation that results in significant bleeding at surgery. In these circumstances, a prophylaxis of thrombotic complications in the early postoperative period can be considered hazardous by many surgeons because of the risk of postoperative bleeding.
In a recent analysis, we characterized specific risk factors for hepatic artery thrombosis (HAT) according to the timing of its development. A donor age greater than 60 years and bench reconstruction of anatomical variants of the hepatic artery were independently associated with cases of HAT that occur in the first 30 days after LT (so-called early HAT), while employment of a donor iliac artery interposition graft to the aorta or graft from a donor who died of a cerebrovascular accident were independently associated with HAT occurring after the 30th postoperative day (so-called late HAT).2 Other possible factors analyzed in that study that were not independently associated with HAT were etiology of recipient end-stage liver disease (related to hepatitis B or C, or nonviral), prior liver transplantation, donor gender, recipient gender and age (≤ 60 years vs. >60 years), preservation solution, cold ischemic time, experience of the surgeon who performed the procedure, intraoperative transfusion requirement of red blood cells and fresh frozen plasma, acute rejection confirmed by liver biopsy, and cytomegalovirus infection.
In this report, we retrospectively analyzed a large series of patients transplanted at a single institution, many of whom received aspirin in the postoperative period as prophylaxis for HAT, to assess whether the incidence of HAT was modified by this strategy, with particular regard to those patients who had specific risk factors for the development of the complication.
LT, liver transplantation; HAT, hepatic artery thrombosis.
PATIENTS AND METHODS
From April 1986 to August 2005, 984 liver transplantations were performed on 864 patients at the Department of Surgery and Transplantation of the University of Bologna. Of these 984 transplantations, 146 were excluded from analysis for the following reasons:
in 127 LTs, patient death (84 cases) or graft loss (43 cases) occurred in the first 30 postoperative days without evidence of HAT (median graft survival, 9 days; range, 0–30 days); 3 LTs were followed by HAT, in which a kinking of the hepatic artery was demonstrated (these complications were excluded from the analysis as they were considered the consequences of a technical failure); 10 LTs employed split-liver grafts from cadaveric donor (split-liver transplantation involves a higher risk of HAT)3; 2 LTs were performed on pediatric recipients (the incidence of HAT is significantly higher in pediatric LT)4; 3 LTs were performed on patients who were diagnosed with paroxysmal nocturnal hemoglobinuria after the loss of 3 consecutive grafts due to HAT5; and 1 LT involved arterial anastomosis done with a Gore-Tex graft. The analysis therefore included 838 adult LTs with a median follow-up of 1,704 days (range, 33–7,017 days) (Table 1).
Table 1. Baseline Characteristics of the Study Population (838 LTs).
No. of cases (%)
Abbreviations: HA, hepatic artery; LT, liver transplantation.
Age > 60 years
Etiology of liver disease
- Post-necrotic liver disease
- Hcc on Cirrhosis
- Alcoholic liver disease
- Fulminant Hepatitis
- Primary Biliary Cirrhosis
- Primary Sclerosing Cholangitis
- Wilson's disease
- Metabolic disease
Reconstruction of anatomical variants of HA
Use of iliac conduit
Age >60 years
Vascular accident as cause of donor death
The arterial anastomosis was usually fashioned with a running 7–0 polypropylene suture (Prolene suture; Ethicon Inc., Johnson & Johnson, Sommerville, NJ) between the celiac or common hepatic artery of the graft and the junction between the hepatic and gastroduodenal arteries of the recipient. When the hepatic artery of the recipient was not suitable for a direct anastomosis, a donor iliac artery graft was used as an interpositional conduit between the recipient's infrarenal aorta and the donor's celiac or hepatic artery.
Duplex Doppler ultrasonography was performed at least twice in the first postoperative week starting from the first postoperative day, at the follow-up visits after discharge from the hospital as a routine test of vascular patency, and whenever indicated by clinical of biochemical findings.
