Immunonutrition with Omega-3 Fatty Acid Supplementation in Severe TBI: Retrospective Analysis of Patient Characteristics and Outcomes

Background Early evidence-based medical interventions to improve patient outcomes after traumatic brain injury (TBI) are lacking. In patients admitted to the ICU after TBI, optimization of nutrition is an emerging field of interest. Specialized enteral nutrition (EN) formulas that include immunonutrition containing omega-3 polyunsaturated fatty acids (n-3 PUFAs) have been developed and are used for their proposed anti-inflammatory and pro-immune properties; however, their use has not been rigorously studied in human TBI populations. Methods A single-center, retrospective, descriptive observational study was conducted at LAC + USC Medical Center. Patients with severe TBI (sTBI, Glasgow Coma Scale score ≤ 8) who remained in the ICU for ≥ 2 weeks and received EN were identified between 2017 and 2022 using the institutional trauma registry. Those who received immunonutrition formulas containing n-3 PUFAs were compared to those who received standard, polymeric EN in regard to baseline characteristics, clinical markers of inflammation and immune function, and short-term clinical outcomes. Results A total of 151 patients with sTBI were analyzed. Those who received immunonutrition with n-3 PUFA supplementation were more likely to be male, younger, Hispanic/Latinx, and have polytrauma needing non-central nervous system surgery. No differences in clinical markers of inflammation or infection rate were found. In multivariate regression analysis, immunonutrition was associated with reduced hospital length of stay (LOS). ICU LOS was also reduced in the subgroup of patients with polytrauma and TBI. Conclusion This study identifies important differences in patient characteristics and outcomes associated with the EN formula prescribed. Study results can directly inform a prospective pragmatic study of immunonutrition with n-3 PUFA supplementation aimed to confirm the biomechanistic and clinical benefits of the intervention.


INTRODUCTION
Traumatic brain injury (TBI) is a signi cant cause of morbidity and mortality that affects all populations, regardless of gender, age, ethnicity, race, and socioeconomic background.Severe TBI results in > 60,000 deaths in the United States (US) annually and over 5 million persons living with TBI-associated disability 1,2 .First-year healthcare costs are estimated to be $2 billion in survivors 2 .Despite the large health and socioeconomic impact, there are still no Level 1 recommendations for medical interventions to improve outcomes in moderate and severe TBI (sTBI) 3 .The pathophysiology of sTBI is complex, with several early pro-in ammatory molecular and cellular changes occurring in the immediate and delayed phases after trauma 4 .Although these pathways are thought to be largely bene cial to support initial reparative processes, they may also contribute to adverse clinical symptomatology, metabolic stress, and disease progression if left untreated.In this context, early in-hospital interventions targeting in ammation and metabolism may prevent secondary brain injury and systemic complications that contribute to poor clinical outcomes after TBI.
Optimization of nutritional status with early (within 48-72H of admission) enteral nutrition (EN) is an emerging therapeutic strategy in critically ill and trauma populations 5,6 .The goal is to both prevent malnutrition and provide essential nutrients to promote cellular and neurologic recovery.After sTBI, malnutrition occurs in up to 76% of ICU patients and is associated with increased mortality and poor functional outcomes [7][8][9] .Its high incidence is multifactorial, with important contributions from dysphagia, increased metabolic demands, altered digestive and metabolic function, and variable feeding practices while in the ICU.Together, critical illness and malnutrition in TBI result in a pro-in ammatory, immunosuppressed and catabolic state that predisposes individuals to complications such as fever and nosocomial infection, which are both independent predictors of poor outcome 10,11 .Early nutritional support with specialized dietary formulas that include immunonutrition with omega-3 polyunsaturated fatty acid (n-3 PUFA) and amino acid supplementation is a novel approach to optimize nutrition and promote recovery after TBI; however, their clinical effect has not been rigorously studied in acute neurologic disease.In pre-clinical models, administration of in ammation-resolving bioactive lipids, including the n-3 PUFAs eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), has robust evidence for neuroprotection, anti-in ammatory properties, and improved cellular energetics 12 .
The purpose of this descriptive retrospective observational study was to compare the population characteristics of sTBI patients who received EN immunonutrition with n-3 PUFA supplementation versus those who received standard, polymeric EN which did not contain the added components of immunonutrition.The supplementation of n-3 PUFAs in TBI is an appealing therapeutic strategy given their proposed anti-in ammatory effect, widespread availability, safety pro le, relative low-cost and general feasibility in implementing into ICU standards of care.When incorporated into universal strategies to optimize nutritional status, such novel approaches might lead to meaningful impact on patient and hospital outcomes.The results of this study are anticipated to inform future prospective clinical trials investigating the clinical effect of immunonutrition including n-3 PUFA supplementation in moderate-to-severe TBI and other acute neurologic diseases.

