Rescue policy for discarded liver grafts: a single-centre experience of transplanting livers ‘that nobody wants’

Abstract

Background

There is a worldwide need to expand the donor liver pool. We report a consecutive series of elective candidates for liver transplantation (LT) who received ‘livers that nobody wants’ (LNWs) in Argentina.

Methods

Between 2006 and 2009, outcomes for patients who received LNWs were analysed and compared with outcomes for a control group. To be defined as an LNW, an organ is required to fulfil two criteria. Firstly, each liver must be officially offered and refused more than 30 times; secondly, the liver must be refused by at least 50% of the LT programmes in our country before our programme can accept it. Principal endpoints were primary graft non-function (PNF), mortality, and graft and patient survival.

Results

We transplanted 26 LNWs that had been discarded by a median of 12 centres. A total of 2666 reasons for refusal had been registered. These included poor donor status (n = 1980), followed by LT centre (n = 398) or recipient (n = 288) conditions. Incidences of PNF (3.8% vs. 4.0%), in-hospital mortality (3.8% vs. 8.0%), 1-year patient (84% vs. 84%) and graft (84% vs. 80%) survival were equal in the LNW and control groups.

Conclusions

Transplantable livers are unnecessarily discarded by the transplant community. External and internal supervision of the activity of each LT programme is urgently needed to guarantee high standards of excellence.

Introduction

Shortage of liver grafts accounts for the principal limitation of liver transplantation (LT). Many surgical techniques have been developed to minimize the effects of this deficit, including reduced-size LT, single or dual graft live-donor LT, splitting a cadaveric organ for two recipients, and domino transplantation. In consequence to the dramatic reduction in the number of live-donor grafts available in developed countries, many transplant programmes have been driven to venture further into innovative strategies, such as those involving the use of donors after cardiac death, severe steatotic liver grafts or organs from so-called ‘expanded criteria donors’ (ECDs), as well as extracorporeal liver perfusion systems for rescuing marginal liver grafts.^1–5

The use of marginal liver grafts has been extensively discussed in the literature.^3,6–8 However, a major concern in the transplant community refers to how we can safely relax deceased donor selection criteria without compromising patient safety.^9–11 In Europe, the Essen group successfully transplanted 10 ‘livers that nobody wants’ (LNWs) outwith the Model for End-stage Liver Disease (MELD)-based allocation system, which defines LNWs as livers discarded by at least three transplant centres before their final acceptance.¹² In Argentina, LT activity was initiated over two decades ago and a MELD-based allocation system was adopted in 2005 for patients listed for LT. In 2006 we initiated our new LT programme on which, forced by low organ donation rates and high mortality on the national waiting list, we decided to adopt an aggressive policy in accepting livers from ECDs for patients listed in our programme. Using a stricter and stronger definition of LNW, we herein report our initial experience of transplanting these otherwise discarded liver grafts allocated using the MELD system.

Materials and methods

Between August 2006 and April 2009, all patients who received LNW transplants were included in the rescued livers group and compared with a control group using a prospectively collected database. To be defined as a ‘liver that nobody wants’, the organ was required to fulfil two major criteria. Firstly, each liver must have been officially offered and refused at least 30 times prior to our final acceptance. In other words, a liver that had been discarded by others was finally accepted by us for an elective candidate whose position on the national waiting list was later than 30th. The second criterion required the liver to have been refused by more than half the active LT programmes in Argentina (i.e. more than eight LT programmes). This study includes only patients with a minimum follow-up time of 6 months after LT.

During the 32 months of the study period, 67 consecutive LTs were performed in our programme. A consecutive series of patients who were electively transplanted with a primary full-size liver were included in this study (n = 51). In order to analyse a homogeneous cohort of patients, the following conditions were excluded: retransplantation (n = 3); emergency LT (n = 7); split-liver transplantation (n = 2), and simultaneous kidney and LT (n = 4). As a control group we used a consecutive series of patients who underwent elective transplantation during the same period, but whose initial positions on the waiting list at the time of liver allocation were prior to 30th. Thus, this group included patients who were more ill and who were better ranked on the national waiting list as a result of their higher MELD scores.

