- Email: [email protected]
Intra-Arterial Thrombolysis for Hepatic Artery Thrombosis following Liver Transplantation
CLINICAL STUDY
Intraarterial Thrombolysis for Hepatic Artery Thrombosis following Liver Transplantation Matthew J. Kogut, MD, David S. Shin, MD, Siddharth A. Padia, MD, Guy E. Johnson, MD, Daniel S. Hippe, MS, and Karim Valji, MD
ABSTRACT Purpose: Hepatic artery thrombosis (HAT) is a major cause of morbidity and death following liver transplantation. The purpose of this study was to evaluate the safety and efficacy of intraarterial thrombolysis (IAT) in liver transplant recipients with HAT. Materials and Methods: Adult liver transplant recipients who underwent attempted IAT for HAT were identified. This included patients with early and late HAT (occurring less than or greater than 30 d after transplantation). Records were reviewed to determine the rates of technical success, complications, surgical revascularization, repeat liver transplantation, and ischemic cholangiopathy. Results: Twenty-six patients underwent attempted thrombolysis, 13 of whom had early HAT. IAT was successfully initiated in 23 patients (88%), with a median IAT duration of 28 hours (range, 12–90 h). Recanalization was achieved in 12 patients (46%). Major complications were observed in 11 patients (42%). The early HAT group showed a trend toward increased major bleeding compared with the late HAT group (50% vs 9%; P ¼ .07). Among 12 patients who had technically successful thrombolysis, five (42%) required surgical revascularization or repeat transplantation within 2 months. At 6 months after thrombolysis, 45% with unsuccessful recanalization avoided surgery or development of ischemic cholangiopathy, similar to the proportion in those who had successful recanalization (42%; P ¼ .88). Conclusions: Posttransplantation hepatic artery thrombolysis yields suboptimal results with a high complication rate, especially in early HAT. Even with successful restoration of flow, clinical outcomes are poor. Although thrombolysis may still be considered in view of the limited treatment options for HAT, awareness of potential complications and suboptimal success rate is essential.
ABBREVIATIONS HAT = hepatic artery thrombosis, IAT = intraarterial thrombolysis, TPA = tissue plasminogen activator
Hepatic artery thrombosis (HAT) occurs in 2.5%–4.9% of patients after adult liver transplantation (1–4). Risk factors for HAT include repeat transplantation, small vessels, donor/recipient size mismatch, cytomegalovirus, complex anatomy, hypercoagulable state, surgical technical errors, and development of hepatic artery stenosis
From the Interventional Radiology Section (M.J.K., S.A.P., G.E.J., K.V.) and Department of Radiology (M.J.K., D.S.S., S.A.P., G.E.J., D.S.H., K.V.), University of Washington, 1959 NE Pacific St., Seattle, WA 98195. Received April 6, 2015; final revision received June 1, 2015; accepted June 2, 2015. Address correspondence to M.J.K.; E-mail: [email protected] From the SIR 2014 Annual Meeting. None of the authors have identified a conflict of interest. Published by Elsevier, Inc., on behalf of SIR. J Vasc Interv Radiol 2015; XX:]]]–]]] http://dx.doi.org/10.1016/j.jvir.2015.06.008
(5,6). Although the clinical presentation and outcomes of HAT are widely variable, this can be a devastating complication with high rates of morbidity, often as a result of biliary necrosis and graft failure. Mortality rates after HAT range from 27% to 58% (5,6). Management of HAT is not standardized; practice varies widely, from repeat liver transplantation and surgical revascularization to endovascular techniques, including pharmacologic intraarterial thrombolysis (IAT), mechanical thrombectomy, balloon angioplasty, and stent placement. HAT diagnosed within 30 days of transplantation is most commonly regarded as early HAT (1–3,5), and HAT diagnosed thereafter is considered late HAT. This temporal distinction is important because the etiology of HAT may differ between the two time periods, and consequently the treatments and natural histories may differ. Early HAT is often related to surgical techniques, as in cases of redundancy and
2
’
Intraarterial Thrombolysis for Hepatic Artery Thrombosis after Liver Transplantation
kinking of the arterial anastomosis (1). Late HAT may be more often related to an underlying stenosis that has developed over time, such as with intimal hyperplasia. In addition, collateral vessel formation may play a more significant role in later periods after transplantation (6,7). The purpose of the present study was to assess the safety as well as short- and long-term efficacy of hepatic arterial thrombolysis in patients with early or late HAT after liver transplantation.
MATERIALS AND METHODS Institutional review board approval was granted for this study, with a waiver of informed consent.
