Objectives: Vascular variations of the extrahepatic artery occur in up to 50% of the population. Exact knowledge of any anomalies is of great significance in hepatobiliary surgery to avoid perioperative com-plications. In fact, in liver transplant, vascular complications are rare but have a major impact on graft function and survival. This study evaluated variations of the extrahepatic artery in donors and recipients as risk factors for vascular complications after liver transplant.
Materials and Methods: From January 2010 until June 2015, 469 liver transplant procedures were performed at our institution. We included 323 patients in our retrospective analysis after exclusion of retransplants, split-livertransplants, and pediatric patients. We analyzed the impact of anatomic variations of recipients and donors on postoperative vascular complications and organ and patient survival.
Results: Of total study recipients, 71.2% had a normal vascular supply according to Michel classification I. However, these patients developed significantly more vascular complications (25.65%) than those with vascular anomalies (15.05%), especially showing higher incidence of arterial stenosis (8.26% vs 2.15%). In contrast, vascular variations in donors and the need for a vascular reconstruction of the graft led to significantly higher mortality (26.76% vs 15.48%). An abnormality of the graft did not influence incidence of postoperative complications or graft survival.
Conclusions: Unexpectedly, recipients with variations of the hepatic artery and grafts with an abnormal arterial supply did not show higher rates of com-plications or mortality. Only vascular reconstruction of the graft before transplant raised the mortality of recipients.
Key words : Arterial supply, Graft failure, Vascular reconstruction
The anatomy of the extrahepatic arterial blood supply shows variations in 19.2% to 49.3% of patients.1,2 An accessory or aberrant arterial vessel can originate from the superior mesenteric artery, left gastric artery, aorta, or other visceral branches. Albrecht von Haller published the first description of an aberrant artery in 1756.3 Since then, variations of the hepatic artery and their incidence have been analyzed by many studies and could be particularly verified after introduction of modern radiologic procedures during the past century.4 In 1966, Michel’s examination of 200 autopsies defined the basic classification of anatomic variations in hepatic arterial supply and served as a benchmark (Table 1).5 Hiatt and associates and Varotti and associates continued with 2 further simplified classifications.6,7
In general, aberrant arteries are of great impor-tance during operations in the upper abdominal cavity and especially in liver surgery and liver transplant, to avoid further complications.8,9 The biliary tree obtains an exclusive arterial supply, and impaired arterial perfusion can cause severe necrosis, biliary abscess formation, and even graft loss. Identification and appropriate reconstruction of anomalies are absolutely necessary to reduce perioperative morbidity and mortality.9
Orthotopic liver transplant remains the criterion treatment standard for end-stage liver disease or fulminant acute liver failure. However, it remains associated with a certain risk for vascular and particularly arterial complications.10 Arterial throm-bosis can occur in up to 12% and arterial stenosis in up to 11% of cases.10 Risk factors include arteriosclerosis, vascular dissection or stenosis, extended ischemic time, ABO incompatibility, and acute rejection; in addition, despite technical and perioperative improvements, patients are at risk due to complex vascular reconstructions.10,11 Our investigation evaluated the influence of extrahepatic arterial variations of donors and recipients on occurrence of vascular complications and thus outcomes after liver transplant.
The primary endpoint of this retrospective study was patient and organ survival after liver transplant depending on the arterial vascular supply before transplant. Secondary endpoints included incidence of vascular or biliary complications during follow-up.
Materials and Methods
This was a retrospective analysis of all orthotopic liver transplant procedures performed at our clinic from January 2010 until the end of June 2015. We included patients older than 18 years with all types of end-stage liver diseases or acute liver failure and who had full graft transplant procedures. Patients with retransplant and split-liver transplant and pediatric patients were excluded. We identified 323 orthotopic liver transplants to be included. Data were collected from our internal digital patient data system. Data collected included sex, age at trans-plant, disease leading to liver transplant, results of preoperative computer tomography (CT) scans, surgical reports, vascular and biliary complications, types of therapies for complications, date of retransplant, date of death, and latest laboratory parameters.
