Parenteral supplementation of omega-3 fatty acid on the prognosis of sepsis: a real world retrospective study

Objectives: The role of omega-3 fatty acids in the treatment of sepsis is always on paradox. we tried to retrieve and download the patients’ data in a certain period through the hospital information system, used data sorting so as to screen out the patients with sepsis so as to nd out the role of omega-3 fatty acids in sepsis. Methods: Through the hospital information system, retrieve and include the patients who were admitted the Department of Shenzhen Hospital from 2016 to June 2019, screen out patients diagnosed with sepsis according to a certain criterion. The patients were grouped by whether they were applicated with omega-3 fatty acid or not. Results: A total of 1733 cases included into analysis, among of whom 303 cases were applicated with omega-3 fatty acid. The amounts and baseline conditions between both groups were imbalance. Severity of omega-3 fatty acid group was higher than that of control group. Chi-square test found that the mortality rate of omega-3 fatty acid was higher than that of control group (p < 0.0001). But age, gender, whether there is abdominal infection, whether there is septicemia, shock, the need for mechanical ventilation, and the need for renal replacement therapy may all affect the prognosis of the patients. If these factors were used as covariates, multiple logistic regression analysis showed that there was no signicant difference in mortality rate between the treatment group and the control group (P = 0.574). Survival analysis showed that the survival rate of treatment group was higher than that of the control group when at the end of total treatment duration (P = 0.035). Conclusion: For patients with more severe sepsis, doctors are more likely to use omega-3 fatty acids in the early stage. Omega-3 fatty acids may improve the long-term prognosis of sepsis, but the conclusion still needs to be accepted carefully.

Parenteral supplementation of omega-3 fatty acid on the prognosis of sepsis: a real world retrospective study Chen

Background
The role of omega-3 fatty acids in the treatment of sepsis is always on paradox. In terms of mechanism, it seems that omega-3 fatty acids can inhibit the in ammatory response [1], while sepsis is just due to the strong in ammatory response induced by serious infection, leading to further failure of multiple organs, which causes the patient's condition inter critical state, and nally dies. Omega-3 fatty acids have antiin ammatory effects, including interaction with leukocyte chemotaxis, adhesion molecule expression, and leukocyte endothelial cell adhesion. And it changes cell phospholipid lipid rafts by replacing n-6 molecules on lipid rafts with n-3, inhibiting in ammatory gene expression and activating antiin ammatory transcription factor peroxidase [2]. However, the outcomes of clinical studies are still quite different. Previous meta-analysis and systematic reviews were used to evaluate the e cacy of omega-3 fatty acids in the treatment of sepsis. Mo et al. Included 721 patients in 12 randomized controlled trials (RCT). It was found that omega-3 fatty acid supplementation can reduce the mortality of sepsis patients and shorten the ICU stay of these patients [3]. However, our team has used the method of randomized controlled trial to evaluate two different series of patients with sepsis and intestinal failure, and found that parenteral supplementation of omega-3 fatty acid has unpredictable outcomes on prognosis of sepsis [4,5]. In particular, in the second randomized controlled trial, we found that omega-3 fatty acids did not reduce the 28-day mortality rate of sepsis patients. However, in 60 days, we can see that the treatment group has a lower mortality rate. The treatment group has a lower total T-lymphocyte of CD3 at the beginning of the treatment, while in 7 days after the treatment, the two groups tend to have the same counts of lymphocyte [5], which can be referred to It is suggested that omega-3 fatty acids have a certain recovery effect on T lymphocytes in sepsis and improve long-term prognosis? Later, we also conducted another meta-analysis, which included more than twice as many randomized controlled trials as Mo's study, to evaluate the effect of omega-3 fatty acid on survival outcome of sepsis and sepsis induced acute respiratory distress syndrome (ARDS). However, it was found that the mortality of sepsis patients did not improve whether they were supplemented with omega-3 fatty acids enterally or parenterally [6].
With the advent of the era of big data, we tried to retrieve and download the patients' data in a certain period through the hospital information system, used data sorting so as to screen out the patients with sepsis, and grouped them according to whether they use omega-3 fatty acids as a parenteral supplement or not, so as to analyze and try to nd out the role of omega-3 fatty acids in sepsis.

