Acute Pancreatitis Induces a Transient Hypercoagulable State in Murine Models

Abstract

Background/Objectives:

Although understudied, risk of venous thromboembolism (VTE) appears to be increased during acute pancreatitis (AP). We aimed to further characterize a hypercoagulable state associated with AP utilizing thromboelastography (TEG), a readily available, point of care test.

Methods:

AP was induced in C57/Bl6 mice using L-arginine and caerulein. TEG was performed with citrated native samples. The maximum amplitude (MA) and coagulation index (CI), a composite marker of coagulability, were evaluated. Platelet aggregation was assessed using whole blood collagen-activated platelet impedance aggregometry. Circulating tissue factor (TF), the initiator of extrinsic coagulation, was measured with ELISA. A VTE model using IVC ligation followed by measurement of clot size and weight was evaluated. After IRB approval and consent, blood samples from patients hospitalized with a diagnosis of AP were evaluated by TEG.

Results:

Mice with AP displayed a significant increase in MA and CI, consistent with hypercoagulability. Hypercoagulability peaked at 24 hours after induction of pancreatitis, then returned to baseline by 72 hours. AP resulted in significantly increased platelet aggregation and elevated circulating TF. Increased clot formation with AP was observed in an in vivo model of deep vein thrombosis. In a proof of concept, correlative study, over two thirds of patients with AP demonstrated an elevated MA and CI compared to the normal range, consistent with hypercoagulability.

Conclusions:

Murine acute pancreatitis results in a transient hypercoagulable state that can be assessed by TEG. Correlative evidence for hypercoagulability was also demonstrated in human pancreatitis. Further study to correlate coagulation measures to incidence of VTE in AP is warranted.

Introduction:

Acute pancreatitis (AP) is one of the most common gastrointestinal diseases resulting in inpatient admissions and can present with a wide range of severity that can result in significant morbidity and mortality ¹. Multiple inflammatory cells and cytokines have been implicated in the pathophysiology of pancreatitis subsequent to the inciting event. The end result of this inflammation is pancreatic tissue injury and activation of a systemic inflammatory cascade that can precipitate a wide variety of clinically significant sequelae, e.g. infected pancreatic necrosis, respiratory failure secondary to acute respiratory distress syndrome, renal failure, etc. Venous thromboembolism (VTE) is another prominent co-morbidity of AP that has been recently recognized but understudied ². Hospitalized patients with acute pancreatitis are noted to have a 2-fold elevated risk of developing VTE compared to age matched hospitalized patients ³. Patients with VTE are also noted to have an increased mortality rate following hospitalization as well as an increased rate of other significant complications ⁴.

Despite evidence of this increased incidence of VTE in patients with pancreatitis, there have been no previous attempts to quantify an induced state of hypercoagulability in pancreatitis. Thromboelastography (TEG) is a viscoelastic assay that measures clotting of whole blood over time measured using a spinning wire probe. TEG has gained increasing clinical prevalence in trauma, transplant surgery and critical care as a means of identifying coagulopathies and guiding transfusion to correct these derangements. However, the potential capacity of this assay to identify a hypercoagulable state has only more recently been explored ^5–7. We aimed to utilize thromboelastography as a clinically translatable assay to evaluate for a hypercoagulable state in a murine model of acute pancreatitis and corroborate these findings in patients with acute pancreatitis.

Materials and Methods:

Murine Models and Treatments

All experimental procedures were approved by the Institutional Animal Care and Use Committee of West Virginia University (Protocol # 1809018204) and performed in accordance with the policies and regulations established by the office of laboratory animal services of West Virginia University. Euthanasia was performed under ketamine or isoflurane anesthesia with cardiac puncture for exsanguination followed by cervical dislocation. Mice were housed in cages provided by the animal care facility within the West Virginia University Health and Science Center with standard housing and care. Immunocompetent C57/BL6 wild type mice aged between 8–14 weeks were purchased from Jackson laboratories and housed within these facilities.

