ORIGINAL ARTICLE ORIGINAL ARTICLE 887. Liver Transplantation AASLD. Received January 25, 2017; accepted April 13, 2017.

Transcription

1 ORIGINAL ARTICLE LI ET AL. Hepatic Artery Reconstruction in Living Donor Liver Transplantation Using Surgical Loupes: Achieving Low Rate of Hepatic Arterial Thrombosis in 741 Consecutive Recipients Tips and Tricks to Overcome the Poor Hepatic Arterial Flow Ping-Chun Li, 1,2 Ashok Thorat, 1 Long-Bin Jeng, 1,3 Horng-Ren Yang, 1,3 Ming-Li Li, 1,2 Chun-Chieh Yeh, 1,3 Te-Hung Chen, 1,3 Shih-Chao Hsu, 1,3 and Kin-Shing Poon 1,4 1 Organ Transplantation Center; 2 Departments of Cardiovascular Surgery; 3 Surgery; 4 Anaesthesiology, China Medical University Hospital, Taichung, Taiwan The reconstruction of the hepatic artery (HA) is the most complex step in living donor liver transplantation (LDLT) because of the smaller diameter of the artery and the increased risk of HA-related complications. Because of the smaller diameter of the HA, many centers use a microsurgical technique with interrupted sutures for arterial anastomosis. The aim of our study was to retrospectively investigate the outcomes after HA reconstruction performed under magnifying loupes using the parachute technique. From August 1, 2002 to August 31, 2016, LDLT was performed in 766 recipients. HA reconstruction for the initial 25 LDLT surgeries was performed using a microsurgery technique (era 1). From May 2007 until the end date, HA reconstruction was performed in 741 recipients by a parachute technique under surgical loupes (era 2). HA reconstruction was performed using surgical loupes in 737 adults (male:female, 526:211) and 4 pediatric patients (male:female, 3:1). The average diameter of the donor graft HA was 2.8 mm (range, mm). The most notable factor in this era was the quick HA anastomosis procedure with a mean time of minutes (range, 5-30 minutes). In era 2, 9 (1.21%) patients developed hepatic artery thrombosis (HAT), whereas 2 patients developed nonthrombotic HA-related complications. Extra-anatomic HA reconstruction was performed in 14 patients due to either primary HA anastomosis failure or a poor caliber recipient HA. The use of magnifying surgical loupes to perform HA reconstruction is safe, feasible, and yields a low incidence of HA-related complications. The parachute technique for HA reconstruction can achieve a speedy reconstruction without increasing the risk of HAT. Liver Transplantation AASLD. Received January 25, 2017; accepted April 13, The arterial reconstruction is a crucial step in determining better graft and recipient outcomes in liver transplantation. Any occurrences of the hepatic artery Abbreviations: ALF, acute liver failure; CT, computed tomography; DDLT, deceased donor liver transplantation; ESLD, end-stage liver disease; HA, hepatic artery; HAT, hepatic artery thrombosis; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; LDLT, living donor liver transplantation; MELD, Model for End-Stage Liver Disease; MHV, middle hepatic vein; PGE1, prostaglandin E1; RGEA, right gastro-epiploic artery; TACE, transarterial chemoembolization; USG, ultrasonography. (HA) flow abnormalities in the immediate postoperative period may lead to fatal complications. In living donor liver transplantation (LDLT), the liver allograft is partial, and the arterial reconstruction is highly technically demanding. (1,2) The risk of hepatic artery thrombosis (HAT) remains high in LDLT due to the smaller caliber of the vessels. HA of a diameter <3 mm was reported to be an independent risk factor for HAT in pediatric liver transplantation using reduced-sized cadaveric liver grafts. (3) One of the most important advances in HA reconstruction for partial ORIGINAL ARTICLE 887

2 LI ET AL. LIVER TRANSPLANTATION, July 2017 grafts was the introduction of microsurgical techniques involving an operating microscope. Tanaka et al. in their initial experience described a favorable longterm outcome after left lobe LDLT when HA anastomosis was done under an operating microscope. They concluded that the patency rates after the microsurgery technique were better, and the risk of HAT significantly decreased. (4) At the beginning of the microsurgical HA reconstruction era, the reported incidence of HAT was as high as 14%-25%. (5,6) However, successful application of the microvascular surgery technique for HA anastomosis has yielded a better outcome with HAT incidence reduced to 1.7%. (7) Generally, a conventional operating microscope that is used for HA reconstruction has a binocular head, counterbalanced pantographic arm, and a fixed floor stand. (8) The use of a large operating microscope has certain disadvantages: it is bulky, requires maneuvering, and presents difficulties for focusing deep in the abdominal cavity because the HA reconstruction field is nearly cm deep from the abdominal wall. Also, the constant liver graft excursion due to diaphragmatic movements greatly limits the speed of arterial anastomosis. As the experience in LDLT increased and inflow reconstruction techniques became more refined, several researchers argued the necessity of microsurgery for HA reconstruction and reported equal or even better results using surgical loupes for HA reconstruction with a magnification of or more. (1,9-11) This has been extended to pediatric