Cryopreserved Iliac Artery is Indispensable Interposition Graft Material for Middle Hepatic Vein Reconstruction of Right Liver Grafts Shin Hwang, Sung-Gyu Lee, Chul-Soo Ahn, Kwang-Min Park, Ki-Hun Kim, Deok-Bog Moon, and Tae-Yong Ha Cryopreserved iliac vein grafts (IVGs) have often been used for reconstruction of middle hepatic vein (MHV) branches in right liver grafts, but their storage pool has often been exhausted in our institution due to the low incidence of deceased donor organ procurement. To overcome this shortage of IVG, we started to use cryopreserved iliac artery graft (IAG). During September and October 2004, we carried out 41 cases of adult living donor liver transplantation, including 29 right lobe grafts with MHV reconstruction. Interposition vessel grafts were autologous vein (n 6), IVG (n 13), and IAG (n 10). IAG was used in 3 (21%) of 13 cases during the first month. For the next month, it was more frequently used (7 [44%] of 16) because handling of cryopreserved IAG was not difficult and its outcome was favorable. On follow-up with computed tomography for 3 months, outflow disturbance occurred in 1 (17%) of 6 autologous vein cases, in 2 (15%) of 13 IVG cases, and in 1 (10%) of 10 IAG cases. Twomonth patency rate of IAG was not lower than that of IVG. In conclusion, we feel that cryopreserved IAG can be used as an interposition vessel graft for MHV reconstruction of right liver graft when cryopreserved IVG is not available. (Liver Transpl 2005;11:644-649.) Abbreviations: IVG, iliac vein graft; MHV, middle hepatic vein; IAG, iliac artery graft; HVC, hepatic venous congestion; RL, right lobe; GSV, greater saphenous vein; LDLT, living donor liver transplantation; SHV, short hepatic vein; CT, computed tomography; V5, segment V hepatic vein; V8, segment VIII hepatic vein. From the Division of Hepatopancreatobiliary Surgery and Liver Transplantation, Department of Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea. Address reprint requests to Sung-Gyu Lee, MD, Division of Hepatopancreatobiliary Surgery and Liver Transplantation, Department of Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, 138-736, Korea. Telephone: 82-2-3010-3485; FAX: 82-2-474-9027; E-mail: sglee2@amc.seoul.kr Copyright 2005 by the American Association for the Study of Liver Diseases Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/lt.20430 Hepatic venous congestion (HVC) from deprivation of middle hepatic vein (MHV) outflow in the right lobe (RL) graft can be prevented by interposition of a vein graft between the major tributaries of MHV and recipient inferior vena cava. 1 In addition to the recipient s autologous vein grafts, such as greater saphenous vein (GSV), left portal vein, and paraumbilical vein, homologous vein grafts such as cryopreserved iliac vein graft (IVG) also have been used for this purpose. 2,3 Recently, cryopreserved IVG became the most preferred vessel graft for MHV reconstruction in our institution, because it did not require time-consuming dissection of recipient tissues and it was usually large and long enough to use for any variant type of MHV branch. However, this IVG preference often exhausted the cryopreserved IVG pool in our situation, because the number of living donor liver transplantations (LDLTs) requiring MHV reconstruction exceeded the low incidence of deceased donor organ procurement. On the other hand, a number of iliac artery grafts (IAGs) always were left in storage because they were never previously used. Although there was no reference to IAG use as a vein graft substitute to our knowledge at that time, we tried to use the cryopreserved IAG in order to overcome the disproportionate shortage of vein grafts. Only after application of several RL grafts did we feel that it was possible to replace IVG with IAG without any serious problem. We herein describe our early experience of MHV reconstruction using IAG for RL grafts. Patients and Methods Recipients and Living Donors During the months of September and October 2004, we have performed 41 cases of adult-to-adult LDLT using 29 RL grafts, 2 left lobe grafts, and 9 dual grafts. As there was one dual LDLT case with combination of RL and left lobe grafts, the total number of RL grafts was 30. Twenty-nine (96.7%) of 30 RL grafts underwent MHV reconstruction (modified RL graft), except for one RL graft that showed an HVC equivalent to only 12% of RL volume as well as graft-recipient weight ratio of 1.3. 4,5 Primary diagnoses of these 29 RL graft recipients were hepatitis B associated liver cirrhosis (n 23), alcoholic cirrhosis (n 2), fulminant hepatic failure (n 2), hepatitis C associated cirrhosis (n 1), and secondary biliary cirrhosis (n 1). Mean age of the recipients was 48 12 (range, 22-62) years, and 21 recipients were male. To date, all of these RL graft recipients are alive after 3 months. Mean age of these 29 living donors was 28 10 (range, 17-53) years, and 18 donors were male. Mean values of graft-recipient weight ratio in 28 single modified RL grafts was 1.02 0.22 (range, 0.73-1.44). All of these RL graft 644 Liver Transplantation, Vol 11, No 6 (June), 2005: pp 644-649