Reasons for transplant were subdivided into cryptogenic and toxic cirrhosis (alcohol, drugs), nonalcoholic steatohepatitis, viral hepatitis, primary biliary and sclerosing cholangitis, secondary sclerosing cholangitis, autoimmune hepatitis, polycystic disease and Caroli disease, acute liver failure, and “other,” which included hemangioendothelioma, Klatskin tumor or cholangiocellular carcinoma, neuroendocrine metastases, Budd-Chiari syndrome, cystic fibrosis, alpha-1-antitrypsine deficiency, Alagille syndrome, amyloidosis, and Wilson and Osler disease. Occurrence of hepatocellular carcinoma was also recorded.
The preoperative vascular supply of recipients and donors was graded per Michel classification.5 Michel classification I was assumed as normal anatomy, and Michel classifications II to XI were considered as having variations in anatomy. Our routine preoperative protocol is for all patients at our clinic to be staged by contrast-enhanced CT scans before transplant. Michel classification of recipients was made using preoperative CT scans, which allowed observation of arterial, portal, and venous phases. Preoperative thrombosis of the portal vein and preoperative described ligamentous stenosis of the celiac trunk were assessed. Michel classification of donors was made using reports of donor grafts.
Data from surgical procedures were extracted from surgical reports. We assumed a normal surgical procedure if the arterial branch patch of the donor’s hepatic artery at the gastroduodenal artery was connected to bifurcation of the recipient’s hepatic and gastroduodenal artery. Deviations from this procedure or back-table reconstructions due to vascular anomalies in the donor were categorized as nonstandard procedures. Data on division of the arcuate ligament and ligature of the splenic artery were also collected.
Vascular complications were defined as post-operative stenosis of hepatic artery, thrombosis of hepatic artery, stenosis or thrombosis of the portal vein, stenosis or thrombosis of the hepatic vein, and other (secondary bleedings from the main hepatic vessels or lienalis steal syndrome). Biliary com-plications were divided into stenosis of biliary anastomosis, ischemic-type biliary disease, biliary leakage, and other (stenosis of papilla of Vater, biliary concrements, hepatic duct necrosis, unknown hyperbilirubinemia, and biliary abscess formation). Vascular and biliary complications were treated by conservative procedures, reconstruction surgery, or endovascular or endoscopic intervention.
Good graft function was defined by normal laboratory results. We collected the latest available laboratory parameters from our data bank as all patients are regularly seen by our outpatient department at least every 3 months. Good graft function was assumed in cases of normal values for bilirubin, gamma-glutamyltransferase, alanine aminotransferase, aspartate aminotransferase, and prothrombin time (excluded if the patient was taking coumarin drugs).
For statistical analyses, we used Excel (version 15.21.1, 2016; Microsoft Corporation, Redmond, WA, USA) with the add-in XLSTAT (version 2016.02.27941; Addinsoft, New York, NY, USA). Differences in nominal variables were tested by the Fisher exact test. Not normally distributed values were tested by the Mann-Whitney U test. P values less than .05 were accepted as significant. Overall survival was approximated by Kaplan-Meyer analysis. Patients were first grouped as those with recipient Michel classification I and those with recipient Michel classification II-XI and analyzed. Patients were then grouped as those with donor Michel classification I and those with donor Michel classification II-XI. In addition, all patients with back-table reconstruction due to arterial variations of the donor were compared with those without these variations. Thirteen patients were withdrawn from the analysis because surgical reports could not be extracted from the database.
From January 2010 to July 2015, 323 patients received a liver transplant at our clinic. Of these, 230 patients had a normal arterial vascular supply according to Michel classification I and 93 had variations in anatomy (Michel classification II-XI) (Figure 1). Among those with variations, 13 had Michel classification II, 16 had Michel classification III, 2 had Michel classification IV, 19 had Michel classification V, 6 had Michel classification VI, 1 had Michel classification VII, 3 had Michel classification VIII, 5 had Michel classification IX, 1 had Michel class-ification X, and 27 had Michel classification XI (which included all patients with other abnormalities). Regarding donor arterial supply, we found 310 grafts had normal arterial supply (Michel classification I) and 13 had abnormal vascular anatomy (9 grafts showed Michel classification V, 3 showed Michel classification VI, 1 showed Michel classi-fication VII, and 1 had Michel classification XI). Distribution of preoperative arterial supply is shown in Figure 1.