Starting from the mid-1990s, antiplatelet prophylaxis with aspirin (100 mg per day) was administered after 237 LTs; the median starting time was 8 postoperative days (range, 0–29 days). No strict criteria were established to allocate patients to the antiplatelet treatment, but the decision was left to the judgment of the surgeon in charge of the patient's management: in general, aspirin administration was considered for those patients who had one or more of the possible risk factors for HAT. Antiplatelet prophylaxis was temporarily withheld from patients whose hemostasis at surgery was problematic and who showed blood loss from the abdominal drains or required blood transfusion in the postoperative period or whose coagulative function was severely impaired. Due to the different levels of confidence in handling hemostatic problems in the perioperative that have been developed with time, it is not possible to define values of one of the coagulative parameters that were employed as a threshold to allocate patients to treatment.
HAT was defined as the complete occlusion of arterial blood flow to the allograft and classified as “early” when it occurred within 1 month after transplantation or as “late” when it occurred after this timepoint. The diagnosis was based on Doppler ultrasonogram findings and confirmed by a hepatic angiogram. The risk factors for HAT that were retrospectively considered are those previously reported and already mentioned in the introduction.2
Continuous variables were reported as medians and ranges. Categorical variables were reported as number of cases and prevalence and differences in subgroups were compared using the Fisher exact test. P values less than 0.05 were considered statistically significant in all the analyses. Statistical analysis was performed using the SPSS for Windows package (Ver. 10.0, SPSS Inc., Chicago, IL).
Among the 838 LTs forming the initial study group, 29 cases (3.5%) of early HAT were observed. Incidence of early HAT (donor age above 60 years and/or bench reconstruction of anatomical variants of the hepatic artery) was significantly higher among those transplants that had risk factors for early HAT (18 HATs per 307 LTs at risk; 5.86%) in comparison to that observed in absence of specific risk factors (11 HATs per 531 LTs; 2.07%) (P = 0.002).
Median time of presentation of posttransplantation early HAT was 5 days (range, 1–28 days). The time of presentation precluded the analysis of a possible effect of antiplatelet prophylaxis on this type of arterial complication. In fact, the median starting time of antiplatelet therapy was 8 days (range, 1 to 29 days) and therefore after the median time of presentation of early HAT. In detail, antiplatelet prophylaxis had already been instituted in only 2 instances of early HAT at the time when they developed. Of the 29 patients who developed early HAT, 19 underwent retransplantation, 4 were successfully revascularized, and 6 died; the latter 10 patients were excluded from the analysis of late HAT.
The study group for late HAT was comprised of 828 LTs (Fig. 1); of them, 498 (60.1%) had specific risk factors for late HAT (LTs performed using graft from donors who died of a cerebrovascular accident and/or using interpositional iliac artery donor grafts). A total of 14 cases (1.7%) of late HAT were observed with a median presentation time of 500.5 days (50–2,405 days). Of 828 LTs, 236 (28.5%) were maintained under antiplatelet prophylaxis. Late HAT occurred in 1 of 236 (0.4%) patients who were maintained under antiplatelet prophylaxis and in 13 of 592 (2.2%) who did not receive prophylaxis (P = 0.049). At the time of the diagnosis, 4 patients were asymptomatic, 4 had mild cholestasis, and 6 had episodes of cholangitis. Of these patients, 11 underwent retransplantation due to widespread destruction of the intrahepatic biliary tree, 1 is presently awaiting retransplantation, and the other 2 whose ischemic damage was confined to the common bile duct were successfully treated with a Roux-en-Y choledochojejunostomy.
Of the 14 cases of late HAT, 13 (92.9%) developed among those 498 patients (2.6%) who had specific risk factors for the development of this complication, while only 1 late HAT was observed among the 330 patients (0.3%) who had no risk factors (P = 0.007). Of the 498 cases at risk, 160 received antiplatelet prophylaxis while the other 338 did not. The incidence of late HAT was significantly higher among those patients who did not receive aspirin (12/338; 3.6%), as compared to those who received the prophylaxis (1/160; 0.6%) (P = 0.037).
Among 330 patients without risk factors, 76 received antiplatelet propyhlaxis while the other 254 did not. In this group, late HAT was observed only among transplant patients who did not receive antiplatelet propyhlaxis (0.4%; P =0.769).