Venous Interposition Using IAG 645 donors recovered uneventfully and were discharged without complication. Indication for Reconstruction of Middle Hepatic Vein Branches and Short Hepatic Veins While preparing to carry out reconstruction of MHV branches or short hepatic vein (SHV), we considered four factors: amount of HVC (from computed tomography [CT] prediction, simulative hepatic vein occlusion, or discoloration after parenchymal transection); size of liver graft (graft volume to recipient s standard liver volume or graft-recipient weight ratio); available material for interposition vessel graft; and technical feasibility of multiple small-sized anastomoses. 4-7 In principle, we reconstructed the major branches of MHV and SHV greater than 5 mm in diameter to make the residual HVC not exceed 10% of RL graft volume. Some MHV or SHV branches approximately 5 mm in diameter were sacrificed to avoid multiple complex anastomoses. Selection of Interposition Vessel Graft The donor surgical team was in close cooperation with the recipient surgical team to determine the most suitable vessel graft for MHV reconstruction. Every possibility of using the recipient s left portal vein, GSV, dilated paraumbilical vein, or cryopreserved IVG was taken into account. Whenever any of these materials were not available, we decided to use cryopreserved IAG during the first month. Thereafter, we no longer hesitated to choose IAG as the interposition graft material. It was not mandatory to match ABO blood group between the cryopreserved vessel and recipient. For largesized MHV tributaries, we preferred a large-calibered vessel graft to GSV. Follow-Up for Patency of the Interposed Vessel Grafts Doppler ultrasonography and dynamic liver CT scans were used for patency follow-up of the interposed vessel grafts. As Doppler ultrasonography was primarily intended to evaluate the hepatic arterial and portal inflows, we have typically used the dynamic liver CT scan as a main tool to evaluate the tissue perfusion and venous outflow of the graft liver. When serum liver enzyme levels rose much higher than their usual values during the first 24 hours after LDLT surgery (e.g., serum aspartate aminotranseferase or alanine aminotransferase 500 IU/L) or routine ultrasonography detected abnormal findings, we often carried out a day-1 CT scan as a screening test. Venous outflow disturbance of the segment V hepatic vein (V5) or segment VIII hepatic vein (V8) was treated by the percutaneous transjugular approach of balloon dilatation and metallic stent insertion. 8 During the first month after LDLT surgery or admission, CT scan was performed weekly or biweekly. Thereafter, it was performed monthly for 2 or 3 times and the follow-up interval became widened. We inspected every CT section and 3-dimensional reconstruction image to evaluate the enhancement state and running course of interposed vessel graft, vessel size discrepancy or stenosis at the liver cut surface, amount and nature of parenchymal perfusion defect, and their interval changes. The significance of these differences was assessed using Fisher s exact test. Differences at P values less than 0.05 were considered to be statistically significant. Results MHV Reconstruction using Autologous Vein Grafts (n 6) Recipient s left portal vein with hilar confluence portion was used for reconstruction of V5 and V8 (n 3). The wall defect of the main portal vein orifice was repaired with a GSV patch to keep the graft length and curvature. Its side branches to the segment 2, 3, or 4 were used for double V5 orifices. Both autologous portal vein and dilated paraumbilical vein were used at the same time (n 1). Autologous GSV was also used alone after Y-shaped anastomosis (n 2). The number and size of actually reconstructed V5 and V8 are summarized in Table 1. These interposition grafts were anastomosed to the