Intraoperative Hypotension during Liver Transplant Surgery is Associated with Postoperative Acute Kidney: A Retrospective Cohort Study

BACKGROUND: Acute kidney injury (AKI) occurs frequently after liver transplant surgery and is associated with signicant morbidity and mortality. While the impact of intraoperative hypotension (IOH) on postoperative AKI has been well demonstrated in patients undergoing a wide variety of non-cardiac surgeries, it remains poorly studied in liver transplant surgery. We tested the hypothesis that IOH is associated with AKI following liver transplant surgery. METHODS: This historical cohort study included all consecutive patients who underwent liver transplant surgery between 2014 and 2019 except those with a preoperative creatinine > 1.5 mg/dl and/or who had combined transplantation surgery. IOH was dened as any mean arterial pressure (MAP) < 65 mmHg and was classied according to the percentage of case time during which the MAP was < 65 mmHg into three groups, based on the interquartile range of the study cohort: “short” (Quartile 1, < 8.6% of case time), “intermediate” (Quartiles 2-3, 8.6-39.5%) and “long” (Quartile 4, > 39.5%) duration. AKI stages were classied according to a “modied” “Kidney Disease: Improving Global Outcomes” (KDIGO) criteria. Logistic regression modelling was conducted to assess the association between IOH and postoperative AKI. The model was run both as a univariate and with multiple perioperative covariates to test for robustness to confounders. RESULTS: Of the 205 patients who met our inclusion criteria, 117 (57.1%) developed AKI. Fifty-two (25%), 102 (50%) and 51 (25%) patients had short, intermediate and long duration of IOH respectively. In multivariate analysis, IOH was independently associated with an increased risk of AKI (adjusted odds ratio [OR] 1.05; 95%CI 1.02-1.09; P < 0.001). Compared to “short duration” of IOH, “intermediate duration” was associated with a 10-fold increased risk of developing AKI (OR 9.7; 95%CI 4.1-22.7; P < 0.001).

CONCLUSION: Intraoperative hypotension is independently associated with the development of AKI after liver transplant surgery. The longer the MAP stays < 65 mmHg, the higher the risk the patient will develop AKI in the immediate postoperative period, and the greater the likely severity.
Trial Registration: Not Applicable Background Acute kidney injury (AKI) is a common postoperative complication following liver transplantation and is associated with increased morbidity, mortality and development of chronic kidney disease. [1][2][3][4][5] One of the most common diagnostic criteria used to classify AKI is the "Kidney Disease: Improving Global Outcomes" (KDIGO) system, which is based on changes in serum creatinine and urine output. [6] However, as urine output is rarely documented accurately in the perioperative setting, increases in serum creatinine are frequently used independently to de ne postoperative AKI ("modi ed" KDIGO classi cation).
Multiple studies have identi ed patient and donor risk factors for AKI following liver transplant surgery including among others, female sex, obesity, diabetes, high model for end-stage liver disease (MELD) score, large amounts of blood loss, use of hydroxyethyl starch solution, perioperative blood glucose variability, cold and warm ischaemia times, donor age and graft sizes. [7][8][9][10][11][12] Haemodynamic variable such as intraoperative hypotension (IOH), most often de ned as a mean arterial pressure (MAP) ≤ 65 mmHg, has been shown to be one of the most important factors associated with postoperative AKI. [13] Numerous large retrospective studies have shown that IOH is associated with postoperative AKI after various types of non-cardiac surgery, [14][15][16][17][18][19] but data on such an association in liver transplantation remain scarce. [20] We therefore conducted a historical cohort analysis to evaluate the association between IOH and the development of postoperative AKI in patients undergoing liver transplant surgery.

Methods
This single centre historical cohort study was approved by our Institutional Review Board on December 14, 2018 under the reference P2018/555 with a waiver of informed consent because of the observational and retrospective nature of the study.
We identi ed all liver transplant patients from 2014 (when the anaesthetic data for our patients started to be computerised) to 2019 with our dedicated operating room softwares (TrackPro® and UltraGenda®, Belgium). We then retrospectively analysed the patients' electronic medical records, which include a continuous intraoperative recording of vital signs (Innovian® Perioperative Care, Dräger, Lübeck, Germany). All patients who underwent a liver transplant between January 1, 2014 and December 30, 2019 were included except those. With a preoperative serum creatinine value > 1.5 mg/dL, and any patient who underwent a combined transplantation procedure (liver-kidney, liver-heart or liver-lung).

