Journal Pre-proofs

Perioperative Omega-3 fatty Acids Fails to Confer Protection in Liver Surgery.

Results of a multicentric, double-blind, randomized controlled trial

Michael Linecker, Florin Botea, Dimitri Aristotele Raptis, Diana Nicolaescu,

Përparim Limani, Ruslan Alikhanov, Pavel Kim, Andrea Wirsching, Philipp
Kron, Marcel A. Schneider, Christoph Tschuor, Patryk Kambakamba, Christian
Oberkofler, Michelle L. De Oliveira, John Bonvini, Michail Efanov, Rolf Graf,
Henrik Petrowsky, Igor Khatkov, Pierre-Alain Clavien, Irinel Popescu

PII: S0168-8278(19)30605-1
DOI: https://doi.org/10.1016/j.jhep.2019.10.004
Reference: JHEPAT 7506

To appear in: Journal of Hepatology

Received Date: 15 July 2019

Revised Date: 1 October 2019
Accepted Date: 2 October 2019

Please cite this article as: Linecker, M., Botea, F., Raptis, D.A., Nicolaescu, D., Limani, P., Alikhanov, R., Kim, P.,
Wirsching, A., Kron, P., Schneider, M.A., Tschuor, C., Kambakamba, P., Oberkofler, C., De Oliveira, M.L.,
Bonvini, J., Efanov, M., Graf, R., Petrowsky, H., Khatkov, I., Clavien, P-A., Popescu, I., Perioperative Omega-3
fatty Acids Fails to Confer Protection in Liver Surgery. Results of a multicentric, double-blind, randomized
controlled trial, Journal of Hepatology (2019), doi: https://doi.org/10.1016/j.jhep.2019.10.004

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier B.V. on behalf of European Association for the Study of the Liver.
 Perioperative Omega-3 fatty Acids Fails to Confer Protection
in Liver Surgery. Results of a multicentric, double-blind,
randomized controlled trial

Michael Linecker1, MD, PhD; Florin Botea2, MD; Dimitri Aristotele Raptis1,3, MD, MSc, PhD;
Diana Nicolaescu2, MD; Përparim Limani1, MD, PhD; Ruslan Alikhanov4, MD; Pavel Kim4, MD;
Andrea Wirsching1, MD, MSc; Philipp Kron1,5, MD; Marcel A. Schneider1, MD;
Christoph Tschuor1,MD, PhD; Patryk Kambakamba1,MD; Christian Oberkofler1, MD;
Michelle L. De Oliveira1, MD; John Bonvini6, MD; Michail Efanov4, MD, PhD; Rolf Graf1, PhD;
Henrik Petrowsky1, MD; Igor Khatkov4, MD;
Pierre-Alain Clavien1, MD, PhD; and Irinel Popescu3, MD, PhD.

1 Department of Surgery and Transplantation, University Hospital Zurich, Switzerland

2 Center of General Surgery and Liver Transplantation, Fundeni Institute Bucharest, Romania
3 Department of HPB- and Liver Transplantation Surgery, University College London, Royal Free
Hospitals, London, UK
4 Department of Liver and Pancreatic Surgery, Moscow Clinical Scientific Center, Russia
5 Department of HPB and Transplant Surgery, St. James's University Hospital NHS Trust, Leeds, UK
6 Department of Anesthesiology, University Hospital Zurich, Switzerland

Part of the OmegavenTM Trial was presented during the 13th Biennial Congress of the
International Hepato-Pancreato-Biliary Association (IHPBA) September 4 - 7, 2018 in the
category clinical trials in progress in Geneva, Switzerland.

Conflict of interest: The authors declare no conflict of interest. The trial is an investigator
initiated trial without competing interest of any pharmaceutical company, including
Fresenius providing the study product.

Financial support: The study was supported by the Clinical Research Priority Program of the
University of Zurich as part of the project ”Non-resectable liver tumors”.

Short running title: OmegavenTM RCT

Trial registration: ClinicalTrial.gov: ID: NCT01884948; Institution Ethical Board Approval:

KEK-ZH-Nr. 2010-0038; Swissmedic Notification: 2012DR3215.

Keywords: Omega3- polyunsaturated fatty acids (Ω3-FA), OmegavenTM, major liver surgery,
Clavien-Dindo classification, Comperative Complication Index (CCI).

1
Abbreviations:
AE: Adverse event; ALT: Alanine transaminase; AST: Aspartate transaminase; CTC: Clinical
trials center; eCRF: Electronic case report from; GFR: Glomerular filtration rate; ICU:
Intensive care unit, IEC: Independent ethics committee; INR: International normalized ratio;
I/R: Ischemia-reperfusion injury; IV: Intravenous; LR: Liver regeneration; RCT: Randomized
controlled trial; ROS: Reactive oxygen species; SAE: Serious adverse event; TG: Triglycerides;
Ω3-FA: Omega3- polyunsaturated fatty acids.

Manuscript: Abstract: 251 words | Manuscript: 3174 words (Introduction through

Conclusion) | References: 31 references | Tables: 2, Figures: 3

Lay Summary: This trial fails to detect an effect on the use of two single Omega-3 fatty acid
infusions before and during liver surgery to reduce complications. Despite strong evidence of
liver-directed anti-inflammatory and pro-regenerative action in various rodent models, this
trial fails to confer liver protection with single Omega-3 fatty acid infusions and is therefore
currently not recommended.

