A, Graft survival by groups. The differences in graft survival between the groups was not statistically significant. Group 1 patients underwent liver transplantation between June 1, 1994, and August 31, 1999; group 2, between September 1, 1999, and September 30, 2001. B, Patient survival by groups. The differences in patient survival between the groups was not statistically significant. Group 1 patients underwent liver transplantation between June 1, 1994, and August 31, 1999; group 2, between September 1, 1999, and September 30, 2001.
Nishida S, Kato T, Levi D, Naveen M, Berney T, Vianna R, Selvaggi G, Buitorago E, Al-Niami A, Nakamura N, Vaidya A, Nery J, Tzakis A. Effect of Protocol Doppler Ultrasonography and Urgent Revascularization on Early Hepatic Artery Thrombosis After Pediatric Liver Transplantation. Arch Surg. 2002;137(11):1279-1283. doi:10.1001/archsurg.137.11.1279
Protocol Doppler ultrasonography of the liver (DUSL) is useful for detecting early hepatic artery thrombosis (HAT). Urgent exploration based on DUSL findings and immediate revascularization of the liver may avoid HAT-related sequelae, namely, biliary complications and retransplantation after pediatric liver transplantation.
Group 1 included 96 liver transplantations performed in 75 pediatric patients from June 1, 1994, to August 31, 1999. Group 2 included 43 liver transplantations performed in 39 pediatric patients from September 1, 1999, to September 30, 2001.
In group 1, DUSL was performed on the first posttransplantation day or on request. Angiographic confirmation of suggested HAT was treated with thrombolysis, angioplasty, or thrombectomy. In group 2, protocol DUSL was performed every 12 hours in the first week and every 24 hours in the second week. The suspicion of HAT warranted urgent surgery without the patient undergoing angiography.
Main Outcome Measures
Incidence of HAT, biliary complications, and retransplantation. Graft and patient survival. Hospital stay, and number of admissions and operations after undergoing HAT.
The incidence of HAT was 10.4% (10 of 96 transplantations) in group 1 and 7.0% (3 of 43 transplantations) in group 2. The incidence of biliary complications after HAT was 100% in group 1 and 0% in group 2 (P = .02). The incidence of retransplantation after HAT was 90.0% (9 of 10 patients) in group 1 and 0% in group 2 (P = .01). Of the 10 patients who experienced HAT in group 1, 5 patients underwent early retransplantation (mean length of time, 13.2 days). All 5 patients who did not undergo early retransplantation had biliary complications. Four of these 5 patients underwent retrasplantation at a later time (mean length of time, 687 days). In group 2, DUSL identified early HAT in 3 patients (7.0%). Emergent thrombectomy and arterial reconstruction were undertaken. All 3 (100%) have their original graft and are alive. None experienced biliary complications. One-year graft and patient survival is 72.0% and 84.0%, respectively, in group 1 and 80.0% and 85.0%, respectively, in group 2. Shorter hospital stay, fewer readmissions, and surgery after HAT were noted in group 2.
Protocol DUSL detects early HAT and urgent revascularization based on DUSL can significantly reduce the incidence of biliary complication and graft loss requiring retransplantation in pediatric liver transplantation.
EARLY HEPATIC artery thrombosis (HAT) remains a major problem after pediatric liver transplantation. Often presenting in the early posttransplantation period, HAT commonly manifests as biliary complications and graft loss requiring retransplantation.1- 5 Early HAT occurs within 2 weeks of liver transplantation.6 Various diagnostic modalities and treatments for HAT have been reported in the literature.1- 14 The optimum diagnostic tool and treatment depends on the condition of the patient, viability of the liver, availability of specific medical expertise, and availability of organs for transplantation. Retransplantation, surgical revascularization, thrombolysis, percutaneous transluminal balloon angioplasty, or hyperbaric oxygen therapy are the therapeutic options described.14- 19 The cadaveric organ shortage, especially the difficulty finding size-matched donors for pediatric recipients makes retransplantation a difficult option. Early detection of HAT and urgent revascularization are of great importance. This approach may lead to graft salvage reducing the need for retransplantation.