Induction of Acute Pancreatitis

Induction of acute pancreatitis was performed using either L-arginine or caerulein injections as previously described ^{8, 9, 10}. The L-arginine solution was prepared under sterile conditions in phosphate buffered saline (PBS), and pH was adjusted to 7. Mice received two intra-peritoneal injections one hour apart at a dose of 2, 3, or 4 g/kg, while control mice received equivalent volume injections of phosphate buffered saline (PBS) ^8,9. This dosing to modify pancreatitis severity has been previously published and histologically validated. A separate cohort of mice received 7 hourly injections of 100μg/kg caerulein diluted in PBS followed by an additional 7 hourly injections 48 hours later ¹⁰. Mice received a sub-cutaneous bolus of 0.5 mL of normal saline and were monitored closely for pain and tolerance of the injection. Mice were administered sub-cutaneous 0.05 mg/kg buprenorphine, TID for up to 3 days, when showing signs of pain or discomfort.

Analysis with Thromboelastography (TEG)

After induction of pancreatitis, mice were monitored for 48 hours prior to euthanasia. Cardiac puncture was performed to obtain 400 μL of native blood into syringes containing 40 μL of 3.4% sodium citrate. This citrated native blood was immediately evaluated with thromboelastography utilizing the TEG 5000^™ (Haemonetics, Boston, MA). The blood (340 μL) was added to Haemonetics TEG 5000 disposable cups and pins (Heamonetics 6211) containing 20 μL of calcium chloride. R time (min), K time (min), α-angle (degrees), maximum amplitude (MA, mm), Ly30 (%), and coagulation index (CI) were recorded. TEG curves are reported on a linear axis with time in minutes on the x axis and amplitude in mm, a measure of clot strength, on the y axis ¹⁰.

A time course experiment analyzing coagulation after pancreatitis induced by 4g/kg L-arginine performed. After induction of pancreatitis, mice were euthanized, and blood was obtained at 6 hours, 24 hours, 48 hours, 72 hours, and 96 hours after induction of pancreatitis. TEG was performed at each time point of interest.

Tissue Factor Assay

To investigate sources of hypercoagulability in acute pancreatitis, we analyzed serum from mice 48 hours after 4g/kg L-arginine treatment as well as control mice. The serum was obtained after the blood was allowed to clot for 30 minutes and then spun at 1000 g for 10 minutes. The serum was diluted 10-fold prior to analysis. We assessed murine serum tissue factor concentrations utilizing ELISA (LS Bio, LS-F14709) according to the manufacturer’s instructions.

Impedance Platelet Aggregometry

Utilizing a platelet aggregometer Model 700 (Chrono-log, Havertown, PA), impedance platelet aggregation was performed on 500 uL of murine whole blood from mice 48 hours after treatment with arginine (4g/kg) or control. Blood was drawn via cardiac puncture at the time of euthanasia as described above and immediately warmed to 37°C within the aggregometer. After warming for 5 minutes, a stir bar was added to the vial containing the blood and the blood was placed within the test position in the aggregometer. 1 μg of collagen (Chronolog) was added to each vial and impedance testing was run for 6 minutes following addition of collagen. The area under the curve and maximum impedance values were obtained.

Deep Vein Thrombosis Model

In order to correlate TEG findings with in-vivo clot formation, we studied a murine model of deep vein thrombosis (DVT) formation that utilizes ligation of the inferior vena cava below the renal veins as previously reported ^{11, 12}. 24 hours after induction of pancreatitis, mice were given intra-peritoneal ketamine (90 mg/kg) and xylazine (10 mg/kg) to induce anesthesia. The abdomen was shaved and prepped with sterile iodine solution. A midline laparotomy incision was performed and the infra-renal IVC was dissected free of the aorta. The IVC and prominent side branches were ligated with a combination of 6–0 silk suture ligature and cauterization. The incision was closed in two layers and the mice were recovered. 48 hours after IVC ligation, mice were euthanized, and the incision was re-opened. The clot was retrieved from the IVC, and clot length and weight were measured.

