1 INTRODUCTION

Over the past few years, the transplant community has made great efforts to increase the pediatric priority in liver organ allocation; however, the mortality of children candidates for liver transplantation (LT), who are disadvantaged because of the lack of size-matched donors, has been steady at approximately 10% every year.1, 2 Split liver transplantation (SLT) was introduced to expand the pool of grafts available for pediatric recipients, providing a left lateral segment (LLS) to a child and an extended-right graft (ERG) to an adult recipient.3 After the initial splitting experience, advances in surgical techniques and a better understanding of recipient/donor matching led to excellent SLT outcomes both in pediatric and adult recipients.4-9

Different organ allocation systems encouraged the implementation of SLT to expand the graft availability from deceased donors for children, without disadvantaging adult LT candidates.10 The SLT rate varies worldwide: in Europe, SLT represents approximately 6% of all LT.11 In the United States, SLT comprises approximately 1% of all LT, despite the fact that it is estimated that 20% of deceased donors meet United Network of Organ Sharing (UNOS) guidelines for split livers.12 In Italy, after the encouraging results of the initial SLT experience of the North Italian Transplant programme,13 since 2015 the National Transplantation Centre (CNT) in collaboration with the Italian College of Liver Transplantation Programmes defined a mandatory-split liver policy (SLP) in order to increase the number of splitting procedures nationwide and to reduce pediatric LT-waiting list mortality.

This study aims to analyze liver allocation in Italy after the introduction of the new SLP and its impact on the pediatric/adult LT-waiting list and on SLT outcomes.

2 MATERIALS AND METHODS

2.1 Liver allocation policy in Italy

In Italy there are 21 LT centers divided into 13 regions governed by the CNT network and grouped into two macro-areas (Figure 1). Since the CNT's establishment, liver allocation policies have seen several modifications. The first national SLT allocation program was approved in 2006; transplant centers could voluntary decide whether or not to participate. At that time, donors aged ≤14 years were preferentially allocated to pediatric recipients (<18 years), whereas donors ≥15 years were allocated to adults. Donors aged 10-50 years with stable hemodynamics, intensive therapy unit (ITU)-stay ≤5 days, transaminases ≤3 times normal, and absence of steatosis on the ultrasound scan were defined as “splittable.” The decision of whether or not to perform the splitting procedure was the choice of the LT center to which the graft was assigned.

In 2015, a consensus redefined the Italian criteria for LT candidate stratification not only based on urgency but also on the principles of utility and transplant benefit. The current liver allocation system is based on the Model for End-Stage Liver Disease (MELD)/Italian Score for Organ Allocation (ISO), which is defined by biochemical MELD and exceptions.14 At present, liver grafts are shared according to the following principles: (1) nationwide, for (a) UNOS status 1 patients; (b) pediatric candidates according to the pediatric LT allocation system15; (2) macro-areas for adult LT candidates with MELD ≥30; and (3) regionally for adult patients with MELD <30.14 Additionally, the age of the donors preferentially assigned to pediatric recipients increased from 14 to 17 years and a national mandatory SLP was adopted. In the new SLP, all deceased donors aged 18-50 years with standard risk (defined as the absence of potential transmissible infections or neoplastic diseases) are mandatorily offered to pediatric transplant centers according to the pediatric national LT-waiting list15 unless a UNOS 1 status or MELD ≥30 adult candidate is on the waiting list. If the deceased donor is “splittable,” the LLS graft is allocated to a pediatric recipient. According to the adult rules, the ERG is then allocated to a recipient not only on the basis of the MELD/ISO score14 but also taking into account clinical parameters and donor-to-recipient size matching (Figure 2).

3 RESULTS

3.2 Outcomes of SLT

3.2.1 Left-lateral segment graft transplantation

Table 5 shows the surgical and recipient characteristics of LLS transplantation. During the study period, the LLS recipients were significantly younger compared to those transplanted in the control period, including 77.3% vs 58.5% children <24 months of age (P = .025).

