Comparison of results on the use of extended criteria liver doners for transplants in Espírito Santo

ABSTRACT Introduction: liver Transplantation is currently the treatment of choice for several terminal liver diseases. Despite the increase in performed transplants, the waiting lists continue to increase. In order to expand the supply of organs, transplantation teams have started to use previously rejected livers for transplants because of an increased risk of unfavorable outcomes. Objective: to evaluate the use of livers of expanded criterion donators. Methods: retrospective study of medical records. The livers were classified as normal or expanded criteria. The groups were divided in low and high MELD. A multivariate analysis was performed through logistic regression. Results: there was no statistical difference regarding early, late and global mortality between the groups. Decreased survival was observed in patients with high MELD (higher or equal to 20) when they received grafts from expanded criterion donators. The association between the occurrence of cardiorespiratory arrest and presence of elevated total bilirubin in donators was associated with higher mortality rates in expanded criterion livers. Conclusion: the overall results are similar, but expanded criteria liver donators was associated with higher mortality in patients with high MELD.

Transplantation teams started using liver grafts that had been previously rejected, the so-called marginal donors, expanded criteria livers or expanded criteria donors (ECD) 8  points to use of vasopressor medication (any dosage of noradrenaline or dobutamine) and serum sodium dosage > 155 mEq/L. We considered a liver with an expanded criteria when its score was greater than or equal to 3 17 .
We excluded Ischemia time (> 13h) and hepatic steatosis from the analysis, since fatty livers are discarded due to the absence of histopathological examination during harvesting. As for ischemia time, the organ is only accepted if the estimated time to implant is less than 12 hours.
The statistical analysis started by characterizing the sample. Concerning the recipient, we recorded the patient's age at the time of transplantation and preoperative MELD. As for the donor, we recorded age, sex, laboratory variables used to distribute the groups, and cause of brain death. Regarding the characteristics of the harvest-to-implant process, we analyzed cold ischemia time, solution used for perfusion and organ preservation, hot ischemia time, and organ harvest site.
We obtained two groups, 56 patients who received grafts from expanded criteria donors (ECD) and 54 patients receiving standard grafts (SG). We performed a descriptive analysis of the recipients with mean, standard deviation, minimum and maximum values referring to age and MELD classification, as shown in Table 1. We also subdivided the SG and ECD groups according to high and low MELD. As for sex in the SG group, 44 were male (81.5%) and 10 female (18.5%); in the ECD group, 44 were male (78.6%) and 10 were female (21.4%).
In the evaluation of donors, we performed an analysis of age, mean time of intubation, AST, ALT, total bilirubin, serum sodium (Na), use of vasoactive medication (VAD), distributed in the SG and ECD groups, as shown in Table 1. Regarding sex, in the SG group 77.8% were male and 22.2% female, while in the ECD 78.6% were male, and 21.4%, female.
During the organ donation process, from the event that caused brain death to the moment of liver  (Table 2).  Among the causes of death in the SG group, the most common was traumatic brain injury (TBI), encompassing the various causes (accidents and gunshot or stab wounds), followed by stroke in both groups, as described in Table 2.  18 . We also presented the Odds Ratio (OR) and the pseudo-r 2 , which is similar to the coefficient of determination (proportion of explanation of the dependent variable as a function of the independent ones) of the linear regression 19 .
The level of significance adopted was 5%, with a 95% confidence interval, and the software used in the analyzes was the IBM SPSS Statistics, version 21.

RESULTS
The survival observed in the SG group in the early period (0 to 30 days) was 94.4%, in the late period (31 to 90 days), 86.3%, the overall survival (0 to 90 days) being 81.5%. In the ECD group, survival was 85.7% in the early period and 87.5% in the late one, the overall survival being 75% (Figure 2 Figure   3).
As for the frequency of primary nonfunctioning, in the SG group it was four cases (7.4%), three of whom died early (less than 7 days); the remaining patient underwent retransplant, but died within 53 days after the operation. In the ECD group, there were also four cases, 7.1%, of whom tow died prematurely and two others underwent retransplant, in which one progressed to early death and the other survived longer than 90 days.   Regarding retransplantation (RTX) in both groups, there were two RTX (3.7%) in the SG group, one due to late hepatic artery thrombosis (after 30 days)

