Activation of complement system
The activation of complement system is normally one of the key events in defensive mechanism against pathogen in sepsis. Its protective function for the host rapidly identifies and eliminates invading pathogen whenever possible. Opsonization of foreign surfaces by covalently attached C3b fulfills three major functions: pathogen clearance by phagocytosis; amplification of complement activation by the formation of surface-bound C3 convertase; and assembly of C5 convertase. Cleavage of C5 induces the formation of the multi-protein pore complex C5b-9 (i.e., membrane attack complex [MAC]), which leads to lysis of pathogen [
21].
Although the major role of complement is protective function for the host through innate immune defense, activated complement could also cause the destructive action to the host endothelium [
22], which impacts the course of sepsis as illustrated in Fig.
1. MAC exerts harmful effects to host’s endothelial cells (ECs) [
22] unless complement regulator CD59 is adequately expressed in ECs and protects them by inhibiting C9 polymerization into MAC [
23,
24].
If CD59 is downregulated in the ECs due to either gene mutation or acquired diseases [
25], activated terminal complement MAC could exert destructive effects to ECs in sepsis, trauma and other critical illnesses [
26‐
28]. When MAC attacks innocent bystander ECs, channel (transmembrane pores) formation could occur on the endothelial membrane [
22] and trigger endothelial dysfunction [
3,
22,
29].
The central role of the endothelium
The endothelium, located at the interface between blood and subendothelial tissue (SET)/ extravascular tissue (EVT), is a monolayer of endothelial cells distributed in the vasculature and organ system of the entire human body. Its vital function is maintaining not only homeostasis of circulatory networks but also providing hemostasis in vascular injury. It is essential to preserve both anatomical and functional integrity of the endothelium at all cost to prevent unneeded intravascular thrombosis from the exposure to TF present in SET/EVT [
30], which could cause macrothrombosis and get involved in vascular and organ damage [
20,
31].
Hemostasis plays a critical role in the pathogenesis of sepsis. Sepsis occurring due to a variety of pathogen causes generalized endothelial injury as a result of disseminated nature of the circulatory system, and leads to systemic endotheliopathy. However, sepsis-induced endotheliopathy triggers functional changes contributing to molecular dysfunction and subsequent partial hemostasis, mediating microthrombogenesis. Unlike localized traumatic intravascular injury, which initiates bleeding and hemostasis due to combined endothelial and SET/EVT damage [
4,
20], anatomic disruption to the vascular endothelium is minimal in sepsis and bleeding is not consequential because endotheliopathy is confined to the endothelium and TF is not exposed [
32]. Thus, complement activation due to pathogen or endotoxin provokes endothelial dysfunction promoting exocytosis of ULVWF and platelet activation, which initiates ULVWF path of hemostasis and leads to microthrombogenesis [
2,
3] as shown in Figs.
2 and
3. This is a very important conception to understand sepsis-associated coagulopathy because sepsis promotes lone activation of ULVWF path without activation of TF path since SET/EVT damage does not occur. This thesis based on microthrombogenesis refutes the current theory that sepsis-associated “DIC” occurs via activation of TF path that leads to fibrinogenesis. Now, “DIC” is reinterpreted to be DIT that occurs via activated ULVWF-initiated thrombogenesis associated with endotheliopathy as illustrated in Fig.
2 [
2‐
4,
20].
“DIC” is still the result of hemostasis even though TF is not involved [
4,
20]. The exocytosis of ULVWF from Weibel-Palade bodies activates ULVWF path and platelets are easily recruited because ULVWF show extremely high affinity to the platelet. This lone activation of ULVWF path in endotheliopathy promotes the formation of microthrombi and lead to EA-VMTD/DIT [
2,
3,
20]. Thus, microthombi of “DIC” are the same as those of TTP and TTP-like syndrome. “DIC” has been inappropriately conceptualized as a fibrin clot disease produced via activated TF/FVIIa-initiated cascade/cell-based coagulation. This ill-founded DIC must be properly renamed as EA-VMTD/DIT, which hematologic phenotype is TTP-like syndrome. On the other hand, consumption coagulopathy in acute promyelocytic leukemia (APL) that occurs due to pathologic activation of aberrant TF path caused by TF released from leukemic promyelocytes should be called true DIC [
20]. True DIC in APL is made of disseminated fibrin clots that occur without vascular injury. Its hematologic phenotype is always characterized by hemorrhagic syndrome [
2‐
4,
20].