Of those with recipient Michel classification I, 149 were men (64.78%) and 81 were women (35.22%). Of those with recipient Michel classification II-XI, 58 were men (62.37%) and 35 were women (37.63%), with no significant differences shown between men and women regarding classification (Table 2). Also, no significant difference was shown with regard to recipient sex versus Michel classification of the graft (Table 2). There were also no significant differences in reasons leading to liver transplant and the prevalence of hepatocellular carcinoma, although there were significantly more patients with toxic cirrhosis with recipient Michel classification II-XI (49.46% vs 32.61%; P = .005) (Table 2). In addition, significantly more grafts with abnormal vascular supply were transplanted in patients with secondary biliary cirrhosis (1.94% vs 15.38%; P = .037) (Table 1). The preoperative occurrence of stenosis of the celiac trunk, thrombosis of the portal vein, and surgical procedure complications did not differ significantly between the recipient classification groups; however, in patients with Michel classification II-XI of graft, the ligamentum arcuatum was significantly more often incised compared with those with Michel classification I of graft (6.13% vs 23.08%; P = .05).
In general, those with Michel classification I of the recipient developed significantly more vascular complications after liver transplant (25.65% vs 15.05%; P = .04) (Table 3). This group also had a greater incidence of stenosis of the hepatic artery (8.26% vs 2.15%; P = .046) and complications that were merged under “other” (11.74% vs 2.15%; P= 0.005) than the recipient Michel classification II-XI group (Table 3). Arterial thrombosis and complications of the portal vein and hepatic veins were not significantly different between groups. Furthermore, “other” biliary complications appeared significantly more often in patients with Michel classification I (4.35% vs 0%; P = .038), although they did not show significantly more biliary com-plications in general (Table 3). Rates of retransplant (6.96% vs 8.60%) and mortality (19.57% vs. 20.43%) were not significantly different between groups (Table 3). Retransplant was performed at a median of 21.94 ± 51.64 days in recipients with Michel classification I and 8.25 ± 5.02 days in recipients with Michel classification II-XI (Table 2).
Overall survival in recipients was not signi-ficantly different between classification groups (1939.52 ± 60.66 vs 1862.61 ± 91.97 days; Table 3 and Figure 2), as well as overall graft survival (398.80 ± 493.51 vs 252.53 ± 370.54; P = .389). Of those with surviving grafts, 45.22% with Michel classification I and 52.69% with Michel classification II-XI showed good long-term graft function (Table 3). When we analyzed complications according to the vascular supply of the graft, we found no significant differences in incidence of complications, whether the graft showed normal or abnormal arterial supply. However, we did observe significantly shorter graft survival in retransplants in grafts with Michel classification II-XI (2.00 ± 0.0 vs 8.25 ± 5.02 days; P = .022) (Table 3).
We also analyzed patients who were grouped as having arterial reconstruction during the transplant procedure due to anatomic variations in donor or recipient (n = 71) and those who only had standard techniques (n = 239) (Table 4). Our analyses showed that vascular complications appeared equally in both groups (23.43% vs 22.54%; Table 4). There were no significant differences between groups in occurrence of postoperative arterial stenosis or thrombosis, portal or venous complications, and complications categorized as “other.” Occurrences of biliary com-plications (26.36% vs 22.54%) and retransplant (6.69% vs 8.45%) were not significantly different between groups (Table 3). Although mortality rate was signifi-cantly higher in the reconstruction group (26.76% vs 15.48%), the difference was not significant. Overall graft survival was significantly longer in the standard technique group (2032.11 ± 55.35 days) versus the reconstruction group (1689.43 ± 112.92 days) (Figure 3). Among living patients, graft function was stable in 50.63% of those who had the standard technique and in 54.72% of those who had reconstruction.
