Use of Endarterectomized Atherosclerotic Artery Allograft for Hepatic Vein Reconstruction of Living Donor Right Lobe Graft

Abstract

When a living donor right lobe graft is implanted into an adult recipient, some vein graft is often necessary to use as an interposition conduit for middle hepatic vein reconstruction or as patch material to facilitate the reconstruction of the main and accessory right hepatic veins. For such purposes, every available kind of vessel allograft has been used; from the recipients, greater saphenous vein, portal vein, paraumbilical vein, and intrahepatic hepatic vein were procured; from the deceased donors, iliac vein, iliac artery, internal jugular vein, and other lengthy vessel segments were collected. Even synthetic vessel was used, although rarely. In practice, we have preferred iliac vein allografts because of their size and shape, flexible handling, and long-term luminal patency. We previously reported that cryopreserved iliac artery graft was successfully used for such purposes.1 On the other hand, a considerable proportion of the deceased donors had various extents of atherosclerosis at the iliac artery. Intimal thickening from moderate to severe atherosclerosis inevitably resulted in marked narrowing of the luminal diameter, so the surgeon had to discard a large proportion of such arterial allografts. Intractable shortage of vessel grafts coupled with rapidly increasing living donor liver transplantation pushed us to increase the use of such suboptimal vessels for conduits.1 We incidentally found that atherosclerotic intima was easily separated from the outer arterial wall after cold storage or cryopreservation. A simple eversion endarterectomy converted a small-luminal atherosclerotic artery into a large-luminal conduit (Fig. 1). It is basically similar to the ordinary carotid endarterectomy. We here present clinical application of the endarterectomized arterial graft, which was prepared by an eversion endarterectomy technique. Eversion endarterectomy technique to remove atherosclerotic core of cryopreserved iliac artery. (A) Iliac artery graft was stretched. (B) Partially atherosclerotic inner core was gently pulled out from outer sheath of tunica media and adventitia. Both everted luminal surface and inner core surface appear smooth and glistening. Endarterectomized iliac artery graft was first used when reconstruction of the graft middle hepatic vein tributaries appeared inevitable, but no large-bored vessel allograft other than atherosclerotic artery was available from the tissue bank. After thawing a long cryopreserved iliac artery, the atherosclerotic core was gently separated, and the outer sheath of tunica media and adventitia was everted and completely peeled off. The everted luminal side was thoroughly inspected to evaluate its smoothness; rough surface may increase the risk of thrombosis. All lining surface was smooth and glistened under surgical microscopy (magnification 10×). The everted sheath was inverted again to place the luminal surface inside. The sheath was then trimmed and used for reconstruction of the middle hepatic vein tributaries at the right lobe graft (Fig. 2). Six-month follow-up with computed tomography and Doppler ultrasonography did not reveal any abnormality of the middle hepatic vein outflow (Fig. 3). Middle hepatic vein reconstruction using completely endarterectomized artery allograft. Segment of severely atherosclerotic iliac artery graft was divided into inner atheroma (yellowish longer segment) and outer sheath (white shorter segment) after application of eversion endarterectomy. Inner diameter of outer sheath is equal to outer diameter of atheroma core. External and internal iliac artery portions were trimmed and anastomosed to 9-mm-sized segment V vein (V5) and 7-mm-sized segment VIII vein (V8), respectively. Computed tomographic follow-up of middle hepatic vein reconstruction. (A) After 1 week, interposed endarterectomized iliac artery conduit (arrow) appeared patent with no attenuation abnormality of liver parenchyma. (B) After 2 months, interposition artery graft (arrow) looked slightly narrowed, but intrahepatic middle hepatic vein branches were well enhanced. At same time, Doppler ultrasonography revealed biphasic or triphasic flow patterns at reconstructed middle hepatic vein tributaries. Endarterectomy of atherosclerotic artery is not a new concept because open carotid endarterectomy has been performed for decades, but to our knowledge, allograft endarterectomy has not yet been reported in the literature. Our experience suggests that an endarterectomized artery can substitute for a vein graft. In a fresh state, or even after cryopreservation or cold storage, normal arterial intima is not easily separated from the outer wall of tunica media and adventitia. When applying forceful endarterectomy to the nonatherosclerotic artery, the luminal surface becomes rough as a result of incomplete detachment of the intima. However, atheroma separation is much easier in cold-stored or cryopreserved arteries with moderate to severe atherosclerosis. It has been reported that cold storage or cryopreservation induces a slight loosening of intimal attachment at the fibroelastic layer, which may facilitate this separation.2 In line with these findings, our preliminary observations suggested that 24- or 48-hour cold storage appeared more effective for atheroma separation than long-term cryopreservation. This newly formed artery allograft has a smooth luminal surface, but it does not have an intimal lining. On the basis of our experience of 500 middle hepatic vein reconstructions performed with various interposition vessel grafts, we think that the intimal integrity of the interposed vessel grafts is not essential to maintain luminal patency because those vessel grafts work as only a simple conduit of hepatic venous outflow.3, 4 Intimal lining of the cryopreserved iliac vein or artery is always damaged at histological examination,2 but such microscopic injury of the lining endothelium does not seem to be associated with thrombogenesis because middle or right hepatic vein drainage is a relatively high-flow system, initial posttransplantation coagulation profile is markedly decreased, and new intimal lining will quickly be formed.5 However, we also found that the endarterectomized lining surface appeared macroscopically rough in a few grafts, implying incomplete separation of the intima (Fig. 4). Since we were afraid of having a risk of thrombus formation, such vessel graft was not used as a vein conduit. It may not be necessary to use antiplatelet agents in the usual thrombocytopenic situations, but they can be administered prophylactically when platelet count is restored or when the interposition graft is in a suboptimal condition. Suboptimal eversion endarterectomy of cryopreserved iliac artery graft. Both everted luminal surface and inner core surface appear rather rough, implying incomplete separation of atherosclerotic intima. Because atheroma core of this artery was friable, some intimal fragments were left at outer sheath. Such artery graft with irregular luminal surface is a poor choice for use as vein conduit. The consistency and thickness of the endarterectomized iliac or femoral artery closely resemble those of the fresh femoral vein, which has a thicker wall than the proximal iliac vein portion. Because such a physical nature of endarterectomized artery allograft was also comparable to those of the fresh greater saphenous vein, endarterectomized artery graft was used as a vein patch material during patch or quilt venoplasty for reconstruction of the main or accessory right hepatic veins.6-8 We used these endarterectomized iliac artery grafts as venous outflow conduit or patch material in 10 right lobe graft recipients during 6 months. Because such experience is limited, further investigation should be conducted before endarterectomized artery graft is generally accepted in practice. In our institution, we have started to store all of the deceased donor iliac arteries at the tissue bank, regardless degree of the atherosclerosis, because the rapidly increasing demand of vessel allografts has resulted in an intractable vessel supply shortage. We think that it is unnecessary to match ABO blood groups for vessel graft selection, as in cardiac valve allograft replacement.9 Considering our data on tissue bank maintenance, endarterectomy may contribute to expand the effective vessel allograft pool by 10-20%. In conclusion, our preliminary experience suggests that an endarterectomized atherosclerotic artery allograft can be used as an alternative vessel material during right lobe graft implantation when other adequate vein graft material is unavailable.