The prevalence of late HAT in patients at high or low risk receiving or not receiving antiplatelet therapy are summarized in Figure 1. No episode of upper or lower gastrointestinal bleeding was observed among patients who received antiplatelet prophylaxis.
Occlusion of the hepatic artery after LT represents a potentially catastrophic occurrence. Collateral arterial flow to the graft that may develop late after LT through peridiaphragmatic vessels6 or through the Roux limb where a choledochojejunostomy is performed at transplantation7 is absent in the first weeks after LT, and the occlusion of the hepatic artery during this time period can produce irreversible ischemic damage to the liver, leading to hepatic necrosis and failure. If early HAT is recognized within hours of its development, salvage maneuvers, beginning with surgical revascularization, can be attempted with a reasonable chance of success. Otherwise, graft loss requiring emergency retransplantation or death of the patient from hepatic failure or sepsis are the outcome of this complication.8 Great effort has been made to institute intensive screening programs based on frequent Doppler ultrasound investigations in the postoperative period to detect early HAT before the development of its clinical signs when the graft can still be rescued.
HAT developing after the 30th postoperative day (so-called late HAT) has a more insidious clinical course and can be asymptomatic or produce initially minimal biochemical abnormalities, thus delaying the time of the diagnosis.9, 10 Moreover, when HAT develops several months or years after transplantation, the chances of detecting it with Doppler screening before the clinical signs of onset are largely reduced by the lower frequency of outpatient controls. Although a quick progression toward hepatic failure is rarely observed in late HAT, the final outcome of this complication in the majority of the cases is graft loss due to widespread necrosis of the bile ducts.
In the largest of the few series published on late HAT, 8 out of 11 patients who experienced this complication required retransplantation, and 1 who was not retransplanted died of recurrent sepsis.10 In this report, despite a frequently mild clinical picture at the time of the diagnosis, 11 out of 14 patients with late HAT eventually lost their graft.
Postoperative pharmacological prophylaxis with antiplatelet agents can reduce the risk of arterial anastomosis thrombosis during vascular surgical procedures.11 However, the efficacy of this strategy in the prevention of HAT after LT has not been adequately investigated. To our knowledge, there is only 1 report available in the literature on antiplatelet prophylaxis after LT. In 1997 Wolf et al. reported a retrospective review of a series of patients transplanted between 1988 and 1993. Although the series included a significant number of pediatric recipients, the authors found no significant difference in the incidence of early HAT between patients who did and did not receive aspirin, while the effect of the prophylaxis on late HAT was not investigated.12
In a recent study, we demonstrated that among the risk factors for HAT assessed with multivariate analysis, donor age greater than 60 years and bench reconstruction of anatomical variants of the hepatic artery were independently associated with early HAT, and employment of a donor iliac artery interposition graft to the aorta or graft from a donor who had died of a cerebrovascular accident were independently associated with late HAT.2 In this report, the effect of antiplatelet prophylaxis on early HAT could not be assessed, as many of the early HATs observed in our series developed before the beginning of the prophylaxis. Due to the severe impairment of the coagulative function often present in the first few days after LT and the high risk of bleeding in those patients who had severe portal hypertension and profuse hemorrhage at surgery, aspirin cannot be safely started in the immediate postoperative period in many patients, and its possible effect in preventing early HAT is therefore difficult to assess.
When the effect of aspirin was evaluated with regard to the prevention of late HAT, however, a lower incidence of this complication was observed in those patients under prophylaxis. The benefit of prophylaxis was even more evident in patients who had specific risk factors for the development of late HAT. Among the 498 patients in our investigation, only 1 case of late HAT occurred among the 160 who had received prophylaxis, while the incidence of the complication was significantly higher in the patients who did not receive aspirin.
Our results seem to indicate that treatment with aspirin is effective in preventing late HAT after LT, that it is not associated with any major complication, and that it should therefore be considered, particularly for those LT recipients at risk for late HAT. As some of the cases of late HAT developed years after LT, we believe that prophylaxis should not be discontinued at any time. It would be advisable, however, to suspend administration of aspirin before invasive procedures, such as liver biopsy or repeated surgery, are carried out. Prospective studies, preferably multicentric and randomized, could provide further confirmation of our results and help to better define the selection criteria for prophylaxis.