LDLT as well, achieving good results after HA reconstruction under loupes of magnification of 63. (12) We previously described the feasibility and better graft outcome after the application of HA anastomosis using surgical loupes. (11) In this retrospective analysis of 766 sequential LDLTs, we intend to describe our experience of the HA reconstruction using surgical Address reprint requests to Long-Bin Jeng, M.D., Organ Transplantation Center, China Medical University Hospital, 2, Yuh-Der Road, Taichung, Taiwan. Telephone: , ext. 1765; FAX: ; otc@mail.cmuh.org.tw Copyright VC 2017 by the American Association for the Study of Liver Diseases. View this article online at wileyonlinelibrary.com. DOI /lt Potential conflict of interest: Nothing to report. loupes of magnification (n 5 741), the detailed surgical technique of HA anastomosis, and institute protocol for the successful management of arterial complications after LDLT. We also aim to describe the challenges and techniques of HA anastomosis in difficult situations such as dual graft hepatic arteries, inability to use the recipient HA, and recipients requiring redo anastomosis due to flow inadequacies. Patients and Methods From August 2002 to August 2016, a total of 766 LDLT surgeries were performed at the China Medical University Hospital, Taiwan. The medical records were retrospectively analyzed for the outcomes and HA-related complications in these patients. The HA reconstruction for the initial 25 LDLT surgeries (from August 2002 to April 2007) was done using a microsurgery technique under an operating microscope of 12 3 magnification by a team of experienced surgeons. We loosely termed the experience gained in this period as era 1 of our institutional liver transplantation program. From May 2007 to August 2016, HA reconstruction was done in 741 recipients by using the parachute technique under surgical loupes. This period of using surgical loupes for HA reconstruction is era 2 of our liver transplantation program. This article highlights the results in era 2 and focuses on the technical details of HA reconstruction using surgical loupes, technical variations in difficult situations, the recipient outcome after transplantation, HA-related complications, and approaches to achieve a low rate of HAT in LDLT recipients. The study protocol received a priori approval by the institutional review board of the hospital. The demographic data, medical records, and pretransplant laboratory data of the recipients from era 2 (surgical loupe era) were assessed for outcomes and arterial flow abnormalities. The complications directly related to HA were recorded and retrospectively analyzed. The donors were evaluated by institute protocol using the pretransplant survey. (13) The donor age must be >18 years, and institutional review board permission is needed for all LDLT surgeries at our institute. The imaging studies that included the computed tomography (CT) scan and magnetic resonance imaging along with the 3-dimensional reconstruction of CT liver images were done to assess the liver segmental anatomy and the approximate volume of both liver lobes. Donor hepatectomy was performed by the 888 ORIGINAL ARTICLE

3 LIVER TRANSPLANTATION, Vol. 23, No. 7, 2017 LI ET AL. rooftop and skeletonization technique of hepatic transection. The details of this surgical technique were reported earlier. (14) We preferred the right lobe as the liver allograft in adult LDLT if the donor liver remnant volume was >30%. However, for pediatric transplant recipients, either the left lateral lobe or the left lobe liver allografts were used. The recipient surgery was done by the standard technique. We preserved the right and left HAs with as much length as possible which was trimmed or shortened during the HA reconstruction of the liver graft as needed for the liver graft HA length and angulation. SELECTION OF THE HA RECONSTRUCTION TECHNIQUE The variations of HA, both in the donor as well as in the recipients, are not uncommon in LDLT surgery. The HA reconstruction was done by the standard technique of end-to-end anastomosis between the graft and recipient HAs. In the presence of dual graft HAs, the second branch of the HA was reconstructed only if it was adequately large and/or if there was a lack of an adequate backflow after the larger branch was reconstructed using various techniques (described later). In the absence of a good caliber recipient HA, an extraanatomical reconstruction of the HA was done (n 5 14). For all the cases in era 2, the reconstruction of the HA was done by cardiovascular surgeons using surgical loupes (described later). Since May 2007, a consistent team of 3 experienced surgeons performed HA reconstruction, and the majority were performed by the chief cardiovascular surgeon. INTRAOPERATIVE DOPPLER FLOWMETER ASSESSMENT OF THE LIVER ALLOGRAFT INFLOW/OUTFLOW, POSTOPERATIVE FOLLOW-UP, AND HAT TREATMENT PROTOCOL Adequacy of the HA and portal vein inflow was measured using a Doppler flowmeter after completion of the vascular reconstruction and was repeated at the end of the recipient surgery before closing the abdomen. If any disturbances were detected, they were addressed immediately. If there was no HA flow during the intraoperative assessment, after a brief period of observation, we