Should endotoxin be a research priority in Gram-negative sepsis and septic shock?

1. Introduction

Endotoxin, from the ancient Greek meaning ‘the poison inside,’ refers to lipopolysaccharide (LPS) that constitutes the Gram-negative bacteria (GNB) main membrane component of the cell wall and it is fundamental for its life annuity. LPS macromolecule contains a toxic lipid A-structure, which is highly conserved across the range of GNB, an inner and outer glycopeptide core structure, and an O-polysaccharide chain along its outer membrane surface, which is greatly species-specific and varies highly between GNB. Endotoxin functions as a pathogen-associated molecular pattern (PAMP) molecule that serves as a ligand for the recognition receptors exposed on the host effector cells of the innate immune system, the Toll-like receptors type 4 (TLR4). Three LPS receptors have been isolated in human cells: (a) membrane-bound or soluble CD14-MD2-TLR4, (b) the macrophage scavenger receptor molecules, and (c) the beta 2 integrins. LPS, in human biological fluids, circulates binding to LPS-binding protein (LBP), that is a hepatic acute phase protein. LBP works as an interlocked molecular assembly able to move LPS to CD14 receptors on the surface of immune effectors cells. CD14 is a glycosylphosphatidylinositol-linked protein recognized principally on the myeloid cell’s membranes; it needs an essential extracellular adapter protein known as myeloid-2 differentiation factor (MD2) to be able to transduce the LPS signal across cell membranes of activated target cells [1]. When the LPS-MD2 complex is presented to TLR4 extracellular domain a series of signaling pathways supervene with successive activation of serine and tyrosine kinases leading to phosphorylation, ubiquitylation and activation of nuclear factor kappa B and other transcription factors. Promoter regions of nuclear factor kappa B binding sites are cytokines, clotting components, complement, nitric oxide synthase, chemokines, and acute phase proteins genes. The inflammatory cytokines (tumor necrosis factor-α, interleukins-1, 2, 8, 12, and 18) and other mediators of inflammation produce generalized inflammation with diffuse microvascular thrombosis, tissue injury, organs failure, and septic shock (SS) after LPS exposure (Figure 1) [2,3]. There is a lot of evidence demonstrating the implication of endotoxin in SS pathogenesis, especially that induced by GNB. Indeed, experimentally, endotoxin can reproduce most of the SS associated pathophysiological features [4]. Animals that do not express the LBP (resistant to endotoxin) are insensitive to the effects induced by GNB infections. Furthermore, the inoculation of a small amount of LPS to animals or human volunteers mimics the vast majority of the pathophysiological features of sepsis. It has shown that the cardiovascular effects of endotoxin are similar to those observed in sepsis: tachycardia, an increase in cardiac index and a decrease in systemic vascular resistance leading to a drop in blood pressure [5]. Endotoxin can be measured in the serum, and as explained above, it could be a useful test for the diagnosis of sepsis due to GNB. Until recently, the Limulus Amoebocyte Lysate (LAL) was the most widely used test in assessing endotoxin in serum. However, this assay is technically challenging, and antibiotics can interfere with the test. The endotoxin rapid test ‘Endotoxin Activity’ (EA) has emerged as a quicker, and therefore more useful, test for the detection of endotoxemia in whole blood. However, the result of this test is not immediately available at the patient’s bedside. High levels of EA are associated with SS, and Multiple Organ Dysfunction (MOD) due to GNB as well as prolonged hospitalization and mortality rates in critically ill patients admitted to intensive care units [6]. Recently, it has been shown that patients with GNB infection, early after SS occurrence, had a high EA prevalence [6]. In addition, endotoxemia was significantly associated with the severity of SS.

Table 1. Characteristics of Studies.

Study	Reference	Year	Time span	Abdominal sepsis	Source control
EUPHAS I	[7]	2009	Dec 2004 – Dec 2007	100%	100%
EUPHAS II	[8]	2016	Jan 2010 – Dec 2014	44%	not clearly reported

Figure 1. A simplified illustration of the relationship between endotoxin and immune host cells. adapted from Marshall JC [3].

MD2: myeloid differentiation factor-2; TLR4: Toll-like receptor 4; MyD88: myeloid differentiation factor 88; IRAK: interleukin receptor-associated kinase; TRAF2: tumor necrosis factor receptor-associated factor 2; NFκB: nuclear factor κB; MAP: mitogen-activated protein; IL-1: interleukin 1; TNF: tumor necrosis factor; PAF: platelet-activating factor.

2. Endotoxin in the diagnosis of sepsis

In the Third International Consensus Definitions for Sepsis and SS (Sepsis-3), lactatemia appeared to be the only reasonable marker of severity as hyperlactatemia is predictive of higher mortality [7]. Lately, in a single-centre prospective observational study, the EA value was evaluated as an early indicator of GNB infection, disease severity, and patient risk in a large cohort of SS patients. In the eighty-two percent of the hundred and seven enrolled patients, an intermediate (≥0.4 units) or high (≥0.6 units) EA values were observed within 24-h of SS onset. A higher mean of EA was found in patients with SS due to GNB compared to patients without GNB cultures [6].

