Serum factor V and arterial lactate levels predict graft survival in liver transplant recipients with aminotransferase above five thousand

Abstract

BACKGROUND

Liver transplantation (LT) is the preferred treatment for end-stage liver diseases. Early allograft failure (EAF) can result in death or retransplantation. One of the key factors predicting EAF is the degree of graft injury, which is typically assessed by elevated alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels. Aminotransferase levels exceeding 5000 U/L within 48 hours of LT are indicative of poor short-term graft survival.

AIM

To investigate outcomes in liver transplant recipients with peak aminotransferase levels exceeding 5000 U/L and to identify predictors of EAF.

METHODS

Adult patients who underwent LT from a deceased (brain-dead) donor between 2011 and 2024 at Hospital de Clínicas de Porto Alegre were screened. Patients with peak AST or ALT levels > 5000 U/L post-LT were included, excluding those with vascular thrombosis. EAF was defined as death or re-transplantation within 90 days. A receiver operating characteristic curve were generated for each EAF predictor to determine the area under the curve (AUC). Sensitivity, specificity, negative predictive value, and positive predictive value were calculated for each predictor’s best cutoff, as defined by the Youden Index. Survival curves were plotted using the Kaplan-Meier method.

RESULTS

Between 2011 and 2024, 341 patients underwent LT. Of these, 29 (8.5%) patients had AST and/or ALT levels exceeding 5000 U/L within the first 48 hours post-LT. Four patients were excluded due to vascular thrombosis, resulting in a study cohort of 25 patients. EAF were also observed in 11 patients. One-year and five-year graft survival rates were 51.7% and 42.6%, respectively. For patients without EAF, one-year and five-year graft survivals were 92.3% and 76.2%, respectively. The key predictors of EAF included serum factor V and arterial lactate levels on postoperative day (POD) 1, with AUCs of 0.936 and 0.919, respectively. The optimal cutoff for EAF prediction were 26.2% for serum factor V and 9 mmol/L for arterial lactate.

CONCLUSION

Aminotransferase levels > 5000 U/L were associated with high EAF risk. However, favorable graft function indicators on POD 1 were associated with long-term survival comparable to that of general LT recipients. Serum factor V and arterial lactate levels emerged as valuable prognostic markers.

Key Words: Aminotransferase; Liver transplantation; Early allograft failure; Factor V; Lactate

Core Tip: Among 341 liver transplant recipients, 25 had an aminotransferase level > 5000 U/L. Early allograft failure (death or re-transplantation within 90 days) occurred in 11 patients. One-year and five-year graft survival rates were 51.7% and 42.6%, respectively. However, approximately 50% of these patients withstood initial graft injury and had a satisfactory long-term prognosis. Therefore, early identification of patients who will recover graft function or undergo graft loss is paramount. Serum factor V and arterial lactate levels can aid in deciding whether to relist patients to retransplantation early in the postoperative period.

INTRODUCTION

Liver transplantation (LT) is the preferred treatment for end-stage liver disease[1]. Over the past few decades, LT outcomes have significantly improved, with 1-year and 5-year patient survival rates reaching 92% and 81%, respectively[2]. However, early allograft failure (EAF) remains a critical issue, often resulting in death or retransplantation[3-5].

A key predictor of EAF is the extent of graft injury, which is typically assessed using alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels[6-8]. Extreme liver graft injury, defined as ALT and/or AST levels exceeding 5000 U/L, is associated with poor short-term graft survival[9]. Thus, early relisting of patients predicted to lose their graft may optimize outcomes[10,11]. However, as some patients with peak aminotransferase levels exceeding 5000 U/L manage to survive the initial post-LT phase, approximately 50% according to previous studies[7,12], rendering retransplantation unnecessary, early distinction between patients who will and will not necessitate retransplantation is paramount.

This study aimed to characterize outcomes in liver transplant recipients with peak aminotransferase levels > 5000 U/L and identify predictors of EAF.

