The hallmark of sepsis is a dysregulated host response to infection. Sepsis is defined as infection-related organ dysfunction, and septic shock is further complicated by refractory hypotension with elevated blood lactate levels (39–41). The pathophysiology of sepsis is extraordinarily complex (42). Various molecules originating from the infecting microorganism, so-called pathogen-associated molecular patterns (PAMPs), or from necrotic cells, the damage-associated molecular patterns (DAMPs) activate the innate immune system through pattern recognition receptors on leukocytes, leading to a signaling cascade eventually resulting in the generation of pro-inflammatory cytokines. This “cytokine storm” is likely responsible for the systemic inflammatory response and the resulting organ dysfunction (induced by both cellular infiltration and ischemia) characteristic of sepsis. As numerous attempts aiming to dampen this cytokine storm have failed in clinical trials (43), considerable challenges remain in the management of sepsis.

Administration of SPM has shown some promising results in animal models of sepsis; however, this approach has not yet been translated into clinical practice. In animal studies, administration of D-series resolvins counter-regulates proinflammatory genes, decreases excessive cytokine production, neutrophil recruitment and infiltration, and enhances phagocytosis of bacteria, reducing tissue damage and improving survival (44–51). Exogenous administration of maresins (52, 53) and lipoxins has similar effects (54–58).

Published human studies to date are mainly observational (Table 1). In addition, some clinical studies investigated aspirin-triggered resolvins and lipoxins. In healthy adults, low-dose aspirin stimulates biosynthesis of anti-inflammatory mediators (69) and, in ICU patients with a severe inflammatory response, it reduces the concentration of proinflammatory mediators (17-HETE, 18-HETE, and 20-HETE) and increases the concentration of the anti-inflammatory mediators 17,18-DiHETE and 14,15-DiHETE (60). However, Dalli et al. reported significantly higher levels of pro-resolving mediators like RvE1, RvD5 and 17R-PD1 in sepsis non-survivors compared to survivors (61). It is therefore arguable that higher levels of SPM might be harmful rather than useful. One possible explanation for this apparent contradiction is that, in sepsis non-survivors, the endogenous increase in SPM may not be sufficient to reverse the inflammatory process or perhaps the time window in which these mediators are produced is critical. Moreover, the increased levels of pro-inflammatory cytokines observed in non-survivors (61) suggest more severe systemic inflammation, where, although increased, SPM levels are not sufficient to resolve the ongoing inflammation. This hypothesis has also been supported by Abdoulnour and colleagues, who found that increased plasma 15-epi-LXA4 levels at baseline were associated with development of ARDS, indicating engagement of counter-regulatory pathways that were ultimately insufficient to prevent the development of ARDS in these patients (63). Finally, many SPM possess dual biological actions and their effect may change over time, as exemplified by the study of Sordi et al. (58): In mice, LXA4 was increased at the beginning of sepsis, contributing to the harmful excessive inflammatory response. However, LXA4 administered in late sepsis was beneficial to the animal, controlling the excessive inflammation. These data suggest that both antagonizing LXA4 actions in the beginning or its administration in later periods could be beneficial in sepsis treatment.

Acute respiratory distress syndrome (ARDS) is characterized by a non-cardiogenic pulmonary edema (70), caused either by pulmonary or extrapulmonary events including severe pneumonia, sepsis, aspiration of gastric content, and trauma. The resulting acute lung injury is driven by excessive inflammation as a consequence of an imbalance of pro-inflammatory and anti-inflammatory cytokines, with release of multiple mediators of inflammation into the alveolar space and into the bloodstream (71). Increased endothelial and epithelial permeability then leads to alveolar fluid accumulation and impaired gas exchange. Resolution of ARDS requires endothelial and epithelial repair and reabsorption of alveolar edema fluid, and SPM are an essential component of the resolution program (72). Despite improvements in clinical management, mortality remains high and there is no specific treatment, nor are there universally agreed-upon biomarkers for survival and outcome in ARDS.

Different types of acute lung diseases have distinct lipid profiles (73) and lipid mediators may represent useful prognostic markers in critically ill patients. LTB4 correlates with lung-injury severity and outcome in patients with ARDS (64, 65) and higher pro-inflammatory mediators like PGF2α and selected pro-resolving mediators like 10S,17S-diHDHA were predictive of ARDS development in patients with sepsis (61). In patients at risk for ARDS randomized to aspirin versus placebo, increased levels of aspirin-triggered lipoxin A4 (15-epi-LXA4) were associated with the development of ARDS (63).

In animal models, administration of SPM has particularly beneficial effects in injured lungs (74). Maresins have organ protective effects, decrease edema, improve lung mechanics and tissue hypoxia (75). RvD1 decreases pulmonary edema, leukocyte infiltration and the release of pro-inflammatory cytokines and alleviates lung injury (76–79) and RvE1 can restore mitochondrial function in human alveolar epithelial cells and accelerates the resolution of experimental lung inflammation (80–82). Moreover, protectin D1 has beneficial effects in influenza-infected mice (83) and 15-epiLXA4 inhibits neutrophil infiltration and enhances pathogen clearance (84, 85).

