Effect of therapeutic plasma exchange on endothelial activation and coagulation-related parameters in septic shock

Sepsis is defined as a life-threatening organ dysfunction caused by a dysregulated host response to infection; if hypotension is refractory to volume resuscitation and serum lactate is elevated, it is termed septic shock [1]. In the absence of a specific intervention other than anti-infective drugs, mortality associated with septic shock is still as high as 60% [2]. The overwhelming host response consisting of cytokine release, attraction of inflammatory cells, and global endothelial activation is the key driver of morbidity and mortality [3]. Endothelial dysfunction leads to systemic aggregation of platelets in virtually all microvascular beds with consecutive consumption of coagulation factors. Clinically, this situation is well known as disseminated intravascular coagulation (DIC) resulting in widespread clotting of the microvasculature system and progressive multi-organ failure [4].

The critical role of acquired protein C deficiency in septic shock with purpura fulminans has been recognized for a long time [5‐8], and some experts suggest to supplement protein C if its activity is severely lowered [8]. Interestingly, the only ever approved “specific sepsis therapeuticum” was an activated protein C preparation (Xigris®) [9], before it was later removed from the market due to a large-scale negative follow-up study [10].

A disintegrin and metalloprotease with a thrombospondin type 1 motif, member 13 (ADAMTS13) is the most important modulator of size and function of the von Willebrand factor antigen (vWF:Ag) in human plasma by cleaving preformed vWF multimer chains [11]. Severe ADAMTS13 deficiency causes accumulation of ultra-large vWF multimers (ULVWF) leading to the clinical picture of severe thrombotic microangiopathy as seen in thrombotic thrombocytopenic purpura (TTP) [12]. More recently, it has been demonstrated that a significant deficiency of ADAMTS13 also appears in sepsis [13‐15]. At the same time, large amounts of vWF:Ag are secreted by the activated septic endothelium leading to both increased platelet aggregation and formation of highly pro-thrombotic ULVWF [16]. In fact, an increased vWF:Ag/ADAMTS13 ratio has been repeatedly associated with the severity of shock and organ failure as well as increased mortality in sepsis [13, 17‐20].

Our group has recently described the use of therapeutic plasma exchange (TPE) against fresh frozen plasma (FFP) as an adjunctive treatment strategy in early (onset < 12 h) and severe (norepinephrine (NE) dose > 0.4 μg/kg/min) septic shock [21]. Although uncontrolled, we observed rapid stabilization of hemodynamics, improvement of fluid balances, and reduction of key pro-inflammatory cytokines and permeability factors. We believe that these positive surrogate effects can be attributed to two important aspects: (1) removal of harmful circulating molecules and (2) replacement of protective plasma proteins consumed by the disease process [22].

Here, we hypothesize that TPE against plasma from healthy donors might (1) remove mediators of increased microvascular clotting (e.g., vWF with large and in many cases ultra-large vWF multimers) and (2) simultaneously compensate for the loss of factors important for regulating coagulation (e.g., activated protein C, ATIII) and fibrinolysis (e.g., ADAMTS13). In this study, we investigated in patients with early and severe septic shock the effect of a single TPE against FFPs on key factors regulating coagulation.

This prospective single-center explorative study examined the effects of early TPE on endothelial activation and coagulation-related parameters in septic shock. Patients included in this investigation experienced a severe form of septic shock as indicated by exceedingly high NE requirement, inflammatory markers, and a high prevalence of multi-organ failure. Low platelet counts with high DIC scores indicated the possibility of overt DIC in these patients. However, traditional DIC scores like the ISTH-DIC Score have been reported to have a poor sensitivity, especially in infectious diseases [29]. The SIC score has been suggested recently to be more specific in sepsis [26] but has been questioned in a validation cohort, as the ISTH score had greater power than the SIC score in predicting ICU mortality in septic patients [30]. We acknowledge the limitation of clinical scores to adequately comprehend DIC in septic patients. In this report, we have followed the current expert opinions that recommend to use the SIC score as a screening tool to identify patients with possible sepsis-associated coagulopathy and subsequentially calculate an overt DIC score like the ISTH score [29]. At the same time, hyper-fibrinolysis as indicated by elevated D-dimer concentrations was present. Patients with septic shock displayed a significant shift of the hemostatic balance, characterized by activation of pro-coagulant proteins and attenuation of anti-coagulant activity at the same time. In this sense, pro-coagulatory factors such as vWF-Ag, ULVWF, and fibrinogen were markedly increased at inclusion while anti-coagulant proteins such as ATIII, protein C, and ADAMTS13 were depleted. These observations are well in line with two previous investigations describing these changes in septic patients [14, 20]. Especially, the vWF-cleaving protease ADAMTS13 was reduced in 94% and severely reduced in 55% of patients. This deficiency of ADAMTS13 was associated with both high vWF:Ag and ULVWF concentrations as well as with thrombocytopenia and increased LDH concentrations. Both changes are characteristic for the presence of thrombotic microangiopathy (TMA) with ongoing hemolysis as a consequence of low ADAMTS13 activity. These cardinal features of TMA by a relative deficiency in ADAMTS13 activity were described previously in septic patients and were associated with worse outcomes [13, 17, 19].