Correspondence:

Pierre-Alain Clavien, MD, PhD

Department of Surgery & Transplantation
University Hospital Zurich
Raemistrasse 100
CH-8091 Zurich
Phone: +41 44 255 3300
E-mail: clavien@access.uzh.ch

2
 ABSTRACT

Objective: To investigate the perioperative use of Omega-3 polyunsaturated fatty acids (Ω3-

FA) to reduce post-operative complications after liver surgery.

Summary Background Data: In a variety of animal models, Ω3-FA disclosed strong protective

effects in alleviating hepatic ischemia/reperfusion injury and steatosis with enhancement of

regeneration after major tissue loss. Given these benefits along with its safety profile, we

hypothesized that perioperative administration of Ω3-FA in patients undergoing liver surgery

may ameliorate the post-operative course.

Methods: A multi-centric, double-blind, randomized, placebo-controlled trial was designed

to test whether two single intravenous infusions of OmegavenTM (Ω3-FA) vs. placebo may

decrease morbidity. The primary endpoints were postoperative complications by severity

(Clavien-Dindo classification) integrated in the Comprehensive Complication Index (CCI).

Results: Between July 2013 and July 2018, 261 patients (132 in the Omegaven and 129 in the

placebo groups) were included in the trial from three centers. Most cases (87%, n=227)

underwent open liver surgery and 56% (n=105) were major resections (≥3 segments). In an

intention-to-treat analysis including the drop out cases, the mortality rate was 4% and 2% in

the OmegavenTM and placebo groups (P=0.447; OR: 95%CI: 0.04-2.51), respectively. Any

complications and major complications (Clavien-Dindo ≥3b) occurred in 46% vs. 43%

(P=0.709) and 12% vs. 10% (P=0.69) in the Omegaven and placebo groups, respectively. The

mean CCI was 17 (±23) vs.14 (±20) (P=0.417). An analysis excluding the drop-outs provided

similar figures.

Conclusions: The routine perioperative use of two single doses of intravenous Ω3-FA (100ml

OmegavenTM) cannot be recommended in patients undergoing liver surgery (Grade A

recommendation).
3
 INTRODUCTION

Omega-3 polyunsaturated fatty acids (Ω3-FA) are essential nutritional components for

the human, whose presence in the body exclusively depends on its exogenous supply. In the

late 1970s, Bang and Dyerberg observed an extraordinary low incidence of coronary heart

disease among Greenland inuits, which was found to be related to a high intake of Ω3-FA

abundant in marine fish; the main food source of this population.1 The liver was

subsequently found to be a central metabolic target of Ω3-FA with growing evidence on their

beneficial effects on fatty liver and metabolic syndrome.2-4 Ω3-FA were found not only to

lower intrahepatic lipid contents5, but also mitigate inflammatory changes in the liver.6, 7

Liver surgery in patients with fatty liver leads to an amplified ischemia-reperfusion injury

(I/R) and impaired regeneration (LR).8 A variety of animal models disclosed significant liver

protection in hepatic surgery (reduction of I/R and improved LR) following treatment with

Ω3-FA.9-11 Of note, these effects were not only observed in steatotic, but also in lean livers.11

For example, in a mouse model, the intravenous (IV) administration of a single dose of Ω3-FA

(OmegavenTM) prior to liver ischemia dramatically mitigated the production of reactive

oxygen species (ROS) upon reperfusion.12 This beneficial effect is mediated via the GPR120

receptor located on hepatic Kupffer cells.12

Based on the accumulating evidence in various animal models of Ω3-FA including the

effective application of single IV doses12, we embarked in a RCT in patients undergoing

hepatic surgery. Considering the safety profile of Ω3-FA in human with its well-established

use to counteract liver steatosis13, we omitted phase I and II trials, and tested putative

benefits of IV Ω3-FA in a phase III trial on early postoperative outcome. We hypothesized,

that such an effect would mostly be related to the prevention of ROS release.12

4
 MATERIAL AND METHODS

Trial design

We designed a prospective, double-blind, multicenter, parallel-group, randomized,

placebo-controlled, phase III trial to test the effect of two-single iv infusions of OmegavenTM

on single (Clavien-Dindo)14 and cumulative (CCI)15 postoperative complications. The

protocol targeted patients undergoing elective liver resection with resection of at more than

one segment. This investigator initiated trial was carried out in three liver centers

(Bucharest, Romania; Moscow, Russia and Zurich, Switzerland).

Patients, study interventions, data collection and monitoring

Eligibility criteria included patients >18 years of age requiring liver resection of at least

one segment or multiple wedge resections (≥3). Detailed inclusion and exclusion criteria are

displayed in Supplementary table 1. Patients were randomly assigned (see below) to the

OmegavenTM or placebo (NaCl 0.9%) group in a 1:1 allocation ratio the day before surgery,

i.e., just after obtaining the informed consent. Eligible study participants received 100ml of

the blinded study product the evening before surgery and the day of surgery starting with

anesthetic induction. One dose of OmegavenTM (100ml) contains 10g of highly refined fish oil

mainly containing eicosapentaenoic and docosahexaenoic acid. Care was taken not to

exceed the maximum infusion rate of 0.05 g fish oil/kg body weight/hour.13 According to

standard operating procedures for anesthesia, Propofol as anesthetic agent was avoided as

it contains concentrated triglycerides (TG). Perioperative source data was collected on site

and entered into the electronic case report form (eCRF) of the clinical trial software

secuTrial® (Interactive System, Berlin, Germany). Data monitoring for all three participating

centers was performed by the Clinical Trial Center Zurich (CTC), Switzerland.