Patients The present retrospective study was conducted at a single institution by querying the hospital’s radiology information system to identify all patients who underwent hepatic arterial angiography of their liver transplants from February 2000 to February 2013. All patients with history of liver transplantation and diagnosis of HAT with attempted endovascular treatment were included. Patients who did not undergo angiography, patients with normal hepatic angiograms, or patients with hepatic artery stenosis without thrombosis were excluded. Baseline demographic data and liver function laboratory values are presented in Table 1. Table 1 . Patient Demographics, Underlying Liver Disease, and Timing of HAT (N ¼ 26) Characteristic
Value
Sex Male Female Age at diagnosis (y) Median Range Time from transplantation (d) Median
18 (69)
’
JVIR
Endovascular Technique Thrombolysis techniques varied slightly among interventional radiologists. However, in general, access into the femoral artery was obtained with insertion of a 5- or 6-F sheath. Diagnostic arteriography was performed (Fig 1a). Attempts to cross the occluded segment were made by using a 0.014- or 0.018-inch guide wire with a microcatheter. If attempts failed, additional attempts were made with a 0.035-inch guide wire. The thrombosed segment was completely crossed in 14 patients, and continuous infusion thrombolysis was then performed through a multiple–side-hole infusion catheter (Fig 1b). Thrombolysis was initiated through an end-hole microcatheter or partially embedded multiple–side-hole catheter in the other nine patients. Thrombolytic agents included tissue plasminogen activator (TPA) and urokinase. Intravenous or intraarterial heparin was concurrently administered. An AngioJet catheter (Possis, Minneapolis, Minnesota) was used in one patient in usual thrombectomy mode on day 1 before continuous infusion of TPA was initiated. In five patients, the sheath was positioned in the celiac artery; the remaining sheaths were positioned in the external iliac artery. Although TPA dosing was variable, in general, thrombolytic dose reductions were not performed specifically for early HAT. Some patients with early HAT had surgical drains in position, which may allow for earlier detection of arterial hemorrhage. Angiography was repeated within 24 hours of initiation of thrombolysis (Fig 1c). Adjunctive techniques after day 1 of continuousinfusion thrombolysis included balloon maceration and/or attempted angioplasty, or stent placement. The decision to continue thrombolysis was made by the interventional radiologist and the referring transplant surgeon based on follow-up angiographic findings and the patient’s clinical status.
8 (31)
Measured Outcomes 55 43–70 35
Range Early (r 30 d)
3–4,985 13 (50)
Late (4 30 d)
13 (50)
Recurrent thrombosis Liver disease before transplantation
4 (15)
Alcohol
7 (27)
Hepatitis B virus Hepatitis C virus
0 17 (65)
Hepatocellular carcinoma
8 (31)
Autoimmune Budd–Chiari syndrome
2 (8) 2 (8)
Hemochromatosis
1 (4)
Cryptogenic/idiopathic
3 (12)
Note–Values in parentheses are percentages. HAT ¼ hepatic artery thrombosis.
Kogut et al
The primary technical outcome of the present study was recanalization of the occluded hepatic artery. The primary clinical outcomes included death, repeat liver transplantation, or surgical revascularization, whichever occurred first. The secondary outcomes measured were intraprocedural hepatic artery dissection and/or perforation, postprocedural bleeding complications, recurrent thrombosis identified on follow-up imaging, and ischemic cholangiopathy. Technical success was defined as restoration of flow within the treated hepatic artery without residual flowlimiting thrombus. Procedure-related complications were recorded and categorized as major or minor based on Society of Interventional Radiology reporting standards (8). A minor complication was defined as a bleeding event that did not require additional management. Major complications included hemorrhage requiring red blood cell transfusion and/or exploratory laparotomy (Fig 2) and hepatic artery dissection and/or perforation leading to
Volume XX
’
Number X
’
Month
’
2015
3
Figure 1. (a) Angiography 6 days after hepatic transplantation shows occlusive thrombus protruding into the recipient common hepatic artery (arrow). (b) Arrowheads point to the radiopaque markers on the multiple–side-hole infusion catheter that was placed across the thrombus. A sheath is in the origin of the celiac artery, and surgical drains are in place. (c) Angiography 18 hours after the initiation of chemical thrombolysis shows restoration of flow through the hepatic arterial system with mild residual irregularity at the arterial anastomosis. Despite a good technical result, the patient received a second transplant 2 months later as a result of the development of ischemic cholangiopathy.
were conducted with the statistical computing language R (version 2.14.1; R Foundation for Statistical Computing, Vienna, Austria). Two-tailed tests were used, with statistical significance defined as P o .05.
RESULTS
Figure 2. On day 2 of thrombolysis, hemodynamic instability developed in this patient after arrival in the angiography suite. Angiography was compromised by motion artifact but showed some flow with significant residual thrombus. CT shows hematoma throughout the abdomen (arrows), with contrast medium from the recent angiography (arrowhead) near the porta hepatis. The patient’s abdominal hematoma was evacuated on an emergent basis, and surgical thrombectomy with anastomotic revision was performed.
surgical revision and/or termination. The time to clinical outcome was calculated as the number of days after thrombolysis was initiated. Except for death, clinical outcomes after repeat transplantation or surgical revascularization were not assessed, as these events were considered clinical endpoints.
