Abstract
OBJECTIVES:
Linoleic acid (LA) is the most abundant polyunsaturated fatty acid in diet, and it is a precursor to inflammatory lipid mediators called oxylipins. The role of LA and its oxylipins in pediatric sepsis and organ injury is uncertain. Recently, pediatric sepsis phenotypes were described, with phenotype D characterized by the highest proportion of acute kidney injury (AKI), multiple organ failure, and risk of death. We aimed to test the hypothesis LA may play a role in sepsis-associated organ dysfunction. We therefore investigated whether increasing plasma LA and LA-derived lipoxygenase oxylipins are associated with sepsis phenotype D and with AKI in a cohort of critically ill children with sepsis.
DESIGN:
We studied a subset of 108 patients from the Phenotyping Sepsis-Induced Multiple Organ Failure Study (PHENOMS) cohort by means of untargeted metabolomics of heparinized plasma samples. Primary outcome was phenotype group. Key secondary outcomes included AKI (defined as both creatinine > 1 mg/dL and oliguria < 0.5 mL/kg/hr), other organ dysfunctions, and hospital mortality. Patients were followed up until discharge or 28 days.
SETTING:
ICU.
PATIENTS:
One hundred eight patients with sepsis.
INTERVENTIONS:
None.
MEASUREMENTS AND MAIN RESULTS:
Higher LA levels were associated with sepsis phenotype D as compared with phenotypes A–C (odds ratio [OR], 1.67; 95% CI, 1.05–2.65; p = 0.03). LA-derived oxylipins 9-hydroxyoctadecadienoic acid and 13-hydroxyoctadecadienoic acid (9-HODE/13-HODE) were also associated with sepsis phenotype D (jointly reported in one variable; OR, 1.26; 95% CI, 1.01–1.57; p = 0.04). Higher LA showed a trend and 9-HODE/13-HODE was associated with AKI (OR, 1.52; 95% CI, 0.97–2.38; p = 0.07 and OR, 1.27; 95% CI, 1.03–1.56; p = 0.02, respectively). Neither LA nor oxylipins were associated with hospital mortality.
CONCLUSIONS:
LA levels and LA-derived lipoxygenase oxylipins are associated with pediatric sepsis phenotype D and AKI. These results support future mechanistic studies to investigate lipid metabolism in the pathophysiology of sepsis.
Keywords: acute kidney injury, linoleic acid, organ dysfunction, oxylipin, polyunsaturated fatty acid, sepsis phenotype
Long-chain polyunsaturated fatty acids (PUFAs) are essential fatty acids broadly categorized as omega-6 and omega-3 based on the position of their double bonds. Much debate and controversy have taken place regarding the role of PUFA in inflammatory regulation and chronic metabolic diseases. It remains unknown whether they have a role in the development of acute organ dysfunction or alter sepsis outcomes. Preclinical sepsis models have found that dietary supplementation with linoleic acid (LA)-rich vegetable oil (omega-6 fatty acid-rich) before induction of experimental sepsis worsens subsequent sepsis mortality (1, 2). A recent Mendelian randomization study suggested that genetically predicted omega-6 to omega-3 ratio is associated with increased risk of death due to sepsis (3).
LA undergoes a series of metabolic transformations, beginning with desaturation by delta-6-desaturase (D6D, encoded by the FADS2 gene) to form arachidonic acid (4–6). This enzyme is rate-limiting and shared between omega-6 and omega-3 pathways (7, 8). Genetic variants in the FADS gene cluster have been linked to various human diseases (6, 9–11). Each of the long chain PUFA (e.g., LA, arachidonic acid, or the omega-3 docosahexaenoic acid [DHA]) can be further metabolized by one of three major pathways: cytochrome p450, cyclooxygenase, and lipoxygenase, to produce oxylipins (12, 13). Oxylipins have diverse effects based on physiologic context, and enzymes that produce oxylipins are common pharmacologic targets for anti-inflammatory agents (e.g., nonsteroidal anti-inflammatory drugs inhibit the cyclooxygenase-2 enzyme). Human studies have found that increased production of arachidonic acid-derived oxylipins is associated with worse sepsis mortality (14, 15).
Recently, sepsis phenotypes have been characterized in both adults and children (16, 17). Pediatric sepsis phenotypes A, B, C, and D were described with phenotype D characterized by high incidence of acute kidney injury (AKI), hyperferritinemia, and high risk of development of multiple organ failure and death (17). Whether PUFA plasma concentrations are associated with development of organ dysfunction and with specific sepsis phenotypes remains unknown. Aside from multiple organ dysfunction, AKI offers an ideal model to investigate these mechanisms because AKI occurs very frequently in sepsis, and because it’s a critical driver of poor outcomes and a major risk factor for chronic organ dysfunction (18).
Therefore, the aim of this study is to test the hypothesis that plasma LA levels and its lipoxygenase-derived oxylipins are associated with pediatric sepsis phenotype D. We further hypothesize that plasma LA levels and its lipoxygenase-derived oxylipins are associated with AKI. Finally, we aimed to assess genetic variations in the FADS cluster to explore the role of genetically determined modulation of PUFA metabolism on the same clinical outcomes.