Methods
Inclusion criteria: Through the hospital information system, retrieve and include the patients who were admitted to the Department of critical medicine of Shenzhen People's Hospital from December 2016 to June 2019, screen according to the following clinical characteristics of patients (TABLE-1), exclude the non-sepsis patients, and then con rm that the patients included in sepsis meet the ICD-9 sepsis diagnosis standard through the repeated checking of the diagnosis by researchers. This study was carried out with the consent of the ethics committee of the hospital. Each patient signed the informed consent form agreeing to accept the obligations of medical treatment, teaching, scienti c research and other aspects of the hospital at the beginning of admission. Shock*** MAP < 60mmHg, or norepinephrine is needed.
Abnormal renal function** SCr > 144umol/L or continuous renal replacement treatment is needed.
Note: * Necessary condition. ** Two positive items plus * may be considered for sepsis diagnosis. *** One positive item plus * may be considered for sepsis diagnosis. **** Reference condition.

Clinical intervention and grouping
The selected patients were diagnosed and treated according to their clinical conditions, Clinical decisions included: microbial culture, antibiotics, uid resuscitation, mechanical ventilation, blood puri cation, etc., and blood samples were taken for examination according to the needs of patients. The blood routine, liver and kidney function, coagulation function and other indicators of the patients were monitored. The doctor would decide whether omega-3 fatty acids were injected to the patients or not by their own judgement. And patients were grouped according to whether they used omega-3 fatty acids or not. At present, omega-3 fatty acids are sh oil fatty acid preparations (re ned sh oil preparations). 10 g / day of sh oil was given intravenously, which was used together with other fat preparations (1:5). The treatment durations of patients were planned by doctors. This study is a retrospective study, and there is no intervention on the start and stop of the treatment plan of patients.

Statistical analysis
The parameters of normal distribution are expressed by means of mean plus minus standard deviation, and the values of non-normal distribution are recorded by means of median. According to whether sh oil is used for grouping, t-test of independent samples or Mann Whitney test is used to assess the differences between the two groups. Multivariate logistic regression analysis was used to evaluate the prognostic effect of other factors on the treatment of sepsis. Draw KM survival curve were drawn and the difference between the two groups of curves were analyzed. The relationship between the total amount of omega-3 fatty acids and the time of ventilator use and the length of stay in ICU was evaluated by tting curve and bubble chart. SPSS 20.0 for window and R software package were used to analyze the data, and P < 0.05 was taken to indicate that the difference between the groups was statistically signi cant.

Results
Patient characteristics and baseline condition Mortality 96 cases died in the omega-3 fatty acid treatment group, with a mortality rate of 31.68%, while 1430 cases died in the control group, with a mortality rate of only 20.00%. The mortality rate of the treatment group was 11.68% higher than that of the control group. Chi-square test shows that the difference between both groups was signi cant (p < 0.0001, see Fig.1A). Omega-3 fatty acid was applicated in early state in most patients (see Fig.1B). However, by observing the Take the death of patients as the clinical outcome, the time that patients transferred out or died were calculated as the nal date of observation. KM curves were made, and compared. The results showed that the survival rate of omega-3 fatty acid treatment group was lower than that of the control group when up to the 30 th day (P = 0.007), while to the 60 th day outcome suggested that the survival rate of omega-3 fatty acid treatment group was slightly overturned (P = 0.062). The survival rate of treatment group was higher than that of the control group when at the end of total treatment duration (P = 0.035).