Endotheliopathy-associated microthrombosis as the crux of the clinical phenotypes
A hallmark of advancing sepsis is microvascular dysfunction [
1,
8,
9] due to disseminated microthrombi composed of platelet-ULVWF complexes in multiorgans, which is promoted via microthrombogenesis [
1,
2]. The pathophysiological mechanism inducing circulatory dysfunction that leads to organ ischemia has been well defined in clinical medicine as vascular microthrombosis [
33,
34] or VMTD [
1‐
4].
In addition to hypoxic microvascular dysfunction, endotheliopathy causes the release of various inflammatory cytokines [
7,
35,
36] and bioactive biomarkers from ECs when infected by different types of pathogen [
37]. Unlike adaptive immune mechanism that produces antigen-specific protective antibody response from each pathogen to each host, sepsis-induced endotheliopathy triggers similar, if not the same, endothelial molecular response without specificity among different types of pathogen (Fig.
3 and Table
1). This endothelial dysfunction causes independent inflammation different from vascular microthrombosis. The manifestations of inflammation and microthrombosis are universally similar among each class and each type of pathogens and display no specificity.
Table 1
Examples of thrombocytopenia (TCIP) in sepsis-associated coagulopathy (VMTD)
Bacteria | Neisseria meningitides | adrenals; meninges | Waterhouse-Friderichsen syndrome |
E. coli O157:H7 | bowels; kidneys; brain | GE; HUS; MODS |
MRSA | lung; multiorgans | ARDS |
Klebsiella pneumonia | lungs; multiorgans | ARDS; TAMOF |
Various bacterial sepsis | multiorgans |
Viruses | Ebola | lungs; liver; multiorgans | ARDS; hepatic necrosis |
H1N1 influenza | brain; lungs; multiorgans | Encephalopathy; ARDS |
MERS-CoV | lungs | ARDS |
SARS-CoV | lungs | ARDS |
Hantavirus | heart; lungs; kidneys | HCPS; HPS; HFRS |
Dengue | adrenals; multiorgans | DSS |
SFTS virus | multiorgans | SFTS |
Hepatitis A virus | liver | Fulminant hepatic failure |
Fungi |
Candida albicans
| multiorgans | ARDS |
Rickettsia | Rickettsia rickettsii | skin; multiorgans | RMSF |
Parasites | Plasmodium falciparum | brain; multiorgans | Cerebral malaria; ARDS |
Plasmodium vivax
| lungs; multiorgans | ARDS |
However, an intriguing question is why then the clinical organ phenotype expression of sepsis is so different among the different hosts and also among the different pathogens. For example, the organ phenotypic features of sepsis could be encephalopathy, myocarditis, pancreatitis, acute respiratory distress syndrome, adrenal insufficiency, fulminant hepatic failure, hemolytic-uremic syndrome (HUS) and others, or any combination of them. This is not due to different kinds of VMTD, but is due to at least two underlying genetic variables governing individual organ susceptibility of the host and affinity of the pathogen to specific organ(s).
Endothelial heterogeneity and organotropism
In endothelial pathogenesis of sepsis, the organ phenotype expression is variable among different hosts by the same pathogen as well as different pathogens. Organ phenotypes are determined by two main endowed biological mechanisms: endothelial heterogeneity of host [
38,
39], and organotropism of pathogen [
40‐
43]. Variable clinical organ phenotypic syndromes occur as seen in the same type of pathogen. Examples are hanta virus, causing cardio-pulmonary syndrome in the heart and lungs, Shiga toxin-producing
E. coli, presenting with HUS (STEC-HUS) in the brain, bowels and kidneys, and Neisseria meningitides, inciting Waterhouse-Friderichsen syndrome and meningitis in the adrenals and meninges. Also, the same organ phenotype can occur in different types of pathogen.