MATERIALS AND METHODS

This retrospective analysis utilized a prospectively collected dataset from the Hospital de Clínicas de Porto Alegre Liver Transplant Program. Adult patients (aged > 18 years) who underwent LT from a brain-dead donor between 2011 and 2024 were screened. Only those with peak aminotransferase (ALT and/or AST) levels > 5000 U/L within the first 48 hours after LT were included in the study. Patients with hepatic artery thrombosis or postoperative portal vein thrombosis were excluded.

Pre-transplant model for end stage liver disease (MELD)-Na score[13] and Child-Pugh-Turcotte classification[14] were calculated based on laboratory and clinical parameters. The etiology of liver disease was evaluated based on clinical features. Pathological reports of liver explants were reviewed for patients with an unknown etiology on clinical grounds. The following laboratory data were obtained on postoperative day (POD) 1: (1) AST; (2) ALT; (3) Total bilirubin; (4) Serum factor V; (5) Arterial lactate; (6) Prothrombin time (international normalized ratio); and (7) Serum creatinine levels. The laboratory parameters on POD 1 were generally obtained between 12 hours and 24 hours after the end of surgery. AST and ALT levels were measured on POD 2. Peak ALT and AST levels were defined as the highest levels of ALT and AST within the first 48 hours post-transplantation. The laboratory MELD score on POD 1 was calculated as described by Malinchoc et al[15]. Doppler ultrasound was used to screen for vascular complications in the first week after LT. Patients with abnormal Doppler ultrasound findings, such as absent arterial or portal flow or an arterial resistive index below 0.5 or above 0.9, underwent computed tomography angiography.

Graft offers were evaluated by consensus between transplant surgeons and hepatologists. Graft biopsy was performed selectively when the donor surgeon judged the liver allograft to have a steatotic appearance. In general, only grafts with < 30% macrovesicular steatosis were used.

EAF was defined as graft failure (patient death or need for retransplantation) on POD 90[16,17]. The cause of graft failure and the most likely cause of peak aminotransferase levels > 5000 U/L were independently reviewed by two investigators (Lazzarotto-da-Silva G and Chaves BM). A receiver operating characteristic (ROC) curve was plotted for each EAF predictor to calculate the area under the curve (AUC). Sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) were calculated for each predictor’s best cutoff, as defined by the Youden Index. Survival curves were plotted using the Kaplan-Meier method. The day of LT was considered day zero.

Ethics

This study complied with the ethical standards and was approved by the Institutional Review Board of the Hospital de Clínicas de Porto Alegre. The requirement for informed consent was waived by the ethics committee because of its observational and retrospective design.

Statistical analysis

Continuous data were expressed as mean and SD or median and interquartile range (IQR) and compared using Student’s t-test or the Mann-Whitney U test, as appropriate. Categorical data were expressed as proportions. All comparisons were two-sided with a significance level of 0.05. All analyses were performed using R for MacOS (version 4.0.3; R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Between 2011 and 2024, 341 patients underwent LT. Of these, 29 (8.5%) patients had AST and/or ALT levels exceeding 5000 U/L within the first 48 hours post-LT. Four patients were excluded because of hepatic arterial thrombosis (n = 2) and postoperative portal vein thrombosis (n = 2), leaving 25 patients for the study. The detailed characteristics of the patients are presented in Table 1. Thirteen patients (52%) were male. The median patient age was 54 years (IQR: 45–60 years). The most common etiology of liver disease was hepatitis C virus viral infection (14 patients, 56%). Three patients (12%) underwent retransplantation. The median MELD-Na score was 13 (IQR: 10–17).

Table 1 Baseline characteristics of adult liver transplantation patients with aminotransferase > 5000 U/L, n (%).