5
 Study endpoints

The primary endpoint was morbidity/mortality, as defined by the Clavien-Dindo

classification of complications (grade I-V)14 and the related Comprehensive Complication

Index (CCI), which integrates the different grades of complications by severity in figures

ranging from 0 (uneventful course) to 100 (death).15 Study end was defined at 30 days after

hospital discharge.14 All patients who signed the consent were followed up until this time

point (Figure 1) . Secondary endpoints included length of hospitalization and intensive care

unit (ICU) stay, surrogate laboratory parameters of safety including serum levels of TG,

platelets, and International normalized ratio (INR). Liver related parameters including ALT

(alanine aminotransferase), AST (aspartate aminotransferase), bilirubin and creatinine were

also collected.

Randomization and blinding

A block randomization was used to reduce a potential center-bias. As both nurses and

surgeons were blinded, one study-independent person was responsible for randomization

and blinding per center. This person allocated patients into the OmegavenTM or placebo

group the day before surgery, and ensured a proper blinding just before the study product

was administered. Blinding was performed using opaque infusion lines and plastic bags to

cover the study product.

Statistical analysis and sample size

Based on an expected mean of CCI of 17 and 65% of total complications in patients

undergoing hepatectomy (cohort of the University Hospital Zurich) we assumed that

perioperative continuous IV administration of OmegavenTM may decrease the CCI and/or

total complications by 30% (α: 0.05, Power (1-β): 0.8, 2-sided test). A 30% reduction of

6
complications and CCI was agreed to be a clinically relevant endpoint. Considering a putative

drop-out rate of 10%, we identified a sample size of 129 patients per randomized group

(total: n=258). An interim analysis for safety was scheduled after 60 cases. Final analysis was

performed in a univariate setting using Fisher's Exact test for categorical variables and

Mann-Whitney-U test for continuous variables. Continuous variables were expressed in

median and interquartile range, categorical variables in absolute numbers and percent (%).

Odds ratios (OR) and 95% confidence interval (CI) were reported when appropriate. P-values

<0.05 were considered significant. All statistical analysis were performed using R Studio

version 1.0.44 (RStudio Inc., Boston, MA), IBM SPSS Statistics version 24 (IBM Corporation,

Armonk, NY) and Graph Pad Prism version 7 (GraphPad Software, Inc., La Jolla, CA).

Trial registration and ethics

This trial was approved by the Ethical Board of the Canton Zurich (Nr. 2010-0038),

notified by Swissmedic (2012DR3215), and registered at ClinicalTrial.gov: ID: NCT01884948.

Local ethics approvals were obtained for Bucharest (Nr. 12832) and Moscow (Nr. 4/2014).

Informed consent was obtained from all subjects and all trial interventions were conducted

according to the Declaration of Helsinki and recent guidelines for Good clinical practice. The

protocol of this RCT was published previously.16

RESULTS

Patient flow and randomization

From July 29, 2013 to July 12, 2018, a total of 261 patients from three centers were

included in the study and randomly assigned to receive either OmegavenTM (n=132) or

placebo (n=129) in two doses the day before and during liver surgery. This patient

7
population depicts the intention-to-treat population. Patients who regularly completed the

study are represented in the per-protocol population (Figure 1).

Comparability of OmegavenTM and placebo groups

Demographics of study participants were comparable between treatment groups

(Table 1). Indication for liver surgery was in the majority of cases malignant liver tumors

(90% and 93%), with colorectal liver metastases (CRLM) as the leading entity in both the

OmegavenTM (27%) and the placebo (34%) groups. Benign tumors were the indication in 14%

and 12%, respectively, and the most common other indications included Echinococcus and

living donor liver transplantation in the OmegavenTM and placebo groups (29% vs. 35% and

25% vs. 30%). Overall, the utilization of inflow-occlusion during transection was similarly

distributed (29% vs. 28%) in the treatment groups with a comparable median Pringle time of

18 vs. 16 minutes (Table 1).

Treatment procedures and patient dropout

After randomization patients were allocated to receive either OmegavenTM (n=132) or

placebo (n=129) (Figure 1). Both, randomization and allocation to the respective trial drug

occurred without errors during the study period, as proper study instruction was performed

beforehand. After infusion start the evening before liver surgery no adverse effects

(anaphylactic reaction, gastrointestinal disorders, headache, rash etc.)13 were recorded.

Likewise, the second infusion starting at anesthetic induction revealed no adverse effects.