Doppler ultrasonography of the liver (DUSL) is a useful modality for detecting HAT.6- 12 Graft rescue after DUSL detection of HAT has been reported after live donor liver transplantation.13 Angiography is a definitive method for confirming the diagnosis of HAT. In addition, a stenotic hepatic artery can sometimes be treated by the interventional radiologist.14 Angiography, however, is an invasive procedure that is time consuming, often delaying attempts at revascularization. Urgent revascularizations after HAT have been reported as a useful option for graft rescue.15- 18 The effect of early detection of HAT and timely intervention for long-term outcomes has not been studied well. The aim of the study is to assess the use of frequent protocol DUSL in the early posttransplantation period in detecting HAT and whether immediate surgical intervention based on the results of DUSL, without angiography, decreases biliary complications, graft loss, and the need for retransplantation due to HAT.
Between June 1, 1994, and September 30, 2001, 139 liver transplantations were performed in 114 pediatric patients at the University of Miami, Miami, Fla. The 114 patients can be divided into 2 historical groups depending on the intervention of the protocol DUSL. Patients in group 1 underwent transplantation between June 1, 1994, and August 31, 1999. This group consisted of 96 pediatric liver transplantations in 75 pediatric patients. Patients in group 2 underwent transplantation between September 1, 1999, and September 30, 2001. This group consisted of 43 transplantations in 39 pediatric patients.
Some details of group 1 include the following: there were 71 cadaveric whole-organ transplantations, 18 reduced-size transplantations, and 7 living-related donor transplantations performed in 75 pediatric patients. Forty-eight hepatic arterial reconstructions (50%) were performed between the donor and recipient hepatic arteries (standard method). Forty-eight hepatic arterial reconstructions (50%) were performed using arterial conduits. Doppler ultrasonography of the liver was performed by a radiologist on postoperative day 1 and subsequently as requested based on the clinical course of the patient. If the findings of the DUSL were suggestive of HAT, angiography was performed for most of the patients. Based on the angiogram, thrombolysis, angioplasty, or thrombectomy was performed.
Group 2 included 36 cadaveric whole-organ transplantations, 4 reduced-size transplantations, and 3 living-related donor transplantations performed in 39 patients. Eighteen hepatic arterial reconstructions (41%) were performed using the standard method. Twenty-five hepatic arterial reconstructions (59%) were performed using arterial conduits. Doppler ultrasonography of the liver was performed by a member of the transplantation team or by a radiologist (S.N., T.B., R.V., G.S., E.B., A.A.N., or T.K.) every 12 hours in the first week and every 24 hours in the next week as per protocol. If the findings from DUSL were suggestive of HAT, emergent laparotomy was performed without a confirmatory angiogram.
Patient demographics and other variables are given in Table 1. The groups were similar except for the length of ischemic time the graft was cold and warm. Doppler ultrasonography of the liver performed by the transplantation team was done using an ultrasound scanner (model SSD 1700; Aloka, Tokyo, Japan). Doppler waveforms of the proper hepatic artery in the porta hepatis and at the right or left hepatic arteries were evaluated. Color Doppler ultrasonography was used to assist in identification of the artery. Measurements obtained by DUSL included the presence or absence of hepatic artery blood flow, the resistive index, and the peak systolic velocity.
We conducted a retrospective case-control comparison of the main outcome measures of these 2 groups. Primary outcome measures were the occurrence of HAT, the occurrence of a biliary complication after HAT, the need for retransplantation after HAT, graft and patient survival, the length of the hospital stay in patients with HAT, the number of subsequent hospital readmissions, and the operative procedures performed in patients with HAT.
Data were collected from medical records and our database. Data are expressed as mean (SD). Statistical analysis was performed using the t test for numeric data, and the χ2 and Fisher exact tests for categorical data. Graft and patient survivals were generated using the Kaplan-Meier method and were compared using the log-rank test. P<.05 was considered statistically significant.
Management and outcome data are given in Table 2. Clinical characteristics of patients with HAT are listed in Table 3. No statistically significant differences between the groups were found for the incidence of HAT, graft survival, or patient survival. The incidence of late biliary complications after HAT (P = .005) and retransplantation after HAT (P = .003) were significantly lower in group 2. A trend toward a shorter hospital stay, fewer hospital readmissions, and fewer operative procedures after HAT was observed in group 2, but this difference was not statistically significant (Table 2, P = .09).