Number	25
Age (years), median (IQR)	54 (45-60)
Male sex	13 (52)
Etiology of liver disease
Hepatitis C virus	14 (56)
Hepatitis B virus	2 (8)
Alcohol	2 (8)
Biliary	2 (8)
Acute liver failure	2 (8)
Polycystic liver disease	1 (4)
Metabolic-dysfunction-associated fatty liver disease	1 (4)
Autoimmune hepatitis	1 (4)
Child-Pugh-Turcotte classification
A	9 (36)
B	9 (36)
C	7 (28)
Model for end stage liver disease-Na, median (IQR)	13 (10-17)
Aspartate aminotransferase	8053 (6730-11340)
Alanine aminotransferase	4135 (2980-6414)
Serum factor V	28 (14-46)
Arterial lactate	4.4 (2.6-11.6)

IQR: Interquartile range.

Graft survival

The graft survival rates at one year and five years were 51.7% and 42.6%, respectively (Figure 1A). EAF were also observed in 11 patients. Only 1 of these patients underwent retransplantation within 90 days post-LT and was alive after one year of follow-up. The cause of EAF was primary non-function in 6 patients, hemorrhagic shock due to intraoperative or postoperative bleeding in 3 patients, and abdominal compartment syndrome due to large-for-size graft syndrome in 1 patient. One patient experienced postoperative bleeding and was found to have biliary necrosis and extensive ischemic areas in the liver during laparotomy, in the absence of arterial complications. Median peak AST was higher in patients who suffered EAF in comparison to patients who did not (10390 U/L vs 7246 U/L, P = 0.05). The median peak ALT level was comparable between the groups (4217 U/L vs 3852 U/L, P = 0.256).

Figure 1 Survival analyses of patients with peak aminotransferase > 5000 U/L. A: Graft survival of overall cohort; B: Graft survival of patients who did not suffer graft loss at postoperative day 90. LT: Liver transplantation.

Among the 14 patients who did not develop EAF, one-year and five-year graft survival rates were 92.3% and 76.2%, respectively (Figure 1B). Only 1 patient in this subgroup underwent retransplantation 302 days after LT.

Predictors of EAF

The prognostic yields of the EAF predictors on POD 1 are shown in Table 2. Serum factor V had the highest AUC (0.936). The optimal cut-off value for serum factor V was 26.2%. The sensitivity, specificity, NPV, and PPV of the cut-off were 90%, 92.8%, 92.8%, and 90%, respectively. Arterial lactate levels also displayed a similar AUC (0.919). The optimal cut-off value was 9 mmol/L. The sensitivity, specificity, NPV, and PPV were 81.8%, 92.8%, 86.7%, and 90%, respectively. All eight patients with both serum factor V below 26.2% and lactate levels above 9 mmol/L had EAF (Figure 2). The ROC curves for serum factor V and arterial lactate levels are shown in Figure 3.

Figure 2 Graft survival in patients with serum factor V > 26.2% and arterial lactate < 9 mmol/L on posoprative day 1. A: Serum factor V > 26.2%; B: Arterial lactate < 9 mmol/L. LT: Liver transplantation.

Figure 3 Receiver operating curve of predictors of early allograft failure on postoperative day 1. AUC: Area under the curve.

Table 2 Performance of predictors of early allograft failure on postoperative day 1.

Predictor	Area under the curve	Cutoff	Sensitivity	Specificity	Negative predictive value	Positive predictive value
Factor V	0.936	26.2%	90	92.8	92.8	90
Lactate	0.919	9 mmol/L	81.8	92.8	86.7	90
Bilirubin	0.672	4.4 mg/dL	45	92.8	68.4	83.3
International normalized ratio	0.825	2.7	90.1	64.3	90	66.7
Laboratory model for end stage liver disease	0.870	27	81.8	78.6	84.6	75
Aspartate aminotransferase	0.669	6947 U/L	90.1	57.1	88.9	62.5
Alanine aminotranferase	0.575	7235 U/L	36.4	100	66.7	100

Although statistically higher in patients with EAF, the AUC of the peak AST level was 0.559, which was considerably lower than that of the aforementioned predictors. However, none of the seven patients with an AST peak of < 6800 U/L had EAF. The most likely cause of peak aminotransferase > 5000 U/L in each patient is listed in Table 3, along with peak AST, ALT, factor V, and serum lactate levels.

