Fibrinolysis Resistance After Injury Is a Risk Factor for a Hospital-Acquired Pneumonia-Like Disease Pattern

Abstract

Background:

Pneumonia is associated with increased morbidity and costs in the intensive care unit (ICU). Its early identification is key for optimal outcomes, but early biomarkers are lacking. Studies suggest that fibrinolysis resistance (FR) after major abdominal surgery is linked to an increased risk of infection.

Patients and Methods:

Patients in a randomized controlled trial for hemorrhagic shock were evaluated for FR. Fibrinolysis resistance was quantified by thrombelastography with exogenous tissue plasminogen activator (tPA-TEG) at 24- and 48-hours post-injury and measuring LY30 (%). A receiver-operating characteristics (ROC) curve analysis was used to identify a cutoff for increased risk of pneumonia, which was then validated in ICU patients at risk for venous thromboembolism (VTE). Multivariable logistic regression was used to control for confounders.

Results:

Forty-nine patients in the hemorrhagic shock cohort had tPA-TEGs at 24- and 48-hours (median ISS, 27; 7% pneumonia). A composite tPA-TEG LY30 of less than 4% at 24 and 48 hours was found to be the optimal cutoff for increased risk of pneumonia. This cohort had a seven-fold increased rate of pneumonia (4% vs. 28%; p = 0.048). Eighty-eight patients in the VTE cohort had tPA-TEGs at 24 and 48 hours post-ICU admission (median ISS, 28; 6% pneumonia). The tPA-TEG LY30 of less than 4% was associated with a 10-fold increased rate of pneumonia (19% vs. 1.5%; p = 0.002). In patients with traumatic brain injury, the same association was found (33% vs. 3.2%; p = 0.006). Adjusting for confounders, the tPA-TEG persisted as a substantial risk factor for pneumonia (adjusted odds ratio [OR], 35.7; 95% confidence interval [CI], 1.9–682; p = 0.018).

Conclusions:

Fibrinolysis resistance quantified by tPA-TEG within 48 hours of ICU admission is associated with an increased risk of pneumonia in patients in hemorrhagic shock and those at risk for VTE. Prospective validation of the tPA-TEG LY30 optimal cutoff for pneumonia and further investigation into whether endogenous FR is a cause of an altered immunity is warranted.

Hospital-acquired pneumonia is a common and resource intensive infection in hospitalized patients, occurring in 1% of all patients and 10% of those mechanically ventilated.¹ Pneumonia is associated with increased morbidity and mortality, particularly for those admitted to the intensive care unit (ICU)² in which mortality can be as high as 30%.³ Moreover, pneumonia is associated with extended hospital stay and increased costs.^4–6 Early identification of pneumonia is essential for optimal outcomes, but this is currently limited by non-specific early clinical signs and symptoms,^7,8 as well as limited biomarkers for determining increased risk or confirming a diagnosis.⁹ This is particularly difficult post-injury because of confounding injuries (e.g., pulmonary contusion). The window between occult and overt infection can result in the loss of intervention time with ultimate progression to multiple organ failure.¹⁰ This gap in knowledge underscores the urgency for new diagnostic tools to aid in earlier recognition and treatment of pneumonia.

For more than 50 years, viscoelastic assays such as thrombelastography (TEG) have been used for resuscitation of critically ill and injured surgical patients, using changes in clot formation and degradation to guide clinical decision making.¹¹ Thrombelastography use in the ICU has identified a low fibrinolytic state with worse outcomes in patients with sepsis.¹² More recent work indicates that resistance to fibrinolysis following major abdominal surgery is associated with a three-fold increased risk in infection, particularly in patients with persistent fibrinolysis resistance (FR).¹³ This association is consistent with animal models that show that plasmin, the major driver of fibrinolysis, modulates the immune response through leukocyte adhesions and migration.¹⁴

Considering the link between FR and infectious complications in critically ill patients, we sought to investigate the relation between FR and pneumonia in trauma patients. Thrombelastography quantification of FR using exogenous tissue plasminogen activator (tPA) has been demonstrated to identify adverse outcomes in both trauma¹⁵ and transplant¹⁶ patients that conventional TEG does not capture. A recent clinical trial treating hemorrhagic shock (HS) with pre-hospital plasma¹⁷ and a prospective observational study for risk for venous thromboembolism (VTE)¹⁸ were completed with tPA-TEGs available at multiple time points after injury. As we previously appreciated that the duration of FR was associated with infections after liver transplantation,¹³ we wanted to evaluate the impact of FR and risk for pneumonia in critically ill trauma patients. We hypothesized that sustained FR within the first 48 hours post-injury is associated with an increased risk of pneumonia.

