Glucocorticoid use and ischemia‐reperfusion injury in laparoscopic liver resection: Randomized controlled trial

Abstract Aim Laparoscopic liver resection (LLR) is increasingly carried out worldwide. However, there are concerns regarding ischemia‐reperfusion injury caused by pneumoperitoneum and the Pringle maneuver. It is not clear whether perioperative use of glucocorticoids lowers the risk of ischemia‐reperfusion hepatic injury in LLR as has been reported for open liver resection. The aim of the present study was to investigate the role of perioperative glucocorticoid use in improving hepatic function and surgical outcomes after LLR. Methods In this double‐blind, randomized controlled trial (UMIN000013823), we enrolled 130 patients who presented to our institution for LLR between April 2014 and October 2018. Six patients were excluded, resulting in 124 patients being randomized to either the glucocorticoid or the control group. Preoperatively, patients in the glucocorticoid group received 500 mg methylprednisolone in saline solution, patients in the control group saline solution only. Surgical outcomes and blood parameters were compared between the two groups. Results The Pringle maneuver could not be carried out in 24 patients, resulting in 50 patients in each group being included in the analysis. Postoperatively, total, direct and indirect bilirubin, and C‐reactive protein and interleukin‐6 levels were significantly lower, albumin levels were significantly higher, and prothrombin time was significantly shorter in the glucocorticoid than in the control group. Surgical outcomes were not significantly different between the groups. Conclusion This first report on preoperative glucocorticoid use in LLR showed that it significantly improved postoperative liver function and thus might enhance the safety of LLR.


| INTRODUC TI ON
Laparoscopic liver resection (LLR) is increasingly carried out worldwide because of recent technical advancements and clinical evidence of better short-term and similar long-term outcomes compared to open liver resection (OLR). [1][2][3][4][5][6] However, several studies have reported that pneumoperitoneum in laparoscopic surgery causes hepatic ischemia-reperfusion injury as a result of the temporary decrease in portal venous blood flow. [7][8][9] As LLR is a relatively new technique, reports on the prevention of hepatic injury during the procedure are scarce.
One of the major determinants of complications after hepatectomy is the extent of intraoperative bleeding. To reduce bleeding, the Pringle maneuver, which involves intermittent vascular clamping of the hepatic hilum, is commonly carried out. However, the resulting temporary hepatic ischemia and subsequent reperfusion leads to activation of complex metabolic, immunological, microvascular, and inflammatory processes that culminate in hepatocellular injury. [10][11][12] This could be further aggravated by additional liver injury caused by pneumoperitoneum during LLR.
For OLR, several studies have shown that perioperative use of glucocorticoids may decrease the cytokine response to the ischemia-reperfusion sequence and thus improve hepatic function and surgical outcomes. [13][14][15][16][17] However, the effect of perioperative glucocorticoid use has not yet been evaluated for LLR. We hypothesized that giving perioperative glucocorticoid would improve postoperative hepatic function and morbidity in LLR. Therefore, our aim in the present study was to investigate the impact of glucocorticoid on hepatic function and surgical outcomes of LLR.

| Participants
Between April 2014 and October 2018, 130 patients who presented to our institution for LLR were enrolled into the trial. The CONSORT flow diagram is shown in Figure 1. Inclusion criteria were: (i) planned pure LLR using the intermittent Pringle maneuver; (ii) Eastern Cooperative Oncology Group performance status 0 or 1; (iii) age ≥20 years; (iv) no pregnancy. Exclusion criteria were: (i) active infection; (ii) uncontrolled diabetes mellitus; and (iii) refusal to participate.

| Randomization
The remaining 124 patients were randomly allocated to either the glucocorticoid (n = 62) or the control (n = 62) group using a 1:1 ratio. Group assignment was done by a statistician who was independent from our study using the minimization method program code in Microsoft Excel and Visual Basic for Applications (Microsoft Corporation). Patients who were allocated to the glucocorticoid group received 500 mg methylprednisolone dissolved in saline solution at the time of anesthesia induction. Patients in the control group received the saline solution only. Anesthesiologists were blinded to the study group to which patients belonged. Based on the random assignment by the statistician, another physician prescribed and prepared the infusion solution and handed it to the anesthesiologist who then gave it to the patient.

| Baseline characteristics
Patient characteristics are reported in Table 1. No significant differences were identified between the two study groups.

| Study endpoints
Primary endpoint was postoperative total serum bilirubin level. .620 Interleukin-6 (pg/dL) 2.5 (1. TA B L E 1 Baseline characteristics of 100 patients undergoing laparoscopic liver resection follow-up blood analyses were carried out once between the third and the fifth postoperative week.

| Sample size
Based on our institutional data on LLR, we expected a standard deviation of the maximum total bilirubin level of 0.5 mg/dL. Sample size was calculated to detect a difference in total bilirubin level of 0.3 mg/dL, 15 with type I error set at 0.05 (two-sided), power at 0.80, and allocation ratio at 1:1. With these parameters, a sample size of 90 patients was required. Considering a rate of loss to follow up (including an impossible or unnecessary Pringle maneuver) of 30%, 130 patients were set as the target for enrolment.

| Statistical analysis
Data were collected for all patients by one surgeon and confirmed by another surgeon to ensure the study protocol was followed.