Anaesthetic protocol
Intraoperative anaesthesia was standardised according to institutional guidelines. Patients arrived in the operating room and were placed under an infrared heating lamp. Several non-invasive monitors were then applied: a 5-lead electrocardiogram (ECG), non-invasive blood pressure, rectal temperature probe, and a frontal electroencephalogram using bispectral index (BIS) monitoring (Aspect Medical System Inc, Natick, MA, USA). A bladder catheter was inserted. 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 not cannulated in case veno-venous bypass was required. A Swan-Ganz catheter (Edwards Lifesciences, Irvine, CA, USA) was inserted and use of haemodynamic agents was guided using continuous cardiac index, mixed venous oxygen saturation, central venous pressure, and arterial pressure. Rapid infusers, perfusion heaters, and a cell saver were ready for use prior to induction. In case of active haemorrhage, anaesthetists typically guided blood product administration using ROTEM monitoring. General anaesthesia was induced with propofol or etomidate. Antinociception was achieved with a remifentanil infusion and anaesthesia was maintained with sevo urane or des urane depending on physician preference. Rapid sequence intubation was performed if patients had not fasted appropriately or if they had abdominal ascites. Neuromuscular blockade was achieved in all patients and controlled with a train-of-four monitor (TOF scan, Idmed, France). The choice of muscular relaxant was left to the discretion of the anaesthetist. Fluid administration consisted of a baseline infusion of balanced crystalloid infusion (Plasmalyte®, Baxter, Belgium) and compensation for blood loss with either Plasmalyte®, 3% modi ed gelatin, or 4% albumin (depending on patient conditions and physician preference).

Surgical procedure
Almost all the liver transplantation were performed by recipient hepatectomy without venous-venous bypass, using the vena cava-sparing technique and piggy-back reconstruction. Liver reperfusion was performed through the portal vein rst followed by subsequent arterial reperfusion. Biliary reconstruction was carried out with an end-to-end choledochocholedochostomy without a T-tube.
Our immunosuppressive regimen comprised primarily tacrolimus with mycophenolate mofetil and prednisone. Tacrolimus trough levels were maintained at 5-10 ng/mL. Steroids were discontinued approximately 3 months after liver transplant surgery.
Measurements and study outcomes MAP was recorded automatically during surgery at 30 second intervals by our anaesthesia information management system (Innovian). We extracted the raw values: all values < 30 and > 150 mmHg were considered to be artifacts and deleted. For each patient, we calculated the mean MAP value during the procedure and the percentage of case time during which the patient was hypotensive, de ned as a MAP < 65 mmHg. IOH was then categorised into 3 levels based on the interquartile range (IQR) values of the study cohort for the percentage of case time during which patients were hypotensive, according to the methodology of Thacker et al [21] : "short duration" of IOH (in the lower 25th percentile), "intermediate duration" (between the 25th and the 75th percentile) and "long duration" (within in the upper 75th percentile).
The primary outcome was the development of stage 1-3 AKI, de ned using serum creatinine-based KDIGO de nitions without taking into account diuresis ("modi ed" KDIGO classi cations) because urine output is rarely documented accurately in the perioperative setting. The three modi ed KDIGO stages are: 1) Mild injury: creatinine increase of at least 0.3 mg/dl within the rst 48-hours or 1.5 to 1.9 times the baseline level during the rst postoperative week; 2) Moderate injury: creatinine increase of 2.0 to 3.0 times the baseline; and 3) Severe injury: creatinine increase of greater than 3.0 times the baseline, creatinine level of at least 4 mg/dl, or dependency on renal replacement therapy.