Clinicians have used the mysterious combined clinical terms such as hepato-renal syndrome [
44], hepatic encephalopathy [
45], and cardio-pulmonary syndrome [
46] often without identifying the involved pathogen even though sepsis might have existed. Perhaps, this oversimplified designation of organ phenotypes has impeded detecting the underlying etiology of organ syndromes. In modern medical literature, for example, many authors claimed one organ phenotypic syndrome causes several other organ dysfunctions such as acute renal failure, encephalopathy, and hepatic failure [
47‐
49]. In sepsis, the concept of endothelial heterogeneity and organotropism for organ localization and the phenotype of VMTD as expression of underlying pathology support that biorgan or multiorgan manifestations are just different phenotypes of the same disease EA-VMTD [
29]. Likewise, additional organ phenotypes that are often designated as extra-organ manifestations (e.g., extra-renal phenotypes in HUS, extra-pancreatic phenotypes in pancreatitis and extra-pulmonary phenotypes in acute respiratory distress syndrome [ARDS]) are expression of the same hemostatic disease EA-VMTD. Indeed, extra-organ phenotypes are misrepresentation.
The organ phenotype expression associated with EA-VMTD is determined by combined mechanism of underlying endothelial heterogeneity of each host [
29,
38,
39] and organ/tissue tropism of each pathogen/toxin [
40]. For examples, the phenotypes of STEC-HUS [
29,
41] are the combined results of genotypic endothelial heterogeneity of the host and genotypic tropism of STEC on the gastrointestinal tracts, kidneys, lungs, brain and others. Likewise, the phenotypes of Waterhouse-Friderichsen syndrome [
50] associated with Neisseria meningitides are expressed as a result of combined endothelial heterogeneity of the host and tissue tropism of the bacteria in the adrenals and meninges. These concepts are very important in the understanding of genesis of MODS.
Molecular pathogenesis based on “two-activation theory of the endothelium”
Although endotheliopathy is proposed to be the main pathology promoting crosstalk mechanism between inflammation and coagulation in sepsis, both the character of blood clots and thrombogenetic mechanism producing coagulopathy could not be comprehensible by this crosstalk theory [
4]. However, because complement activation in critical illnesses affects several molecular functions of ECs, it fits well with endothelial pathogenesis of sepsis.
The “two-activation theory of the endothelium” (Fig.
3) [
1‐
4,
19,
20] has been proposed based on the role of endotheliopathy not only in sepsis, but also in other critical illnesses. Once endotheliopathy occurs in sepsis, endothelial dysfunction promotes the activation of two independent endothelial pathways; one is inflammatory and the other is microthrombotic. In short, two main molecular events are: 1) release of inflammatory cytokines such as interleukin (IL)-1, IL-6, tumor necrosis factor-α, and others [
1‐
3,
7,
35,
36], and 2) activation of the platelet [
51,
52] and exocytosis of ULVWF [
53,
54]. The former triggers inflammation through “activated inflammatory pathway”, and the latter mediates microthrombogenesis via “activated microthrombotic pathway”. Inflammation promotes inflammatory symptoms, but microthrombogenesis mediates thrombosis to produce microthrombi strings composed of platelet-ULVWF complexes [
54‐
56]. Micothrombogenesis is activated when the protease ADAMTS13 is insufficient to cleave the excess of exocytosed ULVWF with or without mild to moderate underlying enzyme deficiency that occurs due to polymorphism or heterozygous mutation of ADAMTS13 [
57‐
60]. These microthrombi strings produce DIT in the smaller and larger vasculatures of various organs [
1‐
4].
Mechanism of sepsis-associated coagulopathy
According to novel “two-path unifying theory” (Fig.
2), in intravascular injury normal hemostasis begins with simultaneous but independent activation of ULVWF path and TF path [
4,
20]. However, in endotheliopathy associated with sepsis, only ULVWF path of normal hemostasis becomes activated because TF is not available in the endothelium [
32]. Endotheliopathy in sepsis triggers exocytosis of ULVWF and activates platelets, and produces microthrombi strings. The microthrombi strings composed of platelet-ULVWF complexes must be the intrinsic character of the blood clots in “DIC” of sepsis that occurs as the result of lone activation of ULVWF path illustrated in Fig.