middle-left hepatic vein stump (n 5) and directly to the retrohepatic inferior vena cava (n 1). One RL recipient of a portal vein graft revealed a perfusion defect at the segment VIII parenchyma on day-1 CT. Anastomotic stenosis of the V8 stump was identified, and was corrected by balloon dilatation and stent insertion at day 1. There was no significant perfusion defect of the RL graft parenchyma except for a small area near the liver cut surface on week-1 CT scans of these six RL grafts. All of V5 and V8 anastomoses at the liver cut surface as well as interposed autologous vessel grafts were identified on month-2 CT scans. MHV Reconstruction using IVG (n 13) Cryopreserved IVG (n 12) and cold-stored IVG (n 1; 2 days in refrigerator) were used for reconstruction of V5 and V8. Most of the cryopreserved IVGs were stored for less than 3 months (ranging from 4 days to 15 months). A branch vein was attached to the side wall of the IVG for double V8 orifices (n 1). When the IVG branch (external or internal iliac vein) diameter was not twice larger than that of V5 or V8, we usually anastomosed it in an end-to-end fashion (Fig. 1). If there was a definite size discrepancy of more than two-fold, it was often anastomosed in an end-to-side fashion after tie-ligation of the IVG branch stump. All of the IVG grafts were anastomosed to middle-left hepatic vein stump without excessive redundancy. In
646 Hwang et al. Table 1. Number and Size of Reconstructed MHV Branches Classified by the Sources of Interposition Vessel Grafts Autologous vein graft (n 6) Iliac vein graft (n 13) Iliac artery graft (n 10) Single Segment V Hepatic Vein (V5) Double Diameter* (mm) Single Segment VIII Hepatic Vein (V8) Double Diameter* (mm) 5 1 8 2 (5-12) 4 2 6 1 (5-7) 10 3 8 3 (5-15) 11 2 7 1 (5-10) 5 5 9 3 (5-15) 6 1 6 2 (5-10) *Values given as mean standard deviation (range). V8 in 3 of 10 right lobe grafts were not reconstructed due to their small size. addition, quilt venoplasty for double and quadruple SHV using a GSV patch was carried out (n 2). 9 There was no significant perfusion defect of RL graft except for a small area on week-1 CT scans in these 13 RL grafts. However, we detected a mottled perfusion defect at the segment V parenchyma and faint visualization of interposition graft on month-1 CT of a recipient who had used an IVG stored for 15 months. We inserted wall stents percutaneously to correct the anastomotic stenosis of V5, after which the perfusion defect disappeared on follow-up CT. In another recipient, enhancement of an interposed vessel graft disappeared on month-3 CT, but there was no noticeable perfusion Figure 1. Reconstruction of single V5 and single V8 using a cryopreserved iliac vein graft. As the orifice of V5 was 10 mm in diameter and the external iliac vein stump was close to 20 mm in diameter, they were anastomosed in an endto-end fashion. The 15-mm sized stump of the internal iliac vein was anastomosed to an 8-mm sized V8 in an end-to-end fashion. Some growth factor was applied during tying of the suture material to prevent anastomotic stenosis. The common iliac vein portion was anastomosed to the middle-left hepatic vein stump. defect at the RL graft. We were able to observe intrahepatic enhancement of V5 and V8 indicating development of collaterals in this recipient CT. In the remaining 11 recipients, all of their anastomoses at the liver cut surface as well as interposed IVG were visualized on month-2 or 3 CT scans. MHV Reconstruction using IAG (n 10) Cryopreserved IAG (n 9) and cold-stored IAG (n 1; 1 day in refrigerator) were used for reconstruction of V5 and V8. All of the cryopreserved IAG were stored for less than 8 months because we had once discarded most of the stored IAG. A branch artery was attached at the side wall of the IAG (n 2) to make an additional orifice of V5 or V8. When the IAG branch (external or internal iliac artery) diameter was not more than 1.5- fold of the V5 or V8 diameter, we usually anastomosed it in an end-to-end fashion (Fig. 2). When there was a definite size discrepancy of more than 1.5-fold, we often reconstructed them in an end-to-side fashion after tie-ligation of IAG branch and oval excision of IAG side wall. All of the IAG grafts were anastomosed to the middle-left hepatic vein stump without excessive redundancy. In addition, quilt venoplasty using a GSV patch for double SHV was applied (n 1). 9 There was no significant perfusion defect of the RL graft except for a small area on week-1 CT scans. However, faint mottling of segment V parenchyma and disappearance of IAG enhancement were detected on month-2 CT of the first IAG case. In the other 9 recipients, all of their anastomoses at the liver cut surface as well as the interposed IAG were identified on month-2 CT scans. The two-month patency rate of the interposed IAG was 90%, which was not lower than that of autologous vein graft or IVG (P 0.1). Besides MHV reconstruction, a