Ferreira-Jr
Comparison of results on the use of extended criteria liver doners for transplants in Espírito Santo that evolved well, and the other due to primary nonfunctioning (PNF), who later died, thus displaying a 50% survival. In the ECD group, there were also RTX, both by PNF, with one favorable outcome and one early death, also representing a 50% survival rate.
occurrence of donor CPA was associated with an increase in the risk of death by 17.89 times (p = 0.043), in addition, MELD greater than or equal to 20 was associated with a death risk increase of 6.99 times (p = 0.046) ( Table   3). Logistic regression for the late period in the SG group indicated an association between prolonged cold ischemia and an increased risk of death. However, in the ECD group, the increase was negligible, 0.99 times (p = 0.049). In addition, the presence of elevated preoperative MELD (≥ 20) was associated with an increase in mortality risk by 10.94 (p = 0.035) ( Table 3).
The logistic regression for the deaths during the evaluated period (0 to 90 days) did not indicate significance for any analyzed variable. In the high preoperative MELD SG group, we found a 5.18-fold increased risk of death, though not significant (p = 0.26). In the ECD group, the presence of elevated preoperative MELD was associated with a 10.81-fold increased risk of death (p = 0.014).
In addition, the donor's elevated total bilirubin (TB) was associated with an increase in the risk of death of 7.11 times (p = 0.022) ( Table 3). There was no statistically significant clinical variable in the SG group, whereas in the ECD group, elevated total bilirubin (TB) was statistically significant, increasing the risk of death by 3.06 times (p = 0.019). The

Ferreira-Jr Comparison of results on the use of extended criteria liver doners for transplants in Espírito Santo
The occurrence of PNF and retransplantation was similar in both groups and the regression analysis did not show any risk or protection factor.

DISCUSSION
Since OLT was instituted as an excellence, definitive, and curative treatment for chronic and terminal liver diseases, there has been a progressive and continuous increase in waiting lists for transplantation. Therefore, alternatives were sought to increase organ supply. The first studies to try to identify factors related to worse results, such as advanced age, length of ICU stay, ischemia (warm and cold) time, laboratory tests, hemodynamic status, and use of VAD, date from 1993 17,20 .
In agreement with these data, we observed that 50.9% (56 cases) of the grafts used for transplantation during the study period were from ECD, which contributed to a significant increase in the number of transplants performed in this period due to the expansion of the liver grafts' acceptance criteria.
Thus, the use of these ECD is, at the moment, one of the most widely used alternatives 21,22 . Several authors have demonstrated that these organs can be safely used when assessing the risk of operative complications compared with prolonged time on the waiting list for OLT, showing cost-effectiveness. Therefore, the vast majority of organs are used, discarding only those whose severe changes would render their utilization unethical 23  Among the factors used to assess the quality of the graft, the only one we found to be associated with increased mortality, both early and overall, was the increase in total bilirubin (> 2), with an increase in the risk of early death by 3.1 times, and overall death, by 7.1 times 26 .
Regarding the frequency of PNF, we found similar rates in both groups, similar to the literature reports 27 , but with rates of retransplantation lower than those published. Therefore, we could not associate PNF and retransplantation with the use of ECD organs.
Another important aspect when evaluating graft quality and possible solutions to increase organs supply is the assistance that potential donors receive when admitted or evaluated in hospitals 28 . Adequate and aggressive support should be given to every patient

Ferreira-Jr
Comparison of results on the use of extended criteria liver doners for transplants in Espírito Santo admitted or evaluated, even those with a poor prognosis, to provide the best care. However, in those cases in which serious injuries are identified that make it impossible for the patient to recover, one should identify potential donors as soon as possible and maintain recovery aggressively, even if ICU support, fluid resuscitation, use of VAD, and hormonal recovery, when indicated, are necessary to maintain their organs 28,29 . Thus, it is possible to decrease the loss of potential donors, improve quality, and decrease family refusal through teams engaged and prepared for such an approach, after adequate care for the victim 28 .
We observed the importance of this aggressive support to maintain hemodynamic stability and prevent CPA. We identified that history of CPA, when associated with other risk factors for ECD graft malfunction, posed a higher risk of complications and mortality, increasing the risk of mortality by 17.9 times before 30 days, further justifying the importance of adequate donor care 28,30 .
Thus, in our survey, we observed the influence on survival of ECD receptors only by elevated bilirubin and a history of CPA, as well as worse results in the subgroup with high MELD score. In a way, this justifies the use of these organs. However, we must emphasize that this is a study with a relatively small sample and new studies with larger numbers of patients may reinforce the information found.
Ultimately, we observed that, in general, there is no optimal graft and the vast majority is utilized as long as there is no clear contraindication.
Expanded criteria donor grafts are widely used and can be safely utilized, with results similar to those of SG, both in the early period, in the late one, and overall.
In those recipients with high MELD (≥ 20), the use of ECD livers has a negative impact on survival. Risk factors that negatively affect survival of ECD liver recipients are elevated bilirubin and CPA. There was no increase in PNF, nor a higher incidence of retransplantation, due to the use of ECD livers.

CONCLUSION
Expanded criteria donor grafts can be safely used, with results similar to those of SG, both in the early period, in the late one, and overall. Risk factors that negatively impact the survival of ECD liver receptors are elevated bilirubin, CPA, and receptors with high MELD (≥ 20). There was no increase in PNF, nor a higher incidence of retransplantation, due to the use of ECD livers.