Domino-liver transplantation: toward a safer and simpler technique in both donor and recipient

Domino (or sequential) liver transplantation (DLT) represents an opportunity to expand the liver allograft pool. The first Swedish experiences with liver transplantation (LT) for hereditary transthyretin amyloidosis, also known as familial amyloidotic polyneuropathy (FAP), the high incidence of this disease in Portugal and the structural normality (except for the production of the mutant TTR) of such livers, led to the development of DLT [1‐3]. The concept is based on the knowledge that non-cirrhogenic, liver-based, metabolic diseases such as FAP, maple syrup disease, hyper-homocysteinemia, methylmalonic acidemia, and hypercholesterolemia are not or slowly transmitted to recipients which do not have these inherited traits [4, 5]. Up to December 2019, 2217 LTs were reported to the FAP World Transplant Registry, the Domino-liver Transplant Registry included 1210 (54%) DLT. DLT is normally only used in elderly and/or cancer patients [6, 7]. The non-use of these potentially excellent liver allografts has four reasons: (a) ethical concerns about putting the domino-donor at risk of a more complex surgical procedure and the domino-recipient for a possible disease transmission (reported in 3.3–21% of domino-recipients); (b) fear of technical complications related to a more difficult arterial inflow and venous outflow reconstruction of the domino-allograft; (c) increased logistics linked to the organization of two simultaneous, LT procedures, and finally (d) the (intra-operative) discovery of advanced liver fibrosis (‘cardiac liver’) caused by the underlying amyloidotic cardiopathy [8‐14].

Over the years, technical developments, routine introduction of inferior vena cava (IVC) sparing hepatectomy techniques, and accumulated experiences in centers specializing in liver-based metabolic diseases overruled these concerns [15‐20]. The original ‘Coimbra-procedure’, described by A. Furtado in 1992, including total hepatectomy with the removal of the IVC, extensive dissection of the supra-hepatic IVC at or above the diaphragmatic ring to assure a sufficiently long supra-hepatic IVC cuff, and the systematic use of the veno-venous bypass (VVB) to overcome hemodynamic instability caused by IVC clamping in neurologically dysregulated FAP patients, clearly disadvantaged the domino-donor compared to the ‘classical’ liver transplant patient [2, 3]. Ten years later, the Lisbon Rui Cabral team, led by JR. Pena and E. Barroso, introduced the ‘double piggy-back’ implantation technique in which the IVC was preserved and the use of VVB avoided in both domino-donor and domino-recipient [16, 21‐23]. The growing confidence in DLT has been exemplified by its extension to split- and living donor DLT [24‐27]. A further modification of the technique, aiming at making DLT safer and simpler, is presented here in detail.

DLT is performed only following informed consent obtained from both domino-donor and domino-recipient patients after a detailed explanation of both donor and recipient transplantation procedures. DLT was approved by the ethical review board of the Université Catholique de Louvain.

During the period February 2000–January 2017, 16 LT were performed for FAP. On three (18.7%) occasions, DLT was not possible due to extended intrahepatic and extrahepatic portal vein thrombosis (1 pat) and advanced fibrosis (2 pats) due to autoimmune cirrhosis and restrictive amyloidotic cardiomyopathy. One Fap liver was split for two small adults. Indications for DLT were hepatocellular cancer (7 pats), primary biliary cholangitis (3 pats), alcoholic cirrhosis with severe encephalopathy (3 pats), and neuro-endocrine metastases (1 pat). IVC-sparing hepatectomy without VVB use was performed in all LT procedures. Median warm ischemia time of the domino-graft was 43 min (range from 30 to 73). In one case, the hepatic artery had to be anastomosed to the splenic artery because of fibrotic changes of the hepatic artery caused by trans-arterial chemo-embolisations; in one right and left, hepatic arteries were anastomosed to similar, corresponding recipient arteries, and in another, the portal vein needed to be prolonged using a free iliac vein to overcome an extended portal vein thrombosis. The day-7 angio-CT scan of the patient in which a dominant inferior RHV was not implanted revealed a normal venous outflow of the right liver. In all domino-grafts, Doppler ultrasound revealed normal, triphasic, outflow patterns, and angio-CT showed a well-vascularized segment I. Two domino-donors presented a right pleural effusion. None of the domino-graft recipients presented after a median follow-up of 40 months (range 1–206) signs of disease transmission. One domino-liver recipient developed an ascitic decompensation 10 years post-DLT due to fibrotic changes of the neo-supra-hepatic cava cuff; a single balloon dilatation resolved this problem. In this case, a stored iliaco-caval graft from another donor had been used to reconstruct the venous outflow. Four domino-recipients died of recurrent cancer (at 7, 23, 41,159 mo post-DLT), two of cardiac complications (at one and six mo), one each of pulmonary fibrosis (at 152 mo), suicide (39 mo), and massive variceal bleeding (1 mo). There was no documented disease transmission in any of the domino-graft recipients.