Statistical Analysis Categoric and continuous variables were compared between groups by Fisher exact test and Mann–Whitney test, respectively. Rates of each clinical outcome were compared between groups by using a Gray test to account for competing risk. All statistical calculations
During the study period, 26 patients underwent angiography and attempted IAT to treat posttransplantation HAT. Table 1 summarizes patient demographics and baseline clinical history before IAT. Thirteen patients (50%) had developed early HAT. Four patients had recurrent thrombosis, and all were previously treated with surgical thrombectomy and arterial anastomotic revision. Of the 26 patients who underwent angiography with the intent to perform IAT, thrombolysis was initiated in 23 (88%). Three patients (12%) did not receive thrombolytic agents because of intraprocedural hepatic artery perforation in one patient and inability to penetrate or traverse the thrombus in two patients. Ten patients received balloon maceration and/or attempted angioplasty. One patient received a stent. Of the 23 patients in whom thrombolysis was initiated, 22 (96%) received TPA (alteplase; Genentech, South San Francisco, California), with doses ranging from 0.25 to 2.0 mg/h. One patient received urokinase, with a dose of 100,000 U/h. Tables 2 and 3 summarize procedural details and treatment outcomes in different subgroups. Of the 23 patients who underwent thrombolysis, the duration of thrombolysis ranged from 12 to 90 hours (median, 28 h). Successful recanalization (defined as restoration of flow without flow-limiting thrombus) was achieved in 12 of 26 patients (46%). There were a total of 11 major complications and one minor complication from the procedures. The minor complication was a groin hematoma at the femoral access site, which was
4
’
Intraarterial Thrombolysis for Hepatic Artery Thrombosis after Liver Transplantation
Kogut et al
’
JVIR
Table 2 . Differences in Patient Characteristics, Procedural Details, and Outcomes between Early and Late HAT Groups (N ¼ 26) Characteristic Male sex
Early HAT (n ¼ 13) 9 (69)
Late HAT (n ¼ 13) 9 (69)
56 43–70
54 44–68
Age at diagnosis (y) Median Range
P Value* 4.99 .56
Recurrent thrombosis
3 (23)
1 (8)
.59
Successfully initiated thrombolysis Lysis time (h)†
12 (92)
11 (85)
4.99 .019
Median
24
43
Range Underwent balloon PTA and/or clot maceration
12–42 5 (38)
17–90 5 (38)
Successful recanalization
8 (62)
4 (31)
Underlying abnormality None identified Stenosis Procedural complications Major
4.99 .24 .70
5 (38)
7 (54)
8 (62)
6 (46)
8 (62) 8 (62)
4 (31) 3 (23)
.24 .11
0
1 (8)
–
Hemorrhage after thrombolysis Ischemic cholangiopathy
6 (46) 2 (15)
1 (8) 5 (38)
.073 .42‡
Surgical revision
4 (31)
1 (8)
.15‡
Repeat transplantation Death
4 (31) 3 (23)
2 (15) 4 (31)
.074‡ .19‡
Minor
Note–Values in parentheses are percentages. HAT ¼ hepatic artery thrombosis, PTA ¼ percutaneous transluminal angioplasty. n Fisher exact test to compare categoric variables and Mann–Whitney test for continuous variables. † Including patients in whom thrombolysis could be initiated (n ¼ 23). ‡ Gray test for comparison of cumulative incidence functions accounting for time to event.
treated conservatively. Major complications included hepatic artery dissection or perforation (n = 4), bleeding at the access site requiring transfusion (n = 2), and intraabdominal hemorrhage requiring transfusion or exploratory laparotomy (n = 5). Patient clinical follow-up ranged from 7 days to 13 years (median, 1,156 d). Five patients (19%) underwent surgical revision of their anastomosis after angiography, all within 4 days after IAT. Three of these five patients had undergone technically successful thrombolysis, but still required surgical revision. Six other patients (23%) underwent repeat liver transplantation, four of which occurred within 42 days (median, 30 d; range, 3–625 d). Differences between those with early and late HAT are described in Table 2. Thrombolysis time was significantly shorter for those with early HAT (median, 24 h vs 43 h; P ¼ .02), yet there was a trend toward a higher rate of hemorrhagic complications in the early HAT group (46% vs 8%; P ¼ .07). No significant differences in clinical outcome or complication rates were seen when comparing patients who had restoration of hepatic arterial flow and those in whom thrombolysis failed (Table 3). Of the 12 patients with initial technically successful recanalization, arterial reocclusion developed in four (33%), as demonstrated by follow-up ultrasound (US) or magnetic resonance imaging,
all within the first year. Three patients (25%) who initially had successful recanalization experienced ischemic cholangiopathy, compared with four (29%) in the group without successful recanalization (P ¼ .90). Seven patients in total (58%) underwent surgical revision (n ¼ 3) or repeat transplantation (n ¼ 4) after successful recanalization. All these surgical revisions were performed within 4 days of recanalization.
DISCUSSION HAT after liver transplantation can be a devastating complication that is a challenge to manage. The natural history of untreated HAT is not fully understood and is therefore unpredictable. One author (9) estimated that one third of HAT cases are asymptomatic, one third will lead to chronic ischemic cholangiopathy, and one third will lead to hepatic necrosis and death. Repeat transplantation has typically been considered the treatment of choice for HAT. However, given the complexity and morbidity of liver transplantation and the scarcity of available organs, salvage of the graft without repeat operation is desirable. These challenges have led to increasing interest in endovascular management.
Volume XX
’
Number X
’
Month
’
2015
5
Table 3 . Differences in Patient Characteristics, Procedural Details, and Outcomes between Successful Recanalization and Failed Thrombolysis Groups (N ¼ 26) Characteristic
Recanalized (n ¼ 12)
Not Recanalized (n ¼ 14)
P Value*
7 (58)
11 (79)
.40 .94
Male sex Age at diagnosis (y) Median
56
54
Range Early HAT
44–70 8 (67)
43–68 5 (36)
.24
Recurrent thrombosis
2 (17)
2 (14)
4.99
32
25
14–90
12–65
5 (42)
5 (36)
5 (42)
7 (50)
7 (58) 5 (42)
7 (50) 7 (50)
.71
4 (33)
7 (50)
.45
Minor Hemorrhage after thrombolysis
1 (8) 3 (25)
0 4 (29)
.67
Ischemic cholangiopathy
3 (25)
4 (29)
.90‡
Surgical revision Retransplantation
3 (25) 4 (33)
2 (14) 2 (14)
.53‡ .61‡
Death
3 (25)
4 (29)
.79
Lysis time (h)† Median Range Underwent balloon PTA and/or clot maceration Underlying abnormality None identified Stenosis Procedural complications Major
.83
4.99 4.99
Note–Values in parentheses are percentages. HAT ¼ hepatic artery thrombosis, PTA ¼ percutaneous transluminal angioplasty. n Fisher exact test to compare categoric variables and the Mann–Whitney test for continuous variables. † Including patients in whom thrombolysis could be initiated (n ¼ 23). ‡ Gray test for comparison of cumulative incidence functions accounting for time to event.