PATIENTS AND METHODS
Study Design, Population, and Setting
This study is a secondary analysis of the multicenter Phenotyping Sepsis-Induced Multiple Organ Failure Study (PHENOMS), approved by the University of Utah Institutional Review Board (IRB) No. 70976 on September 22, 2014, with ongoing analysis approved by the University of Pittsburgh IRB No. 20060235, approval data July 6, 2020. All research procedures were in agreement with the ethical principles outlined in the updated Helsinki declaration (Available at: https://www.wma.net/policies-post/wma-declaration-of-helsinki/. Accessed April 14, 2025). Written informed consent was obtained from one or more parents/guardians for each child. Assent was garnered when the child was able. The PHENOMS prospective cohort included 404 patients with sepsis (17, 19). A previous report using this dataset identified sepsis phenotypes A–D by utilizing k means consensus clustering on 25 bedside variables available for 24 hours (17). Sepsis phenotype D is characterized by a high proportion of AKI, multiple organ failure, and high risk of death (17). The first 27 patients from each phenotype were chosen for untargeted metabolomic analysis and are included in this study. The hypothesis tested in this study was determined before data availability from metabolomics readouts; therefore, the variables extracted were based on the relevant steps of the LA biochemical pathway, rather than statistical relationships within the whole set of analytes.
Metabolomic Analysis
Blood was collected in heparinized tubes within 24 hours of enrollment, samples were centrifuged for plasma extraction, and then frozen and stored at −80°C. Frozen samples were sent to Metabolon (Morrisville, NC) for further analysis. Untargeted metabolomic analyses were performed by ultrahigh performance liquid chromatography-tandem mass spectroscopy. Raw data were extracted, peak values were identified, and the data were processed for quality control using Metabolon’s software. Metabolites were identified by comparison with library entries of purified standards, quantified by area under the curve and reported as batch-normalized values. Missing values were considered below the level of detection, and the lowest detection level was imputed as per Metabolon’s standard procedure (20).
Selection of PUFA and Oxylipins
Our dataset includes levels of the major omega-6 PUFA, LA, arachidonic acid, and major omega-3 PUFA, eicosapentaenoic acid, and DHA. Importantly, it does not include isolated levels of omega-3 α-linolenic acid, as the metabolomics readouts did not differentiate certain species that are closely related structurally but might be otherwise functionally diverse. That is, both α-linolenic acid, an omega-3 PUFA, and the omega-6 γ-linolenic acid were jointly reported in the same variable and were not included in the statistical analyses. Similarly, our dataset does not include the omega-3 docosapentaenoic acid (reported in the same variable as the omega-6 docosapentaenoic acid).
LA and arachidonic acid are closely correlated (due to metabolism through D6D), and eicosapentaenoic acid and DHA are also tightly correlated in our dataset. Thus, to avoid collinearity, we included only one representative of omega-6 and one of omega-3 PUFA in each regression model. We selected LA because this is the most abundant dietary source of omega-6 PUFA, and DHA because it is an omega-3 PUFA generally considered to be an important source of anti-inflammatory mediators (12).
We selected the following lipoxygenase-derived oxylipins based on availability in the dataset and relevance to sepsis in the literature (13, 21): the LA derivatives 9-hydroxyoctadecadienoic acid and 13-hydroxyoctadecadienoic acid (9-HODE/13-HODE, jointly reported as one variable) and the DHA derivatives 14-hydroxydocosahexaenoic acid and 17-hydroxydocosahexaenoic acid (jointly reported). These metabolites were combined in a single variable when reported due to analytical limitations.
Definitions of Organ Dysfunction
Organ dysfunction was assessed daily up to 28 days of follow-up as previously defined (22, 23). Briefly, we used the following definitions by organ system: 1) CNS: Glasgow Coma Scale less than 12 not explained by sedative use; 2) Cardiovascular: Shock was defined as requirement of vasopressor agent for systolic blood pressure less than fifth percentile for age; 3) Pulmonary: Pao2/Fio2 ratio less than 300 requiring mechanical ventilation; 4) Renal: AKI was defined as presence of oliguria (< 0.5 mL/kg/hr) and serum creatinine greater than 1 mg/dL or need for renal replacement therapy; 5) Hepatic: alanine aminotransferase greater than 100 U/L and bilirubin greater than 1 mg/dL; and 6) Hematologic: platelet count less than 100,000/mm3 and international normalized ratio greater than 1.5. We defined the time to organ dysfunction as the number of days from inclusion in the study to the first day qualifying as the organ dysfunction.
Outcomes
Patients were classified into sepsis phenotypes A–D as previously described (17, 19). Our primary outcome was sepsis phenotype D (compared with A–C groups). Key secondary outcomes were development of AKI, all other organ dysfunctions described above, and hospital mortality. Regression models were adjusted for demographics and comorbidities, as well as lactate as a marker of severity.
Genotype Analysis
Whole exome sequencing was performed on DNA extracted from whole blood at the University of Pittsburgh Institute for Precision Medicine on Illumina’s NovaSeq 6000 with a mean coverage of 43.7x. Fast Alignment Sequence Quality format (FASTQ) files were aligned to Homo sapiens reference sequence GRCh38. Resultant files were analyzed in the Fabric Genomics Opal 5.2.2 software (Fabric Genomics, Oakland, CA). Variants were filtered for coverage greater than 10, Phred score greater than 30. We screened PUFA pathway genes in the FADS cluster, and ELOVL1–6 in the first 30 sequential subjects.