Human Pancreatitis Blood Samples

Blood was collected from patients admitted to Ruby Memorial Hospital with acute pancreatitis as part of the protocol approved by the institutional review board of West Virginia University (IRB #1901409584). Blood was drawn at time of admission after obtaining informed consent. Samples were evaluated by TEG using citrated native blood within 1 hour of the sample being drawn. Clinical data, including patient demographics, treatment characteristics and clinical outcomes were collected by chart review.

Statistical Analysis

Data are reported as mean ± standard error of mean. Analysis was performed using Student’s two tailed t-test for comparison of two groups in experimental models. ANOVA was used to assess differences between coagulation profiles across time points for pancreatitis and control cohorts. p < 0.05 were considered significant.

Results

Thromboelastography (TEG) Demonstrates a Hypercoagulable State in Murine Acute Pancreatitis

Our initial goal was to determine if TEG could be utilized to demonstrate and quantify a hypercoagulable state in murine models of induced acute pancreatitis. We performed TEG in wild-type C57/BL6 mice 48 hours after control injections or induction of pancreatitis with either 4g/kg arginine or caerulein (Table I). The R time reflects the time to clot initiation and is a reflection of coagulation factors. Interestingly, AP mice demonstrated a prolonged R time (5.3 ± 0.77 vs 3.0 ± 0.66 minutes, p<0.05, Figure IA) indicating a prolonged time to clot formation. No difference was observed in the K time or the alpha angle, both reflections of clot kinetics (Table I). Maximum amplitude (MA) reflects the total clot strength and is a reflection of platelet activation. TEG demonstrated a significant elevation in MA in mice with L-arginine induced AP as compared with controls (72.86 ± 1.23 vs 62.27 ± 3.08 mm, p <0.001, Figure IB). In addition to clot formation, we evaluated fibrinolysis in mice with acute pancreatitis by obtaining the Ly30 (%), a measure of clot breakdown 30 minutes following attaining the maximum amplitude of clot strength. This value is reported as the percent by which the clot has weakened in the aforementioned timeframe. Lysis at 30 minutes occurred in mice in the control group but not mice in the pancreatitis group (5.44 ± 2.03 vs 0 ± 0 percent, p<0.001, Figure IC). Coagulation index (CI) is a composite measure that evaluates multiple data points provided by the TEG and provides a weighted score of these variables that reflects the overall coagulation status. There was a significantly elevated CI in mice with induced AP compared with controls (4.20 ± 0.25 vs 3.04 ± 0.38, p<0.05, Figure ID), consistent with a hypercoagulable state. Representative TEG curves demonstrating these changes are shown in Figure IE.

Table I.

Mean values for TEG variables in both control and 4g/kg L-Arginine induced pancreatitis groups 48 hours after the induction of pancreatitis.

TEG Variables	R Time (min)	K Time (min)	α -angle (Deg)	Max Amplitude (mm)	Lysis 30 (%)	Coagulation Index
Control (n=10)	3.0 ± 0.65	2.11 ± 0.28	64.44 ± 2.86	62.27 ± 1.67	5.44 ± 2.03	3.04 ± 0.38
Pancreatitis (n=6)	2.95 ± 0.92	2.08 ± 0.73	66.1 ± 7.28	72.83 ± 2.59	0 ± 0	4.75 ± 0.43
p-value	0.762	0.973	0.838	0.020	0.026	0.012

Figure I:

A) R time (minutes) in mice with AP induced by 4g/kg L-arginine (n=16) or caerulein (n=5) was significantly prolonged as compared to controls (n=10) 48 hours post-induction of AP on TEG. B) Maximum amplitude (mm) of clot strength in mice with AP was significantly elevated compared to controls. C) Ly30, reported as the % clot lysis 30 minutes after maximum clot strength, was significantly lower in mice with acute pancreatitis, with pancreatitis mice having no lysis. D) Coagulation index, a composite measure of TEG coagulation parameters, was significantly elevated in mice with L-arginine induced AP compared with controls. E) Representative TEG curves of control (pink), L-arginine induced pancreatitis (green), and caerulein induced pancreatitis (white).