Practical management of HAT varies, stemming from a paucity of available data related to its low incidence and differences in local surgical and interventional radiology expertise. Unfortunately, the existing body of knowledge relevant to endovascular treatment of posttransplantation HAT is problematic; most of the literature is conflicting, and many of the reports evaluate technical success without assessing clinical benefit. One recent metaanalysis of 69 cases of endovascular treatment for HAT from 16 separate series (6) estimated the technical success rate at 68%. However, the definition of success varies among publications, which limits meaningful comparisons across the studies. Although most reports have focused on the immediate technical outcomes of endovascular therapy for HAT, these may not correlate with any clinically meaningful outcomes over time. In addition, recurrent thrombosis has been reported in as many as 33% of cases (10). The present study specifically evaluates technical success and correlates it with clinical outcomes in patients who were primarily treated with IAT. In one series of six patients with HAT (11), half of which were early, the technical success rate of endovascular treatment was low, with restoration of flow achieved in only one patient. Most failures resulted from an inability to cross the occluded vessel, and thrombolysis was initiated in only two patients. In contrast, we failed to
initiate thrombolysis in only a small minority of patients (three of 26; 12%). Restoration of adequate flow was achieved in 12 of 26 patients (46%) in the present series. However, among those who had successful thrombolysis, five of 12 (42%) still required surgical revascularization or repeat transplantation within 2 months. Another five patients (42%) who underwent successful thrombolysis continued to have adequate hepatic function at a median of 828 days despite only two (40%) showing proper hepatic arterial patency on follow-up Doppler US. A slightly smaller proportion of patients with failed thrombolysis (three of 11; 27%) maintained adequate hepatic function at a median of 2,762 days. Our experience suggests that outcomes remain suboptimal even when IAT is technically successful, and any long-term clinical benefit is only questionably attributable to successful recanalization. Recent major surgery is typically cited as a relative contraindication to IAT. The present series found an overall major bleeding complication rate of 30%. The early HAT group showed a trend toward increased major bleeding rate compared with the late HAT group (50% vs 9%; P ¼ .07) despite significantly shorter thrombolysis duration (median, 24 h vs 43 h; P ¼ .02). This would be expected given recent major surgery.
6
’
Intraarterial Thrombolysis for Hepatic Artery Thrombosis after Liver Transplantation
Limitations to the present study include the inherent biases of retrospective investigations. As a result, there are expected differences in operator techniques. During the study period, changes in pharmacology and technology may have also influenced technical success. We were also limited by the small sample size. However, given the relatively low incidence of HAT in the liver transplant population, it would be challenging to report on a large number of patients at a single institution. The present study does not address the cases that were managed successfully with combined thrombolysis and surgical revision, as surgical revision was considered an endpoint. Given that successful recanalization may reveal an abnormality that is best managed surgically, long-term clinical outcomes may be more related to subsequent interventions rather than the thrombolysis. One could argue that technically successful thrombolysis could lead to improved outcomes of surgical revision by revealing underlying abnormalities that could be treated surgically in a semielective fashion. It is possible that patients in this subgroup would have the best long-term clinical outcomes. Although this would be an interesting future study, given our small population, a comparison of this subgroup versus patients who had surgical revision alone for HAT was beyond the scope of the present study. In conclusion, the present series demonstrates a relatively high major complication rate for thrombolysis performed for HAT. Technical success is fairly limited, and even in cases with successful recanalization, this may not translate into long-term hepatic graft viability and clinical benefit. In addition, some patients do well clinically in the absence of successful restoration of hepatic artery flow. As there are limited therapeutic options in this condition, thrombolysis could still be considered in the management decision in conjunction
Kogut et al
’
JVIR
with the transplant surgery team. However, awareness of the high complication rate and limited technical and clinical success is imperative during consultation with other medical providers and during informed consent discussion with the patient. Further research is needed to identify the patients most likely to benefit from aggressive treatment of HAT, as well as patients who may do well without any intervention.
REFERENCES 1. Pareja E, Cortes M, Navarro R, et al. Vascular complications after orthotopic liver transplantation: hepatic artery thrombosis. Transplant Proc 2010; 42:2970–2972. 2. Stange B. Hepatic artery thrombosis after adult liver transplantation. Liver Transpl 2003; 9:612–620. 3. Silva MA, Jambulingam PS, Gunson BK, et al. Hepatic artery thrombosis following orthotopic liver transplantation: A 10-year experience from a single centre in the United Kingdom. Liver Transpl 2005; 12:146–151. 4. Akun E, Yaprak O, Killi R, et al. Vascular complications in hepatic transplantation: single-center experience in 14 years. Transplant Proc 2012; 44:1368–1372. 5. Bekker J, Ploem S, de Jong KP. Early hepatic artery thrombosis after liver transplantation: A systematic review of the incidence, outcome and risk factors. Am J Transplant 2009; 9:746–757. 6. Singhal A, Stokes K, Sebastian A, et al. Endovascular treatment of hepatic artery thrombosis following liver transplantation. Transpl Int 2010; 23:245–256. 7. Wozney P, Zajko AB, Bron KM, et al. Vascular complications after liver transplantation: a 5-year experience. AJR 1986; 147:657–663. 8. Omary RA, Bettmann MA, et al. Quality improvement guidelines for the reporting and archiving of interventional radiology procedures. JVIR 2003; 14:S293–S295. 9. Pinna AD, Smith CV, Furukawa H, et al. Urgent revascularization of liver allografts after early hepatic artery thrombosis. Transplantation 1996; 62 (11):1584. 10. Abdelaziz O, Hosny K, Amin A, et al. Endovascular management of early hepatic artery thrombosis after living donor liver transplantation. Transpl Int 2012; 25:847–856. 11. Sabri SS, Saad WEA, Schmitt TM, et al. Endovascular therapy for hepatic artery Stenosis and thrombosis following liver transplantation. Vasc Endovascular Surg 2011; 45:447–452.
