Background:

Several randomized clinical trials have shown beneficial effects of pre-operative glucocorticoids on post-operative complications.

Studies on the effects of glucocorticoids on the postoperative recovery after liver-resection show significantly lower markers of infection and liver damage, and some studies have shown a shorter hospital stay.

Studies on the effects in the immediate postoperative phase are lacking.

Methods: Randomized, double-blind, controlled trial evaluating incidence of postoperative complications in the immediate postoperative phase (and during admission) after open liver surgery. Participants are randomized to either active treatment (methylprednisolone 10 mg/kg) or control (8 mg dexamethasone), administered just prior to surgery.

All patients undergoing open liver resection at our institution are eligible. Included patients are stratified according to extent of surgery into minor (<3 segments) or major (≥3 segments) group.

Patients in major group participate in Substudy I (markers of endothelial damage).

Patients operated between January and July 2018 participate in Substudy II (delirium).

Post-surgery, patients are traditionally observed and treated in post-anesthesia care units (PACU) until they are discharged to the ward (or directly home) assessed by standardized international discharge criteria.

The research project "Why in PACU?" (Rigshospitalet, Denmark), has since the beginning of 2016 systematically collected and analyzed procedure-related complications in the recovery phase. The complications include pain, nausea/vomiting, circulatory and respiratory problems, orthostatic intolerance and cognitive disorders. Common to all the above-mentioned post-operative problems are the possible links to the inflammatory response caused by the surgical trauma.

Glucocorticoids (GC) can in this context be central for the reduction of acute postoperative organ dysfunctions, caused by the anti-inflammatory effect. In a number of different surgical procedures, single dose, pre-operative glucocorticoids have been shown to reduce post-operative nausea and vomiting (PONV), acute pain and need of opioids as well as accelerate the convalescence. Meta-analyses also showed that single-dose administration of glucocorticoids (methylprednisolone and dexamethasone) for surgical patients is safe as opposed to long-term treatment.

Studies on pre-operative glucocorticoids before liver surgery have shown beneficial effects in regards to markers of liver damage and infection, but studies on clinical outcomes in the immediate post-operative phase are lacking.

The primary aim of this study is to investigate whether high dose pre-operative glucocorticoids reduce complications in the immediate post-operative course.

The investigators will also perform two hypothesis-generating sub studies:

• Sub study I - markers of endothelial dysfunction

The endothelial lining of blood vessels contributes to maintaining haemostasis, and damage can increase risks of cardiovascular and thromboembolic complications. In a recent randomized trial, pre-operative high dose glucocorticoids diminished circulating markers of endothelial damage (after knee arthroplasty). In this study we will investigate whether this also applies liver surgery, and if so, if there is any connection to cardiovascular and thromboembolic complications.

• Sub study II - delirium Studies on delirium after liver surgery show an incidence around 20%. It has not been investigated whether pre-operative glucocorticoids have an effect on this incidence.

The investigators will investigate the incidence of emergence delirium and delirium during the first postoperative day s after liver surgery.

Sample size:

The "Why in PACU?" database shows that complications requiring treatment in PACU occur in up to 40 % of patients after liver surgery. These complications are primarily respiratory and circulatory.

A 50 % reduction in the number of patients with complications requiring treatment is regarded clinically relevant. This will require a sample size of 174 patients, including 10 % dropout (80 % power, 5% level of significans, superiority design). Patients will be stratified according to extent of surgery, into minor or major resection.

The sub studies are hypothesis-generating, and are not subject to power calculations. Sub study I will include all major resections, sub study II will include patients during the first 5 months.

Analysis: Primary end point (complications in the two groups) is compared with chi square test and described with odds ratio (95%CI). Level of significance is p=0,05 Standard statistical analysis will include normally or near-normally distributed variables reported as means and non-normally distributed variables as medians. Means will be compared using the student's t test and medians using the Mann-Whitney U test. Differences in proportions among categorical data will be assessed using Fischer's exact test. A p value < 0.05 will represent statistical significance for all comparisons.

Hypothesis: Preoperative GC administration will decrease the incidence of postoperative complications and overall hospital length of stay following liver surgery. Preoperative GC administration will decrease markers of endothelial dysfunction following major liver surgery. Preoperative GC administration will decrease incidence of delirium following liver surgery.

Data collection:

Data elements to be collected will include, but not be limited to:

• Demographics (age, gender, height, weight, tobacco and alcohol consumption, comorbidities, American Society of Anaesthesiology (ASA) score)
• Preoperative chemotherapy, preoperative use of analgesics and/or other central stimulants
• Preoperative biochemistry
• Diagnosis, procedure, surgery duration
• Blood loss, transfusions, use of drain, hepatic inflow occlusion (length of)
• Postoperative pain, nausea, sedation and vitals, every 30 minutes until transfer to ward
• Postoperative pain, nausea, mood and quality of sleep, self reported, every day until discharge or postoperative day 5 (what comes first)
• Postoperative use of analgesics and anti emetics until discharge or postoperative day 5 (what comes first)
• Complications (hepatic failure, ascites, intraabdominal collection, postoperative bleeding, bile leak, bowel obstruction, wound dehiscence, reoperations, pleural effusion, pulmonary embolus, deep venous thrombosis, infections, cardiac events, cerebral events, other causes of prolonged hospital stay)
• 3-minute Diagnostic Confusion Assessment Method (3D-CAM), postoperative day 0 (all) 1-3 (major resections)
• Endothelial markers (Syndecan-1, soluble thrombomodulin, SE-selectin, vascular endothelial growth factor (VEGF) postoperative day0-3 (major resections)
• Hospital stay, length of stay in PACU
• Mortality (30 days)

The study is not placebo-controlled since the positive effects of dexamethasone 8 mg on PONV have been shown in numerous trials, and is already being administered to all patients at the clinic. It would therefore not be ethically correct to withdraw from this practise.

Inclusion Criteria:

• Age minimum 18
• Planned open liver resection (with or without combined ablation and/or
• cholecystectomy)
• Able to participate (self report pain/nausea)
• Understands danish/english, or has an interpreter during admission
• Signed consent form

Exclusion Criteria:

• ALPPS (Associating Liver Partition and Portal vein Ligation for Staged hepatectomy) procedure
• Combined ventral herniotomy with implantation of mesh
• Combined with operation on tumor in other organs
• Insulin dependent diabetes
• Current (<10 days) treatment with systemic glucocorticoids and/or immunosuppressive treatment (not including inhalations)
• Epidural anesthesia not feasible
• Pregnancy/breastfeeding
• Allergy toward study medication
• Inoperability

40 hours after drug administration

According to GCP definitions