Arterial Lactate Concentration At The End of Liver Transplantation is Independently Associated with One-Year Mortality

BACKGROUND Liver transplant patients who develop hyperlactatemia are at increased risk of postoperative morbidity and mortality, but there are few data on longer-term outcomes. We therefore investigated whether arterial lactate concentration obtained immediately after surgery, at the time of admission to the intensive care unit (ICU), was associated with 1-year mortality. METHODS: In this retrospective study, all Patients who underwent combined transplantation surgery and those (i.e., redo surgery) included. Logistic regression modeling included univariate and multivariate analyses. Receiver operating characteristic (ROC) curves and areas under the curves (AUROCs) were calculated. Lactate thresholds and association with outcome were analyzed for specicity, sensitivity, and Youden’s index. RESULTS: Of 226 patients included, 18.4% died within 1-year of liver transplantation. Immediate postoperative lactate concentration was independently associated with 1-year mortality with an odds ratio (OR) of 1.35 (95% CI: 1.16 to 1.59; p<0.001) per mEq/L increase in lactate and an AUROC of 0.80 (95% CI: 0.72 to 0.87; p<0.001). A lactate concentration of 2.25 mEq/L (cut-off determined using Youden’s index) was associated with increased 1-year mortality with a sensitivity of 0.71 and a specicity of 0.72. CONCLUSION: Increased arterial lactate concentration on admission to the ICU immediately after orthotopic liver transplantation is independently associated with increased 1-year mortality. as tested by the Shapiro-Wilk test. Results for counts are reported as number (% of group). Baseline, intraoperative, and outcome data were rst compared groupwise with a Student’s t-test or a Mann-Whitney U for normal and skewed data, respectively, or a chi-square test for counts.

RESULTS: Of 226 patients included, 18.4% died within 1-year of liver transplantation. Immediate postoperative lactate concentration was independently associated with 1-year mortality with an odds ratio (OR) of 1.35 (95% CI: 1.16 to 1.59; p<0.001) per mEq/L increase in lactate and an AUROC of 0.80 (95% CI: 0.72 to 0.87; p<0.001). A lactate concentration of 2.25 mEq/L (cut-off determined using Youden's index) was associated with increased 1-year mortality with a sensitivity of 0.71 and a speci city of 0.72. CONCLUSION: Increased arterial lactate concentration on admission to the ICU immediately after orthotopic liver transplantation is independently associated with increased 1-year mortality.
Trial Registration: Not Applicable Background Blood lactate concentrations re ect an imbalance between systemic oxygen delivery and tissue oxygen demand, which results in anaerobic metabolism. [1] Because about 60% of lactate is cleared by the liver, blood lactate concentrations also provide an indication of hepatic blood ow and function. [2] Raised blood lactate concentrations may therefore re ect intra-operative tissue dysoxia and graft dysfunction in liver transplant patients. [3][4][5][6] Rarer causes of hyperlactatemia include enzymatic (e.g., pyruvate dehydrogenase) alterations, pharmacologic effects (e.g., metformin), and hypothermia. [1] Hyperlactatemia is often associated with metabolic acidosis, which can alter a patient's hemostasis and hemodynamics, worsen hemorrhage, exacerbate postoperative complications, and ultimately increase mortality. [1,7] Normal arterial lactate concentrations are around 1 mEq/L and values greater than 1.5 mEq/L have been associated with increased mortality in various studies of surgical and critically ill patients.[8-10] Blood lactate levels may therefore represent a useful predictor of complications and outcomes in liver transplant patients. [10] Although the presence of hyperlactatemia at the end of liver transplantation has been shown to be associated with postoperative mortality within the rst [11] and third [4] months after liver transplantation, there are few data on its relationship with long-term mortality. The primary objective of this cohort study was therefore to test the hypothesis that increased lactate concentration on after liver transplantation is associated with increased 1-year mortality.

Methods
The Erasme Hospital Institutional Review Board approved this monocenter study on December 14, 2018 (reference P2018/555), and waived the need for patient informed consent because of the study's retrospective nature. Using our anesthetic electronic records software (Innovian, Drager, Germany), we identi ed all patients who had had an orthotopic liver transplant between September 2013, date of implementation of the software, and December 2019. We reviewed the electronic medical records (MediView, IMMJ Systems, United Kingdom) of all these patients and included those for whom operating room planning dedicated software (TrackPro, UltraGenda, Belgium) had been used. Exclusion criteria included incomplete records, combined surgeries (e.g., simultaneous lung and liver transplant), and previous liver transplantation.