Primary endpoints were primary non-function of the graft (PNF), mortality, and graft and patient survival. Primary non-function of the graft was defined as death or retransplantation within 7 days following LT in the absence of any vascular problems. Secondary endpoints were secondary graft non-function (i.e. vascular complications leading to graft loss within 7 days of LT), primary dysfunction of the graft (PDF), hospital stay and major postoperative complication rate. Primary dysfunction of the graft was assumed when a peak AST (aspartate aminotransferase) level >1.500 IU/l and a prothrombin time ≤50% co-occurred within the first week.^13,14

We analysed national official data from the Instituto Nacional Central Único Coordinador de Ablación e Implante (Unique National Institute for Coordination of Ablation and Transplantation, INCUCAI) to assess procurement and transplantation activity during the study period. We also collected from the official national registry data on the reasons given by transplant centres for refusing liver grafts.

Donor data

Organ procurement was performed as described elsewhere with aortic and portal perfusion using University of Wisconsin preservation solution (Viaspan™; DuPont Pharma, Wilmington, DE, USA).¹⁵ In agreement with others,^13,16–18 the following expanded criteria for donors were identified: cardiac arrest for >15 min or prolonged hypotensive episodes of <60 mmHg for >1 h; donor age >55 years, positive hepatitis B virus (HBV) infection; high vasopressor drug requirement (dopamine dose >10 µg/kg/min or any doses of other amines); hypernatraemia >155 mEq/l; prolonged intensive care unit (ICU) stay (≥5 days with mechanic ventilation); nosocomial infection (identified by positive blood cultures or pneumonia); elevated liver transaminases (AST >170 IU/l or ALT [alanine aminotransferase] >140 IU/l), and prolonged cold ischaemia time (>12 h). Additionally, the body mass index (BMI) of each donor was assessed. Donor obesity was defined by a BMI ≥30 kg/m².

Liver graft data

Pre-transplant donor graft biopsies were selectively performed in the presence of any macroscopic or sonographic sign of steatosis or fibrosis. Such livers were accepted only after histological evaluation of the donor biopsy. After liver graft reperfusion, a core needle biopsy was systematically performed in each recipient. Histological evaluation was performed to assess steatosis and fibrosis using haematoxylin and eosin-stained sections of formalin-fixed, paraffin-embedded liver tissue. Mild (<30%), moderate (30–60%) and severe (>60%) steatosis were defined according to the percentage of the total amount of mixed fat droplets in the biopsy. Liver fibrosis was quantified according to METAVIR score.¹⁹

Recipient data

All recipients were informed in detail about our policy of accepting ECD grafts. Each LT recipient signed written informed consent before surgery. All transplants were performed without veno–venous bypass, as previously described.¹⁵ The following recipient data were collected: age; underlying liver disease; MELD score; BMI; recipient status on the waiting list (elective, emergency); time on the waiting list; transfusion requirements, and operative time.

Postoperative outcome

Liver function was evaluated clinically and through biochemical parameters such as AST, ALT, bilirubin and prothrombin time measured daily during the first week. Liver graft vascular patency was evaluated by echo-Doppler ultrasound during the first day and when clinically indicated. Major complications included complications of grades 3–5 (i.e. requiring surgical intervention or ICU admission or causing death, respectively) according to a validated classification system for postoperative complications.²⁰ After discharge, each patient was followed in the multidisciplinary outpatient clinic.

Statistical analysis

Summary data are presented as median (range). Differences between groups were tested by chi-squared test for categorical and Mann–Whitney U-test for continuous variables. Graft failure was defined by death, retransplantation, documented cirrhosis or advanced fibrosis. All tests were two-tailed. Statistical significance was indicated by P < 0.05. Calculations were performed using spss Version 13.0 (SPSS, Inc., Chicago, IL, USA).