Patients and Methods

Ethics statement

This study was conducted in accordance with guidelines of the Colorado Multiple Institutional Review Board (COMIRB) under approvals #12-1349 and #18-0707.

Test population

An initial analysis was performed on research samples obtained from patients enrolled in the Control of Major Bleeding After Trauma Trial (COMBAT) study,¹⁷ a randomized clinical trial studying the use of plasma in pre-hospital resuscitation of subjects in hemorrhagic shock. From these patients, we included those who had tPA-TEGs available at 24 and 48 hours from injury (post-injury day 1 [PID1] and post-injury day 2 [PID 2]) and dichotomized them based on presence or absence of fibrinolysis resistance, FR and nFR, respectively. These measurements were used in a receiver-operating characteristics (ROC) curve analysis to identify the optimal cutoff (Youden index) of the total tPA-challenged LY30 from both time points (LY30 PID1 + LY30 PID2) associated with an increased risk of pneumonia. We used an ROC to look for an inflection for increased risk of infection in trauma patients, as the degree of FR and associated adverse outcomes are not the same between liver transplant and trauma populations.^13,15

Validation population

To validate the combined LY30 cutoff identified in the test population, a secondary analysis was performed on samples from the Consortium of Leaders in the Study of Traumatic Thromboembolism 2 (CLOTT-2) study,¹⁵ a prospective observational trial of trauma patients at risk of venous thromboembolism (VTE) admitted to the ICU. Similar to the test population, only patients with tPA-TEGs available at 24 and 48 hours were included and dichotomized as FR or nFR, based on the optimal cutoff derived in the COMBAT study patients. Additional a priori defined subpopulation analyses were done on severe traumatic brain injury (TBI, defined as Abbreviated Injury Score ≥3) and recipients of tranexamic acid (TXA) as part of their trauma resuscitation.

Thrombelastography and patient cohorts

For both study populations, blood was collected and stored in citrated tubes and transferred for analysis via trained professional research assistants. Blood samples were assayed with the TEG 5000 Hemostatic Analyzer (Haemonetics, Braintree, MA) according to the manufacturer's recommendations. Fibrinolysis was quantified using the lysis at 30 minutes after maximum clot strength (LY30). The addition of exogenous tPA¹⁹ was used to quantify sensitivity or resistance to fibrinolysis. This was achieved by adding 500 mcL of whole blood into a vial containing lyophilized tPA (Molecular Innovation, Novi, MI) to a final concentration of 75 ng/mL and mixed by gentle inversion. A 340-mcL aliquot of this mixture was then transferred to a 37°C TEG cup, pre-loaded with 20 mcL of 0.2 mol/ L of CaCl₂. The tPA-TEG is not currently Food and Drug Administration (FDA)-approved for clinical practice in the United States, but a modified version is approved for use in Europe.²⁰

Outcomes

The primary outcome of interest was a clinical diagnosis of pneumonia requiring treatment more than 48 hours after injury. Infection was considered present if it was documented in the electronic health medical record, including culture-proven or empirically treated infections.

Exclusion criteria and missing data

Participants with missing tPA-TEG data at pre-defined critical timepoints PID1 or PID2 were excluded from the analysis.

Statistical analysis

Statistical analysis was performed using STATA, version 16.1 (StataCorp, College Station, TX). For the test population, an ROC analysis was performed to identify an optimal cutoff (Youden index) to predict an increased risk of pneumonia. This Youden index was utilized to confirm the definition of total LY30 from both time points (LY30 PID1 + LY30 PID2) for increased risk of pneumonia in the validation population. Multivariable logistic regression was used to control for potential confounders (severe chest injury, severe head injury, age), and to derive adjusted odds ratios (OR) and confidence intervals (CI). A p value <0.05 was considered statistically significant for all comparisons.

Results

Description of study population and primary outcome

A total of 125 patients from the COMBAT randomized trial were reviewed for our test population (Fig. 1A). Seventy-six patients were excluded because of missing tPA-TEG data, leaving 49 (39%) trauma patients who met a pre-hospital definition of hemorrhagic shock (systolic blood pressure [SBP] ≤ 70 mmHg, or SBP of 71–90 mm Hg and heart rate ≥108 beats per minute) to be included in our test cohort analysis (Table 2). For the validation population, a cohort of 240 patients previously enrolled in the CLOTT-2 prospective trial were reviewed. After excluding 152 subjects because of missing tPA-TEGs, 88 (37%) patients were included in our validation cohort analysis (Table 2). The primary outcome of interest was a diagnosis of pneumonia more than 48 hours from injury. The associations between the presence or absence of pneumonia with commonly tracked factors affecting post-injury recovery are shown in Table 3.