The drop-out rate was 15% in the OmegavenTM group and 12% in the placebo group. This

was slightly higher than the assumed drop-out rate of 10% in both groups. Most drop-outs

occurred at the time of laparotomy and exploration of the situs. In 13% and 10% the tumor

could not be resected due to the presence of more diffuse disease including extrahepatic

8
tumor spread. In two and three cases in the respective groups, surgery was converted to an

ALPPS (Associating liver partition and portal vein ligation) procedure. These patients were

excluded in the protocol due to the second stage surgery occurring within the study period.

Overall, in 6 cases the second dose of the study product was not administered

intraoperatively due to a communication error. One patient in each group refused to follow-

up after study product application (Figure 1). Otherwise patients were successfully followed

up until the study end one month after hospital discharge.

Safety monitoring

Restrictive in- and exclusion criteria, particularly focusing on potentially impaired

blood coagulation were an important feature of this study (Supplementary table 1).

Therefore only 13% of all screened patients could finally be included in the study (Figure 1).

An interim analysis was carried out after inclusion of the initial 60 patients (OmegavenTM,

n=31; Placebo, n=29). Mortality and severe complications (≥IIIb) were 2 (7%) and 4 (13%) in

the OmegavenTM and 1 (3%) and 4 (14%) in the placebo group (P=0.60 and P=0.92).

Consistently, CCI and the presence of any complications did not differ between the groups,

22 (±26) and 20 (64%) vs. 20 (±25) and 15 (52%) (P=0.68 and 0.32). The interim analysis

therefore revealed no safety issues enabling completion of the study. As OmegavenTM can

lead to a prolonged bleeding time and an inhibited platelet aggregation patients in rare

cases13 safety blood samples for coagulation parameters including INR and platelets were

obtained one hour after start of the study product infusion to be compared with the

baseline values (Figure 2 c,d).

Mortality and Morbidity

The mortality rate within the study period was 4% and 2% in the OmegavenTM and

9
placebo groups (P=0.45; OR: 95%CI: 0.04-2.51), respectively. Any complications and major

complications (Clavien-Dindo ≥IIIb) occurred in 46% vs. 43% (P=0.709; OR:0.89; 95%CI:0.53-

1.50) and 12% vs. 10% (P=0.69; OR:0.81; 95%CI:0.34-1.89) in the OmegavenTM and placebo

groups, respectively. The mean CCI was 17 (±23) vs.14 (±20) (P=0.42) (Table 2 and Figure 2a).

Most severe complications were cardiac leading to three mortalities in the OmegavenTM

group, followed by three patients experiencing a septic shock and consecutive organ failure

and one patient, who developed a kinking of the portal vein after a right hemi-hepatectomy

leading to lethal liver failure. A code break by the principle investigator for unblinding

mortalities was performed in 4 cases. Overall, biliary, infectious and other complications

were the most common complications. Biliary and infections complications occurred in 10%

and 11% in the OmegavenTM, and 5% and 12% in the placebo groups (P=0.15 and 0.85; Table

2). A Per-protocol analysis (Figure 1) did not reveal a different outcome (Supplementary

table 2).

Subgroup analysis with only patients undergoing major hepatectomy (≥3 segments)

showed a mortality of 7% vs. 3% in the OmegavenTM and placebo group (P=0.26; OR:0.35;

95%CI:0.03-2.25). Any complication and major complications in this subgroup occurred in

57% vs. 53% and 19% vs. 11% in the OmegavenTM and placebo groups (P=0.62; OR:0.83;

95%CI:0.41-1.69 and P=0.24; OR:0.52; 95%CI:0.17-1.46). The subgroup analysis for only

patients with inflow occlusion (Pringle maneuver), as a potential source of I/R, did not reveal

a different outcome for the treatment groups. Mortality was 6 vs. 3% (P=0.609; OR: 0.46;

95%CI:0.01-9.27) in the OmegavenTM and placebo groups. Any complication occurred in 53%

in both treatment groups (P=1.00; OR: 0.99; 95%CI:0.35-2.81), and major complications in

15% and 8% (P=0.47; OR: 0.53; 95%CI:0.076-3.01).

10
 Effects of OmegavenTM on laboratory parameters

After application of the first dose of study medication median triglyceride levels rose

by 45 vs. 9 mg/dl in the OmegavenTM and placebo groups, respectively and stayed in this

range after application of the second dose (P<0.001, Figure 2b). No significant differences

were noted for platelets and INR (Figure 2c,d). Peak median ALT and AST levels appeared

slightly increased, although not reaching statistical significance, in the OmegavenTM

treatment group (332 and 363 vs. 318 and 305 U/l) (Figure 3a,b), bilirubin and creatine levels

appeared unaffected.

Potential OmegavenTM-related side effects

The study represents a clinical phase III trial testing OmegavenTM for a new indication.

But side effects of OmegavenTM are already well characterized.13 OmegavenTM is contra-

indicated for patients with a recent myocardial infarction. Though none of the patient

included in the study experienced a myocardial infarction the last months before surgery,

three mortalities in the OmegavenTM group were attributable to cardiac events after surgery

(Table 3). The placebo group in contrast revealed no lethal cardiac events (P=0.25).