Intraarterial Thrombolysis for Hepatic Artery Thrombosis following Liver Transplantation Matthew J. Kogut, MD, David S. Shin, MD, Siddharth A. Padia, MD, Guy E. Johnson, MD, Daniel S. Hippe, MS, and Karim Valji, MD
ABSTRACT Purpose: Hepatic artery thrombosis (HAT) is a major cause of morbidity and death following liver transplantation. The purpose of this study was to evaluate the safety and efficacy of intraarterial thrombolysis (IAT) in liver transplant recipients with HAT. Materials and Methods: Adult liver transplant recipients who underwent attempted IAT for HAT were identified. This included patients with early and late HAT (occurring less than or greater than 30 d after transplantation). Records were reviewed to determine the rates of technical success, complications, surgical revascularization, repeat liver transplantation, and ischemic cholangiopathy. Results: Twenty-six patients underwent attempted thrombolysis, 13 of whom had early HAT. IAT was successfully initiated in 23 patients (88%), with a median IAT duration of 28 hours (range, 12–90 h). Recanalization was achieved in 12 patients (46%). Major complications were observed in 11 patients (42%). The early HAT group showed a trend toward increased major bleeding compared with the late HAT group (50% vs 9%; P ¼ .07). Among 12 patients who had technically successful thrombolysis, five (42%) required surgical revascularization or repeat transplantation within 2 months. At 6 months after thrombolysis, 45% with unsuccessful recanalization avoided surgery or development of ischemic cholangiopathy, similar to the proportion in those who had successful recanalization (42%; P ¼ .88). Conclusions: Posttransplantation hepatic artery thrombolysis yields suboptimal results with a high complication rate, especially in early HAT. Even with successful restoration of flow, clinical outcomes are poor. Although thrombolysis may still be considered in view of the limited treatment options for HAT, awareness of potential complications and suboptimal success rate is essential.
ABBREVIATIONS HAT = hepatic artery thrombosis, IAT = intraarterial thrombolysis, TPA = tissue plasminogen activator
Hepatic artery thrombosis (HAT) occurs in 2.5%–4.9% of patients after adult liver transplantation (1–4). Risk factors for HAT include repeat transplantation, small vessels, donor/recipient size mismatch, cytomegalovirus, complex anatomy, hypercoagulable state, surgical technical errors, and development of hepatic artery stenosis
From the Interventional Radiology Section (M.J.K., S.A.P., G.E.J., K.V.) and Department of Radiology (M.J.K., D.S.S., S.A.P., G.E.J., D.S.H., K.V.), University of Washington, 1959 NE Pacific St., Seattle, WA 98195. Received April 6, 2015; final revision received June 1, 2015; accepted June 2, 2015. Address correspondence to M.J.K.; E-mail: [email protected] From the SIR 2014 Annual Meeting. None of the authors have identified a conflict of interest. Published by Elsevier, Inc., on behalf of SIR. J Vasc Interv Radiol 2015; XX:]]]–]]] http://dx.doi.org/10.1016/j.jvir.2015.06.008
(5,6). Although the clinical presentation and outcomes of HAT are widely variable, this can be a devastating complication with high rates of morbidity, often as a result of biliary necrosis and graft failure. Mortality rates after HAT range from 27% to 58% (5,6). Management of HAT is not standardized; practice varies widely, from repeat liver transplantation and surgical revascularization to endovascular techniques, including pharmacologic intraarterial thrombolysis (IAT), mechanical thrombectomy, balloon angioplasty, and stent placement. HAT diagnosed within 30 days of transplantation is most commonly regarded as early HAT (1–3,5), and HAT diagnosed thereafter is considered late HAT. This temporal distinction is important because the etiology of HAT may differ between the two time periods, and consequently the treatments and natural histories may differ. Early HAT is often related to surgical techniques, as in cases of redundancy and
2
’
Intraarterial Thrombolysis for Hepatic Artery Thrombosis after Liver Transplantation
kinking of the arterial anastomosis (1). Late HAT may be more often related to an underlying stenosis that has developed over time, such as with intimal hyperplasia. In addition, collateral vessel formation may play a more significant role in later periods after transplantation (6,7). The purpose of the present study was to assess the safety as well as short- and long-term efficacy of hepatic arterial thrombolysis in patients with early or late HAT after liver transplantation.
MATERIALS AND METHODS Institutional review board approval was granted for this study, with a waiver of informed consent.
Patients The present retrospective study was conducted at a single institution by querying the hospital’s radiology information system to identify all patients who underwent hepatic arterial angiography of their liver transplants from February 2000 to February 2013. All patients with history of liver transplantation and diagnosis of HAT with attempted endovascular treatment were included. Patients who did not undergo angiography, patients with normal hepatic angiograms, or patients with hepatic artery stenosis without thrombosis were excluded. Baseline demographic data and liver function laboratory values are presented in Table 1. Table 1 . Patient Demographics, Underlying Liver Disease, and Timing of HAT (N ¼ 26) Characteristic
Value
Sex Male Female Age at diagnosis (y) Median Range Time from transplantation (d) Median
18 (69)
’
JVIR
Endovascular Technique Thrombolysis techniques varied slightly among interventional radiologists. However, in general, access into the femoral artery was obtained with insertion of a 5- or 6-F sheath. Diagnostic arteriography was performed (Fig 1a). Attempts to cross the occluded segment were made by using a 0.014- or 0.018-inch guide wire with a microcatheter. If attempts failed, additional attempts were made with a 0.035-inch guide wire. The thrombosed segment was completely crossed in 14 patients, and continuous infusion thrombolysis was then performed through a multiple–side-hole infusion catheter (Fig 1b). Thrombolysis was initiated through an end-hole microcatheter or partially embedded multiple–side-hole catheter in the other nine patients. Thrombolytic agents included tissue plasminogen activator (TPA) and urokinase. Intravenous or intraarterial heparin was concurrently administered. An AngioJet catheter (Possis, Minneapolis, Minnesota) was used in one patient in usual thrombectomy mode on day 1 before continuous infusion of TPA was initiated. In five patients, the sheath was positioned in the celiac artery; the remaining sheaths were positioned in the external iliac artery. Although TPA dosing was variable, in general, thrombolytic dose reductions were not performed specifically for early HAT. Some patients with early HAT had surgical drains in position, which may allow for earlier detection of arterial hemorrhage. Angiography was repeated within 24 hours of initiation of thrombolysis (Fig 1c). Adjunctive techniques after day 1 of continuousinfusion thrombolysis included balloon maceration and/or attempted angioplasty, or stent placement. The decision to continue thrombolysis was made by the interventional radiologist and the referring transplant surgeon based on follow-up angiographic findings and the patient’s clinical status.