Results

Transplant activity and mortality on the waiting list in Argentina

The national organ donation rate in 2008 was 13.6 per million population (PMP). There was great variation among the country's 24 regions or provinces, with rates ranging from 34.5 PMP in Corrientes to as low as zero in Santiago del Estero and San Luis. Sixteen centres performed LT during the study period, of which 13 were private and three were public hospitals. However, when transplant activity in each centre was evaluated, we found that only seven centres performed more than 20 LTs per year, one centre performed 10–20 LTs and most centres performed under 10 LTs per year (n = 8).

During the study period, 1772 patients were registered on the national waiting list, of whom 161 were listed as emergency patients and 1611 were listed as being in elective conditions. The overall mortality rate on the waiting list was 14.6% (n = 258); mortality among emergency patients was 25.5% (n = 41) and that among elective patients was 13.5% (n = 217). Transplantation activity included 684 deceased-donor and 63 live-donor LTs (54 in paediatric and nine in adult recipients). The overall median waiting list time was 4 days (range 1–19 days) for emergency and 74 days (range 1–1224 days) for elective candidates.

Waiting list characteristics and transplant activity on our LT programme

During the 32-month period, 108 patients were included on our waiting list: of these, 67 patients were successfully transplanted, one died during the operation, one with fulminant liver failure recovered spontaneously, six (5.6%) died while on the waiting list and 33 were still waiting for a deceased liver donor at the time of writing.

A total of 75 liver offers were accepted by our LT team. Of these, three livers were not harvested (two donors suffered cardiac arrest before our procurement team arrived and one harvesting was cancelled because the necessary flight was delayed by bad weather), four donor livers were discarded after procurement (one for incidental ovarian cancer, one for incidental kidney cancer and two for more than three ECD factors in addition to severe steatosis in the frozen liver biopsy) and one patient died on the operation table during transplantation immediately before biliary reconstruction as a result of massive pulmonary thromboembolism (position on the waiting list: 17th; MELD score: 18). Thus, 67 of 75 (89.3%) deceased-donor livers were transplanted to patients on our waiting list.

Characteristics of the LNW group: donor- and graft-related risk factors

We accepted and subsequently transplanted 26 organs that qualified as LNWs. The median recipient position on the national waiting list of this group was 71 (range 31–432). All but one graft had one or more clinical and/or histological risk factors for LT. The only graft discarded by all without any ECD risk factors came from a 14-year-old donor and was discarded for reasons of size mismatch and/or logistical reasons. The other discarded grafts had one (n = 3), two (n = 7), three (n = 13) and four (n = 2) risk factors for LT (Table 1). Our database registered that the most frequent relevant risk factors in LNWs were inherent to the clinical characteristics of the donor and included advanced donor age, hypernatraemia, prolonged ICU stay and cerebrovascular accident as cause of donor death. Donor obesity and alcohol abuse were present in 19% and 11%, respectively. Almost 20% of LNWs had moderate to severe steatosis on liver biopsy.

Table 1.

Donor history and relevant clinical and graft-related risk factors in the ‘livers that nobody wants’ (LNW) group (n = 26)

	LNW group n (%)
Donor history and characteristics
Peripheral vascular diseasea	8 (30.7)
Donor obesity (≥30 kg/m2)	5 (19.2)
Alcohol abuse	3 (11.5)
Diabetes mellitus	2 (7.6)
Clinical donor risk factors
Cerebrovascular accident (cause of donor death)	18 (69.2)
Donor age >55 years	12 (46.1)
Hypernatraemia >155 mEq/l	11 (42.3)
Prolonged intensive care unit stay	7 (26.9)
High vasopressor drug requirement	6 (23.0)
Nosocomial infectionb	4 (15.3)
Cardiac arrest for >15 min or prolonged hypotension	3 (11.5)
Elevated liver transaminases	2 (7.6)
Positive viral infection (HBV)	1 (3.8)
Graft-related risk factors
Steatosis ≥30%	5 (19.2)
Prolonged warm ischaemia time	3 (11.5)
Prolonged cold ischaemia time	1 (3.8)

^aHistory of vascular disease included arterial hypertension, myocardial infarction or previous cerebrovascular accident

^bReferred to positive blood cultures or pneumonia

Official reasons for refusing livers

Collectively, the 26 rescued livers had been consecutively discarded by 298 transplant centres (median per organ: 12 centres, range 9–14) before we finally accepted them. A total of 2666 refusals were officially documented, most of which were related to poor donor conditions, followed by reasons inherent to the LT centre or recipient status (Table 2). The most frequent causes for refusal were poor donor history, poor donor ICU condition and the non-operability of an LT centre. The reasons related to LT centre condition only included: non-operability of the LT programme (n = 305); logistical reasons (n = 45); shortage of blood products (n = 39), and the impossibility of performing a second LT in parallel (n = 9).