Discussion
Sepsis is a serious syndrome. It is estimated that there are 31.5 million cases of sepsis and 19.4 million cases of sepsis in the world, and 5.3 million people may die every year, which has become a heavy burden in the world [7]. Sepsis is newly de ned as fatal multiple organ failure (MODS) caused by severe infection, which weakens the role of systemic in ammatory response syndrome (SIRS) [8]. However, in ammation still plays an important role in the occurrence and development of sepsis. The MODS and death induced by sepsis are mainly attributed to the strong in ammatory response at the initial stage of severe infection and the anti-in ammatory response at the later stage Reaction imbalance, and the complex interaction between in ammation and anti-in ammatory response [9]. It can be seen that the position of in ammation cannot be ignored, and the search for appropriate anti-in ammatory drugs has been the direction of clinical scientists. Among them, omega-3 fatty acid is one of the controversial drugs, but also a nutritional preparation. In this study, 303 patients were treated with omega-3 fatty acids, most of whom were applicated in the early stage of the disease. Although factor analysis suggested that the mortality rate of omega-3 fatty acid patients was higher than that of the control group, multiple factors suggested that the disease severity of these patients was higher than that of the control group. This suggests that doctors may be more inclined to use omega-3 fatty acids in patients with relative crisis. In fact, there was no difference in mortality between the two groups when these factors imbalanced the baseline interacting with application of omega-3 fatty acid. On the contrary, patients treated with omega-3 fatty acids seem to have a better clinical prognosis as time goes on and none of these randomized trials reported safety concerns. From the long-term prognosis improvement of sepsis patients after omega-3 fatty acid treatment, it seems to echo the results of our earlier randomized controlled trials [5].
Eicosapentaenoic Acid (EPA) and docosahexaenoic acid (DHA). Low level of AA is an important determinant of prognosis in sepsis patients [10]. The rats were given diets of homocysteine, homoomega-3 fatty acid group and non-omega-3 fatty acid group respectively before the injection of endotoxin, the rats showed different degrees of weight loss and tissue damage among groups. The liver enzyme and pathological outcomes of the rats receiving omega-3 fatty acid diet were better than those of the other diet groups [11]. N-3 fatty acid docosahexaenoic acid (DHA) was used to treat mouse bone marrowderived dendritic cells, and different toll like receptor (TLR) ligands were used to stimulate them. Flow cytometry was used to detect cell surface maturation markers and intracellular activity. The expression and secretion of cytokines were detected by real-time RT-PCR and ELISA. DHA maintains the immature phenotype of bone marrow-derived DC by preventing the up regulation of MHCII and costimulatory molecules (CD40, CD80 and CD86) and maintaining high level of endocytosis activity. It also inhibited TLR2, 3, 4, and 9 ligand stimulated dendritic cells (DCs) to produce pro-in ammatory cytokines, including IL-12 cytokine family (IL-12p70, IL-23, and IL-27). The inhibit effect of IL-12 expression is through activating PPAR γ and inhibiting NFKAPABP65 translocation. These evidences show that DHA has antiin ammatory effect in vivo [12].
It was found that the DCs were depleted in both the cecal ligation and puncture mice model and the sepsis patients. This process is related to the increase of apoptosis. The loss of DCs caused by sepsis occurs after the activation of CD3+, CD4+ T cells and the loss of lymph nodes. Before the loss of DCs, there is no continuous increase in its mature state. Mature and immature DCs are easy to be lost. CD8+ DCs has priority loss in local and remote lymph nodes [13]. This suggests that DCs plays a key role in sepsis. Recently, it has been found that in the early stage of in ammation, the activated DCs are characterized by the decrease of antigen cross presenting ability of newly discovered antigens and the production of immunogenic cytokines. The immunosuppression induced by sepsis is mainly due to the depletion of mature immature dendritic cells. However, the late development of immune tolerance can lead to the release of tumor necrosis factor inhibitory cytokines by DCs, which is involved in maintaining the local tolerance environment characterized by Treg cell aggregation [14]. By targeting DCs, it is found that the loss of DC quantity and function caused by sepsis is one of the reasons for the de ciency of CD8 T cell immune function, and the treatment of improving the state of DCs after sepsis may be helpful to the recovery of immune function of CD8+ T cell [15]. Pretreatment of DCs with DHA can prevent the maturation of DCs induced by LPS, maintain the low expression of costimulatory molecules and the lack of proin ammatory cytokines (IL-12p70, IL-6 and IL-23). T cells co cultured with DC-DHA expressed high levels of TGF β and Foxp3, but showed no functional Treg phenotype. Similar to the results of in vitro experiments, the bene cial effect of DHA is related to the decrease of CD4+ T cells, which was caused by the decrease of IFN γ and IL-17 production in spleen and central nervous system [16].