preferred to redo the HA anastomosis to establish the arterial inflow. Our protocol is to administer prostaglandin E1 (PGE1) intravenously for all recipients requiring 2 or more attempts of HA anastomosis or initial poor arterial flow. All the recipients were followed up by ultrasonography (USG) on postoperative days 1, 3, and 7 and then weekly thereafter for the first month to assess the graft tissue perfusion, venous outflow, and graft regeneration. Hepatic arterial and/or portal vein complications were suspected when Doppler USG revealed inflow abnormalities with sharply elevated liver enzymes. The complications were further confirmed by a CT angiography. Our guidelines to manage immediate intraoperative HA flow abnormalities are depicted in an algorithm shown in Fig. 1. The HAT that developed after 24 hours of transplantation was initially managed by urokinase therapy. Prompt surgical thrombectomy and revision of HA anastomosis were always our first choice of intervention if HA complications were detected in the immediate postoperative period (within 24 hours) and, also, if urokinase infusion therapy failed (Fig. 2). The standard dose of urokinase used for intraarterial thrombolytic therapy at our institute is as follows: 1.5 million IU of urokinase is added to 250 ml of normal saline. The normal saline is then given 10 ml/hour (60,000 IU of urokinase per hour) through a HA catheter which is placed while performing a CT angiography. The HA angiography should be repeated every 24 hours. Patients should be monitored in the intensive care unit for any hemodynamic instability. We do not administer heparin as a treatment or prophylaxis. TECHNIQUE OF HA RECONSTRUCTION In era 1, the HA anastomosis was done under an operating microscope of magnification 12 3 in an end-toend fashion using interrupted 8-0 Prolene sutures. However, because of the long operating time using microsurgery and the feasibility of using surgical loupes in establishing effective arterial flow, we changed our technical protocol to present the loupe parachute technique protocol. In era 2, the HA reconstructions (anatomical or extra-anatomical) were performed under surgical loupes. For right liver allografts, either the right HA or the common HA was used for anastomosis without causing undue angulation or redundancy of the recipient HA. In the case of a left-sided liver allograft, the left HA was used if the proper angulation was ORIGINAL ARTICLE 889

4 LI ET AL. LIVER TRANSPLANTATION, July 2017 FIG. 1. An algorithm showing the treatment protocol for HA inflow abnormality after LDLT. FIG. 2. Treatment protocol of HAT. maintained. The suture material used was a doublearmed 7-0 Prolene (0.5 metric, 60 cm, 3/8 circle, round-bodied, Blue monofilament, Ethicon, LLC, Somerville, NJ). A Jacobson needle holder and Vickers micro ring forceps were used during HA reconstruction. The timing of the HA anastomosis was calculated from the point of the beginning of the arterial suturing until the establishment of the blood flow. Parachute Technique of HA Anastomosis The anastomosis was done in an end-to-end fashion using 7-0 nonabsorbable Prolene monofilament continuous suturing by a parachute technique under surgical loupes of magnification After graft reperfusion was achieved, a sufficient length of the recipient HA was obtained, and a vascular clamp was 890 ORIGINAL ARTICLE

5 LIVER TRANSPLANTATION, Vol. 23, No. 7, 2017 LI ET AL. FIG. 3. Graphic representation of the parachute technique of HA reconstruction. applied on the recipient side of the HA. The right HA or the common HA was used depending on the available length of the graft HA and the redundancy of the recipient HA. The edges of the graft and recipient HAs were obliquely cut to widen the cross section that increased the ease of the anastomosis. The excess of neurolymphatic tissues around the adventitia were also trimmed. The back-bleeding from the graft HA was controlled by using microvascular bulldog clamps. The double-armed 7-0 Prolene suture was used. The first bite was an in and out suture at the apex of the recipient HA and an out and in suture on the corresponding point of the graft HA (Fig. 3A). This process of continuous suturing was repeated along the posterior walls of the graft and recipient HAs until they were securely oriented (Fig. 3B). The lateral arm of the Prolene suture was used to create a gentle traction force as the posterior wall sutures were completed, and at the same time, the suturing arm of the Prolene was kept under similar traction to distribute the tension equally on the suture line, thus, allowing the arterial walls to parachute together (Fig. 3C). At the end of the posterior wall suturing, the 2 vessel walls were gradually brought together through repeated gentle traction on the both ends of suture arms. Usually, 6-7 continuous stiches are placed to complete the posterior layer of anastomosis. The suturing was then repeated for the anterior wall in a continuous fashion to complete the anastomosis. At completion of the suturing, first, the vascular clamp on the graft side was removed which was followed by removal of the recipient side vascular clamp. The knot was tied without any growth factor. No interrupted sutures were placed on the anterior wall. However, after establishment of the HA blood flow, any arterial bleeding sites were reinforced by 1 or 2 interrupted 7-0 Prolene sutures. The inflow adequacy was confirmed by an intraoperative Doppler flowmeter analysis. Any abnormality of the flow was further assessed, and the anastomosis was repeated in case of a lack of HA flow. Figure 4 shows the intraoperative images of HA reconstruction. Technical Variations for the Arterial Reconstruction HA RECONSTRUCTION IN PRESENCE OF DUAL GRAFT HAS A suitable anastomosis technique needs to be adapted for right liver grafts with multiple HAs. We prefer to reconstruct all the HAs of larger caliber. However, the protocol to reconstruct the smaller HA branches varies. ORIGINAL ARTICLE 891