Statistical Analysis
Avoidance of Sampling Bias.
patient selection and enrollment has been previously described, along with methods employed to balance individual site enrollment (19). Out of the entire cohort, the first 27 patients from each of pediatric sepsis phenotypes A–D were selected for untargeted metabolomic analysis.
Distributional Assumptions.
Variable distributions were verified by inspection of nomograms and by QQ plots. Log transformation was attempted when planned analyses required normality and the log-transformed variable showed a normal or near-normal distribution. All analyses were initially performed in univariate statistics with parametric and nonparametric tests as appropriate.
Primary Analysis.
The primary outcome was sepsis phenotype D. The secondary outcome was AKI. Additional secondary outcomes analyzed included other organ injuries. For the primary outcome of sepsis phenotype group, we dichotomized the sample grouping phenotypes D (n = 27) and all others (n = 81). We built an adjusted logistic regression model that fit the logistic regression assumptions using the following covariates: age, sex, cardiovascular disease, malignancy, and lactate. For the secondary outcomes of organ dysfunction, association between LA or oxylipins with the specific organ dysfunction was estimated using binary logistic regression adjusting for the same covariates as for the primary outcome.
Sensitivity and Exploratory Analyses.
Because we were particularly interested in AKI as an organ injury outcome, we explored linear associations between PUFA and oxylipin exposure and plasma creatinine level obtained from the metabolomics readout, collected at the same time as the PUFA levels within the first day of hospitalization. We used a Cox proportional hazards model to explore the time relationship between LA or oxylipins and development of AKI later in the hospital course (i.e., not present at admission).
Software.
We used R, Version 4.3.1 (R Foundation, Vienna, Austria) for all analyses with R commander graphic user interface with the EZR plugin (24).
RESULTS
Population Characteristics
Baseline characteristics of the selected 108 patients are presented in Table 1. Median age was 5.1 years, 62% male, median Pediatric Risk of Mortality III score 11, with an overall hospital mortality of 11.1% (Table 1). Sepsis phenotype D showed the highest hospital mortality of seven of 27 (25.9%), as well as the highest proportion of patients developing AKI (23/27 patients; 85.2%). The relationships of pediatric sepsis phenotypes with AKI and hospital mortality are shown in Figure 1. Relative levels of LA and 9-HODE/13-HODE among patients with sepsis phenotypes and with or without AKI are shown in Figure 2.
TABLE 1.
Baseline Characteristics
| Sepsis Phenotype | |||||
|---|---|---|---|---|---|
| Variable | Total | A | B | C | D |
| n | 108 | 27 | 27 | 27 | 27 |
| Age, median (IQR) | 5.1 (1.9–11.4) | 2.5 (0.6–4.8) | 4.9 (1.7–11.0) | 8.8 (5.3–14.4) | 6.8 (2.7–12.2) |
| Male sex (%) | 67 (62) | 13 (48.1) | 20 (74.2) | 17 (63) | 17 (63) |
| Race (%) | |||||
| Asian | 4 (3.7) | 0 (0) | 1 (3.7) | 2 (7.4) | 1 (3.7) |
| African American | 17 (15.7) | 5 (18.5) | 4 (14.8) | 4 (14.8) | 4 (14.8) |
| White | 83 (76.9) | 20 (74.1) | 22 (81.5) | 19 (70.4) | 22 (81.5) |
| Not reported | 4 (3.7) | 2 (7.4) | 0 (0) | 2 (7.4) | 0 (0) |
| Ethnicity (%) | |||||
| Hispanic or Latino | 17 (15.7) | 8 (29.6) | 2 (7.4) | 5 (18.5) | 2 (7.4) |
| Not Hispanic or Latino | 88 (81.5) | 18 (66.7) | 24 (88.9) | 21 (77.8) | 25 (92.6) |
| Not reported | 3 (2.8) | 1 (3.7) | 1 (3.7) | 1 (3.7) | 0 (0) |
| Comorbidities (%) | |||||
| Malignancy | 17 (15.7) | 0 (0) | 1 (3.7) | 9 (33.3) | 7 (25.9) |
| Bone marrow transplant | 8 (6.3) | 0 (0) | 0 (0) | 4 (14.8) | 4 (14.8) |
| Solid organ transplant | 1 (0.9) | 0 (0) | 0 (0) | 1 (3.7) | 0 (0) |
| Diabetes | 1 (0.9) | 0 (0) | 0 (0) | 1 (3.7) | 0 (0) |
| Cardiovascular disease | 18 (16.7) | 5 (18.5) | 6 (22.2) | 4 (14.8) | 3 (11.1) |
| Liver disease | 5 (4.6) | 1 (3.7) | 1 (3.7) | 1 (3.7) | 2 (7.4) |
| Steroid use | 21 (19.4) | 3 (11.1) | 2 (7.4) | 9 (33.3) | 7 (25.9) |
| Organ failure (%) | |||||
| Cardiovascular | 92 (85.2) | 18 (66.7) | 26 (96.3) | 24 (88.9) | 24 (88.9) |
| Pulmonary | 85 (78.7) | 24 (88.9) | 27 (100) | 12 (44.4) | 22 (81.5) |
| CNS | 21 (19.6) | 4 (14.8) | 7 (26.9) | 4 (14.8) | 6 (22.2) |
| Hematologic | 20 (18.7) | 0 (0) | 1 (3.8) | 4 (14.8) | 15 (55.6) |
| Hepatic | 32 (29.9) | 2 (7.4) | 4 (15.4) | 10 (37) | 16 (59.3) |
| Renal | 27 (25.2) | 1 (3.7) | 0 (0) | 3 (11.1) | 23 (85.2) |
| Pediatric Risk of Mortality III (25), median (IQR) | 11 (7.0–15.0) | 9 (7.0–11.0) | 13 (9.0–17.0) | 7 (3.0–12.5) | 15 (12.0–20.5) |
| Hospital mortality (%) | 12 (11.1) | 0 (0) | 2 (7.4) | 2 (11.1) | 7 (25.9) |
IQR = interquartile range.