Table 2.

Reasons (n = 2666) for refusal of livers (n = 26) officially documented in the INCUCAI and number of organs discarded for each reason

	Number of refusals n (%)	Number of organs n (%)
Donor-related conditions	1980 (74.2)
Poor donor history	1229 (46)	20 (76.9)
Poor donor ICU condition	521 (19.5)	7 (26.9)
Prolonged cold ischaemia	136 (5.1)	6 (23.0)
Abnormal biochemistry	64 (2.4)	6 (23.0)
Positive HBV serology	30 (1.1)	1 (3.8)
LT centre-related conditions	398 (14.9)
Non-operability of the LT centre	305 (11.4)	19 (73.0)
Logistical reasons	45 (1.6)	4 (15.3)
Shortage of blood products	39 (1.4)	5 (19.2)
Simultaneous LT	9 (0.3)	7 (26.9)
Recipient-related conditions	288 (10.8)
Size mismatch	194 (7.2)	19 (73.0)
Poor recipient conditions	64 (2.4)	19 (73.0)
Immunological factors	21 (0.7)	11 (42.3)
Recipient death	9 (0.3)	9 (34.6)

ICU, intensive care unit; HBV, hepatitis B virus; LT, liver transplant

Interestingly, when the arguments used to discard a liver were carefully analysed, we observed that poor donor history, size mismatch, poor recipient condition and the non-operability of an LT programme were present in more than two-thirds of the liver offers. Surprisingly, 19 livers were discarded (i.e. 19 patients lost the chance to be transplanted) because of the non-operability of one or more LT programmes during the study period.

Clinical status of recipients in the LNW and control groups

Recipient characteristics such as age, gender, time on the waiting list, diagnosis and the incidence of hepatocellular carcinoma in the explanted liver were comparable between groups (Table 3). As expected, the median position on the waiting list at the time of liver allocation was 71 for patients receiving a rescued liver compared with 18 for patients in the control group (P < 0.001). Also as expected, the median recipient MELD score with and without the addition of priority points for MELD exceptions was significantly lower in patients receiving LNWs compared with the control group. In addition, the presence of three or four donor risk factors was more frequently observed in LNWs compared with control group organs (53% vs. 24%; P = 0.03).

Table 3.

Transplant-related characteristics of the ‘livers that nobody wants’ (LNW) (n = 26) and control (n = 25) groups

	LNW group	Control group	P-value
Median recipient age, years (range)	60 (21–73)	56 (20–69)	0.68
Gender, male/female, n	13/13	19/6	0.10
Diagnosis (HCV/alcohol/other), n	8/8/10	8/7/10	0.97
Hepatocellular carcinoma, n (%)	7 (26.9)	7 (28)	0.89
Median time in our LT programme, days (range)	53 (1–696)	42 (1–788)	0.72
Median position on the waiting list (range)	71 (31–432)	18 (2–29)	<0.001
Median MELD score (range)	13 (6–21)	18 (6–34)	0.02
Median corrected MELD score (range)	13 (6–21)	22 (7–34)	<0.001
Donor risk factors (≤2 vs. 3–4 factors)	11/15	19/6	0.03

MELD scores were calculated at the time of transplantation with and without correction for patients with hepatocellular carcinoma according to the United Network for Organ Sharing (UNOS)

HCV, hepatitis C virus; LT, liver transplant, MELD, Model for End-stage Liver Disease

Operative data and transfusion requirements

A piggyback technique was used in only one procedure and the standard technique with vena cava resection was applied in all others. Transitory porto–caval shunt was not used in any patient. A full-size liver was implanted in each recipient; no patients required technical adaptation to avoid intraperitoneal high pressure secondary to a mismatch in size between the donor liver and the recipient's abdominal cavity. Only three grafts had a prolonged cold ischaemia time and four had warm ischaemia time >40 min. The median cold (350 min vs. 350 min) and warm (40 min vs. 35 min) ischaemia time of the liver graft were equivalent in the rescued and control livers (Table 4).

