Introduction
Fibrinolysis, the breakdown of fibrin, is a normal process that maintains blood flow within the vasculature. Due to the progressive decrease in blood vessel diameters, fibrin deposits and small thrombi are filtered into smaller blood vessels, and here, fibrinolysis is required for dissolving such thrombi. Fibrinolysis is triggered following the release of tissue plasminogen activator (t-PA) by platelets and the endothelium, which converts plasminogen to plasmin, that then cleaves fibrin. This process is modulated by the concentrations of t-PA and plasminogen, as well as several inhibitors that target these proteins or modify fibrin to prevent its degradation [
1,
2].
Severe infection and tissue injury induce dynamic and interconnected responses in the inflammatory, coagulation and fibrinolytic systems [
3,
4]. Significant perturbations in fibrinolysis may result in severe haemorrhage (hyperfibrinolysis) or macro- or micro-thrombosis potentially leading to end-organ damage (fibrinolysis resistance) [
5‐
9], with several studies providing evidence that initial hyperfibrinolysis transitions to fibrinolysis resistance due to factor consumption [
9,
10]. In critical conditions, such as sepsis, COVID-19 and non-COVID-19 acute respiratory failure (ARF) and extensive trauma, fibrinolysis resistance has been associated with elevated levels of plasminogen activator inhibitor-1 (PAI-1) [
6‐
8,
11‐
13]. In another study in COVID-19 disease, elevated levels of both PAI-1 and t-PA levels were reported, with hyperfibrinolysis associated with high t-PA levels and a poor outcome [
14]. A consumptive mechanism for the fibrinolysis resistance seen in severe COVID-19 disease was postulated by Medcalf et al. [
15]. The authors propose that levels of fibrin and necrotic material, generated during the infective/inflammatory process in the lungs, consume plasmin leading to pathway exhaustion from deficiency of factors such as plasminogen. Furthermore, that fibrinolysis can be restored by enhancing plasmin formation either via administration of t-PA or its substrate, plasminogen [
15]. This hypothesis is supported by a study demonstrating that low plasminogen levels are associated with increased mortality in severe COVID-19 disease [
16] and by the apparent beneficial effects of direct fibrinolytic approaches using either t-PA [
17] or plasminogen [
18].
Fibrinolysis is a highly localised process, which only happens in the presence of t-PA. This dynamic process is not assessed by any standard coagulation assay. Approaches to evaluate the fibrinolytic response to added t-PA have been repeatedly investigated in recent years, both in plasma and in whole blood. As platelets also play a role in fibrinolysis [
19,
20], functional tests for the fibrinolytic response to t-PA in fresh whole blood, rather than platelet poor plasma, would seem appropriate. Whole blood-based approaches for the functional assessment of the fibrinolytic response to t-PA have usually been performed using viscoelastic testing (VET), a method that allows for a comprehensive assessment of coagulation and fibrinolysis by a continuous assessment of blood clot firmness [
21]. Approaches to perform VET with added t-PA have been described since 2006 [
20,
22‐
26]; however, only recently has a standardised assay been commercially available and approved for diagnostic use. In the ClotPro TPA-test, coagulation is triggered by a combination of recombinant tissue factor and calcium chloride, and fibrinolysis is triggered by a high dose of recombinant t-PA. The assay has been used in various studies that evaluated the decrease of fibrinolysis in critically ill patients following COVID-19 infection [
27‐
30] with several studies correlating VET parameters with the severity of COVID-19 disease [
31‐
35].
This study in critically ill patients with fibrinolysis resistance aimed to evaluate the ability of t-PA and plasminogen supplementation to restore fibrinolysis with assessment using point-of-care (POC) ClotPro viscoelastic testing (VET). We performed prospective observational and interventional point-of-care experiments using ClotPro and its TPA-test in critically ill patients to initially select individuals with fibrinolysis resistance. These results are presented for COVID-19 patients, non-COVID-19 patients and healthy controls to enable a comparison to be made of the severity of fibrinolysis resistance present, given the prominence in the recent literature on fibrinolysis resistance in COVID-19 including novel trials of t-PA and plasminogen administration [
17,
18,
36]. Importantly, these trials have not incorporated individualised assessments of fibrinolysis resistance or treatment responses. Consecutive exploratory VET analyses were conducted in a proportion of fibrinolysis-resistant patients with ex vivo spiking of the TPA-test with additional t-PA with/without plasminogen to assess the requirements for restoration of fibrinolysis. In addition, in a fibrinolysis-resistant patient with bacterial pneumonia causing ARF, alteplase (t-PA) was administered over 24 h and standard ClotPro VET was used to regularly monitor the effect on coagulation and fibrinolysis.