11
 DISCUSSION

In this multicenter, randomized, double-blind, placebo-controlled trial including 261

patients, perioperative administration of Ω3-FA did not exhibit a reduction of postoperative

complications after liver surgery. These results are in contrast to a variety of animal studies

reporting marked reductions of I/R injuries 9, 12 and improved liver regeneration in both

steatotic and lean livers10, 11. Therefore, the current trial does not enable to translate those

exciting experimental observations in patients using markers of postoperative morbidity; the

most meaningful primary endpoint of harm in surgery.

Although liver-defattening effects of Ω3-FA17 including evidence for some anti-cancer

effects11, 18, 19 have been investigated for a while, the clinical use within the scope of liver

surgery is new with little available information. Previous reports have suggested that oral

supplementation of Ω3-FA for 4 weeks reduces intrahepatic lipid content in preparation for

bariatric surgery20 and living donor liver transplantation.21 Just prior to the initiation of this

RCT, another RCT showed that parenteral use of Ω3-FA supplementation for seven

consecutive days after liver transplantation was associated with a reduction of infectious

complications and shorter hospitalization time.22 During the recruitment period of this trial

three RCTs looking on different outcomes after perioperative Ω3-FA supplementation in liver

surgery were published.23-25 One study reported the use of a 20% Ω3-FA emulsion with 7

mL/kg once a day for 2 days before and at the day of living donor hepatectomy, which were

associated with improved liver regeneration one month after surgery, as measured by liver

volumetry.24 The second trial from China reported reduced overall mortality and length of

hospital stay with the use of continuous Ω3-FA parenteral supplementation vs. standard

parenteral nutrition for five days postoperatively in cirrhotic patients undergoing

hepatectomy for cancer.25 Similarly, the third trial, also from China, looked at postoperative

12
complications following a 5 days course of postoperative intravenous Ω3-FA

supplementation after hepatectomy revealing a reduction of total postoperative

complications and length of hospital stay.23 These studies are in contrast with the results of

the current RCT. The major issues with these studies may lie in the lack of consistency

regarding the forms of Ω3-FA application and doses, as well as inhomogeneous patient

cohorts, and eventually different primary endpoints. The study in living liver donors24, for

example, focused on volumetrically measured liver growth rather than complications. The

other study left patients for 5 days on parenteral nutrition with OmegavenTM as

supplementation in the treatment group.23 This might introduce a major bias as parenteral

nutrition (comparator) is well known to harm liver function and OmegavenTM is approved for

reducing this harm. Of note, none of these studies tried to investigate the different effects of

Ω3-FA. It is therefore difficult to interpret those studies, particularly whether the observed

positive effects of Ω3-FA relate to their anti-oxidative vs. defattening properties.

The current study, using 2 perioperative IV doses, focused on the putative anti-

oxidative/anti-inflammatory effects rather than a beneficial effect through defattening,

which obviously could not occur during this short period. We chose this protocol relating to

the strong anti-oxidant effects observed in previous laboratory works from our9, 11, 12, 26 and

other10, 27-29 groups, and accordingly did not restrict the study to patients with liver steatosis.

Perioperative doses also prevent well-known issues with compliance of patients in taking

medication. In addition, patients scheduled for liver surgery mainly present with malignant

tumor (90%); often making difficult to impose a preparation for surgery lasting for weeks.

Similarly to this trial, other RCTs tested different other anti-oxidant/anti-inflammatory

compounds, which showed promising results in experimental studies of liver surgery but

failed to translate its beneficial effect into clinics. For example, the perioperative application
13
of N-acetylcysteine (1 dose), could not reduce postoperative complications after liver

surgery in an RCT including 206 patients.30 More so, the perioperative treatment with

steroids (3 doses) also failed to reduce postoperative complications after liver surgery in an

RCT including 210 patients.31 This data strongly suggests that mitigation of inflammation is

obviously insufficient to reduce the incidence of postoperative complications after hepatic

surgery.

Morbidity and mortality did not show any difference including the subgroup analysis

of major liver surgery and patients with inflow occlusion. Although not reaching statistical

difference, three out of 5 mortalities in the OmegavenTM group were attributable to acute

cardiac events, although pre-existing cardiac disease was an exclusion criterion (Table 2).

Thus, such side effects of OmegavenTM cannot be excluded.13 This is an interesting

observation as the long-term Ω3-FA intake as a food supplement, in contrast, leads to

cardioprotection.1 We could speculate that the documented rise in serum TG levels in

patients subjected to IV Ω3-FA in this study might be related to cardiac events.

The strength of this study is its methodology providing the highest level of evidence.

This placebo-controlled RCT is testing a clear-cut, clinical question and guarantees highest

possible standards to minimize bias. Many patients of the liver resection population,

however, could not be included due to the stringent exclusion criteria for safety reasons

resulting in a final inclusion of only 13 % patients screened (Figure 1). Both, inclusion and

exclusion criteria were chosen restrictively (Supplementary table 1) to meet the highest

possible safety standard for a phase III trial.

There are a number of limitations in this RCT. A positive effect could have been

masked by the heterogeneity of the patients enrolled. A stratification of procedures with

14
different complication profiles could potentially lead to a better discrimination of

OmegavenTM effects. A further problem is that the doses applied might be too low, but

higher doses of OmegavenTM are clearly not recommended due to the risk of side effects.13

The dose recommendation of 10g Ω3-FA per application was strictly followed.13 For this

reason phase I or II trials to define optimal dosing would not have impacted the design of

this study. Bioavailability was documented by the rise of blood TG levels perioperatively .