Table 4.

Intraoperative variables and transfusion requirement in the ‘livers that nobody wants’ (LNW) (n = 26) and control (n = 25) groups

	LNW group	Control group	P-value
Cold ischaemia time, min, median (range)	350 (240–1180)	350 (270–780)	0.77
Warm ischaemia time, min, median (range)	40 (28–56)	35 (28–42)	0.06
Operative time, min, median (range)	248 (150–360)	255 (195–360)	0.14
RBC transfusion, units, median (range)	0 (0–5)	1 (0–4)	0.42
LT without RBC transfusion, n (%)	13 (50)	7 (28)	0.18
FFP transfusion, units, median (range)	8 (0–16)	8 (0–16)	0.95
Platelets transfusion, units, median (range)	0 (0–10)	2 (0–14)	0.92

RBC, red blood cells; LT, liver transplant; FFP, fresh frozen plasma

The median operation time was similar in both groups (248 min in the LNW group vs. 255 min in the control group; P = 0.14). Intraoperative red blood cell (RBC) transfusions were needed in only 61% of the consecutive series. Half of the rescued livers were transplanted without the need of intraoperative RBC transfusions, compared with 28% in the control group, but this difference was not statistically significant (P = 0.18). Intraoperative requirements for RBC, fresh frozen plasma and platelets units were similar in both groups.

Postoperative outcomes

Mechanical ventilation was stopped immediately after LT in the operation room in 23.5% of cases (19% of LNW patients vs. 26% of control cases; P = 0.68). Postoperative biochemical profiles in each group were similar in terms of transaminase peaks, but the bilirubin peak was significantly lower in the LNW group compared with the control group (Table 5).

Table 5.

Comparative analysis of intraoperative outcome and postoperative biochemical profile and outcome in the ‘livers that nobody wants’ (LNW) (n = 26) and control (n = 25) groups

	LNW group	Control group	P-value
Intraoperative outcome
Extubation in the ORa, n (%)	5 (19.2)	7 (26.9)	0.68
Biochemical profileb
Peak total bilirubin, mg/ml, median (range)	3.5 (0.9–14)	6.1 (1.5–25)	0.01
Peak ALT, U/l, median (range)	590 (163–4890)	590 (163–4890)	0.20
Peak AST, U/l, median (range)	533 (162–2870)	895 (240–17600)	0.19
Peak AST > 1500 U/l, n (%)	5 (19.2)	6 (24)	0.94
Prothrombin time < 50%, n (%)	13 (50)	12 (48)	1.00
Primary delayed function, n (%)	4 (15.3)	4 (16)	0.74
Postoperative outcome
Primary non-function, n (%)	1 (3.8)	1 (4)	0.48
Renal failurec, n (%)	2 (7.6)	1 (4)	0.97
Major complications, n (%)	4 (15.3)	5 (20)	0.94
ICU stay, days, median (range)	3 (1–12)	2 (1–27)	0.58
Hospital stay, days, median (range)	8 (5–33)	8 (4–47)	0.64
30-day mortality, n (%)	1 (3.8)	2 (8.0)	0.97
1-year patient survival rate	84%	84%	0.74
1-year graft survival rate	84%	80%	0.94

^aWithdrawal of mechanical ventilation in the operating room

^bPeak serum transaminases and total bilirubin or lower prothrombin time within the first week after liver transplant

^cRequiring haemodialysis or haemofiltration

OR, operating room; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ICU, intensive care unit