6 LI ET AL. LIVER TRANSPLANTATION, July 2017 FIG. 4. Intraoperative images of HA reconstruction. FIG. 5. Back-table technique of unification arterioplasty to form a common orifice. 892 ORIGINAL ARTICLE

7 LIVER TRANSPLANTATION, Vol. 23, No. 7, 2017 LI ET AL. FIG. 6. (A) End-to-side arterioplasty technique. (B) Hepaticcystic arterioplasty. During the retrieval of the right liver allograft with dual HAs, first the smaller branch of the HA is divided, and the presence of the pulsatile backbleeding from the cut end of the HA is checked. If good backflow from the smaller HA is present, the reconstruction of the smaller branch is not necessary. During arterial anastomosis in the recipient, the larger HA should be reconstructed first using the recipient right HA so that flow is established. The initial test for smaller HA back-bleeding should again be confirmed because it signifies the presence of intrahepatic interarterial plexus. Whenever feasible during the back-table procedure, we aim to make a unification arterioplasty to form a single arterial orifice without causing an undue twist and/or angulation of the HA. There are several variations for arterioplasty in the presence of dual HAs that require reconstruction. 1. Side-to-side unification arterioplasty: Back-table unification arterioplasty to form a common arterial orifice by suturing adjacent arterial walls if the extrahepatic course of arteries is small and arteries are in proximity (Fig. 5). 2. End-to-side arterioplasty: It is performed in the presence of longer extrahepatic courses of the dual HAs (Fig. 6A). 3. Hepatic-cystic arterioplasty: The cystic artery of the recipient is anastomosed with the smaller liver graft HA and the larger artery of the graft is anastomosed with the right HA of the recipient. The details of this technique have been described previously. (15) Figure 6B shows the diagrammatic representation of the technique. This technique was applied in 2 recipients. Extra-Anatomic Reconstruction of the HA SUPRACELIAC AORTOHEPATIC CONDUIT An extra-anatomical source for HA flow is used when the recipient HA is beyond utility and a conventional HA anastomosis is not feasible. An innovative supraceliac aortohepatic conduit technique was applied to establish arterial inflow using the recipient s great saphenous vein as the vascular conduit. The technical details have been reported before. (16) RIGHT GASTRO-EPIPLOIC ARTERY AS AN ALTERNATIVE TO RECIPIENT HA The right gastro-epiploic artery (RGEA) can serve as an alternative if the native HA cannot be used. The RGEA should be mobilized over cm from the greater curvature of the stomach and the greater omentum. Although RGEA is a feasible option in certain situations, the size discrepancy and potential for anastomotic kinking can result in poor HA flow. Results In era 2 (surgical loupe era), all recipients (737 adult [male:female, 526:211] and 4 pediatric patients [male:female, 3:1]) underwent HA reconstruction using the parachute technique under surgical loupes. The mean age for adult patients was 54 years (range, years). For adult-to-adult LDLT (n 5 737), right lobe liver allografts were used in 734 recipients, whereas only 3 left lobe liver allografts were harvested. In pediatric LDLT (n 5 4), 2 left lobe liver allografts, 1 right liver allograft, and 1 lateral lobe graft were ORIGINAL ARTICLE 893

8 LI ET AL. LIVER TRANSPLANTATION, July 2017 TABLE 1. Characteristics of the LDLT Recipients That Underwent HA Reconstruction Using Surgical Loupes (Era 2) Era 2 (n 5 741) Sex, male:female Adult-to-adult LDLT, n 526:211 Pediatric LDLT, n 3:1 Age, years, median (range) 54 (21-75) MELD, mean 6 standard deviation (range) (5-47) Indications, n HBV/ESLD 104 HBV/ALF 40 HCV/ESLD 94 HBV/HCV/ESLD 12 Alcoholic ESLD 95 HCC 348 Cholestatic liver disease 7 Others 41 ABO-incompatible LDLT, n 30 TACE for HCC prior to LDLT, n 88 Prior hepatectomy for HCC, n 34 retrieved. The average diameter of the donor graft HA was 2.8 mm (range, mm). The most notable factor in era 2 was the quick HA reconstruction procedure with a mean time of minutes (range, 5-30 minutes) as compared with the longer anastomotic time (>60 minutes) in era 1. In 94.02% of the recipients (n = 693), the arterial flow was adequate after the first attempt of HA anastomosis. Thirty-one recipients required redo anastomosis 2 times due to either failure of the first anastomosis or poor HA flow that was noticed on protocol Doppler flowmeter analysis before closure of the abdomen. In 13 recipients, the HA anastomosis attempts were between 3 and 5 (anastomosis time