In our analyses of 323 liver transplant patients, 28.8% of the recipients had arterial anomalies and 4.0% of the grafts showed variations, which required reconstruction in 26.0% of these cases. Our data correspond to previous literature, as the incidence of Michel classification I has been shown to range between 50.7% and 80.8%.1,2,9 The lower incidence of vascular abnormalities of grafts might be related to documentation errors on the organ report sheet. Michel classifications V (5.9%), III (5.0%), and II (4.0%) of recipients were the most frequent types among our patients; however, 8.4% could not be clearly classified and were grouped as Michel classification XI (Figure 1). This is also similar to the literature, as Michel classifications II, III, and V seem to be the most frequent variations.9 According to Németh and associates, many cases cannot be definitely categorized.9 “Other” has also newly included pentafurcation of the hepatic artery; trifurcation into the left hepatic artery, the right anterior gastric artery, and the gastroduodenal artery with a proximal origination; retroportal right hepatic artery, proximal branching, and anteportal course of the right hepatic artery; and retroportal course of a right posterior hepatic artery.9 In accordance with previous studies, we also included variations directly originating from the preceliac aorta or the celiac trunk.12-14 Nevertheless, nearly one-third of all patients had arterial anomalies, which is of great importance in visceral surgery. Especially in cases of obesity or malignancy, altered anatomic conditions may handicap the correct identification of vascular structures during transplant; therefore, an accurate preoperative evaluation by radiologic methods (eg, contrast-enhanced CT or magnetic resonance imaging) is inevitably necessary.
All of our patients received a complete pre-operative diagnostic examination by a 3-phase CT scan to exclude anatomic anomalies. Hence, patients with variations of the extrahepatic arterial supply were not at a higher risk for vascular or consecutively biliary complications. Rather, we found a higher rate of arterial complications in patients with a preo-perative normal arterial supply (8.26% vs 2.15%; P = .046) (Table 3). These patients also had a higher rate of biliary complications (24.78% vs 21.51%; P = .565) but with no significant differences with regard to graft or patient survival (Table 3). In contrast, Pérez-Saborido and associates, who also found no influences regarding preoperative arterial supply on survival, did observe a higher incidence of arterial complications in recipients with anomalies (20.0% vs 0.0%).15
We also found that abnormal arterial supply of the graft did not significantly influence incidence of postoperative vascular or biliary complications beyond a significantly shorter period between transplant and retransplant in those with Michel classification II-XI of the graft.
The need for bench reconstruction can influence outcomes after liver transplant. Although we did not find significant differences in occurrence of postoperative vascular or biliary complications (Table 4), mortality rate was significantly higher in patients who required reconstruction (26.76% vs 15.48% in those without reconstruction) and overall graft survival was significantly longer those without reconstruction (Table 4 and Figure 2). These findings are similar to the literature. Some studies have shown that particularly the need for bench reconstruction due to donor anomalies can compromise graft outcome.12,15 However, Ferraz-Neto and associates observed a greater number of vascular and even biliary complications among orthotopic liver trans-plant cases with aberrant arterial anatomy but without significance.16
The main limitation of our study was the small number of grafts with abnormal arterial supply, which affected the quality of statistical analyses. Interestingly, we were able to show that vascular abnormalities in recipients or donors did not have a major effect on graft and patient survival. Only vascular supply anomalies in donors with need of back-table reconstruction raised the risk for reduced graft and patient survival.
DOI : 10.6002/ect.2018.0201
From the 1Department of Surgery and the 2Department of Radiology, Charité –
Universitätsmedizin Berlin, Berlin, Germany; and the 3Department of General,
Visceral and Transplant Surgery, Universitätsklinik Münster, Münster, Germany
Acknowledgements: The authors have no conflicts of interest to declare. The author and all coauthors declare that they do not have any financial support from any organization that was or is related to the submitted work; there are no financial relationships with any organizations that might have an interest in the submitted work within the previous 2 years.
Corresponding author: Eva Maria Teegen, Charité Campus Virchow, Department of Surgery, Augustenburger Platz 1, 13353 Berlin, Germany
Phone: +49 30450652215
Table 1. Michel Classification of the Hepatic Artery
Table 2. Demographic Data Distributed According to Michel Classification of Recipient and Donor
Table 3. Complications and Survival After Liver Transplant Depending in Michel Classification of Recipient and Donor
Table 4. Complications and Survival After Liver Transplant Depending on Surgical Procedure of Arterial Anastomosis
Figure 1. Schematic Showing Prevalence of Each Variation According to Michel Classification
Figure 2. Kaplan-Meyer-Analysis of the Cumulative Survival After Liver Transplant of Total Cohort and Depending on Michel Classification of Recipient and Donor (N = 323)
Figure 3. Kaplan-Meyer Analysis of Cumulative Survival After Liver Transplant of Total Cohort and Depending on Surgical Procedure of Arterial Anastomosis (N = 323)