8 (31)
Measured Outcomes 55 43–70 35
Range Early (r 30 d)
3–4,985 13 (50)
Late (4 30 d)
13 (50)
Recurrent thrombosis Liver disease before transplantation
4 (15)
Alcohol
7 (27)
Hepatitis B virus Hepatitis C virus
0 17 (65)
Hepatocellular carcinoma
8 (31)
Autoimmune Budd–Chiari syndrome
2 (8) 2 (8)
Hemochromatosis
1 (4)
Cryptogenic/idiopathic
3 (12)
Note–Values in parentheses are percentages. HAT ¼ hepatic artery thrombosis.
Kogut et al
The primary technical outcome of the present study was recanalization of the occluded hepatic artery. The primary clinical outcomes included death, repeat liver transplantation, or surgical revascularization, whichever occurred first. The secondary outcomes measured were intraprocedural hepatic artery dissection and/or perforation, postprocedural bleeding complications, recurrent thrombosis identified on follow-up imaging, and ischemic cholangiopathy. Technical success was defined as restoration of flow within the treated hepatic artery without residual flowlimiting thrombus. Procedure-related complications were recorded and categorized as major or minor based on Society of Interventional Radiology reporting standards (8). A minor complication was defined as a bleeding event that did not require additional management. Major complications included hemorrhage requiring red blood cell transfusion and/or exploratory laparotomy (Fig 2) and hepatic artery dissection and/or perforation leading to
Volume XX
’
Number X
’
Month
’
2015
3
Figure 1. (a) Angiography 6 days after hepatic transplantation shows occlusive thrombus protruding into the recipient common hepatic artery (arrow). (b) Arrowheads point to the radiopaque markers on the multiple–side-hole infusion catheter that was placed across the thrombus. A sheath is in the origin of the celiac artery, and surgical drains are in place. (c) Angiography 18 hours after the initiation of chemical thrombolysis shows restoration of flow through the hepatic arterial system with mild residual irregularity at the arterial anastomosis. Despite a good technical result, the patient received a second transplant 2 months later as a result of the development of ischemic cholangiopathy.
were conducted with the statistical computing language R (version 2.14.1; R Foundation for Statistical Computing, Vienna, Austria). Two-tailed tests were used, with statistical significance defined as P o .05.
RESULTS
Figure 2. On day 2 of thrombolysis, hemodynamic instability developed in this patient after arrival in the angiography suite. Angiography was compromised by motion artifact but showed some flow with significant residual thrombus. CT shows hematoma throughout the abdomen (arrows), with contrast medium from the recent angiography (arrowhead) near the porta hepatis. The patient’s abdominal hematoma was evacuated on an emergent basis, and surgical thrombectomy with anastomotic revision was performed.
surgical revision and/or termination. The time to clinical outcome was calculated as the number of days after thrombolysis was initiated. Except for death, clinical outcomes after repeat transplantation or surgical revascularization were not assessed, as these events were considered clinical endpoints.