Table 5. Surgical and recipient characteristics of left-lateral segment transplantation

Variables	New split liver policy (n = 66)	Old split liver policy (n = 65)	P value
Surgical variables
In situ/ex situ split	64 (97%) / 2 (3%)	63 (96.9%) / 2 (3.1%)	1.000
Cold ischemic time (hours)	6 (3-10)	5 (1-8)	.108
Warm ischemic time (minutes)	47 (24-110)	40 (30-121)	.238
Recipient variables
Age (years)	1.1 (0.1-11.2)	2.1 (0.1-12.1)	.043
Gender (female)	33 (50%)	36 (55.4%)	.601
BMI	16.4 (13 -23.8)	16.5 (11.3-22.3)	.848
Weight (kg)	8.7 (4-35)	12 (6-30)	.020
Height (cm)	70.1 (53-135)	84 (58-152)	.014
Blood group
0	28 (42.4%)	22 (33.8%)
A	23 (34.8%)	33 (50.8%)	.271
B	12 (18.2%)	7 (10.8%)
AB	3 (4.5%)	3 (4.6%)
PELD score	22 (10-43)	22 (12-39)	1.000
UNOS status
1	8 (12.1%)	7 (10.8%)
1B	1 (1.5%)	6 (9.2%)
2A	5 (7.6%)	7 (10.8%)	.344
2B	10 (15.2%)	10 (15.4%)
3	42 (63.6%)	35 (53.8%)
Indication for transplantation
Acute liver failure	6 (9.1%)	3 (4.6%)
Autoimmune liver disease	2 (3%)	1 (1.5%)
Cholestatic liver disease	43 (65.2%)	36 (55.4%)
Tumor	2 (3%)	9 (13.8%)	.074
Metabolic liver disease	7 (10.6%)	6 (9.2%)
Retransplantation (early/late)	4 (2/2) (6.1%)	10 (1/9) (15.4%)
Other disease	2 (3%)	-
Time on waiting list to transplant (days)	36 (1-530)	35 (1-571)	.634

• ALT, alaninoaminotransferase; AST, aspartatoaminotransferase; GGT, gamma-glutamyltransferase; BMI, body mass index; PELD, pediatric end-stage liver disease; UNOS, United Network of Organ Sharing.

The overall 1-year patient survivals were 84.8% and 87.7% in the study and control period respectively (P = .408); the overall 1-year graft survival was 78.8% for the new SLP and 80% for the old SLP (P = .610) (Figure 3).

In the study group, 10 (15.2%) patients died within 1 year from SLT, of whom 5 (50%) were urgent SLT. Of 7 (10.8%) children who died during the old SLP, only one (14.2%) had urgent transplantation. The retransplantation rate was similar in the two groups (4 [6.1%] vs 5 [7.7%] [P = .712]) (Table 6). The characteristics of the deceased and retransplanted LLS recipients are detailed in Table S2.

Table 6. Causes of death and retransplantation after left-lateral segment transplantation

Variables	New split liver policy (n = 66)	Old split liver policy (n = 65)	P value
Total number of death	10 (15.2%)	7 (10.8%)	.456
Cause of death
PNF	1 (1.5%)	2 (3.1%)	.619
Sepsis	2 (3.0%)	1 (1.5%)	.568
MOF	5 (7.6%)	2 (3.1%)	.077
Tumor recurrence	0 (0%)	1 (1.5%)	1.000
Pulmonary embolism	1 (1.5%)	0 (1.5%)	.496
Biliary complication	0 (0%)	1 (1.5%)	1.000
Cerebrovascular accident	1 (1.5%)	0 (0%)	.496
Retransplantation	4 (6.1%)	5 (7.7%)	.712
Cause of retransplantation
PNF/DNF	1 (1.5%)	3 (4.6%)	.302
Hepatic artery thrombosis	2 (3%)	0 (0%)	.496
Portal vein thrombosis	1 (1.5%)	1 (1.5%)	.991
Chronic rejection	0 (0%)	1 (1.5%)	.312

• DNF, delayed nonfunction; LLS, left-lateral segment; MOF, multiorgan failure; PNF, primary nonfunction.