FIG. 1.

Study population flowchart. FR = fibrinolysis resistance; nFR = no fibrinolysis resistance; PNA = pneumonia; nPNA = no pneumonia; TBI = traumatic brain injury; AIS = Abbreviated Injury Scale; TXA = tranexamic acid

Table 2.

Demographic Characteristics of Test and Validation Cohorts

	COMBAT (n = 49)	CLOTT 2 (n = 88)
Age (y), mean (SD)	39.6 (14.5)	28.2 (6.4)
Male gender, n (%)	41 (83.7)	76 (86.4)
BMI (kg/m2), mean (SD)	27.6 (5.0)	27.1 (7.9)
White race, n (%)	36 (73.5)	49 (55.7)
Injury mechanism, n (%)
Blunt	29 (59.2)	79 (89.8)
Penetration	18 (36.7)	9 (10.2)
Both	2 (4.1)	0 (0.0)
NISS/ISS, mean (SD)	32.4 (19.1)	30.0 (14.6)
PNA, n (%)	8 (16.3)	5 (5.7)
LY30 %, median (IQR)	4.0 (1.95–10.2)	11.1 (4.2–34.6)
FR, n (%)	23 (46.9)	21 (23.9)

COMBAT = Control of Major Bleeding After Trauma Trial; CLOTT-2 = Consortium of Leaders in the Study of Traumatic Thromboembolism 2; SD = standard deviation; BMI = body mass index; NISS = new injury severity score; ISS = injury severity score; PNA = pneumonia; LY30 = lysis at 30 minutes after maximum amplitude; IQR = interquartile range; FR = fibrinolysis resistance.

Table 3.

Associations Between Commonly Tracked Factors Affecting Trauma Patients and the Presence or Absence of Pneumonia

	PNA COMBAT	No PNA COMBAT	p	PNA CLOTT-2	No PNA CLOTT-2	p
Age (y), mean (SD)	37 (15)	40 (15)	0.740	24 (5)	28 (6)	0.157
NISS/ISS, mean (SD)	54 (12)	28 (17)	< 0.001	47 (10)	29 (14)	0.007
BD, mean (SD)	N/A	N/A	N/A	13 (2)	7 (6)	0.011
SBP (mm Hg), mean (SD)	85 (40)	88 (30)	0.515	123 (26)	116 (29)	0.482
RBC (units), mean (SD)	6 (5)	8 (10)	0.801	5 (6)	4 (8)	0.412
FFP (units), mean (SD)	3 (4)	3 (5)	0.700	3 (3)	4 (6)	0.771
FR, n (%)	88%	44%	0.048	80%	21%	0.002

COMBAT = Control of Major Bleeding After Trauma Trial; CLOTT-2 = Consortium of Leaders in the Study of Traumatic Thromboembolism 2; PNA = pneumonia; NISS = new injury severity score; ISS = injury severity score; BD = base deficit; N/A = not available; SBP = systolic blood pressure; RBC = units of red blood cells transfused in 24 h; FFP = units of fresh frozen plasma transfused in 24 h; FR = fibrinolysis resistance.

Table 1.

PICO(T)

P: Critically ill trauma patients at risk of pneumonia

Keywords: trauma; pneumonia; hemorrhagic shock; venous thromboembolism

I: Development of pneumonia and fibrinolysis resistance

Keywords: fibrinolysis resistance; thrombelastography; pneumonia

C: Prophylactic antibiotic therapy, early diagnosis

Keywords: prophylaxis, antibiotic agents; biomarkers

O: Rate of pneumonia in post-injury period

T: Up to 30 days after injury

Fibrinolysis resistance as an independent predictor of post-injury pneumonia-like disease patterns

These tPA-TEG measurements were used in an ROC analysis to identify an inflection point for increased risk of pneumonia. In the test cohort, a composite tPA-TEG LY30 less than 4% at 24 and 48 hours was found to be the optimal cutoff for increased risk of pneumonia and was used to define pFR (Fig. 2). This group of patients with persistent post-injury tPA resistance had a seven-fold increase rate of pneumonia (4% vs. 28%; p = 0.048; Fig. 3A).

FIG. 2.

Test cohort receiver operating characteristic (ROC) curve.

FIG. 3.

Rates of pneumonia with and without fibrinolysis resistance. tPA = tissue plasminogen activator.