Statistical analysis
The normality of continuous data was assessed using a Kolmogorov-Smirnov test. Normally distributed variables were compared using a student's t-test and are expressed as mean ± standard deviation (SD) and those not normally distributed were compared using a Mann-Whitney U-test and are expressed as median [25% − 75%] percentiles. Discrete data were expressed as a number and percentage and compared using a Chi square or a Fisher's exact test when indicated.
We used logistic regression modelling to evaluate the association between IOH and the development of postoperative AKI. Univariate logistic models were used to test for association with AKI using the following independent variables: sex, age, ASA class, weight, body mass index (BMI), Child-Pugh score, baseline serum creatinine and haemoglobin, MELD laboratory score, duration of anaesthesia, duration of surgery, uid volumes (crystalloid, colloid, packed red blood cells, cell saver), estimated blood loss, diuresis, total uid output, net uid balance, use of vasopressors, mean case time with central venous pressure > 8 mmHg, preoperative use of different medications (Table 1), patient comorbidities (Table 1), donor age, donor BMI, postoperative uid balance, use of cardiopulmonary bypass, presence of portal ischaemia or arterial ischaemia and any episodes of MAP < 65 mmHg. Variables signi cantly associated in univariate testing were then included in a multivariate logistic regression to evaluate their association with AKI. Risks of developing AKI based on the model are presented as odds ratios [ORs] and their 95% con dence intervals. Statistical signi cance was determined at the 0.05 level. All analyses were conducted with Minitab (Paris, France) and R (www.r-project.org).

Results
Among the 242 patients who underwent a liver transplantation between January 1st 2014 and December 30th 2019, 205 patients met our inclusion criteria (Fig. 1).
One hundred and seventeen patients (57%) experienced some type of postoperative AKI (stages 1-3  "Total IN" is the sum of crystalloid, colloid, packed red blood cells and cell saver administration and "total OUT" is the sum of estimated blood loss and urine output. Net uid balance is the difference total IN -total OUT. Data are expressed as mean ± standard deviation, median and [25th -75th ] percentiles or number and percentage (%) In univariate testing (Tables 1 and 2), patients who developed postoperative AKI had higher BMI (p = 0.041), were more likely to have received preoperative diuretics (p = 0.039), had higher Child-Pugh (p = 0.0022) and MELD (p = 0.014) scores, had lower preoperative haemoglobin levels (p = 0.014), were more likely to have had prolonged IOH (p < 0.001), portal ischaemia (p = 0.047), or packed red blood cell transfusion (p = 0.013), and had higher total uid input (p = 0.015), estimated blood loss (p = 0.006), and total uid output (p = 0.011) than patients who did not develop AKI.
In multivariable analysis using the perioperative variables shown in Tables 1 and 2, only BMI and IOH (OR = 1.05 [1.02-1.09], p < 0.001), were signi cantly associated with an increased risk of AKI. For every one percent increase in case time spent with a MAP of ≤ 65 mmHg, the risk of AKI increased by about 5%.
Compared to "short duration" IOH, "intermediate duration" IOH was associated with a 10-fold increased risk of developing AKI (OR of 9.7; 95% CI 4.1-22.7; P < 0.0001). "Long duration" IOH was associated with an even greater risk of postoperative AKI (OR 34.6; 95% CI 11.5-108.6; P < 0.0001). Figure 2 shows the three different durations of IOH and their associations with the development of postoperative AKI. This suggests that the observed association between IOH and AKI was a dose-response relationship.