Figure 1.
Alluvial plot showing the relative distribution of patients according to sepsis phenotype, acute kidney injury (AKI), and hospital mortality. Group ID = pediatric sepsis phenotype group category, PS_A-D = pediatric sepsis phenotype A-D.
Figure 2.
Boxplots representing levels of linoleic acid and 9-hydroxyoctadecadienoic acid/13-hydroxyoctadecadienoic acid (9-HODE/13-HODE) among patients with sepsis phenotypes A, B, and C or phenotype D (top row), and among patients with or without acute kidney injury (AKI) (bottom row). Units of measurement were arbitrary units based on batch normalization relative intensity (batch median is represented by 1.0). AKI was defined as serum creatinine greater than 1 mg/dL, oliguria, or renal replacement therapy.
Primary Analysis
Higher plasma LA concentration (odds ratio [OR], 1.67; 95% CI, 1.05–2.65; p = 0.03; Table 2) and LA-derived lipoxygenase oxylipins 9-HODE/13-HODE levels (OR, 1.26; 95% CI, 1.01–1.57; p = 0.04; Table 2) were associated with sepsis phenotype D.
TABLE 2.
Associations With Sepsis Phenotype D Compared With Phenotypes A–C (Binary Logistic Regression)
| Variable | OR (95% CI) | p |
|---|---|---|
| Polyunsaturated fatty acids | ||
| (Intercept) | 0.08 (0.01–0.50) | < 0.01 |
| Age | 1.06 (0.97–1.16) | 0.20 |
| Female sex | 0.80 (0.30–2.15) | 0.66 |
| Cardiovascular disease | 0.48 (0.11–2.18) | 0.34 |
| Malignancy | 2.08 (0.65–6.67) | 0.22 |
| Lactate | 1.38 (0.80–2.38) | 0.25 |
| Linoleic acid | 1.67 (1.05–2.65) | 0.03 |
| Docosahexaenoic acid | 1.16 (0.88–1.55) | 0.30 |
| Lipoxygenase-derived oxylipins | ||
| (Intercept) | 0.21 (0.04–1.19) | 0.08 |
| Age | 1.03 (0.95–1.13) | 0.44 |
| Female sex | 0.72 (0.27–1.95) | 0.52 |
| Cardiovascular disease | 0.64 (0.15–2.68) | 0.54 |
| Malignancy | 2.08 (0.66–6.54) | 0.21 |
| Lactate | 1.03 (0.57–1.85) | 0.92 |
| 9-HODE/13-HODE (from linoleic acid) | 1.26 (1.01–1.57) | 0.04 |
| 13-HDoHE/17-HDoHE (from docosahexaenoic acid) | 1.07 (0.81–1.42) | 0.63 |
HDoHE = hydroxydocosahexaenoic acid, HODE = hydroxyoctadecadienoic acid, OR = odds ratio.
Sepsis phenotypes were determined in this cohort as previously described (17).
Key Secondary Outcomes
LA levels showed a trend, and 9-HODE/13-HODE an association with AKI (OR, 1.52; 95% CI, 0.97–2.38; p = 0.07 and OR, 1.27; 95% CI, 1.03–1.56; p = 0.02, respectively; Table 3).
TABLE 3.
Associations of Levels of Polyunsaturated Fatty Acids and Oxylipins With Acute Kidney Injury (Binary Logistic Regression)
| Variable | OR (95% CI) | p |
|---|---|---|
| Polyunsaturated fatty acids | ||
| (Intercept) | 0.14 (0.02–0.89) | 0.04 |
| Age | 1.09 (1.00–1.19) | 0.05 |
| Female sex | 0.70 (0.26–1.89) | 0.48 |
| Cardiovascular disease | 0.85 (0.22–3.31) | 0.82 |
| Malignancy | 1.52 (0.47–4.90) | 0.49 |
| Lactate | 1.00 (0.55–1.81) | 0.99 |
| Linoleic acid | 1.52 (0.97–2.38) | 0.07 |
| Docosahexaenoic acid | 1.08 (0.81–1.44) | 0.59 |
| Lipoxygenase-derived oxylipins | ||
| (Intercept) | 0.36 (0.06–2.01) | 0.24 |
| Age | 1.07 (0.98–1.16) | 0.14 |
| Female sex | 0.60 (0.22–1.65) | 0.32 |
| Cardiovascular disease | 1.01 (0.27–3.84) | 0.99 |
| Malignancy | 1.55 (0.48–5.00) | 0.46 |
| Lactate | 0.75 (0.38–1.46) | 0.40 |
| 9-HODE/13-HODE (from linoleic acid) | 1.27 (1.03–1.56) | 0.02 |
| 14-HDoHE/17-HDoHE (from docosahexaenoic acid) | 0.97 (0.71–1.33) | 0.85 |
HDoHE = hydroxydocosahexaenoic acid, HODE = hydroxyoctadecadienoic acid, OR = odds ratio.