Venous Interposition Using IAG 647 Figure 2. Reconstruction of single V5 and single V8 using a cryopreserved iliac artery graft. As the orifice of V5 was 11 mm in diameter and the external iliac artery stump was 10 mm in diameter, they were anastomosed in an end-toend fashion. The 13-mm sized stump of internal iliac artery was anastomosed to the 9-mm sized V8 in an endto-end fashion. This size discrepancy of the stump orifices and the elasticity of the arterial wall worked as a pulling force to the edge of V8 orifice. As a result, the orifice of V8 became open spontaneously. The common iliac artery portion was anastomosed to the middle-left hepatic vein stump. short segment of IAG was used as a patch graft to enlarge the cuff of the SHV stump like quilt venopasty (n 1, Fig. 3). 9 Discussion The initial interposition graft material for MHV reconstruction was the recipient s GSV after hydrostatic dilatation. 5 However, its small diameter turned the surgeons attention toward larger calibered autologous veins as well as cryopreserved IVG. 3,10 As the interposed graft vessel works as a simple conduit to drain off the low-pressure hepatic venous outflow, any graft material of adequate diameter and length can be used. Although the use of synthetic material had been presented before, we did not dare to use such material because LDLT is not a totally clean surgery and graft infection can be a fatal complication. Our criteria for the resources to be used as an interposition vein graft was the use of every available vein segment regardless of its origin before adoption of IAG. Currently, this has changed to every available vessel segment because we realized that IAG can replace IVG, although it lacks long-term follow-up results. 2 IAG and IVG are obtainable in one segment during cadaveric donor organ procurement. As the iliac vein itself is too weak to pull out without the iliac artery Figure 3. Benchworks for the interposition of V5 and V8 as well as a patchwork of short hepatic vein (SHV) by using a cryopreserved iliac artery graft. At first, we distended the IAG graft by branch clamping and perfusion solution injection and placed it at the liver cut surface. We designed the lengths and angles of V5 and V8 limbs case by case. Adequate adjustment of V8 limb length must be very important so as not to make a buckling deformity from excessive redundancy. Blank arrow head indicates the orifice of the right hepatic vein. This design should simulate the real situation of graft implantation. In this case, we carried out patchwork for two SHV as in quilt venoplasty: a short segment of IAG was opened and placed over the SHV (arrow) for trimming. Two central openings were made and each opening was anastomosed to each SHV orifice. Peripheral cuff of IAG patch became wrinkled due to dimpling of the central anastomoses. Because this conjoined cuff allowed some redundancy between the right lobe graft and recipient s inferior vena cava, a good operating field was obtained as well. This is a sample figure taken after completion of this study. during harvest, we have to manipulate these two vascular structures altogether. As a result, we have usually obtained IVG and IAG of similar lengths at the same time. The simple concept of using cryopreserved IAG as a substitute for IVG actually made the number of available vessel grafts double without any additional expense. This means that one set of iliac vessels from a cadaveric donor can be used for at least four RL grafts. At first we used IAG only in the troublesome situations where no other suitable vein graft was available. After 2 weeks of follow-up in the first three cases, we were assured that IAGs could cover the shortage of IVGs to some extent. In fact, there were only 3 IAG cases (21%) out of 13 RL LDLTs during the first month. For the next month, it was more frequently used (7 [44%] of 16 RL LDLTs). This frequent use of IAG also alleviated the effort to harvest autologous vein grafts. The short-term follow-up results of IAG were comparable to those of IVG and autologous vein grafts.