One patient in each group developed PNF in the absence of vascular problems and finally died as a result of uncontrolled sepsis on postoperative days 10 and 12 (PNF rate: 3.8% in the LNW group vs. 4% in the control group; P = 0.48). No patient in the series developed secondary graft non-function after LT. The incidences of PDF (15.3% in the LNW group vs. 16% in the control group), major complications (15.3% in the LNW group vs. 20% in the control group) and renal failure (7.6% in the LNW group vs. 4% in the control group) after LT were equal in both groups (Table 5). The median ICU (3 days vs. 2 days) and hospital stay (8 days in both groups) did not differ significantly between the two groups. Although the 30-day mortality rate was lower in the LNW group (3.8% vs. 8.0%; P = 0.97), this difference was not significant. One-year rates for patient (84% vs. 84%) and graft (84% vs. 80%) survival after LT were comparable in the LNW and control groups.

Discussion

The scarcity of organs for LT is the most important factor leading to the deaths of patients on LT waiting lists worldwide. In this setting, we demonstrated that many usable livers are continuously discarded by many LT centres. Our policy of using LNWs yielded a reduction in time on the waiting list, very low waiting list mortality and no drop-out from the waiting list arising from tumour progression. Although most of the livers had been discarded by others for reasons of unacceptable donor status, many refusals were based on LT centre-related conditions, regardless of the quality of the liver graft or the condition of the recipient. It is interesting that we obtained postoperative outcomes and 1-year graft and patient survival rates for transplants of ‘undesirable’ livers into recipients with low MELD scores similar to those obtained in our consecutive series of the most ill patients who received competitive organs allocated under the MELD system.

An ECD graft was recently defined as a liver with an increased incidence of poor graft function, graft failure or transmission of a donor-derived disease.⁷ However, this definition remains subjective and the criteria for accepting a liver vary from country to country and from centre to centre and sometimes depend on whether or not the surgeon on call receives the offer of the liver. In order to properly assess the outcomes of this series of LNW LTs, we must first clarify whether the livers came from ECDs or whether their acceptance was precluded by only recipient- or LT centre-related conditions. Therefore, we carefully revised all ECD risk factors affecting each of the LNWs and demonstrated that the quality of the rescued livers was significantly impaired. As a majority of these livers had three or four risk factors, we speculated that many surgeons discarded the organs because of the cumulative effect of these risk factors.¹⁴ As expected, the official reason most frequently given by a centre for turning down an offer of a liver in this series of LNWs was related to inadequate donor condition. Unfortunately, our national centralized database failed to provide interesting information regarding the surrogate donor-related variables motivating each liver refusal. For example, the exact abnormality in the donor biochemistry or the objective variable that led the centre to assume a ‘poor ICU donor condition’ or ‘poor donor history’ could not be explored.

The most frequent argument used to discard a liver offer was ‘poor donor history’. By contrast, our aggressive policy minimized the weight of previous donor history (e.g. diabetes, peripheral vascular disease, donor obesity, alcohol abuse) and relied mainly on the macroscopic aspect of the liver as assessed by an experienced surgeon during harvesting or in the fresh frozen biopsy as a definitive objective variable.¹ We consider that a donor history of alcohol abuse and donor obesity are unacceptable reasons for refusing a liver. These two conditions can eventually, but do not always, impair liver graft quality. However, when dealing with ECD grafts, some conditions need to be carefully addressed: harvesting should be performed by experienced surgeons; recipient surgery should be carried out with a meticulous surgical technique and minimum blood loss, and maximum efforts should be made to reduce the cold and warm ischaemic time of the graft. Only with this strategy did we obtain postoperative graft outcomes similar to those achieved in the control group.