Acute kidney injury was defined as creatinine > 1 mg/dL with oliguria.
LA, but not oxylipin levels, were associated with hepatic dysfunction (OR, 2.23; 95% CI, 1.35–3.68; p < 0.01; Table S1, https://links.lww.com/PCC/C696) and were negatively associated with pulmonary dysfunction (OR, 0.57; 95% CI, 0.35–0.93; p = 0.02; Table S4, https://links.lww.com/PCC/C696). Neither LA nor oxylipin levels were associated with other forms of organ dysfunction, shock, or hospital mortality. Tables S2, S3, S5, and S6 (https://links.lww.com/PCC/C696) show regression models for hematologic, CNS, shock, and hospital mortality as outcomes. To better understand the role of LA, we compared the cumulative risk of selected key secondary outcomes between patients who had LA levels above and below the median LA for the entire cohort. Supplemental Figure (https://links.lww.com/PCC/C696) shows that patients with above median LA levels had higher risk of hepatic and hematologic dysfunction.
Genotyping
Whole exome sequencing analysis showed no evidence of either rare or common coding variants in the FADS and ELOVL gene clusters. Therefore, we did not further pursue statistical analysis according to patient’s genotype for PUFA pathway enzymes.
Sensitivity and Exploratory Analyses
We analyzed LA, adjusted for demographics, comorbidities, and lactate, along with DHA, arachidonic acid, and eicosapentaenoic acid (i.e., four major PUFA representing both omega-6 and omega-3 series), while recognizing limitation of collinearity among PUFA, but LA remained the only PUFA independently associated with sepsis phenotype D (OR, 1.65; 95% CI, 1.03–2.67; p = 0.04; Table S7, https://links.lww.com/PCC/C696).
We explored a linear relationship between LA and creatinine from metabolomics readout (obtained from the same frozen plasma sample). We found LA to be associated with plasma creatinine in multivariable regression (estimate, 0.23; 95% CI, 0.05–0.41; p = 0.01; Table S8, https://links.lww.com/PCC/C696).
DISCUSSION
We found that elevated LA and its metabolites are associated with sepsis phenotypes. We found that phenotype D and AKI are associated with oxylipins, specifically the LA-derived lipoxygenase 9-HODE/13-HODE. Finally, we found associations of LA plasma concentrations with hepatic and pulmonary dysfunction.
These findings may reflect a metabolic shift favoring the production of pro-inflammatory and bioactive lipid mediators during critical illness, particularly in patients with phenotype D. The mechanistic implications warrant further investigation, as targeted modulation of PUFA metabolism could represent a novel therapeutic avenue for high-risk sepsis subgroups.
Our results are consistent with our overarching hypothesis that LA may contribute to sepsis-associated organ injury. In previous investigations, sepsis phenotype D showed the highest mortality and use of renal replacement therapy, and fewer ICU-free days compared with the other phenotypes (17). Furthermore, phenotype D patients were characterized by hyperferritinemia, thrombocytopenia with propensity to ADAMTS13 deficiency, thrombocytopenia-associated multiple organ failure, and macrophage activation syndrome (17). Our findings linking phenotype D to higher concentrations of lipid mediators are relevant because LA and its derivatives are important sources of inflammatory mediators and, thus, could represent a potentially targetable key mechanism of organ injury in human sepsis. The first and rate-limiting step in the metabolism of dietary LA into arachidonic acid is the enzyme D6D, encoded by the FADS2 gene in chromosome 11. A specific inhibitor of D6D has been shown to have anti-inflammatory effects in a murine model (26). In human observational studies, D6D activity has been associated with inflammation as measured by C-reactive protein (27). Both LA and its downstream D6D product, arachidonic acid, are susceptible to further metabolism by lipoxygenase, cyclooxygenase, or cytochrome P450, producing a complicated array of oxylipin lipid mediators (12, 13). In planning our analysis, we selected levels of precursor dietary-derived PUFA, along with levels of corresponding lipoxygenase-derived oxylipins. Oxylipins from each of the described pathways have been associated with inflammation and sepsis (13, 14). We selected a set of available lipoxygenase oxylipins representative of various PUFA precursors.
Pathogenetic mechanisms have been put forward to explain the complex relationship between PUFA and sepsis-induced organ injury, but no pathway has been conclusively demonstrated to be relevant in human sepsis. Proposed mechanisms of organ injury include inflammation (28, 29), altered energy metabolism through stimulation of peroxisome proliferator-activated receptors (30–32), direct toxicity from reactive metabolites (33–35), and induction of ferroptosis (36). Ferroptosis is a form of regulated cell death triggered by peroxidation of arachidonic acid residues from membrane phospholipids and may be involved in the pathophysiology of both AKI and chronic kidney disease (37). Drugs targeting lipoxygenase inhibition have been shown to modulate ferroptosis in preclinical models (37). We found that levels of both LA, the dietary precursor of arachidonic acid, and its lipoxygenase-derived oxylipins are associated with clinical outcomes. Our results, therefore, agree with proposed injury mechanisms.