Within the first year of SLT, postoperative technical complications were comparable in the two periods (20 [30.3%] vs 24 [36.9%], [P = .529]) (Table S3).

3.2.2 Extended right graft transplantation

The technical and recipient characteristics of ERG transplantation were comparable in the two periods (Table 7). The 1-year patient survival was 93.8% in both groups (P = .538), whereas 1-year graft survival was 86.2% after the introduction of the new SLP and 83.1% in the control period (P = .753) (Figure 4).

Table 7. Surgical and recipient characteristics of extended right graft transplantation

Variables	New split liver policy (n = 65)	Old split liver policy (n = 65)	P value
Surgical variables
In situ/ex situ split	58 (89.2%)/7 (10.8%)	59 (90.8%)/6 (9.2%)	.778
Cold ischemic time (hours)	7 (4-11)	6 (3-15)	.125
Warm ischemic time (minutes)	42 (22-165)	35 (22-80)	.207
Recipient variables
Age (years)	53 (2-71)	53 (8-69)	.714
Gender (female)	29 (44.6%)	38 (58.5%)	.160
BMI	23.2 (13.4-32)	23.5 (14.3-43.3)	.257
Weight (kg)	64 (11-103)	65 (22-109)	.782
Height (cm)	168 (88-183)	165 (86-180)	.279
Blood group
0	30 (46.2%)	21 (32.3%)
A	22 (33.8%)	37 (56.1%)
B	12 (18.5%)	6 (9.2%)	.040
AB	1 (1.5%)	1 (1.5%)
Biochemical MELD/PELD	18 (10-35)	20 (9-42)	1.000
UNOS status
1	2 (3.1%)	3 (4.6%)
2A	5 (7.7%)	8 (12.3%)	.824
2B	38 (58.5%)	35 (53.8%)
3	20 (30.8%)	19 (29.2%)
Indication for SLT
Alcoholic liver disease	10 (15.4%)	8 (12.3%)
Autoimmune liver disease	12 (18.5%)	9 (13.8%)
Cholestatic liver disease	4 (6.2%)	2 (3.1%)
Tumor	16 (24.6%)	24 (36.9%)	.109
Viral-related cirrhosis	16 (24.6%)	7 (10.8%)
Cryptogenic cirrhosis	1 (1.5%)	5 (7.7%)
Metabolic liver disease	5 (7.7%)	4 (6.2%)
Retransplantation	0 (0%)	1 (1.5%)
Others	1 (1.5%)	5 (7.7%)
Time on waiting list to transplant (days)	91 (1-1682)	74 (1-1479)	.764

• ALT, alaninoaminotransferase; AST, aspartatoaminotransferase; GGT, gamma-glutamyltransferase; BMI, body mass index; MELD, model for end-stage liver disease; LFTs, liver function tests; PELD, pediatric end-stage liver disease; UNOS, United Network of Organ Sharing.

In the study group, 4 (6.2%) patients died within 1 year of SLT, of whom 1 (25%) was transplanted with UNOS 1 status; of 4 (6.2%) deceased patients in the control group, none received urgent transplantation. The retransplantation rate was similar in the two groups (5 [7.7%] vs 7 [10.8%] [P = .638]) (Table 8). The characteristics of the deceased and retransplanted ERG recipients are reported in Table S4.