Ongoing efforts aim to clarify how to optimize the matching of donors with recipients.^10,21 Initially, most centres advocated adhering to the notion that ECD organs should be offered only to candidates with extended indications of hepatocellular carcinoma.¹² However, it is now thought that ECD livers may be safely directed to candidates with low MELD scores because healthier recipients may better tolerate a poor initial function of the liver graft or major postoperative complications.^9,22,23 In Argentina, the MELD-based allocation system was adopted in 2005 for patients requiring LT. Unfortunately, some clinical conditions, such as severe encephalopathy, hyponatraemia, refractory ascites, intractable pruritus or recurrent bacterial cholangitis, lead to a diminished quality of life but are not relevant under the MELD allocation policy.^24–26 For these patients who require LT but who are underestimated by the MELD system, these ‘undesirable’ organs may represent a valuable source of liver graft. Our strategy was to simply direct these ‘unwanted’ livers to the candidates on our waiting list whom we considered would benefit most from LT, regardless of the MELD score calculated at the time of the liver offer. Using this allocation strategy for these ECD livers, we achieved postoperative outcomes and graft and patient survival rates similar to those achieved with higher ranked patients transplanted with better organs. Further investigation might define how we can improve transplant benefit to avoid futile LT under the MELD system.

Previous European data have indicated that busy centres (i.e. more than 35 LTs/year) are less selective in their decisions on whether or not to accept a liver compared with lower-volume centres.²⁷ However, we demonstrated that this rule does not necessarily apply in other countries. Our LNW group represented almost half of our LT volume and we do not qualify as a high-volume centre. Despite our surgical volume, and based on our commitment to each patient enrolled on our programme, we decided to accept LNWs and to evaluate the transplant benefit and the risk of dying while waiting for a better organ on an individual basis.

This study raises the question of whether all LT centres should become involved in accepting ECD offers. This policy has an inherent risk of cost-effectiveness and some important financial issues need to be carefully addressed.^22,23 Contractual reimbursement agreements are not indexed by donor quality or the severity of recipient disease, which results in a financial loss for the transplant centre when post-transplant outcome is dismal. Our surgical team failed to transplant 10% of the livers we accepted. Unfortunately, health insurance does not usually cover the costs of recipient hospital admission or organ procurement if the liver is not ultimately transplanted. For example, the cost of the preservation solutions, transporting the liver team by ambulance or by plane and the honoraries of the surgical team involved in the donor surgery are not usually covered. It may be that centres that perform fewer than 10 LTs per year should not economically benefit from this policy. However, our surgical team considered that procuring an ECD graft with the intent of saving a usable liver is always beneficial from an ethical point of view when no other options are available for a patient who may die while on the waiting list. Institutional support and a comprehensive analysis of the cost-effectiveness of the LT programme are paramount before embarking on this aggressive approach. It may be advisable for health insurance companies to share this policy in the future in order to provide better support to experienced programmes that opt to accept ECD grafts.

Finally, there are many non-medical reasons behind the discarding of these livers and thus some political issues need to be carefully addressed. Clearly, there is a wide variety in health care systems, national regulations and transplant entities around the world. In Latin America, the financial and political limitations inherent to many developing countries in the region have negatively impacted the growth of LT. The lack of adequate financial coverage, education and organization are probably the main limiting factors in many countries in the region. Curiously, in Argentina the third most common reason for discarding a liver graft in this series of LNWs was not related to either donor quality or recipient status; it was, instead, related to transplant centre conditions. We must state that during the study period we turned down many offers of livers, but we did not prospectively collect information regarding numbers and reasons. However, we can confirm that although we refused some livers for donor- or recipient-related reasons, we did not discard any organ as a consequence of non-medical issues related to the LT centre. In Argentina, the National Ministry of Health is in charge of overseeing both private and public subsectors of the health care system and is responsible for setting regulations, performing evaluations and collecting statistics in this field. The INCUCAI is the national entity controlling organ allocation and transplantation and must guarantee equal access to LT in our country. Despite the great medical effort expended to develop LT programmes in the country, citing the shortage of blood products, logistical problems or the non-operability of an LT as official reasons for turning down a liver offer alerts us to the need for external monitoring of the quality of health care in this field.²⁸ National entities and medical societies should focus on developing and implementing specific regulations to ensure standards of excellence to provide the highest quality of patient care in the area of LT.

In conclusion, we emphasize that transplantable livers are unnecessarily discarded by the transplant community. Each transplant centre should discuss on a case-by-case basis when to accept an LNW to reduce the mortality and drop-out rates on the waiting list. External and internal supervision of the activity of each LT programme are urgently needed to guarantee a high standard of excellence in the field of transplantation.

Conflicts of interest

None declared.