For the validation cohort, a tPA-TEG LY30 total on PID1 and PID2 of less than 4% was associated with a 10-fold increased rate of pneumonia (19% vs. 1.5%; p = 0.002; Fig. 3B). When adjusting for confounders (age, injury severity score), the tPA resistance emerged as a significant risk factor for pneumonia (adjusted OR, 35.7; 95% CI, 1.9–682; p = 0.02).

Among patients with severe TBI, tPA resistance was associated with a similarly high 10-fold increased risk of pneumonia (33% vs. 3.2%; p = 0.006; Fig. 3C). For those who received a TXA bolus, no association was seen between a low fibrinolytic state and pneumonia (25% vs. 0%; p = 0.267; Fig. 3D), however, only 15 patients received TXA, rendering this comparison underpowered.

Discussion

Our data support that persistent FR is common after trauma. Patients with this prolonged fibrinolytic-resistant state appear to be at increased risk of developing a pneumonia-like disease pattern. This has clinical relevance because the timely diagnosis of pneumonia remains a challenge in polytraumatized patients. A barrier to an early treatment is the reduced effectiveness of commonly used diagnostic algorithms, which are based on clinical symptoms that may be masked early in the post-injury period because of the inflammatory response to injury.²¹ To overcome this, multiple biomarkers,^22,23 machine learning models,²¹ therapeutic interventions,²⁴ and prophylactic antibiotic agents have been proposed.^25,26 Despite these efforts, pneumonia is seen up to four times more frequently in trauma patients than in other critically ill patients.²⁷

Generalized coagulopathy is a well-described complication of severe injury whereas FR in these patients has received less attention.²⁸ Several degrees of decreased fibrinolytic activity may result from major trauma,²⁹ which in itself is associated with a variety of post-injury complications³⁰ and increased mortality.³¹ Fibrinolysis resistance, quantified with the tPA-TEG, has been linked to infections in the post-operative period of liver transplant patients.¹³ Fibrinolysis resistance is also associated with poor outcomes in critically ill patients with sepsis.^12,32,33 Herein we showed that FR is also associated with an increased risk of pneumonia in injured patients. When measured by viscoelastic assays, FR may serve as an early marker of pneumonia risk, especially when it is persistent in the immediate post-injury period.

Although there is no established mechanism for the link between FR and infection, several pathways have been proposed. One such pathway involving a disruption in the plasmin(ogen) axis was first described by Colucci et al.³⁴ in this 1985 report, the authors showed that an acute phase protein that functioned as the main regulator of tPA activity (later described as plasminogen activator inhibitor-1 [PAI-1]), was overexpressed in response to endotoxins. This overexpression subsequently led to a shutdown of fibrinolytic activity. A link between PAI-1 overexpression in response to infection has been further supported by the work of Lorente et al.,³⁵ who showed that persistently elevated PAI-1 levels were associated with worse outcomes in patients with sepsis. In the trauma population, elevated post-injury PAI-1 is associated with increased all-cause mortality.³⁶ However, despite this well-documented association, PAI-1 testing is not standard-of-care and its quantification relies on labor intensive analyses; i.e., enzyme-linked immunosorbent assay (ELISA). Conversely, fibrinolytic activity can be reliably, timely, and affordably measured by viscoelastic assays like TEG, which are already routinely used in trauma resuscitation.^37,38 Furthermore, the tPA-TEG assay has been correlated to PAI-1 levels in a trauma population.³⁹

Other potential mechanisms involving the plasmin(ogen) axis focus on the role of chemokines and their task of generating an immune response in the face of potential infection. One example is the c-c motif chemokine ligand 21 (CCL21), which is cleaved and activated by plasmin⁴⁰ and plays a major role in the adaptive immune system through antigen-presenting cell (i.e. dendritic cells) migration to peripheral tissues to mount an immune response.^41,42 An alternative pathway could be related to systemic fibrin deposition in the microvasculature. This was illustrated by historic studies that show that repeated bouts of endotoxin promote this event.⁴³ In this setting of persistent fibrin deposition, for example, recent data by Silva et al.⁴⁴ suggest that fibrin, in its role as a regulator of neutrophil recruitment through its β₂ integrin receptor and in the setting of impaired fibrinolysis, persistent fibrin an localized inflammation occur.