The observation that levels of LA and 9-HODE/13-HODE were variably associated with different clinical outcomes may point to the heterogeneity of mechanisms of injury in different organs. Alternatively, our results may represent different underlying mechanisms to increase LA levels. For example, LA concentration could conceivably increase in hepatic dysfunction secondary to modulation of de novo lipogenesis pathway enzymes. The liver (along with adrenals) is the major site of D6D expression and plays a central role in systemic fatty acid availability (4, 38). Therefore, we report the intriguing associations of LA levels with liver organ injury, but the interpretation may be more prone to reverse causality bias here than in other organs. The negative association of LA with pulmonary dysfunction was surprising in light of the proposed pathophysiology. We hypothesize that metabolism through D6D may variably impact different organs.
Importantly, in our cohort AKI was defined conservatively by both a substantial creatinine elevation and oliguria. With a median age of 5.1 years, it would be expected that most of our patients’ baseline creatinine levels were around 0.31–0.37 (39). Therefore, our threshold of creatinine greater than 1 mg/dL would approximately correspond to Kidney Disease: Improving Global Outcomes (KDIGO) stage 3 for most patients classified as AKI (we did not have direct KDIGO classification in our dataset) (40). Our data, therefore, provides novel evidence linking PUFA metabolism to renal dysfunction in pediatric sepsis, emphasizing the need for mechanistic studies to elucidate causal pathways and explore interventions targeting lipid mediator synthesis.
Our genetic analysis suggests that the increased LA and derivatives in patients with more severe sepsis and organ dysfunction may not be driven by the presence of specific coding region variants in the FADS cluster in this population. We targeted LA pathway enzymes including D6D (as discussed above), and genes encoding for downstream elongase enzymes ELOVL. Single-nucleotide variants in the FADS gene cluster have been increasingly associated with diverse diseases such as type 2 diabetes, cardiovascular disease, colorectal cancer, and eczema (11, 41–43). While no study to date has investigated the role of FADS cluster variants directly in human sepsis, two recent Mendelian randomization studies utilizing the U.K. Biobank analyzed genome-wide single-nucleotide polymorphisms associated with omega-6 PUFA levels and found significant associations with sepsis mortality (3, 44). Importantly, those studies relied on intron and intergenic, noncoding, variants in the FADS cluster, whereas our analyses were based on whole exome sequencing. It is possible that humans have poor tolerance for FADS coding sequence variation, or that geographic prevalence and small sample size were responsible for our negative results, emphasizing the need for whole genome sequencing in future studies.
The main limitation of our study is its observational nature with limited prospective follow-up. We were also limited by the small cohort size, which we believe explains the lack of robustness of some results to sensitivity analyses (although all associations remained consistent in direction and magnitude, despite borderline p values). Another limitation is the lack of dietary intake data in our cohort. While LA usually is obtained from diet as it is an essential fatty acid, complex factors such as rate of D6D metabolism, and rates of hepatic de novo lipogenesis could modulate observed fatty acid proportions. Finally, due to our study design, we cannot establish causality in our reported associations. Nonetheless, the prospective nature, adequate follow-up, and the ability to analyze metabolites at different steps of biochemical pathways strengthens mechanistic hypotheses previously shown in preclinical models.
In conclusion, we found that levels of LA and its lipoxygenase-derived oxylipins 9-HODE and 13-HODE are associated with phenotype D, and 9-HODE and 13-HODE are associated with AKI in a pediatric sepsis cohort.
Supplementary Material
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/pccmjournal).
RESEARCH IN CONTEXT.
Question:
Linoleic acid is an abundant dietary omega-6 polyunsaturated fatty acid that is precursor to inflammatory lipid mediators. Levels of linoleic acid have been scarcely studied in critically ill patients.
Findings:
In a cohort of pediatric sepsis patients, linoleic acid levels and oxylipins derived from linoleic acid metabolism were associated with higher risk of sepsis phenotype D and acute kidney injury.
Meaning:
High linoleic acid levels are associated with increased risk of organ injury during sepsis. Future studies to identify individual risk factors and investigate causality are warranted.
AT THE BEDSIDE.
Linoleic acid, the most important dietary omega-6 polyunsaturated fatty acid, is a precursor to inflammatory lipid mediators.
In a cohort of 108 children with sepsis, linoleic acid and derived oxylipin levels were associated with sepsis phenotype and with acute kidney injury.
This study strengthens the need to prospectively investigate the role of dietary fatty acids in sepsis-associated organ injury.
Acknowledgments
Dr. Saraiva received funding from the National Institutes of Health (NIH; T32HL007820; T32 Institutional National Research Service Award salary support). Drs. Saraiva, Park, Chen, Berg, Meert, Pollack, Hall, Newth, Harrison, Shanley, and Carcillo received support for article research from the NIH. Dr. Carcillo is the principal investigator, and Dr. Park is a co-investigator on a grant from the NIH (R01GM108618). Dr. Kernan was supported by NIH K23GM148827. Drs. Kernan’s and Carcillo’s institution received funding from the National Institute of General Medical Sciences. Drs. Berg’s, Meert’s, Pollack’s, Hall’s, Newth’s, Harrison’s, and Shanley’s institutions received funding from the NIH. Dr. Hall received funding from AbbVie, Kiadis, the American Board of Pediatrics, Partner Therapeutics, and Sobi. Dr. Carcillo (5U01HD049934-10S1) received funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH, Department of Health and Human Services, and the following cooperative agreements: U10HD049983, U10HD050096, U10HD049981, U10HD063108, U10HD63106, U10HD063114, U10HD050012, and U01HD049934. Dr. Gómez’s institution received funding from Biomerieux, Baxter, and Trilinear Bioventures. Dr. Fuhrman has disclosed that she does not have any potential conflicts of interest.