Of the 96 cases in group 1, 10 (10.4%) developed HAT diagnosed by DUSL. Seven of these patients underwent angiography and 3 were taken directly to the operating room. Median time from the transplantation until the detection of HAT was 4.89 (2.67) days (range, 1-10 days). Five (50.0%) of the 7 angiographically confirmed cases were initially treated with thrombolysis and heparin sodium anticoagulant therapy. The remaining 2 patients (20.0%) underwent thrombectomy and revascularization. In both cases the same inflow was used for the revascularization. One patient was treated by thrombectomy of the hepatic artery and revascularization with the same hepatic artery. The other was treated by thrombectomy of the infrarenal arterial conduit and revascularization with the same conduit. Of the 5 patients initially treated with thrombolysis, 3 were taken to the operating room for thrombectomy and revascularization. The remaining 2 patients were treated nonoperatively, but eventually underwent retransplantation for biliary complications leading to graft failure. Of the 3 patients who underwent thrombectomy and revascularization after attempted thrombolysis, all 3 patients had original hepatic arterial reconstruction using standard methods. All 3 patients were treated by thrombectomy of the hepatic artery and revascularization with the common hepatic artery, splenic artery, or infrarenal arterial conduit. Of these 3 patients, 2 underwent urgent retransplantation owing to acute liver failure. The other patient who did not undergo retransplantation has developed biliary strictures and undergoes frequent radiologically assisted biliary dilatation procedures. Two of the 7 cases originally confirmed by angiography eventually underwent retransplantion after having exploratory surgery that disclosed no trail of thromoblysis. Of the 3 patients (30.0%) in group 1 who were taken directly to surgery based on the DUSL, 2 underwent urgent retransplantation and the other underwent retransplantation later after an attempt was made to salvage the graft by resecting the devascularized portion.
In group 1, nine of 10 patients with HAT underwent retransplantation. Five early owing to acute liver failure and 4 late for complications related to biliary strictures and graft sepsis. One patient has not undergone retransplantation, but has experienced biliary complications. Eight of the patients are alive and 2 patients died. One died of sepsis at 653 days after undergoing transplantation and the other died of a posttransplantation lymphoproliferative disease at 129 days. Management and clinical characteristics are given in Table 2
and Table 3. The mean length of the hospital stay of patients with HAT was 71.5 (47.54) days. The mean number of admissions for patients with HAT was 7.60 (4.55). The mean number of operations for patients with HAT was 12.1 (18.4) (Table 2). One-year graft and patient survival for group 1 is 72.0% and 84.0%, respectively (Figure 1).
Of the 43 cases in group 2, there were 3 instances (7.0%) of HAT. The mean time from transplantation until the detection of HAT was 2.67 (1.15) days. Doppler ultrasonography of the liver was performed frequently according the schedule described earlier.
In all 3 cases, the patient was transferred to the operating room without undergoing angiography within 3 hours after the DUSL was performed. All had infrarenal arterial conduits in the original transplant. All had a fresh clot in the donor hepatic artery. One patient was treated with thrombectomy and revascularization using the recipient splenic artery for inflow. The other 2 patients were treated with thrombectomy and revascularization using the same infrarenal arterial conduit. After revascularization, all 3 patients demonstrated good intrahepatic arterial flow by intraoperative DUSL. All 3 patients have their original graft, are alive, and have normal liver functioning with a median follow-up of 289 days (range, 113-390 days). No late biliary complications have been observed. There was no retransplantation after HAT during this period. The mean length of stay for patients with HAT in group 2 was 18.6 (15.0) days. The mean number of admissions for patients with HAT in group 2 was 2.66 (1.15). The mean number of operations for these patients was 2.33 (0.57) (Table 2). One-year graft and patient survivals for group 2 is 80.0% and 85.0%, respectively (Figure 1).
This study demonstrates the influence of protocol DUSL as a useful test for detecting early HAT in the early postoperative period. Urgent thrombectomy and revascularization based on protocol DUSL findings of HAT significantly decreased late biliary complications, graft loss, and retransplantation after pediatric liver transplantation. The peak incidence of early HAT is during the first 2 weeks after undergoing liver transplantation.6 Urgent revascularization has become an accepted means of managing HAT, offering the potential for graft salvage.15- 18 According to our protocol, we performed DUSL twice a day, every 12 hours in the first week and once a day, every 24 hours in the next week. In cases suggestive of HAT, urgent thrombectomy and revascularization were performed owing to the results of this DUSL instead of radiographic intervention. The accuracy of the DUSL examination and the prompt response to the detection of an arterial problem accounted for the improved results.
Kok et al6 reported on the use of routine DUSL after liver transplantation. They recommended DUSL every 3 days in the early postoperative period. Sakamoto et al13 reported cases of graft salvage in recipients of live donor liver grafts who developed HAT.