Statistical Analysis Categoric and continuous variables were compared between groups by Fisher exact test and Mann–Whitney test, respectively. Rates of each clinical outcome were compared between groups by using a Gray test to account for competing risk. All statistical calculations
During the study period, 26 patients underwent angiography and attempted IAT to treat posttransplantation HAT. Table 1 summarizes patient demographics and baseline clinical history before IAT. Thirteen patients (50%) had developed early HAT. Four patients had recurrent thrombosis, and all were previously treated with surgical thrombectomy and arterial anastomotic revision. Of the 26 patients who underwent angiography with the intent to perform IAT, thrombolysis was initiated in 23 (88%). Three patients (12%) did not receive thrombolytic agents because of intraprocedural hepatic artery perforation in one patient and inability to penetrate or traverse the thrombus in two patients. Ten patients received balloon maceration and/or attempted angioplasty. One patient received a stent. Of the 23 patients in whom thrombolysis was initiated, 22 (96%) received TPA (alteplase; Genentech, South San Francisco, California), with doses ranging from 0.25 to 2.0 mg/h. One patient received urokinase, with a dose of 100,000 U/h. Tables 2 and 3 summarize procedural details and treatment outcomes in different subgroups. Of the 23 patients who underwent thrombolysis, the duration of thrombolysis ranged from 12 to 90 hours (median, 28 h). Successful recanalization (defined as restoration of flow without flow-limiting thrombus) was achieved in 12 of 26 patients (46%). There were a total of 11 major complications and one minor complication from the procedures. The minor complication was a groin hematoma at the femoral access site, which was
4
’
Intraarterial Thrombolysis for Hepatic Artery Thrombosis after Liver Transplantation
Kogut et al
’
JVIR
Table 2 . Differences in Patient Characteristics, Procedural Details, and Outcomes between Early and Late HAT Groups (N ¼ 26) Characteristic Male sex
Early HAT (n ¼ 13) 9 (69)
Late HAT (n ¼ 13) 9 (69)
56 43–70
54 44–68
Age at diagnosis (y) Median Range
P Value* 4.99 .56
Recurrent thrombosis
3 (23)
1 (8)
.59
Successfully initiated thrombolysis Lysis time (h)†
12 (92)
11 (85)
4.99 .019
Median
24
43
Range Underwent balloon PTA and/or clot maceration
12–42 5 (38)
17–90 5 (38)
Successful recanalization
8 (62)
4 (31)
Underlying abnormality None identified Stenosis Procedural complications Major
4.99 .24 .70
5 (38)
7 (54)
8 (62)
6 (46)
8 (62) 8 (62)
4 (31) 3 (23)
.24 .11
0
1 (8)
–
Hemorrhage after thrombolysis Ischemic cholangiopathy
6 (46) 2 (15)
1 (8) 5 (38)
.073 .42‡
Surgical revision
4 (31)
1 (8)
.15‡
Repeat transplantation Death
4 (31) 3 (23)
2 (15) 4 (31)
.074‡ .19‡
Minor
Note–Values in parentheses are percentages. HAT ¼ hepatic artery thrombosis, PTA ¼ percutaneous transluminal angioplasty. n Fisher exact test to compare categoric variables and Mann–Whitney test for continuous variables. † Including patients in whom thrombolysis could be initiated (n ¼ 23). ‡ Gray test for comparison of cumulative incidence functions accounting for time to event.
treated conservatively. Major complications included hepatic artery dissection or perforation (n = 4), bleeding at the access site requiring transfusion (n = 2), and intraabdominal hemorrhage requiring transfusion or exploratory laparotomy (n = 5). Patient clinical follow-up ranged from 7 days to 13 years (median, 1,156 d). Five patients (19%) underwent surgical revision of their anastomosis after angiography, all within 4 days after IAT. Three of these five patients had undergone technically successful thrombolysis, but still required surgical revision. Six other patients (23%) underwent repeat liver transplantation, four of which occurred within 42 days (median, 30 d; range, 3–625 d). Differences between those with early and late HAT are described in Table 2. Thrombolysis time was significantly shorter for those with early HAT (median, 24 h vs 43 h; P ¼ .02), yet there was a trend toward a higher rate of hemorrhagic complications in the early HAT group (46% vs 8%; P ¼ .07). No significant differences in clinical outcome or complication rates were seen when comparing patients who had restoration of hepatic arterial flow and those in whom thrombolysis failed (Table 3). Of the 12 patients with initial technically successful recanalization, arterial reocclusion developed in four (33%), as demonstrated by follow-up ultrasound (US) or magnetic resonance imaging,
all within the first year. Three patients (25%) who initially had successful recanalization experienced ischemic cholangiopathy, compared with four (29%) in the group without successful recanalization (P ¼ .90). Seven patients in total (58%) underwent surgical revision (n ¼ 3) or repeat transplantation (n ¼ 4) after successful recanalization. All these surgical revisions were performed within 4 days of recanalization.
DISCUSSION HAT after liver transplantation can be a devastating complication that is a challenge to manage. The natural history of untreated HAT is not fully understood and is therefore unpredictable. One author (9) estimated that one third of HAT cases are asymptomatic, one third will lead to chronic ischemic cholangiopathy, and one third will lead to hepatic necrosis and death. Repeat transplantation has typically been considered the treatment of choice for HAT. However, given the complexity and morbidity of liver transplantation and the scarcity of available organs, salvage of the graft without repeat operation is desirable. These challenges have led to increasing interest in endovascular management.
Volume XX
’
Number X
’
Month
’
2015
5
Table 3 . Differences in Patient Characteristics, Procedural Details, and Outcomes between Successful Recanalization and Failed Thrombolysis Groups (N ¼ 26) Characteristic
Recanalized (n ¼ 12)
Not Recanalized (n ¼ 14)
P Value*
7 (58)
11 (79)
.40 .94
Male sex Age at diagnosis (y) Median
56
54
Range Early HAT
44–70 8 (67)
43–68 5 (36)
.24
Recurrent thrombosis
2 (17)
2 (14)
4.99
32
25
14–90
12–65
5 (42)
5 (36)
5 (42)
7 (50)
7 (58) 5 (42)
7 (50) 7 (50)
.71
4 (33)
7 (50)
.45
Minor Hemorrhage after thrombolysis
1 (8) 3 (25)
0 4 (29)
.67
Ischemic cholangiopathy
3 (25)
4 (29)
.90‡
Surgical revision Retransplantation
3 (25) 4 (33)
2 (14) 2 (14)
.53‡ .61‡
Death
3 (25)
4 (29)
.79
Lysis time (h)† Median Range Underwent balloon PTA and/or clot maceration Underlying abnormality None identified Stenosis Procedural complications Major
.83
4.99 4.99
Note–Values in parentheses are percentages. HAT ¼ hepatic artery thrombosis, PTA ¼ percutaneous transluminal angioplasty. n Fisher exact test to compare categoric variables and the Mann–Whitney test for continuous variables. † Including patients in whom thrombolysis could be initiated (n ¼ 23). ‡ Gray test for comparison of cumulative incidence functions accounting for time to event.