range, minutes). INTRAOPERATIVE FLOW ASSESSMENT AND MANAGEMENT OF HA FLOW ABNORMALITY The average HA flow during intraoperative Doppler flow assessment was ml/minute (range, ml/minute) with a pulsatile index of The average portal flow measured was ml/ minute (range, ml/minute). All the patients who had initial arterial flow abnormalities were managed by the treatment protocol mentioned earlier (Table 1). Extra-Anatomical Reconstruction of the HA Conventional donor HA-to-recipient HA anastomosis was not possible in 14 of the recipients due to poor TABLE 2. Technical Aspects of HA Reconstruction and Overall HA-Related Complications Era 2 Era 1 (n 5 741) (n 5 25) HA anastomosis time, minutes, (5-30) >60 mean 6 standard deviation (range) Arterial complications (n 5 14) HAT (n 5 11) 9 2 Kinking 0 1 Poor HA flow 1 0 Pseudoaneurysm 1 0 Alternative technique for HA reconstruction Supraceliac aortohepatic conduit 11 0 RGEA-to-graft HA anastomosis 2 1 Death due to HAT 2 1 caliber recipient HAs (n 5 5), probably due to multiple transarterial chemoembolization (TACE) sessions for the hepatocellular carcinoma (HCC), no arterial flow (n 5 6), and development of acute thrombosis (n 5 3) after HA anastomosis. Two of the patients who developed acute thrombosis leading to no flow were subjected to thrombus removal using a Fogarty catheter, and a revision of the HA anastomosis was initially attempted. Supraceliac aortohepatic anastomosis was done in 11 patients using the left great saphenous vein, (17) whereas RGEA was used as an alternative to the recipient HA in 3 patients (1 patient in era 1 and 2 patients in era 2). One recipient with a supraceliac aortic conduit developed a pseudoaneurysm of the saphenous conduit from the aortic side on the second postoperative day and developed an intraabdominal hematoma. The patient underwent exploratory laparotomy, and the anastomosis on the aortic side was securely revised. (16) Two patients who underwent HA reconstruction using the RGEA had poor HA flow on the first postoperative day. Revision of the anastomosis was successfully done for 1 patient, whereas the second patient with RGEA-to-HA anastomosis developed primary graft failure. This patient received emergency deceased donor liver transplantation (DDLT) and recovered well thereafter. HA COMPLICATIONS IN THE POSTOPERATIVE PERIOD Overall, 14 (1.8%) patients developed HA-related complications in our entire series, which also includes the era 1 recipients (Table 2). In era 2, 9 (1.2%) patients developed HAT, whereas 2 patients developed nonthrombotic HA-related complications. One patient developed HAT within 24 hours of LDLT, whereas the remaining 8 recipients 894 ORIGINAL ARTICLE

9 LIVER TRANSPLANTATION, Vol. 23, No. 7, 2017 LI ET AL. HAT* (n 5 11) TABLE 3. Treatment Modalities and Outcome in LDLT Recipients With HAT Type of Graft Urokinase Therapy Revision of Anastomosis Retransplantation Survival Era 1 Patient 1 Right lobe Revision of HA anastomosis Alive Patient 2 Right lobe Attempted, but failed Died of sepsis Era 2 Patient 3 Right lobe Yes, Not successful Revision by supraceliac Alive aortohepatic conduit Patient 4 Right lobe Yes, Not successful Revision attempted DDLT Alive Patient 5 Right lobe Yes, Not successful Revision attempted DDLT Died of sepsis Patient 6 Right lobe Yes, Not successful Died of sepsis *Urokinase therapy was successful in 5 patients achieving complete recanalization without any complications. were diagnosed to have HAT after a minimum of 24 hours of the transplantation. Intra-arterial thrombolytic therapy with a continuous urokinase infusion was given in 9 patients of HAT (Table 3). Five of these patients developed complete recanalization after urokinase therapy without any morbidity in the posttransplant period. After successful recanalization, there was no evidence of arterial stenosis in subsequent follow-up with satisfactory liver graft functions. However, 4 patients were re-explored, and a revision of the HA anastomosis was attempted. One patient underwent HA reconstruction using the supraceliac aortohepatic conduit technique, whereas 2 patients underwent retransplantation by DDLT due to failure of the HA revision and the subsequent deterioration of the patients. One patient died due to overwhelming sepsis before retransplantation could be arranged. One patient who underwent retransplantation with a cadaveric liver graft died due to sepsis (Table 3). In era 2, the nonthrombotic HA-related complications that occurred in 2 patients were as follows: (1) the first patient developed intra-abdominal hematoma on the second postoperative day due to a leak from the pseudoaneurysm that occurred at the aortic side of the aortohepatic conduit (described earlier); (2) the second patient developed poor HA flow without any discernible causes. The revision of the HA anastomosis was attempted in this patient. However, the native HA could not be used after multiple attempts. The liver graft HA intimal dissection was noted, and retransplantation was done by an emergency DDLT. In era 1, 2 (8%) patients developed HAT, whereas 1 patient had poor HA flow due to anastomotic kinking (Table 2). HAT was diagnosed on the second postoperative day in 1 patient for whom a successful revision of the HA reconstruction was done. The second