| RE SULTS
In the control group, 12 patients were excluded from analysis after surgery because the Pringle maneuver could either not be carried out (n = 9) or was not necessary (n = 3). Similarly, in the glucocorticoid Hospital stay (days) 9 (7-14) 9 (7)(8)(9)(10)(11)(12)(13) .615 Surgical outcomes are reported in Table 2. No significant differences were identified in these parameters between the two groups including the rate of PHLF (4.0% in both groups). However, the rate of morbidity and the CCI scores tended to be lower in the glucocorticoid group than in the control group. Median, 75th percentile, and 90th percentile levels of CCI were 0, 0, and 25.8 in the glucocorticoid group; and 0, 14.4, and 36.4 in the control group, respectively (P = .080).
Time course of total, direct, and indirect bilirubin levels before and after surgery is shown in Figure 2. Total, direct, and indirect bilirubin levels on the second postoperative day were significantly lower in the glucocorticoid than in the control group. Time course of other blood analyses is shown in Figure 3. The lowest level of albumin (3.2 g/dL vs 2.9 g/dL, P = .0002) was significantly higher, the

| D ISCUSS I ON
This is the first randomized controlled trial to report the impact of giving preoperative glucocorticoid on liver function and surgical outcomes after LLR. Previous studies reported that glucocorticoid use in OLR had a positive effect on postoperative complications and blood parameters. [13][14][15][16][17] In a meta-analysis of glucocorticoid use for OLR, methylprednisolone was used in five studies and hydrocortisone in one study. 14 All six studies showed positive impacts of giving glucocorticoid. We chose methylprednisolone for our study because it has less mineralocorticoid activity and a longer half-life than hydrocortisone. Our study showed that a single preoperative i.v. bolus of methylprednisolone in LLR significantly lowered postoperative total bilirubin level, increased albumin, reduced CRP and IL-6 levels, and shortened prothrombin time, confirming our hypothesis that it would improve postoperative liver function. There was also a trend toward better surgical outcomes.
Post-hepatectomy liver failure is one of the most serious complications of liver surgery, with postoperative total bilirubin levels and prothrombin time being indicative of PHLF. 21,22 In the present study, the rate of PHLF was generally low and comparable between the glucocorticoid and control groups. However, we believe that the improvement in postoperative total bilirubin level and prothrombin time owing to glucocorticoid use may have a positive clinical impact on patient outcome.
The intermittent Pringle maneuver, a temporary clamping of the hepatoduodenal ligament, is a safe and useful procedure to reduce blood loss during liver resection. However, the intermittent Pringle maneuver does carry a risk of ischemia-reperfusion injury to the liver, which is defined as an inflammatory response and organ damage, triggered by ischemia inducing hypoxic stress and reperfusion inducing oxidative stress. 23,24 Numerous factors contribute to Association between laparoscopic surgery and a postoperative decrease in liver function has previously been reported. 9,29 This decrease in liver function likely results from pneumoperitoneum pressure, which is usually higher than the normal portal blood pressure, leading to a reduction in portal blood flow. 7,30 The inflation and deflation of the pneumoperitoneum may also induce an ischemia-reperfusion injury. Thus, there are two possible causes of hepatic ischemia-reperfusion injury during LLR, one being the intermittent Pringle maneuver and the other the pneumoperitoneum.
Therefore, greater care is required to prevent ischemia-reperfusion hepatic injury in LLR than in OLR. Additionally, the operative time is often longer for LLR than for OLR, 5,6 resulting in a longer overall duration of the Pringle maneuver. Thus, giving glucocorticoid might have a greater effect in LLR than in OLR.
Regarding the postoperative complications, we had hypothesized that glucocorticoid use would decrease postoperative morbidity through its attenuation of the inflammatory response after major abdominal surgery, which is a main contributor to postoperative morbidity and delays recovery. 31,32 The absence of a significant difference in postoperative morbidity between the glucocorticoid and control groups in our study might be related to the relatively small number of patients (n = 50) in each group. However, we did observe a trend towards lower morbidity rates and CCI scores in the glucocorticoid as compared to the control group. Lower levels of postoperative CRP and IL-6 may further contribute to decreasing the incidence of postoperative complications.
However, increase in blood glucose levels is an important adverse effect of glucocorticoid use that must be considered. On the first postoperative day, among patients in the glucocorticoid group, blood glucose level was higher than normal, and higher than the blood glucose values in the control group, returning to normal levels by the second postoperative day. Therefore, in the present study, the adverse effect of the methylprednisolone bolus on blood glucose levels was considered to be transient and limited.
Limitations of the present study must be considered in the interpretation of the results. This was a single-center trial with a relatively small number of patients and, therefore, selection bias cannot be excluded. Because of the small sample size, subgroup analyses to evaluate potential effects of the extent of hepatectomy, liver cirrhosis and the length of the Pringle maneuver on postoperative blood parameters could not be carried out. Therefore, larger scale, multicenter studies are required to confirm our findings.
In conclusion, this is the first report on the effect of giving a single bolus of preoperative methylprednisolone on hepatic function and surgical outcomes after LLR, showing a benefit when assessing postoperative bilirubin and albumin levels, and prothrombin time.
Based on our results, we believe that preoperative glucocorticoid use could enhance the safety of LLR, which is increasingly carried out worldwide.

ACK N OWLED G EM ENTS
The authors would like to thank the staff of the Department of Anesthesiology at Iwate Medical University School of Medicine for their assistance. We would like to thank Editage (www.edita ge.com) for English language editing.

D I SCLOS U R E
Funding: There were no funding supports for this study.
Conflicts of Interest: Authors declare no conflicts of interest for this article.
Author Contribution: All authors: conception, design, interpretation of data, surgery, and acquisition of data. Y.H. and T.T.: data analysis.