We hypothesised, based on the existing literature, that fibrinolysis-resistant patients would display significant variation in the degree of fibrinolytic compromise that could be corrected through supplementation with additional t-PA and/or plasminogen. We also hypothesised that the ClotPro TPA-test would offer rapid point-of-care capability to monitor the effect on coagulation and fibrinolysis of intravenous t-PA administration in hypofibrinolytic patients, thus permitting a personalised dose that maximised benefit while minimising risk.
Methods
ClotPro® (enicor GmbH, Munich, Germany) was used for viscoelastometry. ClotPro is a new-generation viscoelastometry analyser, which has shown good agreement with the widely used ROTEM delta device [
37,
38].
In all individuals, blood was collected into 3.2% buffered sodium citrate tubes and analysed by ClotPro within 30 min of collection using the EX-test®, FIB-test® and TPA-test® (standard Clotpro assays) as per the manufacturer’s instructions. In all three assays, coagulation is activated by recombinant tissue factor, CaCl2 is used to recalcify the sample and polybrene is used to block heparin effects. In the FIB-test, the platelet contribution to the clot is blocked by a combination of cytochalasin D (an actin polymerisation inhibitor) and eptifibatide, a small molecular inhibitor of the glycoprotein α2bβ3 receptor (GpIIb/IIIa). In the TPA-test, recombinant t-PA (650 ng/mL) is used to trigger fibrinolysis and the lysis time (LT) is calculated by the associated software that indicates the time taken for the maximum clot firmness to be reduced by 50%. Blood fibrinogen concentrations will influence the volume of the fibrin clot formed and endogenous blood levels of the fibrinolytic proteins, t-PA and plasminogen, and their inhibitors will influence the rate of clot lysis. Standard ClotPro assays were performed by ICU nurses, registrars, residents, consultants and research staff trained on the ClotPro device, with a subset of medical and research staff carrying out the experimental tPA/plasminogen supplementation assays.
The manufacturer’s normal ranges were used for the EX-test and FIB-test. The platelet contribution to the clot, platelet A10, was calculated by subtracting the FIB-test A10 value from the EX-test A10 value. For the determination of fibrinolysis resistance, we used TPA-test lysis times of 300 s or greater, this value aligning with the 90th percentile of a blood donor population (
n = 60; 304 s) [
29], and being > 4 standard deviations above the mean LT obtained in our healthy staff controls (
n = 20; mean 180.4 ± SD 28.6 s) (Fig.
1D).
A total of 105 COVID-19-infected and non-COVID-19-infected patients admitted to the intensive care unit were screened using standard ClotPro assays. Fifty-eight individuals were identified as fibrinolysis-resistant (TPA-test LT > 300 s). Of these 58, 32 were included in the ex vivo supplementation experiments (described below). Availability of reagents and staff trained in the experimental procedures determined which patients were included in these supplementation experiments. Twenty controls (healthy ICU staff members) were also studied to establish normal fibrinolysis parameters (Additional file
1: Fig. S1).
Ex vivo VET supplementation experiments
Repeat viscoelastometry was consecutively performed on the same blood sample in fibrinolysis-resistant patients with ex vivo spiking of the TPA-test with additional t-PA (650–1300 ng t-PA/mL blood) by passing the blood through 1 or 2 additional TPA-test tips (enicor GmbH, Munich, Germany). Such supplementation would also double the concentrations of tissue factor, CaCl2 and polybrene being added to the blood. To study the effect of increased concentrations of these reagent components on coagulation, sub-studies were performed using blood samples from two patients which were passed through 1 and 2 EX-test tips prior to the commencement of the assay. These experiments demonstrated a < 5% difference in the clot amplitude at 10 min (A10), maximum clot firmness (MCF) and lysis index at 30 min (CLI30).
