Outcome Analysis in Adult-to-Adult Living Donor Liver Transplantation Using the Left Lobe Yuji Soejima, Mitsuo Shimada, Taketoshi Suehiro, Shoji Hiroshige, Mizuki Ninomiya, Satoko Shiotani, Noboru Harada, Ijichi Hideki, Yusuke Yonemura, and Yoshihiko Maehara Graft size problems remain the greatest limiting factor for expansion of living donor liver transplantation (LDLT) to the adult population. The result of adult-to-adult LDLT using the left lobe with special reference to graft size has not been fully evaluated to date. In this study, we evaluated the outcome of adult-to-adult LDLT using the left lobe and also analyze the impact of using small-for-size grafts on outcome. Thirty-six recipients who underwent adult-to-adult LDLT using the left lobe (n 14) or left lobe plus caudate lobe (n 22) were included in the study. Variables including preoperative and operative data, patient and graft survival, complications, and causes of graft loss were studied. Furthermore, the incidence of small-for-size syndrome and its impact on graft survival were studied. Mean graft volume (GV) was 420 85 g (range, 260 to 620 g), which resulted in 38.2% 8.1% (range, 22.8% to 53.8%) of the recipient standard liver volume (SLV). Overall 1-year patient and graft survival rates were 85.7% and 82.9%, respectively. Seven grafts were lost. Small-for-size syndrome occurred in 7 of 16 patients (43.8%) with cirrhosis and only 1 of 20 patients (5.0%) without cirrhosis (P.005). Recipients who developed small-for-size syndrome had inferior graft survival to those who did not (P.07). In conclusion, adult-to-adult LDLTs were found to be feasible without affecting patient or graft survival. Small-for-size syndrome developed more frequently in patients with cirrhosis. Minimum GV in adult-to-adult LDLT should be 30% less than the recipient s SLV in patients without cirrhosis, whereas 45% less was required in patients with cirrhosis. (Liver Transpl 2003;9:581-586.) Living donor liver transplantation (LDLT) was initiated first in children in 1989 in response to a severe organ shortage of pediatric donors. 1,2 LDLT now is an accepted and established alternative to cadaveric donor liver transplantation, especially for small children. This innovation has dramatically reduced pretransplantation mortality for children in Western countries and has become the only source of grafts in patients for whom a cadaveric liver transplantation from braindead donors is not indicated. Since 1996, indications for this modality have been extended to adult recipients, especially in countries where the availability of brain-dead donors is severely restricted. 3 At the initiation of adult-to-adult LDLT, a left lobe with or without the caudate lobe had been the only option available because of donor safety. However, a left-lobe graft could provide only 30% to 50% of the required liver volume for adult recipients and thus tended to be too small for adult recipients to sustain their metabolic demands. 4 Recently, right-lobe donation has emerged to overcome the graft size problem, with good initial results reported in both donors 5 and recipients. 6,7 In the context of these size problems, the concept of left-lobe donation for adult recipients has now been almost completely abandoned, especially in Western countries. Therefore, very scarce information on results of adultto-adult left-lobe liver transplantation has been available to date. Small-for-size grafts are characterized clinically by a combination of prolonged functional cholestasis, intractable ascites, and delayed recovery of both prothrombin time and encephalopathy. 4 However, characteristics of this syndrome have not been fully examined with a large series of patients. In this study, we analyzed the outcome in adult-toadult LDLT using the left lobe and tried to clarify the safety and risk of a left-lobe donation. Furthermore, we tried to clarify the incidence and results of small-for-size syndrome in this population, with special reference to the existence of underlining cirrhosis. Patients and Methods Patients Between October 1996 and April 2001, a total of 51 consecutive LDLTs were performed at Kyushu University, Fukuoka, Japan, after obtaining approval from the Ethics and Indication Committee of Kyushu University. Among them, 36 adult recipients (age 19 years) who underwent LDLTs From the Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan. Address reprint requests to Yuji Soejima, MD, Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka 812-8582, Japan. Telephone: 81-92-642-5466; FAX: 81-92-642-5482; E-mail: ysoejima@surg2.med.kyushu-u.ac.jp Copyright 2003 by the American Association for the Study of Liver Diseases 1527-6465/03/0906-0008$30.00/0 doi:10.1053/jlts.2003.50114 Liver Transplantation, Vol 9, No 6 (June), 2003: pp 581-586 581