Table 3 Laboratory parameters, outcome and most likely cause of peak aminotransferase of each patient.

Peak aspartate aminotransferase (U/L)	Peak alanine aminotransferase (U/L)	Factor V POD 1	Lactate POD 1	Early allograft failure	Most likely cause of peak aminotransferase > 5000 U/L
14256	4440	33.9	2.3	No	Long cold ischemia time (540 minutes) and warm ischemia time (56 minutes)
6491	2980	27.3	4.1	No	Arterio-portal fistula secondary to core needle biopsy of the allograft
7763	4479	67.9	2.58	No	Intraoperative and/or postoperative bleeding
8511	5545	14.6	8.9	No	Intraoperative and/or postoperative bleeding
8053	3030	46.3	4.4	No	Unclear/liver with ischemic areas at relaparotomy
5151	1736	35.6	11.5	No	Intraoperative and/or postoperative bleeding
5415	4135	27.6	1.41	No	Unclear
6730	2740	46.3	5.8	No	Long warm ischemia time (60 minutes)
5138	2661	35.8	1.6	No	Biopsy-proven graft steatosis
11340	5188	28.5	1.96	No	Low initial portal inflow
10120	6860	57.4	3.5	No	Right hepatic vein thrombosis causes ischemia in right posterior liver section
6166	3568	66	1.1	No	Long cold ischemia time (645 minutes)
6450	2830	54	1.2	No	Severe postoperative hemodynamic instability
13394	6414	63.2	3.7	No	Abdominal compartment syndrome due to large-for-size-graft
10995	8971	Not applicable	11.1	Yes	Primary nonfunction
7164	4135	10	17.4	Yes	Primary nonfunction
10390	1655	9.9	18.1	Yes	Intraoperative and/or postoperative bleeding
27840	10320	37.5	3.61	Yes	Abdominal compartment syndrome due to large-for-size-graft
12628	8166	17	14.8	Yes	Intraoperative and/or postoperative bleeding
8276	3997	6.8	14.7	Yes	Primary nonfunction
7543	3355	13.3	11.6	Yes	Primary nonfunction
13900	6589	3.7	16	Yes	Primary nonfunction
7857	4217	13	9.2	Yes	Primary nonfunction
7708	2681	25.2	4.1	Yes	Intraoperative and/or postoperative bleeding
15115	7610	14.2	18	Yes	Intraoperative and/or postoperative bleeding

POD: postoperative day.

Fourteen patients had serum factor V > 26.2% on POD 1. Only one those patient had EAF, which occurred on POD 3. One-year and 5-year graft survival rates in patients with factor V levels above the cutoff were 85.1% and 68.8%, respectively. Conversely, 15 patients had arterial lactate levels < 9 mmoL on POD 1. Two of these patients developed EAF on POD 3 and 15. One-year and 5-year graft survival rates were 79.4% and 64.2%, respectively.

DISCUSSION

Elevated aminotransferase levels in the early postoperative period are considered an ominous sign in LT. Most definitions of early graft dysfunction or poor initial graft function include AST and/or ALT in their criteria[3,6,18,19]. Additionally, an aminotransferase level elevation above 5000 U/L has been suggested as an indication for early retransplantation due to the poor prognosis of these patients[9].

In the present study, 25 patients with aminostransferase exceeding 5000 U/L within the first 48 hours after LT in the absence of vascular inflow thrombosis were evaluated for short-term and long-term outcomes. Forty-four percent of the patients died or underwent retransplantation by POD 90, which is consistent with previous reports on graft survival in this population[7,9,12]. Despite the initial poor outcome, our findings indicate that patients who withstand the initial insult and survive beyond POD 90 exhibit 1-year and 5-year graft survival rates similar to those of patients who are not exposed to severe hepatocellular injury. Our findings also showed that serum factor V and arterial lactate levels are adequate tools for predicting EAF in this population.