To validate these findings in another model of murine pancreatitis , we used TEG to quantify the coagulation profile in caerulein induced pancreatitis. Compared to control injections, we again found an elevated R time (6.52 ± 1.10 vs 3.0 ± 0.66 minutes, p<0.05, Figure IA), MA (77.64 ± 1.66 vs 62.27 ± 3.08 mm, p <0.0001, Figure IB), and CI (4.63 ± 0.41 vs 3.04 ± 0.38, p<0.01, Figure ID). The Ly30 was also found to be decreased in caerulein induced pancreatitis compared to controls (0 ± 0 vs 5.44 ± 2.03 percent, p<0.01, Figure IC). We again found no difference the clot kinetics as reflected by the K time and alpha angle.

In the clinical setting, patients present with a range of pancreatitis severity. L-arginine at 4g/kg simulates severe pancreatitis but other doses have also been reported that induce mild (2 g/kg) to moderate (3 g/kg) pancreatitis ^{12, 13}. Given our initial findings of hypercoagulability following 4g/kg L-arginine, we investigated these lower doses. The R time remained prolonged at 3g/kg (7.63 ± 0.76 minutes, p<0.001) and 2g/kg (6.67 ± 1.10 minutes, p<0.01) as compared to controls (Figure IIA). Similarly, the MA was elevated at 3g/kg (69.46 ± 1.92 mm, p<0.05) and 2g/kg (70.17 ± 1.55, p<0.01) (Figure IIB). There was also noted to be no fibrinolysis after 30 minutes after either 3g/kg or 2g/kg (Figure IIC). No difference as compared to controls was noted for the K time, alpha angle, or CI after treatment with 3g/kg and 2g/kg L-arginine (Figure IID).

Figure II:

A) Mild (2g/kg, n=9), moderate (3g/kg, n=10) and severe (4g/kg, n=16) pancreatitis was induced using L-argingine. R time (minutes) was elevated in AP mice regardless of pancreatitis severity as compared to controls (n=10). B) Maximum amplitude (mm) of clot strength remained consistently elevated across all pancreatitis severities. C) No fibrinolysis, as demonstrated by the Ly30, occurred at any stage of pancreatitis. D) The coagulation index, a composite measure of the coagulation profile, was significantly elevated after induction of severe pancreatitis with 4g/kg L-arginine but not after mild or moderate pancreatitis. E) A time course of the hypercoagulable state associate with acute pancreatitis, as measured by an elevated coagulation index (CI) on TEG. CI is significantly elevated compared to controls beginning 24 hours after induction and continuing to 48 hours, but eventually returning to baseline by 72 hours

After our initial studies demonstrated a significantly elevated MA and CI at 48 hours, we sought to determine a temporal relationship between the onset of pancreatitis and hypercoagulability. TEG was performed at multiple time points following induction with 4g/kg L-arginine. There was a significant difference between the pancreatitis group and control across time points (p=0.03). CI becomes significantly elevated beginning at 24 hours, extending to 48 hours, but returning to baseline within 72 hours after induction of pancreatitis (Figure IIE).

Elevated Tissue Factor and Enhanced Platelet Aggregation Drive Hypercoagulability in Murine Acute Pancreatitis

In addition to TEG, we demonstrated evidence of hypercoagulability with corroborative serum measures. Tissue factor is released due to tissue injury and is the initiator of the clotting cascade. In mice with acute pancreatitis, there were increased circulating levels of serum tissue factor (352 ± 10 pg/ml vs 273 ± 34 pg/ml, p<0.05, Figure IIIA) 48 hours after pancreatitis induction.