582 Soejima et al Table 1. Patient and Graft Demographics Recipient Age (yrs) 45.3 1.0 (22-65) Gender (Women/Men) 14/22 Height (cm) 159.1 7.6 (147.7-174.5) Weight (kg) 54.7 9.2 (41.3-82.0) SLV (ml) 1,106 110 (913-1,386) Indications Fulminant hepatic failure 15 Primary biliary cirrhosis 9 Hepatitis C 5 Retransplantation 2 Others 5 Donor Age (yrs) 36.5 12.4 (19-60) Gender (women/men) 10/26 Graft Graft type Left lobe 14 Left lobe caudate lobe 22 GV (ml) 420 85 (260-620) GV-SLV ratio (%) 38.2 8.1 (22.8-53.8) Graft-recipient weight ratio 0.79 0.19 (0.41-1.11) Blood type compatibility (n) Identical 26 10 Compatible NOTE. Values expressed as number of patients or mean SD (range). using left-lobe grafts from adult donors were included in the study. Left-lobe grafts included a left lobe with (n 22) or without (n 14) a caudate lobe. Living donors consisted of a mother (n 1), fathers (n 2), sons (n 13), daughters (n 2), brothers (n 10), sisters (n 5), husbands (n 2), and a wife (n 1). Indications for LDLTs included fulminant hepatic failure (n 15; 41.7%), primary biliary cirrhosis (n 9; 25%), hepatitis C (n 5; 13.9%, including 2 patients with hepatocellular carcinoma), retransplantation (n 2), and others (n 5). Demographics of recipients and donors are listed in Table 1. Donor Selection and Informed Consent The primary selection criterion for a living donor was voluntary. Acceptance criteria for donors included age between 20 and 65 years, relationship within second degree of consanguinity with the recipient or spouses, blood group compatibility, and negative serological test results for hepatitis B, hepatitis C, human immunodeficiency viruses, and other transmissible viruses. All donors had normal liver function test results and no history of liver disease. Eligible donors proceeded to the imaging studies, including chest and abdominal X-ray and 3-mm slice computed tomographic (CT) scan for excluding any unrecognized intraabdominal pathological states. A CT scan also was used for volumetric analysis of graft size matching, delineation of the vascular anatomy, and evaluation of degree of fat content. A percutaneous liver biopsy was not routinely performed. After confirming adequacy for donors, informed consent was obtained from the donor twice independently by the surgical team and members of the ethical committee. The donor was informed that he or she might withdraw at any time. For surgical safety, 400 to 1,200 ml of autologous blood was stored before surgery. To identify any vascular abnormality and facilitate the surgical procedure, donor hepatic angiography was performed routinely. Evaluation of Graft Size Standard liver volume (SLV) of recipients was calculated according to the formula by Urata et al. 8 Graft volume (GV) was predicted from the CT volumetric analysis. Actual GV was measured at the back table. GV-SLV ratio was used for graft size matching. Basic criteria to decide the graft type for adult recipients were based on the expected GV-SLV ratio. Until November 2000, left-lobe grafts were used exclusively despite the expected GV-SLV ratio. From December 2000, we have liberally used right-lobe grafts if the expected GV- SLV ratio was less than 30%. Donor Surgical Procedure The surgical procedure has been described in detail elsewhere. 9 Briefly, the abdomen was entered through a bilateral subcostal incision with a midline extension. A cholecystectomy was performed, and intraoperative cholangiography was performed through the cystic duct to delineate the bile duct anatomy. The left hepatic duct then was transected sharply. Particular care was taken not to injure the right posterior hepatic duct, which sometimes drains into the left hepatic duct. Intraoperative ultrasonography was performed to delineate hepatic vein anatomy and verify the transection plane. Parechymal transection was performed using the Cavitron Ultrasonic Surgical Aspirator (CUSA System 200; Valleylab Inc, Boulder, CO) and electrocautery without any vascular inflow interruption on either side of the liver. The transection plane was always on the right side of the middle hepatic vein. The graft was perfused in situ with cold University of Wisconsin solution (Viaspan; Dupont, Wilmington, DE). At the back table, an additional 1,000 ml of University of Wisconsin solution was used for portal flushing. Venoplasty was performed at the back table, if necessary. The Cellsaver (BLAT2; Cobe Cardiovascular, Inc, Arvada, CO) was routinely used to reduce blood loss during surgery. The biliary leakage test was performed through a cannula inserted in the cystic duct using indocyanine green solution to detect leakage from the cut surface. After experiencing bile duct stricture in a donor, cholangiography was routinely performed to check the consistency of the bile duct before closing the abdomen. Recipient Surgical Procedure The recipient total hepatectomy was performed basically without venovenous bypass with preservation of the inferior