Another limitation is that the study end was set 30 days after hospital discharge14, which

could be insufficient to detect the whole spectrum of complications. Certainly, most early-

occurring complications, which may result from I/R and impaired LR may be recorded in this

time period. In addition, center volume and slightly different surgical techniques for hepatic

resection could potentially impact on results although a block randomization was performed

to distribute equally between the centers.

In summary, this multicenter, randomized, placebo-controlled trial failed to reduce

postoperative complications using two IV doses of Ω3-FA (100ml OmegavenTM). This finding

is in contrast to experimental and clinical findings disclosing many beneficial effects. Like

other anti-inflammatory compounds (e.g. N-acetylcysteine and steroids) IV use of Ω3-FA is

not recommended perioperatively in liver surgery. This, however, does not apply to the liver-

defattening properties in oral long-term use of Ω3-FA.

15
 FIGURE LEGENDS

Figure 1. Patient flow diagram

According the CONSORT statement 2010 the process from patient screening to allocation of

the study product until final follow-up is displayed. Only 13% fulfilling the inclusion criteria

and ruling out exclusion criteria were included in this RCT. These patients represent the

intention-to-treat population. In the OmegavenTM group n=20 patients and in the placebo

group n=16 patients were excluded, mainly for a non-resectable tumor situation

intraoperatively, forming the per-protocol population.

Figure 2. Complications, circulating TGs and coagulation parameters

Panel A displays mean (±SD) CCI values (0-100) for OmegavenTM and Placebo groups. Panel B

shows the increase of TG levels one hour after start of first and second infusion in the

OmegavenTM group labeled by an arrow. Besides TG, laboratory safety parameters including

platelets and INR were collected at this timepoint revealing no difference in the respective

groups (Panel C and D). Analyses were performed for the intention-to-treat population.

Figure 3. Transaminases, bilirubin and creatinine during study intervention

Liver-related laboratory parameters including ALT, AST and bilirubin as well creatinine was

monitored over the first five postoperative days. No significant differences were noted,

however peak median ALT and AST levels appeared slightly increased in the Omegaven TM

treatment group (332 and 363 versus 318 and 305 U/l). Analyses were performed for the

intention-to-treat population.

16
 ACKNOWLEDGEMENT

The authors wish to thank the study coordinators Karin Petterson and Astrid Hirt for

their excellent support in performing this trial. Further, we would like to thank Lisette

Paratore-Hari, PhD from the CTC for programming the eCRF and managing the secuTrial ®

software and the monitoring team of the CTC led by Jan Michel, PhD for his indispensable

work on monitoring data input in three study centers. We would also like to thank all the

residents/fellows who helped with informed consents and sample/data collection. Last, but

not least we would like to thank Fresenius for kindly providing the study product.

17
 REFERENCES

1. Dyerberg J, Bang HO, Stoffersen E, et al. Eicosapentaenoic acid and prevention of thrombosis
and atherosclerosis? Lancet 1978; 2(8081):117-9.
2. Budowski P. Review: nutritional effects of omega 3-polyunsaturated fatty acids. Isr J Med Sci
1981; 17(4):223-31.
3. Kobatake Y, Hirahara F, Innami S, et al. Dietary effect of omega-3 type polyunsaturated fatty
acids on serum and liver lipid levels in rats. J Nutr Sci Vitaminol (Tokyo) 1983; 29(1):11-21.
4. Storlien LH, Kraegen EW, Chisholm DJ, et al. Fish oil prevents insulin resistance induced by
high-fat feeding in rats. Science 1987; 237(4817):885-8.
5. Romero-Gomez M, Zelber-Sagi S, Trenell M. Treatment of NAFLD with diet, physical activity
and exercise. J Hepatol 2017; 67(4):829-846.
6. Okada L, Oliveira CP, Stefano JT, et al. Omega-3 PUFA modulate lipogenesis, ER stress, and
mitochondrial dysfunction markers in NASH - Proteomic and lipidomic insight. Clin Nutr 2018;
37(5):1474-1484.
7. Calder PC. Omega-3 polyunsaturated fatty acids and inflammatory processes: nutrition or
pharmacology? Br J Clin Pharmacol 2013; 75(3):645-62.
8. Vetelainen R, van Vliet A, Gouma DJ, et al. Steatosis as a risk factor in liver surgery. Ann Surg
2007; 245(1):20-30.
9. El-Badry AM, Moritz W, Contaldo C, et al. Prevention of reperfusion injury and
microcirculatory failure in macrosteatotic mouse liver by omega-3 fatty acids. Hepatology
2007; 45(4):855-63.
10. Marsman HA, de Graaf W, Heger M, et al. Hepatic regeneration and functional recovery
following partial liver resection in an experimental model of hepatic steatosis treated with
omega-3 fatty acids. Br J Surg 2013; 100(5):674-83.
11. Linecker M, Limani P, Kambakamba P, et al. Omega-3 Fatty Acids Protect Fatty and Lean
Mouse Livers After Major Hepatectomy. Ann Surg 2017; 266(2):324-332.
12. Raptis DA, Limani P, Jang JH, et al. GPR120 on Kupffer cells mediates hepatoprotective
effects of omega3-fatty acids. J Hepatol 2014; 60(3):625-32.
13. https://compendium.ch/mpro/mnr/8980/html/de#7100. accessed, February 2019.
14. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal
with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;
240(2):205-13.
15. Slankamenac K, Graf R, Barkun J, et al. The comprehensive complication index: a novel
continuous scale to measure surgical morbidity. Ann Surg 2013; 258(1):1-7.
16. Linecker M, Limani P, Botea F, et al. "A randomized, double-blind study of the effects of
omega-3 fatty acids (Omegaven) on outcome after major liver resection". BMC Gastroenterol
2015; 15:102.
17. Yan JH, Guan BJ, Gao HY, et al. Omega-3 polyunsaturated fatty acid supplementation and
non-alcoholic fatty liver disease: A meta-analysis of randomized controlled trials. Medicine
(Baltimore) 2018; 97(37):e12271.
18. D'Eliseo D, Velotti F. Omega-3 Fatty Acids and Cancer Cell Cytotoxicity: Implications for Multi-
Targeted Cancer Therapy. J Clin Med 2016; 5(2).
19. Inoue-Yamauchi A, Itagaki H, Oda H. Eicosapentaenoic acid attenuates obesity-related
hepatocellular carcinogenesis. Carcinogenesis 2018; 39(1):28-35.
20. Iannelli A, Martini F, Schneck AS, et al. Preoperative 4-week supplementation with omega-3
polyunsaturated fatty acids reduces liver volume and facilitates bariatric surgery in morbidly
obese patients. Obes Surg 2013; 23(11):1761-5.
21. Clavien PA, Oberkofler CE, Raptis DA, et al. What is critical for liver surgery and partial liver
transplantation: size or quality? Hepatology 2010; 52(2):715-29.