REFERENCES
- 1.Kume A, Suganuma K, Umemiya-Shirafuji R, et al. : Effect of vegetable oils on the experimental infection of mice with Trypanosoma congolense. Exp Parasitol 2020; 210:107845. [DOI] [PubMed] [Google Scholar]
- 2.Svahn SL, Grahnemo L, Palsdottir V, et al. : Dietary polyunsaturated fatty acids increase survival and decrease bacterial load during septic Staphylococcus aureus infection and improve neutrophil function in mice. Infect Immun 2015; 83:514–521 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lei P, Xu W, Wang C, et al. : Mendelian randomization analysis reveals causal associations of polyunsaturated fatty acids with sepsis and mortality risk. Infect Dis Ther 2023; 12:1797–1808 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Brenner RR: The desaturation step in the animal biosynthesis of polyunsaturated fatty acids. Lipids 1971; 6:567–575 [DOI] [PubMed] [Google Scholar]
- 5.Marcel YL, Christiansen K, Holman RT: The preferred metabolic pathway from linoleic acid to arachidonic acid in vitro. Biochim Biophys Acta 1968; 164:25–34 [DOI] [PubMed] [Google Scholar]
- 6.Glaser C, Lattka E, Rzehak P, et al. : Genetic variation in polyunsaturated fatty acid metabolism and its potential relevance for human development and health. Matern Child Nutr 2011; 7(Suppl 2):27–40 [Google Scholar]
- 7.Sprecher H: Interactions between the metabolism of n-3 and n-6 fatty acids. J Intern Med Suppl 1989; 731:5–9 [DOI] [PubMed] [Google Scholar]
- 8.Bezard J, Blond JP, Bernard A, et al. : The metabolism and availability of essential fatty acids in animal and human tissues. Reprod Nutr Dev 1994; 34:539–568 [DOI] [PubMed] [Google Scholar]
- 9.Khamlaoui W, Mehri S, Hammami S, et al. : Association between genetic variants in FADS1-FADS2 and ELOVL2 and obesity, lipid traits, and fatty acids in Tunisian population. Clin Appl Thromb Hemost 2020; 26:1076029620915286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jager S, Cuadrat R, Hoffmann P, et al. : Desaturase activity and the risk of type 2 diabetes and coronary artery disease: A Mendelian randomization study. Nutrients 2020; 12:2261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhang B, Jia WH, Matsuda K, et al. ; Genetics and Epidemiology of Colorectal Cancer Consortium (GECCO): Large-scale genetic study in East Asians identifies six new loci associated with colorectal cancer risk. Nat Genet 2014; 46:533–542 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dyall SC, Balas L, Bazan NG, et al. : Polyunsaturated fatty acids and fatty acid-derived lipid mediators: Recent advances in the understanding of their biosynthesis, structures, and functions. Prog Lipid Res 2022; 86:101165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Shearer GC, Walker RE: An overview of the biologic effects of omega-6 oxylipins in humans. Prostaglandins Leukot Essent Fatty Acids 2018; 137:26–38 [DOI] [PubMed] [Google Scholar]
- 14.Jones TN, Janani L, Gordon AC, et al. : A novel role for cytochrome P450 epoxygenase metabolites in septic shock. Crit Care Explor 2022; 4:e0622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sulaiman D, Wu D, Black LP, et al. : Lipidomic changes in a novel sepsis outcome-based analysis reveals potent pro-inflammatory and pro-resolving signaling lipids. Clin Transl Sci 2024; 17:e13745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Seymour CW, Kennedy JN, Wang S, et al. : Derivation, validation, and potential treatment implications of novel clinical phenotypes for sepsis. JAMA 2019; 321:2003–2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Qin Y, Kernan KF, Fan Z, et al. : Machine learning derivation of four computable 24-h pediatric sepsis phenotypes to facilitate enrollment in early personalized anti-inflammatory clinical trials. Crit Care 2022; 26:128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zarbock A, Nadim MK, Pickkers P, et al. : Sepsis-associated acute kidney injury: Consensus report of the 28th Acute Disease Quality Initiative Workgroup. Nat Rev Nephrol 2023 [Google Scholar]
- 19.Carcillo JA, Berg RA, Wessel D, et al. ; Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network: A multicenter network assessment of three inflammation phenotypes in pediatric sepsis-induced multiple organ failure. Pediatr Crit Care Med 2019; 20:1137–1146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ford L, Kennedy AD, Goodman KD, et al. : Precision of a clinical metabolomics profiling platform for use in the identification of inborn errors of metabolism. J Appl Lab Med 2020; 5:342–356 [DOI] [PubMed] [Google Scholar]
- 21.Gabbs M, Leng S, Devassy JG, et al. : Advances in our understanding of oxylipins derived from dietary PUFAs. Adv Nutr 2015; 6:513–540 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Carcillo JA, Halstead ES, Hall MW, et al. ; Eunice Kennedy Shriver National Institute of Child Health and Human Development Collaborative Pediatric Critical Care Research Network Investigators: Three hypothetical inflammation pathobiology phenotypes and pediatric sepsis-induced multiple organ failure outcome. Pediatr Crit Care Med 2017; 18:513–523 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Doughty L, Clark RS, Kaplan SS, et al. : sFas and sFas ligand and pediatric sepsis-induced multiple organ failure syndrome. Pediatr Res 2002; 52:922–927 [DOI] [PubMed] [Google Scholar]