Adult-to-Adult Left-Lobe Living Donor Liver Transplantation 583 Table 2. Surgical Data Donor Surgical time (hr) 7.75 (5.3-11.7) Blood loss (ml) 1,100 (350-3,500) Recipient Surgical time (hr) 11.5 (8.8-19.9) Blood loss (ml) 4,600 (1,105-63,110) Transfusion Packed red blood cells (units) 14 (4-180) Fresh frozen plasma (units) 5 (0-80) NOTE. Values expressed as mean (range). vena cava. Graft implantation started with the end-to-end anastomosis of the recipient and graft conduit of the left and middle hepatic vein using 5-0 PDS-II running sutures (Ethicon, Inc, Somerville, NJ). During hepatic vein anastomosis, the graft was flushed through a cannula in the portal vein with 250 ml of plasma protein fraction. The donor left portal vein was anastomosed end to end to the recipient s left portal vein or main portal vein using 6-0 PDS-II running sutures with a growth factor. Hepatic artery anastomosis was performed using the microvascular technique under magnification with the operative microscope. 10 Usually, the donor left hepatic artery was anastomosed to the recipient left or right hepatic artery in an end-to-end fashion using interrupted 8-0 Prolene sutures (Ethicon, Inc). The bile duct was reconstructed by an end-to-side hepaticojejunostomy over an internal stent for the first 10 cases, then without a stent for the remaining cases. Color Doppler ultrasound examination was routinely performed before closing the abdomen. Postoperative Management Color Doppler ultrasound examination was routinely performed twice a day for 1 week postoperatively and then once a day for the next week. Induction and maintenance immunosuppression basically consisted of a double regimen of oral tacrolimus and steroid. OKT3 was administered for steroidresistant rejection. Antibiotics were administered for 5 days postoperatively, including the day of surgery. Serial CT scans were obtained postoperative days (PODs) 7, 14, and 28 to assess regeneration of the graft. Ascites production is defined as the amount of ascites through indwelling drains (with or without leakage through the drain orifice) per day. Definition of Small-for-Size Syndrome To determine the impact of a small-for-size graft on outcome, we defined small-for-size syndrome as the presence of both prolonged functional cholestasis and intractable ascites. Prolonged functional cholestasis is defined as total bilirubin level greater than 5 mg/dl POD 14, without other definitive causes for cholestasis. Intractable ascites is defined as a daily production of ascites more than 1 L POD 14 or more than 500 ml POD 28. Statistical Analysis A survival analysis was performed using the Kaplan-Meier method, and groups were compared by means of the log-rank test. Continuous variables were compared using a two-tailed unpaired t-test for independent samples. Categorical data were compared using Chi-squared test. P less than 0.05 is considered significant. All statistical analyses were performed using StatView 4.5 software for Macintosh (Abacus Concepts Inc, Berkeley, CA). Results Patient Demographics Table 1 lists demographics of recipients, donors, and grafts. Graft types consisted of 14 left-lobe grafts without the caudate lobe and 22 left-lobe grafts with the caudate lobe. Average GV was 420 g, which resulted in an average GV-SLV ratio of 38.2%. Surgical Data Table 2 lists surgical data. Median donor operative time was 7.75 hours (range, 5.3 to 11.7 hours), with a median blood loss of 1,100 ml (range, 350 to 3,500 ml). No banked blood cells were transfused. In recipients, median operative time was 11.5 hours (range, 8.8 to 19.9 hours), with a median blood loss of 4,600 ml (range, 1,105 to 63,110 ml). Fourteen units of packed red blood cells (range, 4 to 180 units) and 5 units of fresh frozen plasma were transfused. Patient and Graft Survival Figure 1 shows overall patient and graft survival. Overall patient and graft survival rates at 1 year were 85.7% and 82.9%, respectively. Three-year patient and graft survival rates were 85.9% and 79.3%, respectively. If Figure 1. Overall 1- and 3-year patient (85.7% and 85.7%, respectively) and graft survival rates (82.9% and 79.3%, respectively).