Anesthesia protocol
Anesthesia was standardized according to our institutional protocol. After arrival in the operating room, patients and were placed under an infrared heating lamp. Standard non-invasive monitoring consisted of a 5-lead EKG, non-invasive blood pressure, temperature, frontal processed electroencephalogram, diuresis, and neuromuscular blockade monitoring. Vascular access consisted of one or two large bore peripheral venous catheters, right femoral artery and vein catheters, and right jugular vein catheter. The left femoral and internal jugular veins were left at the surgeon's disposal for possible veno-venous bypass. A pulmonary artery (Swan-Ganz) catheter was inserted, and hemodynamic management was guided using calibrated cardiac index, central venous oxygen saturation, inferior vena cava pressure, central venous pressure, and arterial pressure. Hemodynamic targets were left to the discretion of the attending anesthesiologist. Rapid infusers, perfusion heaters, and an intraoperative blood product salvage system were ready for use prior to induction. In case of massive bleeding, it was recommended that the anesthesiologist should use ROTEM technology to guide blood product administration. Anesthesia was induced with propofol or etomidate and neuromuscular blockade with rocuronium or succinylcholine, depending on the patient's history, and maintained with cisatracurium. Anti-nociception was maintained with remifentanil and unconsciousness with sevo urane or des urane. Rapid sequence intubation was performed if patients had not fasted or had abdominal ascites.

Surgical procedure
The standard approach for orthotopic liver transplantation was recipient hepatectomy with a vena cavasparing technique, piggy-back reconstruction, and no venous-venous bypass. Liver reperfusion began with the portal vein and then arterial reperfusion. End-to-end choledochectomy without a T-tube assured Results A total of 276 liver transplant procedures were performed during the study period; 28 patients had incomplete anesthesia medical records, 10 had had combined surgery, and 10 had had a previous liver transplant. Accordingly, 228 patients were included in the analysis (Fig. 1). Forty-two patients (18.6%) patients died within the year after transplantation The American Society of Anesthesiology (ASA) score, Child-Pugh score, Model for End-stage Liver Disease (MELD) score, duration of anesthesia, duration of surgery, crystalloid volume, blood product transfusion, platelet transfusion, total infused volume, uid balance, portal ischemia time, mean end tidal CO 2 , and immediate postoperative lactate concentration were associated with 1-year mortality in biliary reconstruction. Portal ischemia (i.e., cold ischemia) time was de ned as the time from donor aortic clamping to portal unclamping of the recipient. Arterial ischemia (i.e., warm ischemia) was de ned as the time from recipient portal unclamping to hepatic artery unclamping. Immunosuppression consisted of tacrolimus with mycophenolate mofetil and prednisone. Tacrolimus levels were targeted at 5-10 ng/mL and steroids were maintained for the rst 3 months post liver transplantation.
At our institution all patients are transferred to the ICU immediately after liver transplantation.

Data collection and outcomes
Preoperative demographics, intraoperative and postoperative (at patient arrival in the ICU) arterial lactate concentrations, postoperative major and minor complications, 30-day mortality, and 1-year mortality were collected from the medical records for each patient.
The primary objective was to test the hypothesis that postoperative lactate concentration at ICU admission was associated with long-term mortality (i.e., death within one year of transplantation).
Secondary objectives included the association between lactate concentration and 30-day mortality and between lactate and major postoperative complications (de ned in Appendix 1).

Statistical Analysis
Data are reported as mean ± standard deviation or median [25th -75th percentiles] depending on the normality of the distribution as tested by the Shapiro-Wilk test. Results for counts are reported as number (% of group). Baseline, intraoperative, and outcome data were rst compared groupwise with a Student's t-test or a Mann-Whitney U for normal and skewed data, respectively, or a chi-square test for counts. Univariate testing identi ed potential contributors to mortality. All variables with a p value < 0.10 were introduced into a multivariate analysis and a backwards feature selection process removed variables with a p value > 0.05 in the model.
Receiver operating characteristic (ROC) curves were established for 1-year mortality, 30-day mortality, and postoperative complications; 95% con dence intervals were calculated using the DeLong method. Lactate concentration thresholds and the mortality or morbidity risk were calculated, with 2000 bootstrapped Monte-Carlo samples for speci city, sensitivity, and Youden's Index univariate analysis (Tables 1 and 2). Postoperative events associated with 1-year mortality were: sepsis, postoperative bleeding requiring transfusions, atrial brillation, encephalopathy, acute respiratory distress syndrome, reoperation, renal replacement therapy, and acute kidney injury (stages II and III) ( Table 3).    (Fig. 2), and a lactate concentration of 2.25 mEql/L (cut-off determined using Youden's index) was associated with 1-year mortality with a sensitivity of 0.71 and a speci city of