In addition, omega-3 fatty acids also have effects on other in ammatory cells. It has been found that it can improve the pathogens clearance ability of neutrophils [17], which is mainly related to DHA. EPA can inhibit the migration of neutrophils to the focus. Endothelial cells can participate in the migration of neutrophils by producing prostaglandin (PG) D2. When PG D2 is combined with receptor DP-1 on neutrophils, it will lead to the adhesion and migration of neutrophils. However, the endothelial cells pretreated by EPA may reduce the production of PG D2 and increase the production of PG D3, thus inhibiting the migration of neutrophils [18]. At the same time, omega-3 fatty acids have multiple double bonds in their carbon chain. Because each double bond causes the carbon chain to bend, the accumulation of polyunsaturated fatty acids in the cell membrane cannot be as tight as that of saturated fatty acids. This increases the uidity of immune in ammatory cells, reduces brittleness, and plays an important role in prolonging cell life [19]. This effect can also appear in other different kinds of immune cells, and it can adjust the immune mechanism of patients. Because leukocyte chemotaxis, migration and pathogen clearance are mainly in the early stage of in ammatory response, omge-3 fatty acids may affect this cell behavior in the early stage.
In the long run, omega-3 fatty acids may inhibit DNA methylation of in ammatory cells. Fatty acids can be modi ed by DNA methylation in vitro, while in vivo studies showed that total DNA methylation and PDK4 speci c DNA methylation were positively correlated with eicosapentaenoic acid and arachidonic acid. There was a negative correlation between HDAC4 methylation and arachidonic acid [20]. One study found that 174 patients with Alzheimer's disease (AD) received 1.7g DHA and 0.6g EPA or placebo every day for 6 months. Two of the four CpG sites of peripheral blood leukocytes (PBLs) were signi cantly reduced in methylation. Hypomethylation at CPG2 and CPG4 sites was negatively correlated with the change of plasma EPA concentration, but not with the change of plasma DHA concentration [21]. However, it is not clear whether omega-3 fatty acids can also play a role in DNA demethylation of blood cells in in ammatory stage of sepsis. However, from bioinformatics research, it was found that the methylation of cg01770232 in IL-6 promoter was related to the increase of IL-6 concentration, while the higher concentration of omega-3 fatty acids inhibited the methylation of cg01770232 in IL-6 promoter, thus also inhibited the expression of IL-6. The relationship between n-3 polyunsaturated fatty acids and cg01770232 methylation depended on rs2961298 genotype [22]. IL-6 is an early indicator of in ammatory reaction, which is caused by severe infection. Patients can soar within 24 hours, then rapidly decline. It is a sensitive and speci c indicator for early diagnosis of sepsis [23]. Therefore, the increase of serum concentration of omga-3 fatty acids can possibly inhibit DNA methylation process and early in ammation.
Although omga-3 fatty acids may improve the prognosis of sepsis patients, the dosage and regime are not clear. In our department, the formula of n3: n6 = 1:5 is generally used. The total dose of omega-3 fatty acids used by each patient is different. This study is the rst attempt to use the hospital information system to obtain data, collate and analyze, so there is also a serious imbalance of amounts of patients between the control group. In this retrospective study, the treatment plan could not be intervened, and the existing data could not be used to evaluate the patients and the side effects. Nevertheless, we still try our best to sort out and evaluate the patients' condition from the aspects of organ function and auxiliary treatment. Due to the different severity of the patients' condition, only statistical methods can be used to adjust the data of the patients. Pairing study may be helpful to obtain more accurate results and conclusions.

Conclusion
For patients with more severe sepsis, doctors are more likely to use omega-3 fatty acids in the early stage. Omega-3 fatty acids may improve the long-term prognosis of sepsis, but the conclusion still needs to be accepted carefully. Because the clinical condition of patients is of difference, the results of matching research may be able to draw more accurate conclusions.

Declarations
Ethics approval and consent to participate fatty acid treatment. In most patients, omega-3 fatty acid was applicated in early state of treatment. Survival analysis (KM curves) (2A up to the 30th day 2B up to the 60th day 2C at the end of total treatment duration.