18
22. Zhu XH, Wu YF, Qiu YD, et al. Liver-protecting effects of omega-3 fish oil lipid emulsion in
liver transplantation. World J Gastroenterol 2012; 18(42):6141-7.
23. Gong Y, Liu Z, Liao Y, et al. Effectiveness of omega-3 Polyunsaturated Fatty Acids Based Lipid
Emulsions for Treatment of Patients after Hepatectomy: A Prospective Clinical Trial.
Nutrients 2016; 8(6).
24. Ibrahim ES, Saleh SM, El Hoseeny M, et al. Effect of omega-3 on hepatic regeneration in adult
living donors undergoing hepatic resections for liver transplantation: A randomized
controlled trial. J Crit Care 2016; 31(1):157-62.
25. Zhang B, Wei G, Li R, et al. n-3 fatty acid-based parenteral nutrition improves postoperative
recovery for cirrhotic patients with liver cancer: A randomized controlled clinical trial. Clin
Nutr 2017; 36(5):1239-1244.
26. El-Badry AM, Jang JH, Elsherbiny A, et al. Chemical composition of hepatic lipids mediates
reperfusion injury of the macrosteatotic mouse liver through thromboxane A(2). J Hepatol
2011; 55(6):1291-9.
27. Ali FF, Rifaai RA. Preventive effect of omega-3 fatty acids in a rat model of stress-induced
liver injury. J Cell Physiol 2018.
28. Dossi CG, Gonzalez-Manan D, Romero N, et al. Anti-oxidative and anti-inflammatory effects
of Rosa Mosqueta oil supplementation in rat liver ischemia-reperfusion. Food Funct 2018;
9(9):4847-4857.
29. Li Z, Zhao F, Cao Y, et al. DHA attenuates hepatic ischemia reperfusion injury by inhibiting
pyroptosis and activating PI3K/Akt pathway. Eur J Pharmacol 2018; 835:1-10.
30. Grendar J, Ouellet JF, McKay A, et al. Effect of N-acetylcysteine on liver recovery after
resection: A randomized clinical trial. J Surg Oncol 2016; 114(4):446-50.
31. Hayashi Y, Takayama T, Yamazaki S, et al. Validation of perioperative steroids administration
in liver resection: a randomized controlled trial. Ann Surg 2011; 253(1):50-5.

19
20
21
22
Highlights

 This RCT fails to confer liver protection with Omega-3 fatty acids IV perioperatively
 No differences in complications were evident in the Omegaven vs. Placebo groups
 Subgroup analyses likewise did not detect differences in outcome after liver resection

23
24
Table 1. OmegavenTM RCT Patient and Surgery Characteristics

Omegaven Placebo
Variable P
n=132 n=129

Demographics
Age, years 59 (44-68) 60 (50-68) 0.583
Gender; male, n (%) 62 (47%) 67 (52%) 0.457
BMI, kg/m2 24 (22-27) 25 (23-28) 0.042

Indication for liver surgery

Liver tumor, n (%) 108 (82%) 109 (85%) 0.622
Benign tumor, n (%) 18 (14%) 16 (12%) 0.769
Malignant tumor, n (%) 90 (68%) 93 (72%) 0.502
CRLM, n (%) 35 (27%) 44 (34%) 0.225
HCC, n (%) 17 (13%) 13 (10%) 0.562
Cholangiocarcinoma, n (%) 22 (17%) 15 (12%) 0.288
Gallbladder cancer, n (%) 3 (2%) 3 (2%) 1.000
Other, n (%) 13 (10%) 18 (14%) 0.249
Other indications, n (%) 24 (18%) 20 (16%) 0.565