Discussion
Even a moderate increase in blood lactate concentration early after orthotopic liver transplantation was associated with increased long-term mortality. Previous studies have reported an association between postoperative hyperlactatemia and mortality after liver transplantation, [4,11,12] but none evaluated longterm mortality. Our study is consistent with the prior observations after liver transplantation [4] liver surgery, [12] major abdominal surgery, [13] cardiac surgery [14][15][16], and intensive care admission. [9] There are several potential causes of increased lactate concentration at the end of liver transplant surgery. Liver transplant patients can develop tissue dysoxia and graft failure, both of which have been associated with increased mortality during major abdominal surgery and after liver transplantation. [5,6] However, the incidence of graft failure in our cohort was similar in survivors and non-survivors, suggesting that tissue hypoperfusion was the predominant cause of the higher lactate concentrations in the non-survivors. Other factors contributing to hyperlactatemia could be compromised hemodynamic status during complex surgery, re ected by increased surgical duration and portal ischemia. Hemodynamic alterations can lead to tissue hypoxia with subsequent development of anaerobic metabolism [17] and followed by multiple organ dysfunction.
[18] The signi cantly higher rates of organ failure, including encephalopathy, acute respiratory distress syndrome, kidney injury, and sepsis in patients who did not survive the rst year after transplantation compared to those who did, further suggests that intraoperative tissue dysoxia contributed to the increased lactate concentrations. Patients who died within one year of transplantation had higher preoperative risk (i.e., higher ASA, MELD, and Child-Pugh scores). These scores re ect patient frailty and are linked to hepatic and other organ dysfunctions, which increase the burden of perioperative patient care. [19][20][21] The ASA score, for example, has been consistently linked with increased postoperative complications and mortality. [21] Our results con rm this nding, as the ASA score was independently associated with postoperative mortality.
Several other factors were independently associated with postoperative mortality, including duration of anesthesia and infused crystalloid volume. Excessive uid administration has been linked to increased postoperative complications either because this is an indicator of hemodynamic instability or because of an intrinsic effect of uid. [22][23][24] These factors attest to the hemodynamic and anesthetic challenges of these patients and strengthen the association between unstable hemodynamics, decreased perfusion, tissue hypoxia, increased lactate, and postoperative mortality. Increased postoperative lactate concentration was also associated with major complications, with good speci city. In addition to tissue hypoxia, other causes of postoperative morbidity after liver transplant, include coagulopathy and biliary complications,) [25,26] for which increased lactate concentrations may have lower sensitivity.
Liver transplant patients with postoperative hyperlactatemia should be carefully investigated to determine the cause of hyperlactatemia and administer effective treatment. A perioperative goal-directed hemodynamic strategy may help to prevent intraoperative tissue hypoxia and its associated postoperative complications. [27][28][29] Maintaining optimal hemodynamic goals (e.g.,preload, arterial pressure, and cardiac output), is essential, but clinicians must still recognize that hyperlactatemia may be due to other causes during liver surgery. [12] Regardless of its etiology, hyperlactatemia is linked to increased short-and long-term mortality and its presence should promote close observation and prophylactic measures to prevent postoperative complications and death.

Limitations
This study had several limitations. First, it was a single center, retrospective study, with a relatively small number of patients. Second, although a single arterial lactate concentration was strongly associated with 30-day and 1-year mortality, evaluating the time course of blood lactate levels may give even more information; this should be further studied. Third, our study did not include donor data, which could have an impact on survival. Although invasive hemodynamic monitors were used, future studies should assess the impact of goal-directed hemodynamic strategies on postoperative morbidity and mortality in liver transplant patients. Our study, however, re ects standard of care in an academic center with expertise in liver transplantation anesthesia.

Conclusions
In patients who underwent orthotopic liver transplantation, an increased lactate concentration immediately after surgery was independently associated with increased 1-year mortality. OD is a consultant for Medtronic (Dublin, Ireland) The other authors have no con icts of interest related to this article