Figure III:

A) Plasma tissue factor, the initiator of extrinsic coagulation, is elevated in mice with L-Arginine induced pancreatitis mice compared with control (n=4/group). B) Platelet aggregation, as demonstrated by AUC obtained from impedance platelet aggregometry, is significantly elevated in mice with acute pancreatitis (n=4/group).

Because the MA was significantly elevated on TEG and is a measure of platelet activity, we confirmed the role of platelets in pancreatitis induced hypercoagulability using impedance platelet aggregometry. Collagen stimulated whole blood in mice with acute pancreatitis had significantly increased platelet aggregation (AUC 32.8 ± 1.17 vs 23.8 ± 2.67, p<0.001) compared to controls (Figure IIIB).

In-vivo Clot formation in Acute Pancreatitis

Given the demonstration of a transient, hypercoagulable state in acute pancreatitis, we next sought to demonstrate whether this hypercoagulable state translated to increased clot formation in vivo. Inferior vena cava (IVC) ligation is an established murine model for evaluating deep vein thrombosis as a way to quantify the propensity for clot formation . Due to the findings demonstrating the greatest fold change in CI to be 24 hours after induction of acute pancreatitis, IVC ligation was performed at this time. Mice with AP had a significantly increased clot weight (12 ± 3.46 vs 3.17 ± 1.33 mg) (Figure IVA) and clot length 4.67 ± 0.67 vs 1.5 ± 0.74 mm) (Figure IVB) compared with controls. These findings suggest an increased propensity for clot formation as a result of a hypercoagulable state induced by acute pancreatitis.

Figure IV:

Inferior vena cava ligation was performed in mice with and without acute pancreatitis as a model of deep vein thrombosis. Mice were then euthanized and the clot was harvested for weight and length to assess coagulability. A) Mean clot weight is significantly elevated in mice with AP (n=7) as compared with controls (n=8). B) Mean clot length is significantly increased in mice with induced AP compared with controls. These data suggestion that mice with acute pancreatitis are more prone to clone formation.

Patient Correlative Assessment of Pancreatitis Induced Hypercoagulability

To corroborate our murine findings, we evaluated the prevalence of hypercoagulability in a small, prospective cohort of patients with AP. Human blood samples were obtained at time of admission from patients admitted with a primary diagnosis of acute pancreatitis. Patients presented at varying times from onset of symptoms to admission and had diverse clinical characteristics (Supplemental Table I). The majority of these patients had mild to moderate pancreatitis without necrosis, and the preponderance had gallstone induced pancreatitis. The median number of days between TEG and onset of symptoms was 4 (Range: 2–21). TEG demonstrated a majority of patients to have a hypercoagulable state as compared to the reference ranges provided by the manufacturer, with over two-thirds of patients having elevations in both MA (Supplemental Figure IA) and CI (Figure IB) compared with normal reference range ¹⁴. Representative TEGs from patients with pancreatitis are shown demonstrating these changes (Figure IC).

Discussion

Acute pancreatitis is a common gastrointestinal disease with a varying severity and the potential for significant mortality. There are many serious sequelae of this disease, but one of the more prominent and clinically significant is venous thromboembolism (VTE). VTE can be seen as deep vein thrombosis in the upper or lower extremity, pulmonary embolism or mesenteric vein thrombosis, which is common in the setting of acute pancreatitis due to localized inflammation around the portal and superior mesentic veins associated with the disease. Although the rate of VTE and mesenteric vein thrombosis during episodes of acute pancreatitis has not been widely reported, several recent studies demonstrate increased risk for VTE during episodes of acute pancreatitis ^{2–4, 15–21}, with the incidence ranging from 7 to 40% during necrotizing pancreatitis . Furthermore, formation of VTE in patients with AP is an independent variable associated with increased morbidity and mortality ⁴. However, despite these reports, a hyercoagulable state in acute pancreatitis has not been well characterized or quantified.