584 Soejima et al Figure 2. Graft survival according to GV-SLV ratio. graft survival was analyzed based on graft size, then graft size did not influence graft survival (Fig. 2). There were seven extra-small grafts, which accounted for a GV- SLV ratio of less than 30%. Six of seven grafts in this group have survived. Complications and Causes of Graft Loss Table 3 lists the incidence of posttransplantation complications. Despite routine use of the microvascular technique for hepatic artery anastomosis, there were two cases of hepatic artery thrombosis PODs 8 and 39, which resulted in a patient death (POD 10) and retransplantation (POD 44), respectively. The latter had biliary leakage that was being managed conservatively when he developed hepatic artery thrombosis. Biliary complications, including seven biliary leaks (19.4%; five leaks from the bile duct anastomosis, two leaks from the cut surface) and eight anastomotic strictures (22.2%), were the most frequent complications in Table 3. Complications Bleeding 5 (13.9) Rejection Acute 9 (25.0) Chronic 1 (2.7) Vascular Hepatic artery thrombosis 2 (5.6) Biliary Leak 7 (19.4) Stenosis 8 (22.2) Cholangitis 10 (27.8) Others Intractable ascites 11 (30.1) NOTE. Values expressed as number (percent). Figure 3. Association between the incidence of small-forsize syndrome and existence of cirrhosis. Horizontal bars indicate the safe limit of GV-SLV ratio, free from smallfor-size syndrome in each group. this cohort. Cholangitis also was common (27.8%). Most cholangitis probably resulted from anastomotic stricture, which often required radiological intervention (e.g., percutaneous balloon dilation) or surgical revision. Five patients with biliary leakage were treated with conservative drainage, whereas 2 patients required relaparotomy. Of note, 11 patients developed intractable ascites, which required a large amount of daily fluid and albumin replacement. Causes of graft loss included sudden-onset intra-abdominal bleeding leading to cardiac arrest in 1 patient. Other causes included hepatic artery thrombosis (n 2), chronic rejection (n 1), possible humoral rejection (n 1), sepsis (n 1), and adult T-cell leukemia (n 1). There was no graft lost to primary graft nonfunction directly attributed to the small-for-size graft. Small-for-Size Syndrome Eight recipients developed small-for-size syndrome (22.2%). Average GV of this group was 374.4 g (range, 260 to 500 g), which resulted in an average GV-SLV ratio of 34.4% (range, 22.8% to 42.4%). Figure 3 shows the association between incidence of small-forsize syndrome and existence of cirrhosis. As shown, there was no significant difference in GV-SLV ratios between recipients with and without cirrhosis. However, 43.8% of recipients with cirrhosis developed small-for-size syndrome compared with only 5% in recipients without cirrhosis (P.005). These results indicate that the safe limit of GV-SLV ratio for recipients could differ according to indication for LDLT. The safe limit for patients with cirrhosis may be approx-
Adult-to-Adult Left-Lobe Living Donor Liver Transplantation 585 Figure 4. Graft survival according to the development of small-for-size syndrome. imately 45%, whereas 30% could be adequate for recipients without cirrhosis. As shown in Figure 4, development of small-for-size syndrome adversely affected graft survival, although the difference was not statistically significant (P.07). Discussion With the increasing success of liver transplantation as a therapy for patients with end-stage liver disease, the demand for this procedure continues to grow. Currently, there are almost 17,000 patients on the waiting list in the United States for liver transplantation, yet there are only 5,000 livers available annually. 11 Although there are ongoing attempts to increase donation rates and use novel techniques, including split-liver transplantation, 12,13 this will still leave a shortage of livers for transplantation. The concept of living donation was first reported in Brazil, 1 and the first successful case was reported in Australia. 