Study Population
This is a single-center retrospective observational study of sTBI conducted in Los Angeles, California.The single study site, Los Angeles General Medical Center (formerly named LAC + USC Medical Center at the time of patient admission) is a high-volume Level 1 trauma center and the largest safety net hospital in the Western United States.This descriptive study was considered hypothesis-generating, aimed to characterize differences in patient populations based on interventions provided as clinically indicated, with no primary hypothesis testing.
The target population were patients with sTBI who received EN in the ICU.Study subjects were identi ed from the institutional trauma registry that is maintained by the Division of Trauma and Acute Care Surgery.A ve-year period between July 2017 and July 2022 was studied.Sample study subjects met the following inclusion criteria: age ≥ 18 years, diagnosis of sTBI (history and radiographic ndings consistent with TBI, Glasgow Coma Scale score [GCS] ≤ 8), ≥ 14 day ICU length-of-stay (LOS), prescribed and received EN during the majority (> 7days) of the 14-day ICU period.No time cut-off for initiating EN was used.Patients who presented with GCS = 3 with dilated pupils, or those who did not tolerate planned EN initiation due to gastric dysmotility were excluded from the analysis.These inclusion and exclusion criteria were chosen to ensure that study subjects were su ciently exposed to the intervention of interest.The study population included both isolated head trauma (IHT) and polytrauma with TBI.It was preplanned to study subgroups of IHT and polytrauma to determine how these populations might differ in their baseline characteristics and response to EN.

Intervention
The intervention of interest was EN using an immunonutrition formula containing n-3 PUFAs from sh oil.
Those who received EN with immunonutrition, either via gastric or post-pyloric feeding tube, were compared to those who received standard polymeric EN formulas not classi ed as immunonutrition.
Immunontrition formulas have been developed for their proposed anti-in ammatory properties, to promote immune system function, and support biosynthesis and protein needs in high-risk critically ill and trauma patients 13 .At the study site, EN products are supplied by Abbott Nutrition®.Peptide-based products classi ed as immunonutrition that were prescribed to study subjects were Vital High Protein™ (containing EPA = 2.3g/L, DHA = 0.9 g/L, vitamin D, C and E), Vital AF 1.2 Cal™ (containing EPA = 2.7g/L, DHA = 1.1 g/L, vitamin D, C and E) and Pivot™ (containing EPA = 2.6 g/L, DHA = 1.1 g/L, arginine, glutamine, vitamin C and E).In addition to n-3 PUFAs, formulations also contain essential vitamins and amino acids as listed to support immunomodulation, antioxidant activity and cellular peptide biosynthesis.Given the retrospective, observational nature of this study, the choice of speci c EN formulation used, time to initiation, dose and duration were determined by clinical judgement.At the study site, a clinical registered dietitian is involved in the nutritional management of ICU patients who require EN, including determining the formulation prescribed, goal dose rate to meet caloric needs, and reinitiation schedule after pauses in EN delivery.Although not protocolized at our institution, immunonutrition formulas are often chosen for patients with high protein requirements due to critical illness or wound healing as described.

Data Collection and Study Variables
The institutional trauma registry was queried to identify patients meeting inclusion criteria.The sample population consisted of remaining subjects after inclusion and exclusion criteria were applied.Data that was not contained in the trauma registry was retrospectively collected from the electronic medical record (EMR) by study personnel.Demographic and baseline characteristics including co-morbidities were obtained, as well as the EN product used and time-to-initiation.For outcome measures, clinical markers of in ammation were recorded, including peak 14-day values of the common acute phase reactants Creactive protein (CRP, mg/L) and white blood cell count (WBC, x10 9 /L), and measures of fever occurrence.
Both the number of calendar days with fever (any occurrence of temperature > 37.8 o C) and cumulative fever burden (arithmetic sum of daily peak fever > 37 over the 14-day study period) were collected.This measure of fever burden has been validated as an independent predictor of outcomes after TBI 10 .Patient outcomes, including incidence of culture-positive infection during the 14-day ICU period, ICU and hospital LOS, early GCS recovery (post-TBI day #7 GCS), and discharge status were analyzed.The outcome measures were chosen based on data that could be feasibly collected through retrospective EMR review.This study was designated IRB exempt by the University of Southern California Institutional Review Board.