Major trauma is a leading cause of morbidity and mortality around the globe (86, 87). Severe traumatic injury has a considerable impact on the immune and metabolic system (88, 89) and leads to a posttraumatic cascade of inflammatory changes (90–93). Therefore, lipid mediators have been proposed as prognostic markers in trauma patients (67, 68, 94). In patients with traumatic brain injury (TBI), cerebrospinal fluid concentration of free fatty acids is significantly elevated and correlates with clinical outcomes (66).

Accumulating evidence from animal studies suggests that various lipid mediators may have a role as therapeutic agents in cerebral and spinal cord injury. Elovanoids are derivatives from very long chain PUFAs and have neuroprotective properties in animal models of TBI and ischemic stroke (95, 96). In other animal models of TBI, RvD1 promotes functional recovery and halts glial activation and neuronal death, and RvE1 modulates the inflammatory response (97, 98). Moreover, parenteral or enteral administration of DHA reduces lesion size and axonal injury in rodents with TBI (99–101). The effect of DHA administration in rats with spinal cord injury has recently been summarized in a systematic review and meta-analysis (102). The reported studies suggest that, in rats, DHA can promote motor functional recovery after spinal cord injury. This effect appears limited to administration of DHA, and is not observed with EPA (103). Finally, Maresin 1 also improves neurological outcomes after experimental spinal cord injury (104). Although these findings are encouraging, further validation with adequate animal models are needed, taking into consideration the dose, target specificity and central nervous system penetration of tested compounds.

Ischemia/reperfusion injury is a major determinant of poor outcome in patients with ischemic stroke and cardiac arrest survivors (105). In cardiac arrest, global cerebral ischemia alters cell metabolism and the balance of cerebral vasodilator/vasoconstrictor eicosanoids, rendering the cells susceptible to further damage after reperfusion: Vasoconstrictor eicosanoids are increased, and inhibition of 20-HETE synthesis (a potent vasoconstrictor) improves cortical perfusion and short-term neurologic outcome in a rat model of cardiac arrest (106).

In ischemic stroke, various in vitro and in vivo studies demonstrated that SPMs reduce leukocyte infiltration and neuronal injury, enhance efferocytosis and decrease both the production of inflammatory cytokines and oxidative stress (107). Cerebral artery occlusion and reperfusion causes significant reduction in endogenous RvD2 levels, and treatment with RvD2 reduces cerebral infarction, inflammatory cytokines, edema and neurological dysfunction (108). In another animal model, RvD1 promotes functional recovery, reduces neuroinflammation and prevents neuronal cell death (109). Neuroprotectin D1 (NPD1) down-regulates apoptosis and promotes cell survival (110, 111), and the administration of its precursor DHA has similar beneficial effects in experimental stroke (112–114). Additional administration of aspirin leads to cerebral synthesis of aspirin-triggered NPD1 (AT-NPD1), which reduces infarct size and significantly improves neurological scores in rats (110).

As with stroke, ischemia/reperfusion plays a pivotal role in the pathophysiology of myocardial infarction and contributes to up to 50% of the final infarct size (115). A crucial aspect is the balance between vasoconstrictive and vasodilatory metabolites of arachidonic acid (116). Vasodilating epoxyeicosatrienoic acids (EETs) have cardioprotective effects (117, 118), and increasing EETs via administration of selective soluble epoxide hydrolase inhibitors shows beneficial effects in animal models of ischemia/reperfusion injury (119–122). Moreover, lipoxin administration post myocardial infarction improves left ventricular ejection fraction in mice (123). RvD1 promotes the resolution of acute inflammation initiated by myocardial infarction and has renoprotective effects, delaying the onset of heart failure and cardiorenal syndrome (124, 125). Finally, RvE1 prevents apoptosis in cardiac myocytes exposed to ischemia/reperfusion and decreases infarct size in rats (126). These experimental data suggest a potential for therapeutic use of SPMs in patients with myocardial infarction, however, no clinical studies have been published to date.

Acute kidney injury (AKI), a frequent complication of critical illness, occurs in more than 50% of ICU patients (127). As management of AKI is largely supportive, early identification of patients at risk is of paramount importance. Several novel biomarkers for early detection of kidney damage have been identified (127), but limitations in specificity and sensitivity have prevented their clinical application. As early lipid changes are involved in the pathogenesis of AKI (128, 129), lipidomic analysis offers—once more—a promising approach for identifying diagnostic and prognostic biomarkers (130). Moreover, SPM have been studied as potential therapeutic agents in AKI due to their organ-protective properties in ischemia/reperfusion (131). In mice, administration of RvD or PD1 before an ischemic insult results in reduced functional and morphological kidney injury (132, 133) and aspirin-triggered resolvin D1 down-regulates the inflammatory response and protects against endotoxin-induced AKI (134).

In summary, analysis and characterization of specific lipid mediator profiles has the potential to improve diagnostic and prognostic accuracy in various conditions commonly encountered in the ICU. Numerous experimental studies provide a theoretical basis for therapeutic administration of lipid mediators in specific circumstances. However, translation from bench to bedside is still in its infancy.