case of HAT was diagnosed after 1 month (Table 3). However, the patient developed sepsis and died of multiorgan failure. For the patient with poor HA flow due to kinking, an exploratory laparotomy was performed, and a revision of the HA anastomosis was attempted. However, because of the failure of the HA anastomosis, an extra-anatomic HA reconstruction was successfully done using RGEA. Discussion The reconstruction of the HA is a crucial step in LDLT, and any HA-related complications are detrimental to graft outcome in the posttransplant period. Although application of the microsurgical technique in LDLT reduced the arterial complication rates, the HA anastomosis under an operating microscope is always time consuming, and a long learning curve is required. A continuous end-to-end suturing technique using magnifying loupes appears more feasible and yields similar or even better results. (1,10,11,17) This large-scale experience of HA reconstruction under magnifying loupes in 741 consecutive LDLT recipients shows the safety and feasibility of the procedure. The average time taken for the completion of the HA anastomosis was minutes (range, 5-30 minutes) using loupes as compared with the prolonged operating time for HA reconstruction in era 1. (24 adult-to-adult LDLT (Male:Female, 18:6; mean age, years) were done whereas single pediatric LDLT (Female, 12 years) was performed). A right liver allograft was harvested for all the recipients in era 1. HA reconstruction was performed by a plastic surgeon using a microsurgery technique. One patient developed HAT, whereas 1 patient had poor HA flow. The HA anastomosis was unusually prolonged more than 60 minutes for all patients. Two patients died in this era directly related to HA complications in the posttransplant period. The increased rate of HA-related complications in era 1 can be attributed to the initial learning curve of the ORIGINAL ARTICLE 895

10 LI ET AL. LIVER TRANSPLANTATION, July 2017 liver transplant program. However, we changed our protocol to the currently used surgical loupe parachute technique because we thought that the use of the microsurgery technique was unnecessary and that similar or even better results could be achieved if HA reconstruction was performed under surgical loupes. The HA-related complications occurred in 1.5% (n 5 11) of the era 2 patients. Marubashi et al. (18) reported superior results after HA reconstruction using loupes of magnification with an advantage of reduced operative time. In this retrospective study, no incidence of HAT was reported, whereas 6 (5.5%) patients of their study cohort had interventional radiology confirmed HA stenosis. However, the false-positive rate of HA stenosis diagnosis was high until 3 months after transplantation (72.8% specificity). Initially a continuous suturing technique for arterial anastomosis for the partial liver allografts was argued because it can lead to an anastomotic stenosis with potential biliary complications. Coelho et al. (19) reported increased incidence of arterial complications while using a continuous suturing technique in HA anastomosis. However, a recently conducted study by Tzeng et al. showed the safety without incidence of HA stenosis or thrombosis. (1) The reported HA stenosis rate after liver transplantation ranges from 5% to 11%. (20,21) In our present cohort, there was no incidence of symptomatic HA stenosis leading to liver allograft dysfunction, and the continuous suturing technique, thus, has been proven to be safe and feasible. This can be attributed to our refinement of the HA anastomotic technique. To avoid major caliber discrepancies of the graft and recipient HAs, we obliquely cut the HAs or slit open the anterior wall of the graft HA to some extent and the corresponding point on the posterior wall of the recipient HA to increase the cross section, which can be twice as wide as the diameter of the smaller vessel. This facilitates the continuous suturing of the HAs. Because of the discrepancies of the graft and recipient HAs, however, there may be a false impression of the anastomotic narrowing of the HA on follow-up CT angiography. However, this is correlated with the Doppler USG of the abdomen and the clinical picture of the patient as well as the laboratory parameters for liver functions. A resistive index of the HA of <0.4 and/or a peak arterial velocity of <15 cm/s at the proximal part of HA on Doppler USG are considered abnormal, and in the presence of elevated liver enzymes, actangiographyofthehamustbeperformed.the symptomatic HA stenosis can be treated by endovascular intervention such as angioplasty and stenting. (22,23) Although the risk factors for the arterial complications in the post-ldlt period are not yet clarified, there are several conditions that can lead to poor HA flow such as kinking, undue angulation, spasm, or the development of acute thrombosis at the anastomotic site. If the cause of low HA flow is arterial spasm, we advocate a brief waiting period for the spasm to relieve and to rule out any anatomical factors mentioned above. In a HA spasm, portal vein clamping for 5-10 seconds does help by compensatorily increasing the arterial flow which may relieve the spasm. We administer PGE1 in all the patients who had a HA spasm. The kinking or