Statistical Analysis
Data normality for continuous variables was assessed by histogram and Shapiro-Wilk test.Univariate analyses for baseline differences by study groups were conducted by Chi-square test for categorical variables and independent t-test or Wilcoxon rank sum test for continuous variables, depending on data normality.The comparison of inpatient mortality rate was conducted using multivariate Poisson regression with log link function and adjusted for confounders which were de ned as baseline variables with large differences between groups that were clinically important but not necessarily reaching a p < 0.05 threshold.There were three variables that met the de nition of confounder to the intervention effect: age, gender, and occurrence of non-central nervous system (non-CNS) acute surgery.Using the log link function for Poisson modeling, the reported associations were interpreted as relative risks (RRs).The overdispersion assumption for Poisson regression was investigated through the goodness of t test using Pearson Chi-square/DF ratio.When this ratio was larger than 1, a negative binomial distribution was used.Differences in hospital and ICU LOS were tested by multivariate generalized linear models and adjusted for potential confounders as described above.The interaction between IHT and polytrauma subgroups and intervention groups was tested in both models.A strati ed intervention effect by IHT was illustrated for trend of effect modi cation.Missing data for laboratory values was excluded from the analysis.SAS 9.4 was used for all data analyses.

RESULTS
A total of 151 patients with sTBI were analyzed (Figure 1), of which 61 patients had isolated TBI and 90 patients had polytrauma with associated TBI.Baseline characteristics are shown in Table 1.In the study population, the average age was 43 years, 85% were male and 60% were classi ed as Latinx/Hispanic.94 patients (62.3%) received EN with immunonutrition containing n-3 PUFA supplementation and 57 (37.7%) patients received standard, polymeric EN formulations.Compared to the standard EN group, the treatment group who received immunonutrition with n-3 PUFA supplementation were slightly younger (41.2 v 46.5, p=0.07) and were more likely to be male and Latinx/Hispanic.The immunonutrition group more commonly sustained motor vehicle accident as the mechanism of injury and were signi cantly more likely to have polytrauma and require emergent non-CNS surgery.There were no statistically signi cant differences in baseline co-morbidities between the primary study groups and initial GCS was nearly identical.Vital AF TM was the most prescribed immunonutrition formula and was used in 47.9% of patients in the treatment group.Baseline weights and time-to-initiation of EN between study groups was similar (49.6H in the immunonutrition group v 52.8H in the standard group).
In unadjusted univariate analysis, peak 14-day ICU WBC and CRP values were similar between study groups (Table 2).Both 14-day cumulative fever burden (15.1 v 15.2, p=0.75) and calendar days with fever (8.4 v 8.6, p=0.74) were no different in those prescribed immunonutrition versus standard EN formulations.The observed culture-positive infection rate was lower in the immunonutrition group compared to the standard group, but this did not reach statistical signi cance (80.9% v 86.0%, p=0.42).Individually, there was no signi cant difference in rates of cerebrospinal uid, respiratory, blood, urine or wound infections.The presence of ileus on ICU day 14 was similarly rare in both groups, and the number of calendar days with NPO status was nearly identical (3.7 days with immunonutrition v 4.0 with standard EN).Those receiving immunonutrition with n-3 PUFA supplementation gained on average 1.4kg over the 2-week period versus a 0.1kg weight loss in the standard EN group, but this difference was not statistically signi cant (p=0.41).Early GCS recovery (day #7 GCS) was similar between study groups (6.0 v 5.8, p=0.58).Both ICU and hospital LOS were decreased in the immunonutrition group, but this was not statistically signi cant.In-hospital mortality was lower in the immunonutrition group, but this was also not statistically signi cant (16.0%versus 19.3%).

