In this study, we conducted a causal inference analysis on the association between coagulation-related factors and sepsis-related risks. The results do not support a causal effect of coagulation factor levels on the incidence of sepsis, ICU admission, or 28-day mortality risk in European populations. However, the expression of TMEM173 shows a positive causal effect on the risk of ICU admission in sepsis patients.
Blood coagulation is initiated through different mechanisms involving the intrinsic and extrinsic pathways, ultimately converging on the common pathway, leading to the generation of thrombin and fibrin. TF is a transmembrane glycoprotein primarily located on the surface of extravascular cells under normal conditions and serves as the initiator of the extrinsic coagulation pathway. When blood vessels are damaged, tissue factor is exposed to the blood, where it immediately binds to plasma coagulation factor VII, thereby initiating the coagulation cascade[28]. During infection, coagulation is primarily triggered by the exposure of TF to circulating coagulation factors FVII[29]. And in this situation, mainly derived from monocytes, TF was considered as a response to pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs) [30]. In our results, the plasma levels of TF and FVII did not affect sepsis-related risks. Given the different sources of TF under normal and pathological conditions, this suggests that the regulation of coagulation during infection involves more complex “superparameters.” Moreover, the differences in TF levels under normal conditions are likely insignificant compared to the substantial mobilization and release of TF during infection.
The entire process from coagulation activation to DIC includes the excessive activation of thrombin, the substantial consumption of coagulation factors, the activation of the fibrinolytic system leading to an increase in fibrin degradation products, and the formation of obstructive microthrombi[29]. The anticoagulant and fibrinolytic systems play crucial roles in the amplification phase following the initiation of coagulation. Among the most important components are AT and plasmin. Other significant components include TFPI, protein C, protein S, ADAMTS13, and TM. ADAMTS13 primarily functions by cleaving vWF, thereby regulating its function and indirectly exerting an anticoagulant effect [31]. Protein C is a natural anticoagulant that inhibits coagulation by inactivating factors Va and VIIIa [32]. Protein S acts as a cofactor for Protein C, enhancing its anticoagulant effect [33]. TM is an endothelial cell receptor that binds thrombin on the endothelial surface and activates Protein C, thereby exerting an anticoagulant effect [34]. In some clinical studies, supplementation with TM has been beneficial, and recombinant TM has been shown to reduce mortality in patients with sepsis and DIC [35]. In existing GWAS studies, there are no SNPs associated with plasma levels of Antithrombin (AT) and TFPI that meet the threshold requirements for instrumental variables. Additionally, in the eQTLGen database, there are no cis-eQTLs for the relevant genes that could be used as instrumental variables. The remaining anticoagulant molecules—Protein C, Protein S, ADAMTS13, and TM—showed no presumed causal relationships between their plasma levels and the risks of sepsis occurrence, ICU admission, or 28-day mortality. This situation may imply that the response and function of coagulation factors differ within the dynamic time window from normal conditions to the late stages of sepsis. Identifying the optimal timing and window for intervention could lead to more effective management and treatment of sepsis.
In sepsis, there is a complex interplay between the inflammatory and coagulation systems. Excessive expression of cytokines in DIC, such as TNF, IL-6, and IL-10, can either promote or counteract the coagulation process [36–38]. Inflammatory cytokines can also influence coagulation status by downregulating the activity of the anticoagulant system, allowing endothelial damage, and modulating fibrinolytic activity [39, 40]. TF, FVIIa, FXa, thrombin, and fibrin can induce pro-inflammatory signaling, particularly through the activation of members of the G protein-coupled protease-activated receptor family [41]. Plasma kallikrein is a serine protease whose zymogen form is activated by binding to high-molecular-weight kininogen (HMWK) and subsequent activation by factor XI. It plays an essential role in various physiological processes, including coagulation, fibrinolysis, vasodilation, inflammation, and regulation of vascular permeability [42]. PAI-1, an acute phase response protein and a significant inhibitor in the fibrinolytic system, is upregulated in cardiovascular diseases, metabolic syndrome, venous thrombosis, cancer, and inflammation [43–46]. In sepsis patients, elevated PAI-1 levels are positively correlated with disease severity and mortality [47]. Preclinical models also suggest that PAI-1 is a critical factor in acute kidney injury in septic mice [48]. In this study, our results do not support a presumed causal relationship between the plasma concentration levels of these molecules and sepsis-related risks.