The current findings suggest that there is a transient, hypercoagulable state induced by AP in two murine models of acute pancreatitis. While mild and moderate acute pancreatitis induced changes in coagulation, there were more significant derangements toward hypercoagulability in severe pancreatitis. Further, this hypercoagulability is demonstrated utilizing a readily available, point of care test, TEG, that is poised for clinical translation. TEG has been most widely utilized in clinical practice to provide information of coagulopathies in critically ill, trauma and surgical patients and guide blood product transfusions ^22–25. However, this data adds to a growing body of literature demonstrating that it can also be used to reproducibly quantify a hypercoagulable state in multiple diseases and settings ^26–30. Supporting the TEG findings, we also demonstrate an in-vivo difference in clot formation in mice with acute pancreatitis as demonstrated increased clot size and weight following inferior vena cava ligation. These findings suggest that our TEG data is recapitulating a true physiologic difference in propensity to form clot and a demonstrable hypercoagulable state in AP. Although IVC ligation is an established technique for thrombosis research, it is an invasive procedure that can induce physiologic stress ^{11, 12}. However, mice have a much lower propensity for spontaneous clot formation than humans and spontaneous clots are difficult to observe, making further induction of thrombosis necessary in this model.

Given these initial findings, we sought preliminary, proof-of-concept corroboration of this hypercoagulability in human acute pancreatitis. We obtained blood samples from inpatients admitted with a primary diagnosis of new onset acute pancreatitis. TEG at time of presentation in these patients demonstrated findings similar to our murine samples, with evidence of hypercoagulability affecting a majority of patients. Although higher rates of VTE have been seen in necrotizing pancreatitis ², most of these patients had mild to moderate pancreatitis without necrosis, with very few of them requiring advanced intervention such as a cystgastrostomy or percutenaous drainage and no patients undergoing surgical necrosectomy. The current study is highly limited by a small sample size that precludes drawing conclusions from these corroborative results. These preliminary limited findings warrant further investigation into the use of TEG to potentially predict VTE in patients with pancreatitis.

We have further investigated the transient nature of hypercoagulability with a time course analysis demonstrating a peak in CI after 24 hours with resolution within 72 hours of induction of acute pancreatitis in our murine model. Importantly, the two murine models of arginine and caerulein induced acute pancreatitis studied here do not reflect the varied etiologies of human acute pancreatitis. This time course is likely specific to our murine model and may vary signficantly with different experimental models and across the spectrum of patients with acute pancreatitis. Interestingly, although our murine TEG data suggest a plateau in CI at 24 hours with return to baseline within 72 hours, our preliminary study of patients had evidence for hypercoagulability that occurred despite a wide range in the duration from the patient’s onset of symptoms to the timing of the TEG, with many patients beyond 96 hours of symptoms before presenting to our facility. While in mice, we were able to correlate TEG findings with pathophysiologic clot burden, this was not possible in patients due to the small cohort of patients studied. Only one patient (5%) was diagnosed with a mesenteric venous thrombosis. The small number of patients with VTE is a clear limitation of this study, however, the current work supports further study to correlate TEG findings to VTE events, perhaps in a multi-institutional fashion.

There are likely to be multiple mechanisms responsible for the hypercoagulable state demonstrated in the current work. In both our murine model and patients, we identified an elevated maximum amplitude on TEG, confirmed with platelet aggregometry, suggesting the propensity to form clot as a result of acute pancreatitis is likely mediated by increased platelet activation. Indeed, several studies have confirmed enhanced platelet activity and function in response to the inflammation associated with pancreatic injury during acute pancreatitis ^31–35. Consistent prior literature, we also demonstrate elevation in the initiator of extrinsic coagulation, tissue factor, with acute pancreatitis ^{36, 37}. Tissue factor is likely released from damaged endothelial cells and responding immune cells during pancreatic necrosis and links the inflammatory response with abberations in coagulation ⁴⁰. Neutrophil extracellular traps (NETs), where neutrophil release their intracellular contents in the circulation or tissue, have been implicated in the pathogenesis of acute pancreatitis ^{8, 38–41}. NETs have been strongly associated with a prothrombotic state in pancreatic cancer and other disorders ^42–45 and promote thrombosis through enhanced platelet aggregation and release of tissue factor ⁴². NETs are therefore a promising target for further mechanistic inquiries regarding hypercoagulability and acute pancreatitis. Interestingly, we found a prolonged R time, meaning increased time to clot formation, in mice treated with 4g/kg arginine. An elevated R time is often associated with hypocoagulopathy due to a lack of coagulation factors. However, the elevated MA and CI in our murine pancreatitis mice indicate a persistent hypercoagulable state despite this prolonged R time. This may be a similar phenomenon to disseminated intravascular coagulation (DIC) where extensive thrombosis causes a depletion of coagulation factors and subsequent bleeding from the inability to form clot appropriately. In fact, TEG analysis of patients shown to have DIC shows similarly prolonged R times as seen in our severe arginine induced pancreatitis model ^{49, 50}.