In August 1999, we had reviewed our data. Mean days of HAT diagnosis was 4.89 (2.67) in group 1 (range, 1-10 days). In September 1999, based on these observations, we began performing more frequent DUSL examinations in the early posttransplantation period after patients had undergone pediatric liver transplantation. Additionally, we adopted the policy that DUSL findings suggestive of arterial abnormalities would warrant urgent exploratory surgery and revascularization of the liver graft. The goal was early recognition and correction of HAT, thus reducing its morbidity.
The standard diagnostic tool for HAT is angiography. However, this procedure is invasive, labor intensive, and time consuming. Doppler ultrasonography of the liver is noninvasive, portable, repeatable, and readily available. By performing serial examinations at frequent intervals during the first 2 weeks after the patient has undergone transplantation, HAT can be discovered when it is not yet suggested on clinical grounds. This saves precious time. In our experience, the time posttransplantation in which HAT was diagnosed was significantly shorter in group 2 (mean, 2.67 [1.15] days; range, 2-4 days). We believe that HAT occurs early in the postoperative period. Significant delay in diagnosis is the cause of morbidity in group 1. On the other hand, in group 2, based on the findings of the protocol DUSL, the patients who had suspected HAT were taken directly to the operating room for thrombectomy and revascularization. The use of DUSL in this role was accurate. Our patients had no false-negative or false-positive results. Evaluation of the liver using the DUSL in this patient population is technically straightforward, easily performed by a nonradiologist. Thus, we believe that DUSL may replace angiography as the diagnostic option for pediatric patients.
The best treatment for HAT is changing. In the past, early HAT after undgergoing liver transplantation was considered uniformly fatal if the patient did not undergo urgent retransplantation. Recently, the importance of urgent thrombectomy and revascularization has been reported.15- 18 However, biliary complication, graft loss, and late retransplantation have tempered enthusiasm for this approach.15- 18 To date, patients with HAT in group 2 have experienced no biliary complications and graft loss. The success of urgent revascularization clearly depends on early diagnosis and prompt intervention before the development of irreversible hepatic or biliary ischemia.15- 17 Once the diagnosis of HAT has been made, revascularization must be undertaken without delay. Revascularizations were achieved within 12 hours of the onset of HAT in group 2. Muiesan et al20 reported that arterial conduits provided an effective and reliable method of revascularization in patients at higher risk of arterial thrombosis. Other treatment modalities including radiological intervention and even hyperbaric oxygen treatment have been reported with limited success.14,19 However, long-term outcomes and large numbers of cases of treatment for HAT by radiological intervention after undergoing pediatric liver transplantation have not been reported and are unclear. We believe that radiological diagnosis and treatment, if performed without delay, may help to improve the results. The most effective way is to diagnose the condition promptly and to treat it without delay. Late biliary complications and graft sepsis are serious issues after the rescue of the graft from HAT. Ischemic time has an effect on graft function and late biliary complications.21,22
In our experience, the results were far better when HAT was detected early by DUSL and it was addressed immediately. In group 1, nine of 10 cases of HAT resulted in graft loss regardless of intervention. In group 2, there were 3 cases of HAT but no graft loss or biliary complications. The patients with HAT in group 1 required more interventions, readmissions, and surgical procedures than those in group 2, and still faced retransplantation. We had to perform multiple procedures including catheter insertion and drainage of the bile duct for these patients. Graft survival of group 2 was better than that of group 1 (not significant) owing to prevention of the graft loss from HAT. We did not have any graft loss from HAT in group 2. This is particularly important when facing an increasingly limited organ supply. Patient survival is the same in both groups. Patient survival of those in group 1 was improved by retransplanation.
This study demonstrated that frequent DUSL examinations as a protocol during the early postoperative period within 2 weeks is valuable for detecting HAT after pediatric liver transplantation. Urgent thrombectomy and revascularization based on the DUSL findings provide a real potential for graft salvage and in doing so avoid the high morbidity associated with this complication.
Accepted for publication May 25, 2002.
We thank Jun Kadono, MD, for his assistance in initiating the DUSL protocol and the staff of the Department of Pediatrics at the University of Miami for their important role in caring for the patients who underwent transplantation.
Reprints: Seigo Nishida, MD, Highland Professional Bldg, 1801 NW Ninth Ave, Suite 514, Miami, FL 33136 (e-mail: email@example.com).
This article was corrected for errors on September 3, 2015.