- 24.Kanda Y: Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 2013; 48:452–458 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pollack MM, Dean JM, Butler J, et al. : The ideal time interval for critical care severity-of-illness assessment. Pediatr Crit Care Med 2013; 14:448–453 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Obukowicz MG, Welsch DJ, Salsgiver WJ, et al. : Novel, selective delta6 or delta5 fatty acid desaturase inhibitors as anti-inflammatory agents in mice. J Pharmacol Exp Ther 1998; 287:157–166 [PubMed] [Google Scholar]
- 27.Saito E, Okada T, Abe Y, et al. : Abdominal adiposity is associated with fatty acid desaturase activity in boys: Implications for C-reactive protein and insulin resistance. Prostaglandins Leukot Essent Fatty Acids 2013; 88:307–311 [DOI] [PubMed] [Google Scholar]
- 28.Lankinen MA, Fauland A, Shimizu BI, et al. : Inflammatory response to dietary linoleic acid depends on FADS1 genotype. Am J Clin Nutr 2019; 109:165–175 [DOI] [PubMed] [Google Scholar]
- 29.Su H, Liu R, Chang M, et al. : Dietary linoleic acid intake and blood inflammatory markers: A systematic review and meta-analysis of randomized controlled trials. Food Funct 2017; 8:3091–3103 [DOI] [PubMed] [Google Scholar]
- 30.Chawla A, Repa JJ, Evans RM, et al. : Nuclear receptors and lipid physiology: Opening the X-files. Science 2001; 294:1866–1870 [DOI] [PubMed] [Google Scholar]
- 31.Crossland H, Constantin-Teodosiu D, Greenhaff PL: The regulatory roles of PPARs in skeletal muscle fuel metabolism and inflammation: Impact of PPAR agonism on muscle in chronic disease, contraction and sepsis. Int J Mol Sci 2021; 22:9775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Van Wyngene L, Vanderhaeghen T, Timmermans S, et al. : Hepatic PPARalpha function and lipid metabolic pathways are dysregulated in polymicrobial sepsis. EMBO Mol Med 2020; 12:e11319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Esterbauer H, Schaur RJ, Zollner H: Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic Biol Med 1991; 11:81–128 [DOI] [PubMed] [Google Scholar]
- 34.Michiels C, Remacle J: Cytotoxicity of linoleic acid peroxide, malondialdehyde and 4-hydroxynonenal towards human fibroblasts. Toxicology 1991; 66:225–234 [DOI] [PubMed] [Google Scholar]
- 35.Lorente L, Martin MM, Abreu-Gonzalez P, et al. : Sustained high serum malondialdehyde levels are associated with severity and mortality in septic patients. Crit Care 2013; 17:R290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Yamane D, Hayashi Y, Matsumoto M, et al. : FADS2-dependent fatty acid desaturation dictates cellular sensitivity to ferroptosis and permissiveness for hepatitis C virus replication. Cell Chem Biol 2022; 29:799–810.e4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Bayir H, Dixon SJ, Tyurina YY, et al. : Ferroptotic mechanisms and therapeutic targeting of iron metabolism and lipid peroxidation in the kidney. Nat Rev Nephrol 2023; 19:315–336 [DOI] [PubMed] [Google Scholar]
- 38.Cho HP, Nakamura MT, Clarke SD: Cloning, expression, and nutritional regulation of the mammalian Delta-6 desaturase. J Biol Chem 1999; 274:471–477 [DOI] [PubMed] [Google Scholar]
- 39.Fitzgerald JC, Basu RK, Fuhrman DY, et al. ; Pediatric Organ Dysfunction Information Update Mandate (PODIUM) Collaborative: Renal dysfunction criteria in critically ill children: The PODIUM consensus conference. Pediatrics 2022; 149:S66–S73 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Khwaja A: KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract 2012; 120:c179–c184 [DOI] [PubMed] [Google Scholar]
- 41.Koletzko B, Reischl E, Tanjung C, et al. : FADS1 and FADS2 polymorphisms modulate fatty acid metabolism and dietary impact on health. Annu Rev Nutr 2019; 39:21–44 [DOI] [PubMed] [Google Scholar]
- 42.Kröger J, Zietemann V, Enzenbach C, et al. : Erythrocyte membrane phospholipid fatty acids, desaturase activity, and dietary fatty acids in relation to risk of type 2 diabetes in the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam Study. Am J Clin Nutr 2011; 93:127–142 [DOI] [PubMed] [Google Scholar]
- 43.Li SW, Lin K, Ma P, et al. : FADS gene polymorphisms confer the risk of coronary artery disease in a Chinese Han population through the altered desaturase activities: Based on high-resolution melting analysis. PLoS One 2013; 8:e55869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Zhang Z, Yin Y, Chen T, et al. : Investigating the impact of human blood metabolites on the Sepsis development and progression: A study utilizing two-sample Mendelian randomization. Front Med (Lausanne) 2023; 10:1310391. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.