2 This modality has helped diminish waiting list mortality for children, and the technique has been established. Inevitably, the critical shortage in livers for the adult population has drawn great interest in using a portion of adult livers for adult recipients. The first adult-toadult LDLT series using a left lobe was reported from Japan, 14 with excellent short-term results. However, its application to adult patients was severely limited to selected recipients because of the size of grafts available from left-lobe donation. Therefore, especially in Western countries, left-lobe donation was soon abandoned, and its application rapidly evolved to right-lobe donation, which requires a hepatectomy of almost two thirds of the donor liver, which would provide the recipient with sufficient GV. There have been increasing reports 6,7,15 published on results of adult-to-adult LDLT using the right lobe; however, scarce data remain available regarding LDLT using the left lobe. We herein fully evaluated for the first time results of our largest series of adult-to-adult LDLTs using the left lobe. In these cases, we frequently encountered recipients who developed small-for-size symptom, which sometimes resolved spontaneously over time. We define small-for-size syndrome as the presence of both functional cholestasis and intractable ascites. Before conducting the current study, it was difficult to anticipate the development of small-for-size syndrome before transplantation. Some recipients of grafts of less than 30% of SLV did not develop small-for-size syndrome, whereas some developed the syndrome even with grafts larger than 40%. In the current study, we first found and clearly showed that the presence of cirrhosis or pretransplantation portal hypertension was an important predictor of developing small-for-size syndrome with a small-for-size graft. Results indicated that patients with cirrhosis might require a larger graft than those without cirrhosis. The mechanism of small-for-size syndrome remains unknown. When transplanted under conditions of portal hypertension, small grafts are supposed to be exposed to relatively excessive portal perfusion and portal pressure compared with grafts under normal portal pressure. Experimental data suggest that hyperperfusion of the liver was detrimental, and improved results have been observed with portal decompression of small grafts. 16 In addition, gut-derived endotoxin and substrates, including fatty acids, may further deteriorate small grafts after reperfusion. 4 Clinically, some groups have advocated the use of a temporal portacaval shunt to reduce or minimize the influence of such substances accumulating during portal clamping. 17 Our unpublished data suggested an association between smallfor-size syndrome and amount of portal flow after transplantation. Therefore, our current approach in managing the problem with small-for-size grafts is to reduce the relatively excessive portal flow by ligating the proximal splenic artery or performing splenectomy, if necessary. The Mount Sinai group reported that pretransplantation disease severity of recipients was proposed to be one of the important factors in developing small-forsize syndrome. 18 They analyzed 22 adult recipients of left (n 10) or right lobes (n 12) from living donors. Child s class B or C recipients of small grafts (graftrecipient weight ratio 0.8%) showed an inferior graft survival rate of 25% (1 of 4 grafts), whereas graft survival of Child s class A recipients was 100% (8 of 8
586 Soejima et al grafts) without showing symptoms related to small-forsize graft. They concluded that pretransplantation disease severity significantly affected graft survival of recipients of small-for-size grafts. Which graft should be used for adult-to-adult LDLT, the left lobe or right lobe? Clearly, all forms of living donor transplantation are subject to varying degrees of complications and death, depending on the complexity of the procedure, and despite good intentions and experienced hands, there can be no argument that a finite risk for death to the donor exists. Almost all reports from the United States, 6,19 Hong Kong, 20 and Kyoto, 7 Japan, advocate the superiority of right-lobe donation in view of recipient safety. However, in view of donor safety, it is obvious that left-lobe donation is safer than right-lobe donation because the remnant volume of the liver might be larger, although no data have been published to date. Right-lobe donation tends to be associated with significant hyperbilirubinemia after surgery, which usually resolved spontaneously without adverse sequelae. 21 However, even in left-lobe LDLT, the possibility of donor mortality is real, and at least one such death has been reported. 