Table 8. Causes of death and retransplantation after extended right graft transplantation

Variables	New split liver policy (n = 65)	Old split liver policy (n = 65)	P value
Total number of death	4 (6.2%)	4 (6.2%)	1.000
Cause of death
Sepsis	2 (3.1%)	1 (1.5%)	.559
Myocardial infarction	1 (1.5%)	0 (0%)	.315
Tumor recurrence	1 (1.5%)	1 (1.5%)	1.000
MOF	0 (0%)	2 (3%)	.496
Retransplantation	5 (7.7%)	7 (10.8%)	.638
Cause of retransplantation
PNF/DNF	3 (4.6%)	3 (4.6%)	1.000
Hepatic artery thrombosis	2 (3.1%)	2 (3.1%)	1.000
Biliary complications	0 (0%)	2 (3.1%)	.154

• DNF, delayed nonfunction; ERG, extended right graft; MOF, multiorgan failure; PNF, primary nonfunction.

Postoperative complications were observed in 15 (23%) ERG recipients transplanted in the new SLP and in 24 (36.9%) cases in the control group (P = .085) (Table S5).

3.2.3 Univariate and multivariate analysis

By log-rank test, no significant differences in 1-year patient or graft survivals for LLS and ERG transplantations were observed, according to the centers’ split liver procurement volume (Table S6).

Among LLS transplantation, re-transplantation vs first transplantation (HR = 3.349, P = .006), recipient body weight >20 kg vs ≤20 kg (HR = 4.088, P = .001), and donor-to-recipient weight ratio (DRWR) ≤4 vs >4 (HR = 1.380, P = .018) were risk factors for graft failure by univariate analysis (Table S7). In the multivariate analysis, recipient body weight >20 kg vs ≤20 kg (HR = 5.113, P = .048) and retransplantation as an indication of SLT (HR = 2.641, P = .035) were predictors of graft failure (Figures S1, S2).

In ERG transplantation, cold ischemic time (CIT) >8 hours vs ≤8 hours (HR = 2.574, P = .039) and donor ITU-stay >5 days vs ≤5 days (HR = 1.946, P = .046) were identified as risk factors for graft loss according to the univariate analysis (Table S8), and only CIT >8 hours vs ≤8 hours (HR = 2.475, P = .048) in the multivariate analysis (Figure S3).

4 DISCUSSION

Improving SLT programs has attracted great interest in recent years as a method of reducing pediatric LT-waiting list mortality.10 Although the SLT rate constantly increased in Europe (>400 SLTs/year since 2002),18 a change in donor demographics resulted in a reduction of deceased liver donors suitable for split procedures. The main issues limiting the expansion of SLT programs worldwide include (1) the need for a national/international organ allocation system with an “intention-to-split” policy, and the cooperation between pediatric and adult LT centers; (2) the lack of international donor selection criteria for SLT; (3) split grafts have been generally perceived by centers as carrying greater morbidity and mortality compared to whole grafts.10, 19, 20

In Italy, the steady increase in average age of deceased donors and the reduction in the number of SLTs causing a lengthening of waiting times for pediatric LT drove the analysis of the Italian SLT data21-27 and the definition in 2015 of a mandatory SLP, which includes three major steps: first, the CNT offers all donors aged 18-50 years at standard risks to the pediatric transplant centers for SLT; second, the pediatric transplant centers decide if the LLS is suitable for pediatric LT candidates; and third, the ERG, which is considered ab initio fit for transplantation as the pediatric donor selection criteria are more narrowing than those adopted for adults, is allocated to the best-matching adult recipient. Subsequently, the graft “splittability” remains only a technical aspect (ie, vascular anomalies).