During sepsis, there is significant bidirectional regulation between the innate immune system and the coagulation system, mediated by platelets and neutrophils. Neutrophil elastase and cathepsin G can promote coagulation by enhancing the functions of TF and FXII, as well as by activating the local proteolytic mechanism of the coagulation inhibitor TFPI, thus stimulating coagulation and intravascular thrombus growth [8]. Activated platelets can form complexes with neutrophils, lowering the threshold for the release of neutrophil extracellular traps (NETs). NETs are web-like structures composed of DNA, histones, and various antimicrobial proteins released by neutrophils in response to pathogens, playing essential roles in immune defense, inflammation, autoimmune diseases, and cancer [49]. NETs promote intravascular coagulation and facilitate platelet adhesion, activation, and aggregation [50, 51]. Moreover, the extent of NET formation can predict the occurrence of DIC and mortality in sepsis patients [52]. Currently, there are no effective GWAS studies identifying SNPs that influence platelet levels. In the process of NET formation, critical proteins include myeloperoxidase, peptidylarginine deiminase 4, elastase, and cathepsin G. We examined the cis-eQTLs of ELANE and CTSG, the genes encoding elastase and cathepsin G, which are directly related to coagulation factors, and found that they also do not have presumed causal relationships with sepsis-related risks.
Recent studies indicate that pyroptosis, an intersection of immune and inflammatory responses, plays a crucial role in the activation of the immune and coagulation systems during infectious diseases. Both caspase-1 and caspase-11 cleave GSDMD, which can form pores in the plasma membrane, promoting the release of IL-1β, IL-18, and inducing cell death [53–55]. The increase in cell membrane calcium, dependent on TMEM173, drives the cleavage and activation of GSDMD, leading to the release of tissue factor [27]. Previous research has also shown that upregulation of the TMEM173 pathway is associated with the severity of disseminated intravascular coagulation and mortality in sepsis patients. In our study, the expression of TMEM173 showed a positive presumed causal relationship with the risk of ICU admission in sepsis patients, which is the only positive result. The regulation of coagulation during infection appears to overlap only partially with normal coagulation pathways downstream. Given the extensive systemic regulation of the coagulation system and the multi-organ damage seen in sepsis, further exploration of treating coagulation abnormalities under infectious conditions may need to be redirected beyond the known coagulation regulatory mechanisms.
Over the past decade, more than 200 clinical studies have failed to identify methods that can consistently and specifically rescue sepsis patients[56]. One significant reason is the complex regulatory mechanisms of the immune and coagulation systems, and the cascading reactions within the last one. Our Mendelian randomization study revealed that abnormalities in coagulation factors do not increase the risk of sepsis incidence, ICU admission, or 28-day mortality. This result suggests that: (1) Immune system-dependent abnormal activation of the coagulation system is relatively minor during the onset of sepsis. (2) The impact of the cascading activation of the coagulation system in sepsis far exceeds the influence of abnormalities in individual coagulation or coagulation-related factors. (3) The regulation intensity and types of molecules involved in the coagulation system under infectious conditions may differ significantly from normal conditions, consistent with the findings related to TMEM173. Our study suggests that systemic responses may be more crucial than the action of individual factors in treating coagulation disorders in sepsis. It is essential to consider the immune system’s response first. Additionally, given the minor degree of immune system-dependent abnormal activation of the coagulation system during the early stages of sepsis, further research is needed to identify the timing for recognizing and intervening in coagulation system disorders. The timing of coagulation abnormalities could be used as a parameter for subclassifying sepsis patients.
The limitations of this study include that the samples were from European populations, and more data is needed to validate these findings in other populations. Existing studies lack data on certain important coagulation molecules (e.g., platelets, PAD4, TFPI, D-dimer) that meet the analysis threshold criteria.