22 Ethical issues of LDLT have always centered on the balance of risks and benefits for both the recipient and donor. We thus currently think that with appropriate size matching and careful recipient selection (cirrhosis versus noncirrhosis), adult-to-adult LDLT can be successful with either a left or a right lobe. In conclusion, adult-to-adult left-lobe LDLTs were found to be feasible without affecting patient and graft survival. Small-for-size syndrome occurred in 22% of patients who underwent left-lobe LDLT. Recipients with cirrhosis more frequently developed this syndrome. Proposed minimum GV in adult-to-adult LDLT should be a GV-SLV ratio of more than 30% for recipients without cirrhosis and more than 45% for patients with cirrhosis. References 1. Raia S, Nery JR, Mies S. Liver transplantation from live donors. Lancet 1989;21:497. 2. Strong RW, Lynch SV, Ong TN. Successful liver transplantation from a living donor to her son. N Engl J Med 1990;322:1505-1507. 3. Kawasaki S, Makuuchi M, Matsunami H, Hashikura Y, Ikegami T, Nakazawa Y, et al. Living related liver transplantation in adults. Ann Surg 1998;227:269-274. 4. Emond JC, Renz JF, Ferrel LD, Rosenthal P, Lim RC, Roberts JP, et al. Functional analysis of grafts from living donors. Ann Surg 1996;224:544-554. 5. Sakamoto S, Uemoto S, Uryuhara K, Kim I, Kiuchi T, Egawa H, et al. Graft size assessment and analysis of donors for living liver transplantation using right lobe. Transplantation 2001;71: 1407-1413. 6. Marcos A, Fisher RA, Ham JM, Shiffman ML, Sanyal AJ, Luketic VAC, et al. Right lobe donor liver transplantation. Transplantation 1999;68:798-803. 7. Inomata Y, Uemoto S, Asonuma K, Egawa H, Kikuchi T, Fujita S, et al. Right lobe graft in living donor liver transplantation. Transplantation 2000;69:258-264. 8. Urata K, Kawasaki S, Matsunami H, Hashikura Y, Ikegami T, Ishizone S, et al. Calculation of child and adult standard liver volume for liver transplantation. Hepatology 1995;21:1317-1321. 9. Nishizaki T, Ikegami T, Hiroshige S, Hashimoto K, Uchiyama H, Yoshizumi T, et al. Small graft for living donor liver transplantation. Ann Surg 2001;233:577-580. 10. Mori K, Nagata I, Yamagata S, et al. The introduction of microvascular surgery to hepatic artery reconstruction in living donor liver transplantation Its surgical advantages compared with conventional procedures. J Am Coll Surg 1996;183:74-76. 11. Annual Report of the US Scientific Registry of Organ Transplantation and the Organ Procurement and Transplantation Network Transplant Data: 1998-2001. UNOS, Richmond, VA, and the Division of Transplantation, Bureau of Health Resources and Services Administration, US Department of Health and Human Services, Rockville, MD, 2001. 12. Bismuth H, Houssin D. Reduced-sized orthotopic liver graft in hepatic transplantation in children. Surgery 1984;95:367-370. 13. Otte JB, de Ville de Goyet J, Alberti D, Balladur P, de Hemptinne B. The concept and technique of the split liver in clinical transplantation. Surgery 1990;107:605-612. 14. Hashikura Y, Makuuchi M, Kawasaki S, et al. Successful living related partial liver transplantation to an adult patient. Lancet 1994:343;1233-1234. 15. Wachs EW, Bak TE, Karrer FM, Everson GT, Shrestha R, Trouillot TE, et al. Adult living donor liver transplantation using a right hepatic lobe. Transplantation 1998;66:1313-1316. 16. Ku Y, Fukumoto T, Nishida T, et al. Evidence that portal vein decompression improves survival of canine quarter orthotopic liver transplantation. Transplantation 1995;59:1388-1392. 17. Kawasaki S, Hashikura Y, Matsunami H, et al. Temporal shunt between right portal vein and vena cava in living related liver transplantation. J Am Coll Surg 1996;183:74-76. 18. Ben-haim M, Emre S, Fishbein T, Sheiner PA, Bodian CA, Schwartz ME, et al. Critical graft size in adult-to-adult living donor liver transplantation: Impact of the recipient s disease. Liver Transpl 2001;11:948-953. 19. Wachs ME, Bak TE, Karrer FM, Everson GT, Shrestha R, Trouillot TE, et al. Adult living donor liver transplantation using a right hepatic lobe. Transplantation 1998;66:1313-1316. 20. Lo CM, Fan ST, Liu CL, Wei WI, Lo RJ, Lai CL, et al. Adultto-adult living donor liver transplantation using extended right lobe graft. Ann Surg 1997;226:261-270. 21. Fujita S, Kim ID, Uryuhara K, Asonuma K, Egawa H, Kiuchi T, et al. Hepatic grafts from live donors: Donor morbidity for 470 cases of live donation. Transplant Int 2000;13:333-339. 22. Singer PA, Siegler M, Whitington PF, et al. Ethics of liver transplantation with living donors. N Engl J Med 1989;321:620-622.