In the study period, donors within SLT-criteria accounted for 26% of all liver donors, of which 60% were offered for SLT and of these, 21% were effectively used for SLT. Consequently, the introduction of the new SLP significantly increased the number of donors offered for split procedures and SLT rate represented 8.4% of all LT activity, which is higher compared to the recent split data from the Eurotransplant Area (5%).28 Despite the expansion of SLT donor criteria, most of the split procedures were performed in young donors, proving that pediatric centers had more opportunities to choose the best donor for children candidates to LT, with high rate of acceptance.29

Among donors refused for SLT, 40% were not accepted due to the donor quality and 39% due to the absence of suitable recipients, mainly related to donor/recipient size matching. Interestingly, the donors refused due to the lack of size matching recipients were similar to those accepted for SLT for all clinical variables, except for higher body weight. Therefore, almost 40% of potentially “splittable” donors are currently not used for standard splitting procedure, but might represent a consistent donor pool for full-right/full-left SLT or for international networks of organ sharing.20, 30

A mandatory SLP has been adopted also in other countries; thus their split liver donor selection criteria are much more strict when compared to these currently used in Italy.10 In United Kingdom all brain-death donors <40 years, with weight >50 kg, and <5 days of ITU stay are mandatory offered for splitting in absence of superurgent or multivisceral candidates31; in France, donors aged <30 years are first proposed to pediatric LT candidates.32 Contrarily, in the UNOS network splitting is not mandatory and only donors between 18-40 years of age with no more than a single vasopressor, transaminases ≤3 times normal, and body mass index ≤28 are considered for SLT, but are not primarily offered to pediatric centers2; the result is that only 6.3% of liver donors meet the SLT criteria and only 3.8% of these are utilized for SLT.33

Of the criteria used to select donors suitable for split procedures, donor age >50 years and <10 years has been recognized as a crucial risk factor for SLT graft failure.18 In our series, donors aged >50 years were used for SLT in only four UNOS 1 status children because of the impossibility of waiting longer for a graft; thus the “adapted” allocation of ERGs guaranteed good outcomes in all adults.

Notwithstanding the promising results on the use of partial grafts from pediatric donors,24 presently in Italy split liver procedures in donors <18 years are not mandatory. However, our data confirmed that SLTs from young donors have favorable outcomes,34 suggesting that split procedures should be mandatory also in pediatric donors (at least for body weight >40 kg), limiting the use of reduced grafts to selected cases.

Our experience showed that the presence of a national organ-exchange organization and the collaboration between pediatric and adult centers were key factors to maximize the use of donors without compromising outcomes.19, 35-37 The availability of surgeons trained to perform split procedures in most of the Italian transplant centers was also essential to implement the national SLT programme.20 The standardization of surgical techniques, consisting of in situ split procedures38 and rules for vascular structures’ division, resulted in limited CIT (organs were shipped from the donor hospital to the recipient center) and low rates of vascular complications, which represent the most common issues in SLT.8, 28, 39-41 These good outcomes justify the enormous logistical efforts of the in situ split technique related to the prolonged donor surgery42; hence, a previous Italian experience reported that standard in situ split procurement added 156 ± 33 minutes to the donor operation, whereas using full-right/full-left split procedure required 185 ± 50 more minutes.27

The introduction of the new SLP did not have an impact on SLT morbidity and mortality, which were comparable to the control period. Children transplanted with LLS showed 1-year patient survival of 85% and graft survival of 79%, in line with the European Liver Transplant Registry18 (patient survival: 89.1%; graft survival: 83.3%) and UNOS43 data (graft survival: 75.5%). Adults receiving ERG had a 1-year patient survival of 94% and a graft survival of 86%, similar to those reported in literature.8

In children receiving LLS, retransplantation and recipient body weight >20 kg were the only risk factors of graft failure, confirming that the recipient status significantly influences SLT outcomes and adequate donor/recipient size matching is essential to avoid “small-for-size syndrome.”44 In ERG transplantation, only CIT >8 hours increased the risk of graft failure in agreement with the UNOS data.43