DISCUSSION
In this retrospective descriptive cohort study of sTBI conducted at a single Level-I trauma center, those who received immunonutrition with n-3 PUFA-containing formulas for EN were slightly younger and more likely to be male, Hispanic/Latinx, have sustained non-CNS injury in addition to TBI, and have undergone emergent non-CNS surgery due to their injuries.The greater likelihood of being prescribed n-3 PUFAcontaining immunonutrition in patients with polytrauma and in those requiring non-CNS surgery is not unexpected, given the high protein requirements believed to be necessary to support increased metabolic demand and promote wound healing after general traumatic injury and surgery 8,14 .Further, the results also suggest that in clinical practice, the decision to prescribe EN with immunonutrition is not based on the presence of trauma alone.Speci cally, older patients, female gender, and those with isolated TBI may not be considered to require similar nutritional interventions compared to younger male patients with polytrauma.Particularly for isolated TBI, this might be an incorrect assumption, as sTBI patients may need up to 200% of their baseline metabolic requirements 15 , and with similarly high rates of malnourishment (68-76%) reported in both polytrauma and TBI populations 8,16 .In addition, given in ammation and infectious risk are universal regardless of trauma type, further clinical study is needed to better de ne the nutritional needs and clinical targets for TBI patients, especially in those at high-risk for malnutrition and poor functional outcomes.
Immunonutrition with n-3 PUFA supplementation compared to standard, polymeric EN formula was not associated with any statistically signi cant differences in clinical markers of in ammation and immunity.Speci cally, peak 14-day WBC and CRP, 14-day fever burden and calendar days with fever were nearly identical between study groups.Although the incidence of culture-positive infection was lower in those receiving immunonutrition, this did not reach statistical signi cance (80.9% versus 86.0%, p = 0.42).Given the proposed mechanism of immunonutrition with n-3 PUFA supplementation is to attenuate systemic in ammation and promote immune function, these results might be interpreted as unexpected; however, several circumstances should be considered.These outcome measures were chosen because they are assessable from retrospective medical record review but are likely non-speci c and crude measures of relevant anti-in ammatory activity that may be altered by several confounding factors.The study population was also severely injured (average GCS = 4.8), which may have blunted the effect of the nutritional intervention.
The culture-positive infection rate found in this study is higher than many hospital-acquired infection rates reported in moderate-severe TBI 17,18 and hospitalized trauma patients 19 .Both trauma and TBI are thought to increase the risk of in-hospital infection 20,21 , while multiple surgical interventions further increase infection risk 22 .Given the features of our study population, subjects were at high infection risk.Lastly, rates of infection in TBI populations are based on early retrospective report, while true infection risk is likely underestimated.Our reported rate of lower respiratory infection (78.2%) is comparable to recent literature described in a prospective surveillance study (72.2%) 23 .Given nosocomial infection after TBI is associated with worse outcomes 17,21 , the results also highlight the clinical problem of infection risk following TBI and the need to identify novel approaches to prevent this common complication.
Despite no identi ed differences in clinical markers of in ammation and immunity, there were important differences in clinical outcomes observed between the two study groups.In multivariate linear regression analysis, EN with immunonutrition and n-3 PUFA supplementation resulted in 10.7 fewer hospital days compared to standard EN (-19.3, -2.0, p = 0.02).In pre-planned subgroup analysis, this appeared to be primarily driven by differences in those with polytrauma (-28.3,-4.9, p = 0.006).Similarly, ICU LOS was signi cantly shorter in the polytrauma subgroup.Although there was no statistically signi cant difference with in-hospital mortality between study groups in multivariate logistic regression analysis, there was a 0.7 rate ratio for mortality with treatment compared to standard nutrition (-0.6, 1.9, p = 0.53).From this study, it is unclear why immunonutrition with n-3 PUFA supplementation is associated with positive outcomes, but the results are su ciently promising to warrant prospective study in clinical e cacy trials in this subgroup of the trauma population.It is possible that immunonutrition with n-3 PUFAs positively impacts energetics and optimizes nutritional status in ways that we were not able to capture in a retrospective study.Particularly for n-3 PUFA supplementation, the pre-clinical data is robust.In rodent models of TBI, eicosanoid n-3 PUFA therapy has been demonstrated to be neuroprotective against in ammation-mediated oxidative stress and pro-apoptotic signaling after TBI [24][25][26] , and can reduce proin ammatory signaling, maintain mitochondrial integrity, and reduce infarct size in stroke models [27][28][29] .Additional amino acid components of immunonutrition, speci cally arginine and glutamine, have been shown to support the synthesis of several cellular proteins vital to oxidative metabolism, immune function, and cellular repair [30][31][32] .
Although this was not a prospective study of early EN (< 48-72H from admission), prescription of enteral immunonutrition initiated with a mean of 50.8H after admission in our study population appeared safe.
Although a comprehensive evaluation of ICU complications was out of the scope of this study, the use of early EN immunonutrition therapy was not associated with weight loss, increased NPO days, day 14 ileus, or increased infection risk compared to standard therapy.This was an important nding as results from the general ICU population have been con icting regarding the safety of early EN, partly due to a concern for aspiration-related respiratory infections 33 .The capability to resolve in ammation without compromising immune function is one of the positive aspects of n-3 PUFA supplementation and immunonutrition, and because these components are naturally occurring with no known serious adverse effects, their supplementation is considered safe with a large therapeutic window 12 .Early EN may be a clinically important intervention in TBI.It is estimated that every 10 kcal/kg/day increase in energy intake is associated with a 30-40% reduction in mortality risk 34 .
Important limitations to this study should be considered.In our retrospective analysis, data quality was dependent on the accuracy of electronic documentation.Given this anticipated limitation, study outcome measures were selected that were objective and were commonly documented as part of standard practice.This study lacked more precise measures of nutritional status, which could not be accurately measured by chart review.Additionally, due to the retrospective nature of the study, we were unable to control all potential sources of bias and effect modi cation.The justi cation of selecting one nutrition intervention versus another was not standardized or documented.Nutrient components also differ between each EN formula, and we could not accurately control for % caloric, protein, fat and carbohydrate needs met in retrospective analysis.Also, those who received EN with immunonutrition including n-3 PUFA supplementation may have received differential management compared to those receiving standard therapy based on the patient characteristics discussed.As with any observational study, correlation does not prove causation.Given the stringent inclusion criteria, the study population was relatively small.A larger study sample would provide a greater power in detecting clinically meaningful differences, especially in subgroups.Future blinded, randomized trials are needed to elucidate the direct impact of immunonutrition formulas on the in ammatory response and short and long-term patient outcomes after moderate-severe TBI.Regarding generalizability, our US-based study population represented a very severe TBI cohort with a large proportion of Latinx and male subjects and may not be fully generalizable to other patient populations and regions of practice.The practicality of the intervention may also be in uenced by the availability of speci c enteral nutrition products in different sites; however, n-3 supplementation products are widely available in many countries and are relatively low-cost.
The emerging science in both pre-clinical and human populations of both n-3 PUFA supplementation and other amino acid immunonutrition warrants further study in prospective, pragmatic, randomized clinical e cacy trials.Additional correlations with biomarkers will be helpful in describing the molecular mechanisms as how immunonutrition improves clinical outcomes and to optimize proposed nutritional interventions.If mechanistic activity and positive impact on clinical outcomes can be demonstrated in clinical study, routine incorporation of n-3 PUFA supplementation and immunonutrition with EN would be a novel approach to optimize ICU nutritional status and promote recovery after moderate or severe TBI in thousands of patients annually.Products that provide enriched amounts of n-3 PUFAs and amino acids are already in clinical practice, are widely available, and are relatively low-cost, supporting their rapid translation and implementation into standards of ICU practice if bene t can be shown.Guideline recommendations do provide some support for these interventions in general ICU populations 35,36 , but further rigorous study is needed in patients with TBI 37 .