angulation, however, needs revision of the HA anastomosis. The HAT is the most dreaded complication after LDLT and can lead to graft loss and mortality if not diagnosed at an early stage. In the literature, the reported incidence of HAT varies from 4% to 25%. (5,6,21,23,24) In era 2 (HA anastomosis using magnifying loupes), HAT occurred in 9 (1.2%) patients, whereas 2 (8%) patients from era 1 were diagnosed to have HAT. Although interrupted sutures for HA anastomosis were suggested initially to reduce the arterial complications, (19) we employed running sutures to all HA reconstruction procedures using 7-0 Prolene under magnifying loupes that yielded equal or even better results than previously described conventional microsurgery procedures. There was no correlation between the pretransplant TACE procedure and postoperative HAT occurrence. However, patients with multiple TACE sessions were more prone to native HA intimal dissection. Five of the 14 recipients who required extra-anatomical HA reconstruction had multiple TACE sessions for HCC prior to the transplantation. The therapeutic treatment options for HAT management include medical treatment or surgical intervention. If HAT develops in the immediate postoperative period, our protocol is to re-explore immediately and revise the anastomosis if possible. However, if the patient is clinically stable, intraarterial thrombolytic therapy can be the initial modality of treatment. In the present cohort, 9 patients with HAT (in era 2) received initial medical treatment using urokinase as a continuous hepatic arterial infusion. A complete recanalization of the HA was achieved in 5 (55.5%) patients without any deleterious consequences. Urokinase infusion was given for 24 hours at the rate of 60,000 IU/hour, and hepatic angiography was repeated at the end of infusion. Any clinical deterioration or failure to reanalyze were the deciding factors for re-exploration. The good 896 ORIGINAL ARTICLE

11 LIVER TRANSPLANTATION, Vol. 23, No. 7, 2017 LI ET AL. outcome after using urokinase infusion as initial therapy for the HAT has been reported earlier. (25,26) If HAT is suspected in the immediate postoperative period or when intra-arterial thrombolytic therapy fails, urgent re-exploration is warranted. The HA anastomosis should be revised and risk factors such as kinking or angulation should be corrected by excising the redundant portion of the HA. In the presence of the recipient HA intimal dissection, an extraanatomical arterial reconstruction must be established. The extra-anatomic HA revascularization can be achieved by using a recipient right gastric artery, RGEA, left gastric artery, or splenic artery without an interpositional vascular conduit and from the aorta (either supraceliac or infrarenal) using a vascular graft. (16,27,28) In our experience, 11 of the patients required a supraceliac aortohepatic conduit using the left great saphenous vein, (16) whereas RGEA was used for HA reconstruction in 3 patients. One patient with an aortohepatic conduit, however, suffered from pseudoaneurysm at the aortic side that caused hematoma and hemodynamic instability. This complication was rectified by emergency exploratory laparotomy and revision of the anastomosis. Occurrence of pseudoaneurysm arising from HA after LDLT is not uncommon and can successfully be treated by endovascular treatment. (29) Our experience highlights that HA reconstruction in LDLT using surgical loupes can be done safely and that the continuous suturing technique does not lead to increased arterial complication. Factors such as smaller arterial diameter, anastomotic method, and use of magnifying loupes were not associated with an increased risk of arterial complications. Use of magnifying loupes for HA reconstruction increases the ease of anastomosis and appears more feasible to overcome the excursion of the diaphragm during the anastomosisasthehasarepresentdeepintheabdominalcavity and create technical difficulties if an operating microscope is used. With a low HA-related complication rate of 1.5%, we state that the parachute technique with running sutures even in smaller diameter HAs is a reliable technique and can easily be applied by a well-experienced surgeon. REFERENCES 1) Tzeng YS, Hsieh CB, Chen SG. Continuous versus interrupted suture for hepatic artery reconstruction using a loupe in livingdonor liver transplantation. Ann Transplant 2011;16: ) Uchiyama H, Hashimoto K, Hiroshige S, Harada N, Soejima Y, Nishizaki T, et al. Hepatic artery reconstruction in living-donor liver transplantation: a review of its techniques and complications. Surgery 2002;131(suppl):S200-S204. 