Practical management of HAT varies, stemming from a paucity of available data related to its low incidence and differences in local surgical and interventional radiology expertise. Unfortunately, the existing body of knowledge relevant to endovascular treatment of posttransplantation HAT is problematic; most of the literature is conflicting, and many of the reports evaluate technical success without assessing clinical benefit. One recent metaanalysis of 69 cases of endovascular treatment for HAT from 16 separate series (6) estimated the technical success rate at 68%. However, the definition of success varies among publications, which limits meaningful comparisons across the studies. Although most reports have focused on the immediate technical outcomes of endovascular therapy for HAT, these may not correlate with any clinically meaningful outcomes over time. In addition, recurrent thrombosis has been reported in as many as 33% of cases (10). The present study specifically evaluates technical success and correlates it with clinical outcomes in patients who were primarily treated with IAT. In one series of six patients with HAT (11), half of which were early, the technical success rate of endovascular treatment was low, with restoration of flow achieved in only one patient. Most failures resulted from an inability to cross the occluded vessel, and thrombolysis was initiated in only two patients. In contrast, we failed to
initiate thrombolysis in only a small minority of patients (three of 26; 12%). Restoration of adequate flow was achieved in 12 of 26 patients (46%) in the present series. However, among those who had successful thrombolysis, five of 12 (42%) still required surgical revascularization or repeat transplantation within 2 months. Another five patients (42%) who underwent successful thrombolysis continued to have adequate hepatic function at a median of 828 days despite only two (40%) showing proper hepatic arterial patency on follow-up Doppler US. A slightly smaller proportion of patients with failed thrombolysis (three of 11; 27%) maintained adequate hepatic function at a median of 2,762 days. Our experience suggests that outcomes remain suboptimal even when IAT is technically successful, and any long-term clinical benefit is only questionably attributable to successful recanalization. Recent major surgery is typically cited as a relative contraindication to IAT. The present series found an overall major bleeding complication rate of 30%. The early HAT group showed a trend toward increased major bleeding rate compared with the late HAT group (50% vs 9%; P ¼ .07) despite significantly shorter thrombolysis duration (median, 24 h vs 43 h; P ¼ .02). This would be expected given recent major surgery.
6
’
Intraarterial Thrombolysis for Hepatic Artery Thrombosis after Liver Transplantation
Limitations to the present study include the inherent biases of retrospective investigations. As a result, there are expected differences in operator techniques. During the study period, changes in pharmacology and technology may have also influenced technical success. We were also limited by the small sample size. However, given the relatively low incidence of HAT in the liver transplant population, it would be challenging to report on a large number of patients at a single institution. The present study does not address the cases that were managed successfully with combined thrombolysis and surgical revision, as surgical revision was considered an endpoint. Given that successful recanalization may reveal an abnormality that is best managed surgically, long-term clinical outcomes may be more related to subsequent interventions rather than the thrombolysis. One could argue that technically successful thrombolysis could lead to improved outcomes of surgical revision by revealing underlying abnormalities that could be treated surgically in a semielective fashion. It is possible that patients in this subgroup would have the best long-term clinical outcomes. Although this would be an interesting future study, given our small population, a comparison of this subgroup versus patients who had surgical revision alone for HAT was beyond the scope of the present study. In conclusion, the present series demonstrates a relatively high major complication rate for thrombolysis performed for HAT. Technical success is fairly limited, and even in cases with successful recanalization, this may not translate into long-term hepatic graft viability and clinical benefit. In addition, some patients do well clinically in the absence of successful restoration of hepatic artery flow. As there are limited therapeutic options in this condition, thrombolysis could still be considered in the management decision in conjunction
Kogut et al
’
JVIR
with the transplant surgery team. However, awareness of the high complication rate and limited technical and clinical success is imperative during consultation with other medical providers and during informed consent discussion with the patient. Further research is needed to identify the patients most likely to benefit from aggressive treatment of HAT, as well as patients who may do well without any intervention.
REFERENCES 1. Pareja E, Cortes M, Navarro R, et al. Vascular complications after orthotopic liver transplantation: hepatic artery thrombosis. Transplant Proc 2010; 42:2970–2972. 2. Stange B. Hepatic artery thrombosis after adult liver transplantation. Liver Transpl 2003; 9:612–620. 3. Silva MA, Jambulingam PS, Gunson BK, et al. Hepatic artery thrombosis following orthotopic liver transplantation: A 10-year experience from a single centre in the United Kingdom. Liver Transpl 2005; 12:146–151. 4. Akun E, Yaprak O, Killi R, et al. Vascular complications in hepatic transplantation: single-center experience in 14 years. Transplant Proc 2012; 44:1368–1372. 5. Bekker J, Ploem S, de Jong KP. Early hepatic artery thrombosis after liver transplantation: A systematic review of the incidence, outcome and risk factors. Am J Transplant 2009; 9:746–757. 6. Singhal A, Stokes K, Sebastian A, et al. Endovascular treatment of hepatic artery thrombosis following liver transplantation. Transpl Int 2010; 23:245–256. 7. Wozney P, Zajko AB, Bron KM, et al. Vascular complications after liver transplantation: a 5-year experience. AJR 1986; 147:657–663. 8. Omary RA, Bettmann MA, et al. Quality improvement guidelines for the reporting and archiving of interventional radiology procedures. JVIR 2003; 14:S293–S295. 9. Pinna AD, Smith CV, Furukawa H, et al. Urgent revascularization of liver allografts after early hepatic artery thrombosis. Transplantation 1996; 62 (11):1584. 10. Abdelaziz O, Hosny K, Amin A, et al. Endovascular management of early hepatic artery thrombosis after living donor liver transplantation. Transpl Int 2012; 25:847–856. 11. Sabri SS, Saad WEA, Schmitt TM, et al. Endovascular therapy for hepatic artery Stenosis and thrombosis following liver transplantation. Vasc Endovascular Surg 2011; 45:447–452.