Given the inferior results of liver retransplantation compared to primary LT[20-22], accurately identifying recipients who are likely to recover graft function, not only to ensure judicious utilization of liver grafts, but also to prevent unnecessary risks to the patient. However, once the inevitable evolution to graft failure is identified, retransplantation should be attempted before irreversible shock and multi-organ failure ensue. Therefore, it is crucial to correctly identify which patients are at an increased risk of EAF and which are likely to recover their graft function as soon as possible. In this study, we explored the performance of several laboratory parameters of liver function on POD 1 in predicting EAF. Serum factor V and arterial lactate levels on POD 1 were the two best predictors of EAF. Two previous publications from our group have shown an association between serum factor V on POD 1, mortality and graft failure[23,24]. Additionally, a high post-LT arterial lactate concentration has been correlated with poor graft survival[25-28], although a recent large cohort study did not corroborate these findings[29].

When confronted with a condition for which the only available effective treatment is a high-risk procedure, such as retransplantation, clinicians should strive to minimize overtreatment. In other words, they should avoid indicating the treatment to patients who do not require it. Therefore, a test with adequate specificity and NPV is necessary to minimize the number of false positives. Importantly, both serum factor V and arterial lactate levels on POD 1 showed a specificity > 90%. Furthermore, patients with serum factor V levels > 26.2% and lactate levels < 9 mmol/L on POD 1 had 1-year graft survival rates of 85% and 79.4%, respectively, which were significantly better than those of liver retransplantation[21,30]. Thus, we suggest that these tests should be incorporated into the decision-making process for patients with aminotransferase levels > 5000 U/L.

In previous publications as well as in the present study, an aminotransferase cutoff of 5000 U/L was set arbitrarily[7,9,12]. In our findings, although peak AST level was not among the best predictors of EAF, no patient with peak AST between 5000 U/L and 6800 U/L had EAF. A similar trend was not observed for ALT levels. Additionally, some patients achieved long-term graft survival despite a peak AST over 10000 U/L. Therefore, our findings suggest that although aminotransferase above 5000 U/L in the early posttransplant period should be regarded as a “red flag”, the elevation per se is insufficient to determine which patients will suffer graft loss. The decision on whether a patient needs to undergo retransplantation should rely on liver function parameters instead of hepatocellular injury findings.

Heterogeneity in the cause of elevated aminotransferase levels may be considered a confounding factor in our study. In at least seven patients, peak aminotransferase levels > 5000 U/L were attributed to either intraoperative or postoperative bleeding. In these cases, excessive bleeding was likely a manifestation of coagulopathy that could be ascribed to graft dysfunction or delayed graft function, both of which are closely related to the degree of graft injury. Thus, we believe that the cause of aminotransferase elevation in most cases was ultimately graft injury in some of the following phases: (1) Pre-preservation (e.g., steatosis); (2) Preservation (e.g., cold-preservation injury and rewarming injury); and (3) Reperfusion (e.g., ischemia-reperfusion injury).

Our study had some limitations. First, this was a retrospective study with a small sample size. Second, we did not have data on whether the patients received fresh frozen plasma, which could potentially influence serum factor V levels. However, in a previous study by our group, patients with low serum factor V levels received more fresh frozen plasma intraoperatively, demonstrating that the transfusion of plasma clotting factors had little impact on the interpretation of serum factor V results[24]. Third, as few centers use serum factor V in the post-transplant period, extrapolating our results to other transplant services might be difficult. Ideally, the results of this study should be evaluated in a large prospective multicenter study.

CONCLUSION

In conclusion, aminotransferase levels exceeding 5000 U/L in LT recipients was associated with a high rate of graft loss within 90 days. However, patients demonstrating adequate graft function on POD 1 exhibited long-term graft survival comparable to general LT recipients, despite severe hepatocellular injury. Both serum factor V and arterial lactate concentrations appear to be effective indicators for evaluating prognosis in the early phase of LT.