3) Mazzaferro V, Esquivel CO, Makowka L, Belle S, Kahn D, Koneru B, et al. Hepatic artery thrombosis after pediatric liver transplantation--a medical or surgical event? Transplantation 1989;47: ) Tanaka K, Uemoto S, Tokunaga Y, Fujita S, Sano K, Nishizawa T, et al. Surgical techniques and innovations in living related liver transplantation. Ann Surg 1993;217: ) Broelsch CE, Whitington PF, Emond JC, Heffron TG, Thistlethwaite JR, Stevens L, et al. Liver transplantation in children from living related donors. Surgical techniques and results. Ann Surg 1991;214: ) Mori K, Nagata I, Yamagata S, Sasaki H, Nishizawa F, Takada Y, et al. The introduction of microvascular surgery to hepatic artery reconstruction in living-donor liver transplantation--its surgical advantages compared with conventional procedures. Transplantation 1992;54: ) Inomoto T, Nishizawa F, Sasaki H, Terajima H, Shirakata Y, Miyamoto S, et al. Experiences of 120 microsurgical reconstructions of hepatic artery in living related liver transplantation. Surgery 1996;119: ) Ohdan H, Tashiro H, Ishiyama K, Ide K, Shishida M, Irei T, et al. Microsurgical hepatic artery reconstruction during livingdonor liver transplantation by using head-mounted surgical binocular system. Transpl Int 2007;20: ) Guarrera JV, Sinha P, Lobritto SJ, Brown RS Jr, Kinkhabwala M, Emond JC. Microvascular hepatic artery anastomosis in pediatric segmental liver transplantation: microscope vs loupe. Transpl Int 2004;17: ) Banshodani M, Tashiro H, Onoe T, Ide K, Ohdan H. Longterm outcome of hepatic artery reconstruction during livingdonor liver transplantation. Transplant Proc 2011;43: ) Li PC, Jeng LB, Yang HR, Lee CC, Poon KS, Chen TH, et al. Hepatic artery reconstruction in living donor liver transplantation: running suture under surgical loupes by cardiovascular surgeons in 180 recipients. Transplant Proc 2012;44: ) See comment in PubMed Commons below Tannuri AC, Monteiro RF, Santos MM, Miyatani HT, Tannuri U. A new simplified technique of arterial reconstruction in pediatric livingdonor liver transplantation: a comparison with the classical technique. J Pediatr Surg 2014;49: ) Jeng LB, Thorat A, Li PC, Li ML, Yang HR, Yeh CC, et al. V-Plasty technique using dual synthetic vascular grafts to reconstruct outflow channel in living donor liver transplantation. Surgery 2015;158: ) Jeng LB, Thorat A, Yang HR, Yeh CC, Chen TH, Hsu SC, et al. Rooftop and skeletonization technique of hepatic transection to include or exclude the middle hepatic vein during donor hepatectomy in living donor liver transplantation: solving the middle hepatic vein controversy experience in 397 sequential live donors. Med Sci Tech 2016;57: ) Lee CC, Jeng LB, Li PC, Yang HR, Lu CW, Chen TH, et al. Innovative technique to reconstruct two branches of the right hepatic artery in living donor liver transplantation. Transplant Proc 2008;40: ) Li PC, Thorat A, Jeng LB, Yang HR, Li ML, Yeh CC, et al. Successful application of supra-coeliac aorto-hepatic conduit using saphenous venous graft in right lobe living donor liver transplantation. Liver Transpl 2017;doi: /lt ) Enne M, Pacheco-Moreira L, Balbi E, Cerqueira A, Alves J, Valladares MA, et al. Hepatic artery reconstruction in pediatric ORIGINAL ARTICLE 897

12 LI ET AL. LIVER TRANSPLANTATION, July 2017 living donor liver transplantation under 10 kg, without microscope use. Pediatr Transplant 2010;14: ) Marubashi S, Kobayashi S, Wada H, Kawamoto K, Eguchi H, Doki Y, et al. Hepatic artery reconstruction in living donor liver transplantation: risk factor analysis of complication and a role of MDCT scan for detecting anastomotic stricture. World J Surg 2013;37: ) Coelho GR, Leitao AS Jr, Cavalcante FP, Brasil IR, Cesar- Borges G, Costa PE, et al. Continuous versus interrupted suture for hepatic artery anastomosis in liver transplantation: differences in the incidence of hepatic artery thrombosis. Transplant Proc 2008;40: ) Kim BS, Kim TK, Jung DJ, Kim JH, Bae IY, Sung KB, et al. Vascular complications after living related liver transplantation: evaluation with gadolinium enhanced three-dimensional MR angiography. AJR Am J Roentgenol 2003;181: ) Tamsel S, Demirpolat G, Killi R, Aydin U, Kilic M, Zeytunlu M, et al. Vascular complications after liver transplantation: evaluation with Doppler US. Abdom Imaging 2007;32: ) Wakiya T, Sanada Y, Mizuta K, Urahashi T, Ihara Y, Yamada N, et al. A comparison of open surgery and endovascular intervention for hepatic artery complications after pediatric liver transplantation. Transplant Proc 2013;45: ) Gad EH, Abdelsamee MA, Kamel Y. Hepatic arterial and portal venous complications after adult and pediatric living donor liver transplantation, risk factors, management and outcome (A retrospective cohort study). Ann Med Surg (Lond) 2016;8: ) Tian MG, Tso WK, Lo CM, Liu CL, Fan ST. Treatment of hepatic artery thrombosis after orthotopic liver transplantation. Asian J Surg 2004;27: ) Sevmis S, Karakayali H, Tutar NU, Boyvat F, Ozcay F, Torgay A, Haberal M. Management of early hepatic arterial thrombosis after pediatric living-donor liver transplantation. Transplant Proc 2011;43: ) Jeon GS, Won JH, Wang HJ, Kim BW, Lee BM. Endovascular treatment of acute arterial complications after living-donor liver transplantation. Clin Radiol 2008;63: ) Ikegami T, Kawasaki S, Hashikura Y, Miwa S, Kubota T, Mita A, et al. An alternative method of arterial reconstruction after hepatic arterial thrombosis following living-related liver transplantation. Transplantation 2000;69: ) Dokmak S, Aussilhou B, Landi F, Dondero F, Termos S, Paugam-Burtz C, et al. The recipient celiac trunk as an alternative to the native hepatic artery for arterial reconstruction in adult liver transplantation. Liver Transpl 2015;21: ) Thorat A, Lee CF, Wu TH, Pan KT, Chu SY, Chou HS, et al. Endovascular treatment for pseudoaneurysms arising from the hepatic artery after liver transplantation. Asian J Surg 2017;40: ORIGINAL ARTICLE