CONCLUSIONS
This study identi es patient characteristics of those who received immunonutrition EN with n-3 PUFA supplementation compared to standard, polymeric formulas.The use of immunonutrition was associated with important differences in hospital LOS and ICU LOS; however, the cause of this observed difference cannot be concluded from this study.Together, this suggests that there are potentially meaningful differences in patient characteristics and outcomes associated with the type of enteral feeding provided after severe TBI.A prospective, pragmatic study of the intervention is warranted to con rm the biologic mechanisms of treatment and elucidate their impact on patient outcomes.Our results directly inform the future design of such studies in TBI and other acute neurologic disorders.Given the burden of disease and limited medical treatment options after moderate-to-severe TBI, novel approaches are urgently needed.

Declarations Table 2: Unadjusted Outcomes by Study Group
Results are displayed as means+SD for continuous variables and % for categorical variables.CRP was analyzed in 63 patients with data available, 38 in the immunonutrition group and 25 in the standard group.WBC: white blood cell count, 10 9 /L; CRP: C-reactive protein, mg/L; NPO: nothing-by-mouth, no enteral nutrition given; Wt: weight; GCS: Glasgow Coma Scale score; SNF: skilled nursing facility; LOS: length-of-stay; ICU: intensive care unit.

Table 3: Multivariate Regression Analysis of Outcomes in Both the Whole Study Population and Subgroups
Results are displayed as rate ratio (95% con dence interval).LOS: length-of-stay; TBI: traumatic brain injury; ICU: intensive care unit.

Figures Figure 1 Flow
Figures