Type of liver surgery*

Open procedures, n (%) 102 (91%) 105 (93%) 0.612
Laparoscopic procedures, n (%) 10 (9%) 8 (7%) 0.612

Extent of resection*
Major liver surgery* 70 (63%) 76 (67%) 0.383
Minor liver surgery 42 (37%) 37 (33%) 0.383
Extended right hepatectomy, n (%) 7 (6%) 9 (8%) 0.615
Right hepatectomy, n (%) 26 (23%) 31 (27%) 0.455
Extended left hepatectomy, n (%) 5 (4%) 4 (4%) 1.000
Left hepatectomy, n (%) 15 (13%) 8 (7%) 0.190
Bisegmentectomy, n (%) 15 (13%) 23 (20%) 0.162
Central hepatectomy, n (%) 2 (2%) 1 (1%) 1.000
Segmentectomy, n (%) 6 (5%) 6 (5%) 1.000
Enucleation/Paracystectomy, n (%) 7 (6%) 5 (4%) 0.769
Wedge resections, n (%) 29 (26%) 26 (23%) 0.877

Intraoperative times*
Operation time, min 315 (240-415) 330 (249-405) 0.890
Inflow occlusion, n (%) 33 (29%) 32 (28%) 0.851
Inflow occlusion time, min 18 (15-30) 16 (14-38) 0.983

Table 1 represents an Intention-to-treat analysis except intraoperative data (*)

*, refers to a Per-protocol analysis excluding drop-outs (OmegavenTM, n=112; Placebo (n=113).
**, refers to resections ≥3 segments.
Abbreviations: BMI, Body mass index; CRLM, Colorectal liver metastasis; HCC, Hepatocellular
carcinoma. RCT, Randomized controlled trial.

25
Table 2. OmegavenTM RCT Outcome Characteristics

Omegaven Placebo
Variable P
(n=132) (n=129)

Postoperative Complications
Mortality, n (%) 5 (4%) 2 (2%) 0.447
Any complication, n (%) 61 (46%) 56 (43%) 0.709
No complications, n (%) 71 (54%) 73 (57%) 0.709
Major complications* (>IIIa), n (%) 29 (22%) 20 (16%) 0.206
Major complications* (>IIIb), n (%) 16 (12%) 13 (10%) 0.695
CCI** 17 (±23) 14 (±20) 0.417

Grade I 8 (6%) 12 (9%) 0.360

Grade II 24 (11%) 24 (19%) 1.000
Grade IIIa 14 (11%) 7 (5%) 0.172
Grade IIIb 9 (7%) 7 (5%) 0.798
Grade IVa 1 (1%) 3 (2%) 0.366
Grade IVb 0 (0%) 1 (1%) 0.494
Grade V 5 (4%) 2 (2%) 0.447

Type of complications
Biliary 13 (10%) 6 (5%) 0.152
Bleeding 4 (3%) 1 (1%) 0.370
Cardiac 4 (3%) 3 (2%) 1.000
Ileus 1 (1%) 3 (2%) 0.366
Infection 14 (11%) 15 (12%) 0.845
Liver failure 4 (3%) 2 (2%) 0.684
Pulmonary 4 (3%) 5 (4%) 0.747
Renal 2 (2%) 0 (0%) 0.498
Other 15 (11%) 21 (16%) 0.284

Transfusion, RBC units 0 (0-0) 0 (0-0) 0.868

ICU stay, days 0 (0-1) 1 (0-1) 0.578

Hospital stay, days 7 (7-9) 7 (7-10) 0.545

Peak AST, U/l 363 (207-697) 305 (206-582) 0.343

Peak ALT, U/l 332 (166-690) 318 (207-505) 0.855

Table 2 represents an Intention-to-treat analysis.

*, defined by the Clavien-Dindo classification of surgical complications.
**, Median 0 (0-27) and 0 (0-23), P=0.417 of the OmegavenTM and placebo groups.
Abbreviations: ALT, Alanine transaminase; AST, Aspartate transaminase; CCI, Comprehensive
complication index; ICU, Intensive care unit; RBC, Red blood cells; RCT, Randomized controlled trial.

26
Table 3. OmegavenTM RCT Mortalities

Hepatectomy Gender Age Onset Cause of death Cardiac

Patients
(Year) (POD) event

Omegaven
Pat. 1 Right Hx Male 85 8 Cardiac Unclear
Pat. 2 Right Hx Male 82 2 Cardiac Arrhythmia
Pat. 3 Right Hx Male 73 27 Liver failure
Pat. 4 (portal kinking)
Pat. 4 Central Hx Male 58 5 C. difficile
infection
Pat. 5 Right Hx Male 73 3 Cardiac Ischemia

Placebo
Pat. 6 Right Hx Male 71 26 Sepsis
Pat. 7 Right Hx Male 78 45 Unclear

Table 3 represents an Intention-to-treat analysis.

Abbreviations: POD, Postoperative day; Hx, Hepatectomy.

27