Domino LT represents a rare opportunity to enlarge the liver donor pool [2]. The extensive supra-hepatic IVC dissection aimed at obtaining a long supra-hepatic venous cuff (but possibly causing pleural and pericardial effusions) and the systematic use of VVB aiming at overcoming hemodynamic instability when clamping the IVC (but possibly causing cutaneous (wound infection and lymphocele), venous (pulmonary and peripheral embolism), and neurological (neuropraxia, paresthesia) complications disadvantaged the domino-donor compared to the ‘classical’ liver recipient. The domino-recipient was also disadvantaged by a more challenging graft implantation due to insufficient vessel lengths. Technical improvements in both domino-donor and domino-recipient procedures allowed the ethical concerns concerning the applicability of DLT to be overcome [6]. Routine use of IVC-sparing hepatectomy techniques and attention to the domino-graft outflow reconstruction played a major role in this [15, 18]. Many variants have been reported to optimize the domino-graft venous outflow (Table 1; Fig. 4). The extension of the supra-hepatic vena cava cuff by a free venous graft, reported by Azoulay et al. in 1999, was a first step to make the DLT easier [29]. In 2001, Nishida et al. proposed the end-to-side infra-hepatic cavo-cavostomy as a means to overcome the difficulty of the supra-hepatic anastomosis; the supra-hepatic IVC was closed with a vein patch [30]. At the 2001 ESOT congress, the Lisbon team reported for the first time the use of FAP-livers with IVC preservation in the FAP-donor and extension of the domino-graft with a free iliaco-caval graft [21, 22]. Pinto-Marques et al. confirmed, in a series of 260 DLT, the feasibility without VVB use and with outflow tract reconstruction (“neo-supra-hepatic cuff”) of the domino-allograft using IVC, iliaco-caval, reno-caval, or pulmonary veins. It is of note that eight patients presented a venous outflow obstruction and that six patients needed to be re-transplanted [16]. The fact that 16 more technical modifications have been reported during the period 2001–2019 indicates the need for further standardization and simplification of the technique [8, 28, 31‐45]. The venous outflow reconstruction of the domino-graft, such as described here, is in line with this. The use of a 180° rotated and anatomically-based modified free iliaco-caval vein graft allows the procedure to be simplified. When adequately tailored and incised, this venous graft does not need any supplementary gesture. In contrast to the literature, none of our domino-recipients developed a post-transplant venous outflow obstruction related to the design of the neo-supra-hepatic venous cuff. Interference with renal and thoracic organ procurement teams is eliminated as there is no need to take the pulmonary vein nor reno-caval venous confluence and the piggy-back implantation of the domino-graft is easy. The implantation of a post-mortem whole or right split liver graft in the domino-donor is also easy when performing a side-to-side cavo-caval implantation under partial clamping of the recipient IVC. Our extensive experience with this technique also learned that implantation of right accessory, even dominant, HVs looks not to be necessary for full-size LT [17].

The technical simplification of DLT described here addresses the challenge of the outflow reconstruction, the avoidance of VVB use in both domino-donor and domino-recipient, and possible interference with thoracic and renal procurement procedures related to venous graft harvesting. The modified outflow reconstruction of the domino-graft should represent a standardized approach to use in domino-liver transplantation, aiming thereby at optimal use of these precious grafts.