2. Therefore, the concept of “DIC” in sepsis-associated coagulopathy is due to activated TF path producing “fibrin clots” is an incorrect interpretation. This interpretative mistake has been very costly. Thus, according to “two-path unifying theory”, “DIC” is the same to EA-VMTD/DIT and acute “DIC” is EA-VMTD/DIT with hepatic coagulopathy/hemorrhagic syndrome. EA-VMTD/DIT without hemorrhagic syndrome had been called chronic “DIC” as previously detailed [
2‐
4,
20].
The endothelium positioned at the interface between the blood and SET/EVT functions as the initiating site of the foundry producing thrombus following vascular injury [
4]. Because sepsis-induced vascular injury is typically limited to the endothelium, only ULVWF path becomes activated and causes vascular microthrombosis in the smaller and larger vasculatures [
3]. Sepsis-associated coagulopathy has shown to express the changes in antithrombotic and prothrombotic markers, which include TF path-related markers such as thrombin, antithrombin, thrombin-antithrombin complexes, activated protein C (APC), tissue factor pathway inhibitor (TFPI), and thrombomodulin (TM). These expressed TF path markers have indirectly supported the role of activated TF path in the pathogenesis of “DIC”. However, these TF path-related markers can be explained by alternative mechanisms, including acute DIC with hepatic coagulopathy, combined micro-macrothrombotic syndrome, concomitant vascular injury as a result of surgery and vascular access, and perhaps more commonly MODS resulting in tissue necrosis, especially of the liver, kidneys and lungs, in advancing severe sepsis. In retrospect, expressed TF path markers in sepsis can be ascribed to secondary events that are unrelated to primary pathogenesis of activated ULVWF path as illustrated in Fig.
2.
In patients with severe sepsis, thrombin generation, APC, TFPI and TM are not the true markers of “DIC”, but are interpreted as secondary markers of TF path. The true markers of “DIC” are activated platelet, increased FVIII activity, and increased VWF from, excessive release of ULVWF, and upregulated collagen binding that result from activated ULVWF path [
4,
20]. It is no wonder why clinical trials using antithrombin agents, APC, recombinant TFPI and recombinant TM was unsuccssful in sepsis-associated coagulopathy. It was likely due to the fact that sepsis-associated “DIC” is not true DIC, but is EA-VMTD (i.e., TTP-like syndrome) caused by activated ULVWF path. Further, endothelial injury alone cannot induce fibrinogenesis, but only provokes partial hemostasis, leading to microthrombogenesis, and produces EA-VMTD. On the other hand, APL-associated DIC that occurs due to activation of aberrant TF path is true DIC [
20].
In sepsis-associated coagulopathy, this author believes neutrophil extracellular traps (NETs) participate in thrombosis in the unifying stage of macrothrombus, but the problem of NETosis is with the logic of hemostatic principle and inconsistent characters of thrombosis in vivo and animal models in the reported literature. The essential logic in the formation of thrombosis is normal hemostasis must be accompanied by intravascular injury [
4,
20]. Without it, no thrombosis can be formed because activation of ULVWF path and/or TF path is sine qua non in thrombogenesis. Further, only three exceptions among thrombotic disorders occur without vascular injury. They are (1) TTP, which is aberrant thrombosis because vessel is not injured, but still utilizes “ULVWF” path in ADAMTS13 deficiency, (2) APL, which is aberrant thrombosis because vessel is not injured, but still utilizes “TF” path due to overexpressed TF from leukemic promyelocytes, and (3) heparin-induced thrombocytopenia with thrombosis syndrome, which is pathologic thrombosis because vessel is not injured, and also utilizes neither ULVWF path nor TF path [
4,
20]. However, we know cellular and molecular traps, and perhaps adhesion molecules participate in the hemostatic plug and thrombus formation, which is illustrated in Fig.
2, where blood cells/molecules are trapped in comingling process during blood clots formation (macrothrombosis) in the unifying stage of microthrombi strings and fibrin clots. This author believes that NETosis is not active hemostatic processes, but is a passive one associated with secondary event trapping blood cells and molecules such as DNAs and histones in the process of thrombogenesis.
Now, supported by “two-activation theory of the endothelium” and “two-path unifying theory of hemostasis” in the understanding of sepsis and septic shock, we can easily define all clinical phenotypes of the sepsis-associated syndromes via two independent endothelial pathways promoting inflammation and microthrombogenesis.