After the new SLP, the number of liver offers to LT pediatric candidates significantly improved (proportionally to the Italian donation rate), giving more opportunities to children to receive SLT; in consequence the median pediatric LT-waiting list time significantly decreased (from 7 months to less than 3 months), being considerably shorter compared to other series (ie, in the US >40% of children spend over a year on the waiting list).45 Likewise, the pediatric LT-waiting list mortality rate was 2.5%, remarkably lower compared to other reports (7-12%),1, 2, 10 and in half of the cases death occurred in children awaiting urgent re-LT. Because in Italy the majority of children are transplanted with split grafts from deceased donors, we believe that the low pediatric LT-waiting list mortality is mainly related to the pediatric priority in organ allocation. Additionally, in the observational period the LDLT activity reduced, thus the paucity of cases and the different centers’ policy on the use of living-related resource do not allow to interpret data. Hence, the SLT rate is influenced not only by the adoption of mandatory-SLP and donation rate but also by the pediatric LT-waiting list demand (which is variable over the time) and the use of LDLT resource.

On the adult side, the prioritization of UNOS 1 status and MELD ≥30 LT candidates ensured that the mandatory-SLP did not have a negative impact on the adult LT-waiting list mortality and dropout rate. The decreased adult LT-waiting list mortality rate observed in the study period was not related to the new SLP, but it was caused by the higher number of adult LT performed and the introduction of the new liver allocation model (MELD/ISO score), which includes MELD exception and hepatocellular carcinoma priorization.14 In the current system, transplant centers accepting ERG could choose the adult recipient based not only on the MELD/ISO score but also on the recipient clinical status and the donor/recipient size matching, resulting in an increased rate of partial grafts accepted for adults. Moreover, the presence of urgent/MELD ≥30 adult candidates eliminates the obligation of splitting but does not prohibit the possibility for those to receive a SLT (as occurred in 2 cases in our study). ERGs were allocated to a relatively low biochemical MELD score (18 [10-35]), being in line with other series.46 Limitations of our study consist in matching the new SLP with an historical control group and to not compare ERG with whole graft; thus, similar 10-years outcomes of ERG and whole LT performed in Italy were recently reported.9

In conclusion, the current is the first report of a national mandatory SLP. The first donor offer to pediatric LT candidate increases the number of children receiving SLT, resulting in low LT-waiting list dropout and mortality rate, which is mainly limited to urgent transplantation. The prioritization of organ allocation to UNOS 1 status and MELD ≥30 adult candidates ensure that mandatory SLP does not harm the adult LT-waiting list. Split liver donor criteria can be safely expanded, providing optimal graft and patient SLT outcomes. Thus, a significant proportion of potentially “splittable” donors is currently not used for conventional split procedures but might be employed for adult-to-adult SLT or for international organ networks in order to optimize liver allograft resources.

AUTHOR CONTRIBUTIONS

Angelico R: designed the work, analyzed and interpreted data, draft the work, approved the final version, and reviewed the accuracy and integrity of the work.

Trapani S: designed the work, analyzed and interpreted data, drafted the work, approved the final version, and reviewed the accuracy and integrity of the work.

Spada M: designed the work, analyzed and interpreted data, revisited critically the manuscript for intellectual content, approved the final version, and reviewed the accuracy and integrity of the work.

Colledan M: substantial contribution to the conception of the work, interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, and reviewed the accuracy and integrity of the work.

de Ville de Goyet J: substantial contribution to the conception of the work, acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Salizzoni M: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

De Carlis L: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Andorno E: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Gruttadauria S: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Ettorre GM: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Cescon M: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Rossi G: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Risaliti A: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Tisone G: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Tedeschi U: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Vivarelli M: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Agnes S: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

De Simone P: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Lupo LG: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Di Benedetto F: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Santaniello W: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Zamboni F: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Mazzaferro V: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Rossi M: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Puoti F: analyzed and interpreted data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Camagni S: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Grimaldi C: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Gringeri E: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Rizzato L: acquisition and interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Nanni Costa A: substantial contribution to the conception of the work, interpretation of data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

Cillo U: designed the work, interpreted data, revisited critically the manuscript for intellectual content, approved the final version, reviewed the accuracy and integrity of the work.

DISCLOSURE

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.