648 Hwang et al. There was only 1 case (10%) of delayed occlusion among 10 RL grafts using IAG. Of course, we can expect that such occlusion of late onset will occur more frequently as time goes by, but we are sure that it may not lead to significant impairment of graft function. Based on our experience of 400 RL grafts with MHV reconstruction, maintenance of V5 and V8 outflows for the first month would be sufficiently long to prevent HVC-associated graft failure. We recognized some advantageous aspects of IAG: first, IAG has a thick wall, by which its length, axis, and shape did not change after restoration of MHV blood flow. This characteristic of IAG means that morphologic- and hemodynamic-based designs for MHV reconstruction become much simpler in IAG use than in IVG use (Figs. 2 and 3). In reality, both innate and cryopreserved IVGs have very thin walls, by which their length and graft axis are vulnerable to alteration after full expansion (Fig. 1). Second, IAG can keep the smallsized orifice of V5 or V8 open spontaneously because the elasticity of the arterial orifice will pull the periphery of the vein stump spontaneously (Fig. 2). This feature may be beneficial to decrease the incidence of early anastomotic stenosis. For end-to-side or side-to-side anastomosis, we excised some side walls of the IAGs to use this advantage and to prevent slitlike anastomotic narrowing. Briefly, the surgical technique for IAG is much easier than that for IVG. One theoretical disadvantage of IAG use is its limited distensibility at the lower hepatic vein pressure. In fact, the image of the interposed IAG looked smaller calibered on follow-up CT than it appeared during surgery (Fig. 4). However, on Doppler ultrasonography, a monophasic waveform was detected only in 1 IAG case, while the other 9 IAG cases revealed a biphasic or triphasic pattern. 11 We presume that the low distensibility or compliance of the IAG wall to such a low venous pressure may not affect the intraluminal flow pattern. We have often observed atherosclerotic change of the iliac artery during deceased donor vessel procurement or benchwork. We think that IAG with a smooth thin plaque is acceptable as an adequate interposition vessel graft for MHV reconstruction. At this time, the general rule of arterial stitch from the intima to adventitia should be observed to prevent vessel layer separation. We also used a short segment of IAG for patch plasty of the SHV. As the wall of the IVG was usually too weak for such a patchwork, we devised a quilt venoplasty using a GSV patch. 9 At that time, we felt that the characteristics of the IAG patch were much similar to Figure 4. Three-dimensional reconstruction of week-3 CT scan to show the full length of interposed iliac artery graft. Common iliac artery portion was buried in the expanding hematoma (open arrowhead), but its intraluminal flow was well maintained. The inner diameter of the external iliac artery portion seemed to be smaller than we expected, but there was no definite anastomotic stenosis of V5 (arrow). those of the GSV patch. This implies that quilt venoplasty can be achieved simply by using a wide IAG patch rather than a patchwork of narrow GSV segments (Fig. 3). Based on our early experience, handling of IAG was easier than we expected and its short-term patency rate was not lower than that of IVG. Therefore, we think that cryopreserved IAG can be liberally used as an interposition vessel graft for MHV reconstruction of an RL graft when cryopreserved IVG is not available. References 1. Lee SG, Park KM, Hwang S, Lee YJ, Choi DN, Kim KH, et al. Congestion of right liver graft in living donor liver transplantation. Transplantation 2001;71:812-814. 2. Lee KW, Lee DS, Lee HH, Joh JW, Choi SH, Heo JS, et al. Interposition vein graft in living donor liver transplantation. Transplant Proc 2004;36:2261-2262. 3. Cattral MS, Greig PD, Muradali D, Grant D. Reconstruction of middle hepatic vein of a living-donor right lobe liver graft with recipient left portal vein. Transplantation 2001;71:1864-1866. 4. Hwang S, Lee SG, Park KM, Kim KH, Ahn CS, Lee YJ, et al. Hepatic venous congestion in living donor liver transplantation: preoperative quantitative prediction and follow-up of its sequences using computed tomogram. Liver Transpl 2004;10: 763-770. 5. Lee SG, Park KM, Hwang S, Kim KH, Choi DN, Joo SH, et al. Modified right liver graft from a living donor to prevent congestion. Transplantation 2002;74:54-59. 6. Hwang S, Lee SG, Kim KH, Park KM, Lee YJ, Ahn CS, et al. Intraoperative assessment of hepatic venous congestion with direct clamping of the hepatic vein trunk for living donor liver transplantation. Transplant Proc 2004;36:1462-1465. 7. Sano K, Makuuchi M, Miki K, Maema A, Sugawara Y, Imamura
Venous Interposition Using IAG 649 H, et al. Evaluation of hepatic venous congestion: proposed indication criteria for hepatic vein reconstruction. Ann Surg 2002;236:241-247. 8. Ko GY, Sung KB, Yoon HK, Kim JH, Song HY, Seo TS, et al. Endovascular treatment of hepatic venous outflow obstruction after living-donor liver transplantation. J Vasc Interv Radiol 2002;13:591-599. 9. Hwang S, Lee SG, Park KM, Kim KH, Ahn CS, Moon DB, et al. Quilt venoplasty using recipient saphenous vein graft for reconstruction of multiple short hepatic veins in right liver grafts. Liver Transpl 2005;11:104-107. 10. Dong G, Sankary HN, Malago M, Oberholzer J, Panaro F, Knight PS, et al. Cadaver iliac vein outflow reconstruction in living donor right lobe liver transplantation. J Am Coll Surg 2004;199:504-507. 11. Ko EY, Kim TK, Kim PN, Kim AY, Ha HK, Lee MG. Hepatic vein stenosis after living donor liver transplantation: evaluation with Doppler US. Radiology 2003;229:806-810.