3. Endotoxin as a prognostic factor in sepsis

LPS heterogeneity significantly reduces the interpretation of quantitative endotoxin assays and limits an accurate comparison of endotoxin levels among individuals. On the other hand, an endotoxin removal developed using polymyxin B (PMX) that bind and neutralize endotoxin seems to show encouraging results suggesting a role of endotoxin in the sepsis pathophysiology. Patients with intra-abdominal infections, mostly caused by Gram-negative bacteria, represented the majority of the population studied in the EUPHAS studies (I and II): PMX hemoperfusion (PMX-HP) added to standard treatment was associated with improvement of hemodynamic and organ dysfunction parameters (mean arterial pressure [MAP], vasopressor requirement, oxygenation, Sequential Organ Failure Assessment [SOFA] score), and with lower death rate (28-day mortality) (Table 1) [7,8]. The populations in which the extracorporeal technique showed clinical benefits, associated with the removal of endotoxin, were highly heterogeneous in terms of the source of infection and severity scores at the time of admission. Therefore, in these populations, there were both patients with only SS and septic multiple organ failure due to GNB [7,8]. Similarly, in a retrospective study on patients with SS due to intra-abdominal or GNB infection, Lee et al. showed that PMX-HP was associated with a significant benefit on death rate (28-day mortality: odds ratio [OR] 0.18, 95%CI 0.04–0.92), inotropic and Acute Physiology and Chronic Health Evaluation II (APACHE II) scores [9]. Furthermore, a recent meta-analysis showed a reduction of the overall mortality among patients treated with PMX-HP (risk ratio 0.81; 95%CI, 0.70–0.95). When subgroups of patients with sepsis and SS were analyzed, PMX-HP was associated with a lower mortality in the intermediate (risk ratio, 0.84; 95%CI, 0.77–0.92) and high-risk (risk ratio, 0.64; 95%CI, 0.52–0.78) groups, but not in the low-risk group (risk ratio, 1.28; 95%CI, 0.89–1.84) [10]. In contrast, a multicenter randomized controlled study demonstrated a non-significant increase in mortality (28-day mortality: OR 1.59, 95%CI 0.86–2.94 and 90-day mortality: 1.61, 95%CI 0.90–2.87, respectively) and no improvement in organ failure with PMX-HP therapy than the standard treatment of peritonitis-induced SS [11]. A randomized study – the EUPHRATES trial – comparing PMX-HP with standard therapy did not show any difference in 28-day mortality (relative risk, 1.09; 95%CI, 0.85–1.39), or reduction in EA between groups or from baseline, all participants randomized in the EUPHRATES trial had a high (≥ 9 points) MOD Score and elevated EA assay level (≥ 0.6) [12]. However, a recent post hoc analysis of the EUPHRATES trial suggested that, in selected patients with baseline EA assay level between 0.6 and 0.89 (no extreme endotoxemia), PMX-HP was associated with an absolute 28-day mortality benefit, improvement in blood pressure with a trend of a lesser vasopressors requirement, and higher mechanical ventilation-free days compared to the control group with no PMX-HP [13].

4. Expert opinion

Endotoxin appears to be present in most patients who meet standard clinical criteria for sepsis, particularly in GNB sepsis. This high prevalence of endotoxemia in this setting suggests that the EA assay might also be a useful surrogate marker of sepsis severity. Conceptually a quick and extensive removal of endotoxin might limit the inflammatory response, which characterizes SS, with a better prognosis in specific categories of patients. Despite the perception of the beneficial effects of extracorporeal endotoxin removal have been known for more than 20 years, the guidelines of the Sepsis Surviving Campaign suggest only the measurement of procalcitonin, as biomarker, either to shortening or to stopping the antibiotic treatment, both with weak recommendation because of the low quality of evidence [14,15]. The major limitation in endotoxin use is that no method is available for a rapid endotoxin detection at patient’s bedside when it would be more relevant in therapeutic management, and only recently, nanomedicine is responding to this need. The endotoxin removal in the more advanced phases of sepsis, SS, septic multiple organ failure and extreme endotoxemia at baseline is probably the reason why the patients have a less favorable prognosis. In fact, in the EUPHRATES trial (in which PMX-HP did not show a reduction in 28-day mortality in patients with extreme endotoxemia) participants had a higher Multiple Organ Dysfunction Syndrome (MODS) score compared to the previous studies that showed an advantage of extracorporeal endotoxin removal in a higher heterogeneous population in terms of severity scores at admission [7–9]. Therefore, a timely extracorporeal endotoxin removal might be beneficial if it is performed in less severe SS patients, in terms of severity scores at admission and endotoxemia level at baseline (EA assay 0.6 to 0.89) with a high grade of suspicion of GNB infections (abdominal and genitourinary tract). However, this point needs to be validated in a randomized controlled trial. The guidelines of the Sepsis Surviving Campaign suggest several preventative maneuvers (some of low-level evidence) such as venous thromboembolism prophylaxis and stress ulcer prophylaxis, which minimally contribute to the SS therapy. In contrast, the detection of endotoxin and its removal to prevent multiple organ failure and reduce mortality is not considered. This omission from the current SS guidelines might potentially deprive our less severe SS patients with a GNB infection from one of the effective SS approaches, in critical care medicine, with is an evidence-based, at worst, of grade 2B, which goes back 10 years [7–9,15]. In an era of multi-resistance, where the antibiotic option for the treatment of severe infections due to GNB in intensive care patients, especially of abdominal origin, is very limited, the research into different strategies seems to be inalienable [16].

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.