Importantly, mice in both murine models of acute pancreatitis are under pathophysiologic stress, partially immobilized by illness and have decreased oral intake with some element of dehydration. Within the limitations of our current study, it is challenging to causally demonstrate that the biology of acute pancreatitis is responsible for the hypercoagulable changes demonstrated apart from these clinical sequelae of the disease. Indeed, several sterile inflammatory models generate hypercoagulability and this phenomenon is likely not unique to pancreatitis alone . Future, mechanistic studies may be able to isolate factors specific to the pathophysiology of acute pancreatitis and their respective contributions to this hypercoaguble state.

The clinical implications of this quantifiable hypercoagulable state are not clear at this time, particularly given the strong limitations in sample size of this corroborative data. However, given the potential for clinical application, these findings warrant future investigation. Management or prophylaxis for thrombosis during episodes of acute pancreatitits in the clinical setting are potentially complicated by the prevalence of bleeding complications and conversion to hemorrhagic pancreatitits. Inflammation and exposure to pancreatic fluid can result in erosion of the arteries surrounding the pancreas, which can result in life-threatening hemorrhage from a pseudoaneurysm. There is currently no precise clinical tools utilized to know which patients are at increased bleeding risk. There are a limited number of series demonstrating the safety of thromboprophylaxis during acute pancreatitis , but patients frequently fail to receive adequate prophylaxis, in part due to concern over bleeding risk ^{46, 47}. As a result, risk stratification to support a personalized approach to thromboprophylaxis regimens is an important goal ⁴⁸. Predictive nomograms have been developed to assess risk of DVT ⁴⁹ and splanchnic vein thrombosis ⁵⁰ and provide useful data towards this goal. Incorporating a point of care test such as a TEG could provide quantifiable patient-specific data that could be useful in weighing thromboprophylaxis or sugggest the need for VTE screening with ultrasound or CT. Further research is necessary to catalogue the thrombotic implications of the hypercoagulable state measured by TEG. Although our study demonstrated a large proportion of AP patients show signs of hypercoagulability, we were underpowered to correlate this information with incidence of venous thromboembolism in our patients. Larger, multi-institutional studies should be performed to further investigate this interesting finding and to assess the risk/benefit ratio of anti-coagulation during the treatment of acute pancreatitis.

In conclusion, we demonstrate a transient hypercoagulable state in murine acute pancreatitis. TEG is a novel tool to quantify and characterize this hypercoagulable state that warrants further prospective evaluation in patients with pancreatitis. This propensity to clot is further validated by increased in-vivo clot formation in a murine deep vein thrombosis model. Further study into the mechanism(s) of this hypercoagulable state is warranted and could potentially provide novel therapeutic targets to better treat this unique subset of patients.

Supplementary Material

1

Supplemental Figure I: TEG was performed on inpatients admitted with acute pancreatitis. A) The proportion of patients with acute pancreatitis with MA values above the normal range (n=21). B) The proportion of patients with acute pancreatitis with Coagulation Index (CI) values above the normal range. C) Representative TEG curve from patient with acute pancreatitis (green) versus control (white).