In this analysis, we observed higher DIC scores in patients with low ADAMTS13. In contrast, an earlier study by Peigne et al. found no correlation between the DIC score and ADAMTS13 [18]. A possible explanation concerning these contradictory results might be the differences in the severity and duration of septic shock as well as the severity of organ failures between these two cohorts. Peigne et al. described the lower need for both mechanical ventilation and renal replacement therapy. Additionally, inflammatory activity indicated by IL-6 concentration was more than ten times higher in our cohort with a median (IQR) IL-6 of 10,810 pg/mL (2538–27,570) compared to 895 pg/mL (330–1843) in the study by Peigne et al.

The clinical relevance of reduced ADAMTS13 is well known from a disease termed TTP in which the formation of ULVWF multimers due to reduced ADAMTS13 leads to thrombocytopenia, hemolysis, AKI, and (often severe) neurological symptoms. Supplementation of ADAMTS13—that is found in high concentration in the plasma from healthy donors—by TPE can (in some cases within minutes) positively affect these symptoms.

The question arises if it is possible to influence the slightly more complex dysbalance of the hemostatic system in clinical sepsis scenario. Treatment with activated protein C (APC) has demonstrated efficacy in selected patients, e.g., with purpura fulminans [7, 8], but has failed in larger more general sepsis populations [10]. No recombinant ADAMTS13 is available until now, and regular renal replacement regimens have no ability to reduce pro-coagulant factors. We hypothesized that the exchange of septic shock plasma with fresh frozen plasma (FFP) from healthy donors does not only lead to a reduction of harmful proteins but also rather replenishes protective factors that had been consumed by sepsis analogously to TTP. In fact, a key finding of our analysis is that TPE might have the potential to partially attenuate this dysbalance by removing pro- and simultaneously replacing anti-coagulant factors.

A high vWF:Ag/ADAMTS13 ratio has been associated previously with the development of TMA, higher severity of shock, degree of organ failure, and worse outcomes in patients with septic shock [20]. Here, we not only can confirm the elevation of this ratio in septic shock but we can also demonstrate a putative therapeutic strategy to recalibrate this disequilibrium in the coagulation process.

This study has important limitations, mainly its small sample size affecting the power of the results, the single-center setting, and the lack of a clinical correlate that reflects the actual microcirculatory state. Currently available methods in clinical use for microvascular perfusion monitoring such as biomarkers of endothelial damage/activation biomarkers), videomicroscopic techniques, tissue oxygenation evaluation techniques, and partial pressure of carbon dioxide (pCO₂)-based evaluation techniques were not employed in this study. Unfortunately, the size of this study was too small to demonstrate significant differences in ADAMTS13, vWF:Ag, or the presence of ULVWF between survivors and non-survivors or patients with and without shock reversal by TPE, respectively. As this study was a single-center study, there is a potential for patient selection bias. In addition, the intervention was administered at a fixed dose, which precludes us from providing data on the effects at different dosages or time frames.

In this prospective study, septic shock was associated with activation of pro-coagulant pathways such as vWF:Ag and ultra-large VWF multimers and consumption of anti-coagulant factors such as ATIII, protein C, and ADAMTS13 activities. Early TPE treatment partially attenuated this dysbalance by removing pro- and replacing anti-coagulant factors. It has yet to be determined by an appropriately powered RCT if TPE might improve the outcome in septic patients.