Clinical significance

Although studies identifying the specific mechanisms that link FR to infection are lacking, our data add to the evidence of their relation. Specifically quantifying FR beyond the initial injury stage, and perhaps up to 48 hours after injury, could provide useful data to risk-stratify patients for risk of pneumonia. With a seven-fold increased risk of pneumonia in patients who presented with hemorrhagic shock, and up to 10-fold increase in those at risk of VTE, biomarkers such as C-reactive protein, pro-calcitonin, and neutrophil gelatinase-associated lipocalin (NGAL)⁴⁵ could play a role in guiding early antibiotic therapy in patients with FR. Identifying and treating an infection as early as possible has a survival benefit in critically ill trauma patients.⁴⁶

Furthermore, in the setting of impaired fibrinolytic activity, excess fibrin deposits and subsequent neutrophil overactivation, as described above, may be a contributor to the development of persistent inflammation, immunosuppression, and catabolism syndrome (PICS). This syndrome has been associated with an increased risk in nosocomial infections secondary to myelodysplasia in critically ill patients.⁴⁷ Such association, among others described in the literature showing a relation between low fibrinolytic activity and poor outcomes, may suggest a role for readily available treatments focusing on activating the fibrinolytic response. One such treatment is tPA, which has been used in the acute phase of ischemic stroke for more than 25 years.⁴⁸ More recently, its application in critically ill patients was studied during the coronavirus disease 2019 (COVID-19) pandemic, where the Study of Alteplase for Respiratory Failure in SARS-CoV-2 COVID-19 (STARS) trial⁴⁹ combined tPA with systemic anticoagulation to counteract the deleterious effects of microvascular thrombi on respiratory function. The STARS investigators concluded that tPA boluses followed by therapeutic anticoagulation was safe and showed improvements in oxygenation and trends towards reduced overall mortality.

Limitations

This study is limited by its retrospective nature in which there is an inability to control for confounders such as site-specific treatment and testing guidelines and their effects on diagnosis and treatment of the primary outcome. Furthermore, the studies from which our analyses gathered data were not designed to quantify pneumonia as an outcome. This is confounded by the fact that there were no established standard criteria to diagnose and treat pneumonia, which limits our ability to state with certainty that all patients included had analogous infections. In addition, as discussed in the rationale for this study, confirming a diagnosis of pneumonia in a severely injured and critically ill patient is a difficult task. On top of the already insidious onset of hospital-acquired pneumonia, its presentation in a polytraumatized patient may be further compounded by pulmonary contusions, pulmonary edema, and leukocytosis in the setting of a generalized inflammatory response.^50,51

All these processes, as well as other common conditions associated with trauma can contribute to a presentation similar to pneumonia. Given the lack of standardized diagnostic and treatment criteria across the two trials analyzed in this study, we cannot rule out the possibility of having included patients whose clinical presentation resembled that of pneumonia, but not necessarily met a textbook definition of an acute infection of the pulmonary parenchyma.

Moreover, the cohort analyzed by this study is also limited by a convenient sample of patients that had tPA-TEGs available at 24 and 48 hours after injury. Furthermore, the diagnostic criteria for FR was also based on a ROC curve to find an inflection point for pneumonia and warrants validation in a prospective manor. Overall, these results provide useful preliminary data that require prospective validation in a large, multicenter population, properly powered to find statistically significant differences in outcomes. The applicability of this assay, however, is limited by the lack of a commercially available, off the shelf tPA-TEG kits in the United States. This may change in the future, as a new off the shelf tPA-based viscoelastic assay, ClotPro (Haemonetics, Braintree, MA), has proven effective in quantifying FR in critically ill patients outside the United States.⁵²

Conclusions

Fibrinolysis resistance quantified by the tPA-TEG during the first 48 hours of ICU admission is associated with an increased risk of pneumonia in both patients with hemorrhagic shock and those at risk for VTE. These findings warrant prospective validation of the tPA-TEG LY30 optimal cutoff for pneumonia, and further investigation into whether endogenous FR is a cause of an altered immune response.

Author's Contributions

Conceptualization: Rodriguez, H. Moore. Data curation: Sauaia. Formal analysis: Sauaia, H. Moore. Methodology: H. Moore. Writing—original draft: Rodriguez. Writing—review and editing: Rodriguez, Saben, Moore, Knudson, P. Moore, Pieraacci, Sauaia, H. Moore. Supervision: Saben. Funding acquisition: P. Moore, Knudson, H. Moore.

Funding Information

This study was supported in part by National Heart Lung and Blood Institute: R00-HL151887 (H.B.M.), The University of Colorado's Academic Enrichment Fund (H.B.M.), and the Department of Defense via the Medical Research and Development Program W81XWH-1- 17-1-0673 (M.M.K.) and United States Army Medical Research Acquisition Activity W81XWH-12-2-0028 (E.E.M.).

Author Disclosure Statement

H.B.M. and E.E.M. have an issued patent for the tPA-TEG that is licensed through the University of Colorado. The remaining authors have no relevant conflicts to disclose.