Discussion
The presence of IOH was associated with an increased risk of developing postoperative AKI after liver transplantation and this association was independent of potential perioperative confounders. Moreover, the longer a patient spent with a MAP < 65 mmHg during the liver transplantation procedure, the greater the risk he or she had of developing AKI in the immediate postoperative period. These ndings con rm that IOH is of real clinical importance and should not be overlooked during the intraoperative period.
Multiple large retrospective studies have shown an association between IOH and postoperative AKI, [13][14][15][16][17][18][19][20]22] and others have reported an association between the duration of IOH and cardiac, renal and neurological adverse events. [13,17,23,24] However, this association remains poorly de ned in the context of liver transplantation. [20] To our knowledge, only one study has assessed the relationship between IOH and the risk of AKI in this patient population. [20] In that study, the authors demonstrated that severe IOH, de ned as a MAP < 50 mmHg was strongly related to the development of moderate and severe AKI (stage 2-3). Patients undergoing liver transplant surgery frequently experience IOH as a result of various factors, including, among others, the duration of surgery, the severity of bleeding, the severity of the ischaemic reperfusion syndrome and the severity of the end-stage liver disease, characterised by a hyperdynamic state (high cardiac output and low systemic vascular resistance). However, most studies, that have assessed predisposing factors for AKI after liver transplant surgery, focused mainly on preoperative factors, which are often not modi able. Perioperative risk factors, such as IOH, are, in contrast, potentially modi able, and may be minimised by close a collaboration between the surgeon and the anaesthetist. Our results suggest that avoiding or at least minimising the duration of IOH may be a valuable target to reduce the development of postoperative AKI.
Importantly, our hospital has no any strict MAP targets for liver transplant surgery (except to avoid a MAP < 65 mmHg) and MAP management is left to the discretion of the anaesthetist in charge of the patient.
Two large randomised controlled trials have demonstrated that targeting a higher arterial pressure during surgery (well above 65 mmHg) was associated with a lower incidence of postoperative AKI. [25,26] In the rst, there was a lower incidence of organ dysfunction in the group of patients managed using a targeted systolic arterial pressure closer to the patient's baseline value compared to the control group in which the same blood pressure target was used for all patients. [25] In the second study, targeting a MAP level between 80-95 mmHg in chronically hypertensive patients reduced the occurrence of postoperative AKI compared to two other MAP targets (65-79 and 96-110 mmHg). [26] French national guidelines recommend maintaining of MAP > 70 mmHg in patients with chronic hypertension (which is the case in 60% of our study cohort) in order to prevent AKI. [27] It naturally follows that targeting a strict MAP goal of 65 mmHg can potentially be awed as a strict de nition of IOH is quite challenging. While some authors use a reduction from baseline value" (e.g. a 20-30% reduction from the patient's preoperative MAP value), others continue to use the well-known "absolute" threshold value of 65 mmHg to de ne IOH. We decided in this study to choose the latter as this is the most common practice at our institution. The validity of this threshold can of course be challenged, but Salmasi and colleagues demonstrated that management based on an absolute MAP threshold of 65 mmHg in all patients was equivalent to management targeting relative reductions in MAP from preoperative values in terms of incidence of myocardial and kidney injury. [13] Additionally, although the results of a large randomised controlled study supported the individualization of arterial pressure targets in order to reduce the incidence of organ dysfunction (including a reduction in AKI), [25,28] it is important to remember that such an approach can be extremely challenging to apply in patients undergoing liver transplant surgery, as higher values may potentially increase bleeding, making surgical conditions more challenging. As always, the risk-bene t ratio should be carefully assessed and future investigation into an optimal de nition of IOH is urgently required for liver transplant recipients.
This study has several additional limitations that should be taken into consideration when interpreting our ndings. Firstly, it was observational, retrospective, single-centre and included a relatively small sample size. Therefore, a causal relationship cannot be established and our results may not be generalisable to other hospitals with different perioperative haemodynamic and anaesthetic management. Secondly, our ndings may be biased by unmeasured confounding parameters at both the patient and hospital levels. Thirdly, as urine output was not taken into account for the classi cation of AKI, this may have led to a slight "underestimation" of the incidence of postoperative AKI in our study cohort. Fourthly, per KDIGO de nitions, we de ned AKI as the change in creatinine value between the preoperative value and the highest value during the rst postoperative week. This might introduce timevarying confounding or mediating factors, which limit interpretation of the study nding. Fifthly, postoperative hypotension was not taken into account as MAP was less frequently measured in the intensive care unit or on the oor than in the operating room. Sixthly, although all patients had a pulmonary catheter, data on mixed venous oxygen saturation (SvO 2 ) and cardiac index were not linked to our electronic medical records and thus, could not be assessed in the present study. However, it is important to note that a recent manuscript demonstrated that decreased SvO 2 was associated with postoperative AKI after liver transplantation. [29] Seventhly, we had no data on the occurrence of postreperfusion syndrome and its importance on IOH duration. Finally, it is important to note that we reported the odds ratio for a frequent outcome (AKI), and the odds ratio can overestimate the risk in this situation.

Conclusions
Our ndings indicate that IOH is independently associated with the development of AKI after liver transplant surgery. The longer the MAP stays < 65 mmHg, the higher the risk the patient will develop AKI in the immediate postoperative period, and the greater the likely severity. Avoidance of IOH during liver transplant surgery may thus help reduce the incidence of this severe postoperative complication.
Prospective studies are needed to assess whether targeting a higher MAP during this complex surgical procedure can reduce the risk of postoperative AKI. VdL: Statistical analysis of the data and edited the nal manuscript.