Plasminogen (Hyphen-Biomed, France) was initially added to the ClotPro cup prior to the commencement of the assay using freshly thawed plasminogen reconstituted in deionised water (stock 2 µg/uL) to achieve a final concentration in the ClotPro analysis cup of 59 µg/mL blood, about a third of reported blood concentrations [
39]. Subsequently, to streamline the process for point-of-care testing and to enable an increased dose to be tested without diluting the blood, preloaded pipette tips were prepared with lyophilised plasminogen that could be used on the ClotPro automated pipette and that achieved a concentration of 147 µg/mL blood, this value being on the lower end of reported normal blood concentrations (150–200 µg/mL). The blood was drawn up through the plasminogen-loaded tip and placed into the cup with a mixing step, prior to the blood then being drawn up from the cup through the TPA-test tip (containing CaCl
2) with the usual mixing step. The assay was then commenced after this second step.
In vivo study
In a fibrinolysis-resistant patient with ARF entered into a clinical trial of VET-guided treatment using low-dose alteplase, the EX-test, FIB-test and TPA-test were repeated regularly throughout the 24-h infusion and one hour after cessation and analyses commenced within 10 min of blood collection.
Data analysis
The data were analysed using PRISM V9.0 software (GraphPad, CA). Results are reported as medians including the interquartile range within brackets [IQR], since non-normally distributed as per the Shapiro–Wilk test. Comparisons were made using the Mann–Whitney test for non-normally distributed data (D’Agostino and Pearson test) and reported including the median difference and its 95% confidence interval (95% CI). Correlations were analysed by Spearman’s correlation coefficient (rho) and reported including the 95% CI. Two-sided p-values are reported for all analyses. Statistical significance was set at a two-sided p-value of < 0.05.
Discussion
These data suggest that (i) fibrinolysis resistance is associated with many critical conditions requiring intensive care, and occurs with equal severity in COVID-19 and non-COVID-19 patients, (ii) a TPA-test lysis time of > 300 s appears to distinguish patients with fibrinolysis resistance from patients with slowed fibrinolysis due to elevated fibrinogen levels, (iii) ex vivo supplementation of t-PA on its own, or in combination with its substrate, plasminogen, is able to correct the majority of fibrinolysis-resistant cases, (iv) a TPA-test lysis time of > 1000 s appears to predict the beneficial response to combined t-PA and plasminogen supplementation, and (v) ClotPro VET and the associated TPA-test can potentially be used to monitor the response and guide the dose of in vivo t-PA supplementation that is delivered with the intention of restoring fibrinolysis.
In the normal healthy state, the fibrinolytic system is tightly regulated, as demonstrated by the low variation observed in our healthy control population. Significant recent attention has focused on fibrinolysis resistance that occurs in a proportion of patients with severe COVID-19 disease. Our study is a reminder that equally severe fibrinolysis resistance occurs in critically ill patients with a range of non-COVID-19-related diagnoses that have a systemic inflammatory response as a common feature.
The amplitude of the fibrin clot formed in the FIB-test closely correlates with plasma fibrinogen levels [
40]. Correlation analyses between the amplitude of the fibrin clot (FIB-test A10) and the TPA-test LT in patients with a LT ≤ 300 s and those with a LT > 300 s (Fig.
2B) demonstrated that in patients with a LT > 300 s, the degree of fibrinolysis resistance was not due to the formation of larger fibrin clots, as was observed in healthy controls and patients with a LT ≤ 300 s (Fig.
2A). These data suggest that fibrinogen is the leading determinant of TPA-test lysis times when fibrinolysis is occurring normally; however, this relationship is lost, most likely due to a change in the balance of other factors, such as pro- and anti-fibrinolytic protein activity. The extent of fibrin cross-linking and the presence of neutrophil extracellular traps, bacterial or platelet polyphosphates, von Willebrand factor (vWF) and reactive oxygen species within the blood also negatively impact fibrinolysis [
41‐
45].
The data obtained in the ex vivo supplementation experiments identified two states of fibrinolysis resistance: (i) one that could be corrected with t-PA supplementation alone, thus suggesting reduced t-PA activity such as due to an excess of PAI-1, and (ii) the other that required the addition of t-PA and plasminogen, thus suggesting reduced plasmin activity on its own, due to excessive consumption or inhibition by α2-antiplasmin, or in combination with reduced t-PA activity. As part of the acute phase inflammatory response, endothelial cells and platelets release t-PA and plasminogen activator-1 (PAI-1), the principal inhibitor of t-PA [
46]. Thus, an imbalance of t-PA:PAI-1 levels may rapidly develop resulting in fibrinolysis resistance that correlates with multi-organ dysfunction syndrome and death in bacterial and viral infection [
13,
47,
48]. Moreover, in a study of 29 patients critically ill with COVID-19, the ClotPro TPA-test LT was found to significantly correlate with plasma PAI-1 levels (
r = 0.70;
p < 0.0006) [
27].
Reductions in plasmin activity also occur in systemic inflammation and may be due to insufficient t-PA to convert plasminogen to plasmin, high levels of plasmin inhibitors and/or plasmin consumption through intravascular fibrin degradation, tissue repair and several aspects of the immune response [
39,
49]. In ARF patients, inhibition of plasmin activity was reported in bronchoalveolar lavage samples and was partially attributed to increased levels of α2-antiplasmin [
50]. Additionally, in a study of patients with severe sepsis and septic shock, reduced levels of plasminogen and the coagulation inhibitors, anti-thrombin III (ATIII) and Protein C, were measured in comparison to patients with less severe sepsis, in whom fibrinolysis was strongly activated and coagulation inhibited by ATIII. The authors concluded that the consumption of plasminogen and coagulation inhibitors was the principal mechanism leading to fibrinolysis resistance in the more severely ill patients and they demonstrated that the same mechanisms occurred in sepsis due to several different pathogens [
10]. Additional evidence for plasminogen consumption causing fibrinolysis resistance was obtained in burns patients with reduced plasminogen levels associating with the extent of burn injury and development of organ dysfunction [
9]. Also contributing to depleted plasminogen levels is the finding that the protease released from activated neutrophils, neutrophil elastase, is capable of degrading plasminogen [
51]. Several lines of evidence indicate, therefore, that the systemic inflammatory response following severe infection or injury maybe associated with fibrinolytic changes that are dynamic and characterised by early stage hyperfibrinolysis which transitions to fibrinolysis resistance due to factor consumption (Additional file
1: Fig. S3).
The ClotPro TPA-test and the novel exploratory extensions of this test described herein permit identification of fibrinolysis-resistant patients and potentially the corrective treatment required. The results of this study demonstrate the significant variation in the degree of fibrinolysis resistance that occurs between patients, as measured by the TPA-test LT, and in the amounts of t-PA ± plasminogen required to restore fibrinolysis ex vivo. These data imply that a personalised approach to the correction of fibrinolysis resistance is required rather than uniform protocols.
The dichotomous response to plasminogen supplementation may reflect endogenous plasminogen levels. A baseline LT > 1000 s may indicate reduced plasminogen levels; thus, supplementation in the presence of t-PA enhanced clot lysis. In contrast, where the baseline LT is < 1000 s, supplemented plasminogen combined with sufficient endogenous levels may result in the inhibition of clot lysis. We are unaware of this effect being previously described, although a previous study demonstrated inhibitory effects of high t-PA levels on plasmin lysis of fibrin [
52]. Further ex vivo experiments are planned to correlate viscoelastometry results with laboratory-based measures of fibrinolytic proteins and to investigate the mechanism of the observed inhibitory effect.
The utility of the ClotPro TPA-test to monitor in real time the effect on fibrinolysis of systemic fibrinolytic protein administration, was demonstrated, in this case with alteplase. The working hypothesis for the VETtiPAT-ARF trial (NCT05540834) is that the infusion of low-dose t-PA over days will correct reduced t-PA activity resulting in the restoration of fibrinolysis but without compromising coagulation. The unchanged EX-test and FIB-test parameters measured during the alteplase infusion demonstrated that coagulation in response to tissue factor was preserved throughout. A similar method could also be used to monitor plasminogen administration or the combination of both proteins. This novel method mimics the use of activated partial thromboplastin clotting time (aPTT) to monitor therapeutic heparin administration, with the contemporary capacity to dose the supplemented fibrinolytic protein/s according to the TPA-test LT. Additionally, this approach enables the dose of the fibrinolytic protein to be tailored to the patient’s requirements, thus potentially increasing efficacy while reducing thrombosis and bleeding risk, as well as cost.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.