SCCM Resource Library-A Statewide Analysis on the Origins of de Novo Pulmonary Embolism: Does the Liver Play a Role?

Video Transcription

Hello, my name is Ryan Cohen. I'm a fourth year surgery resident at Einstein Medical Center in Philadelphia.  I was born and raised in Los Angeles, California and attended medical school at Boston University. I am pursuing a fellowship in minimally invasive surgery.  I will be presenting this work on behalf of the surgery department at Einstein Medical Center. A statewide analysis on the origins of de novo pulmonary embolism.  Does the liver play a role? I have nothing to disclose.  Our traditional understanding of DVT and PE stems from the ideas of Virchow and his famous triad.  Namely, that stasis, endothelial injury, and hypercoagulability lay at the heart of DVT formation.  Virchow postulated that PE forms as a result of these thrombi in the periphery propagating and lodging in the lungs.  Moreover, this theory has provided the framework for the modern management of PE and DVT. Assuming Virchow's theory is true, we would expect a DVT present with every case of PE.  Paradoxically, studies over the last 30 years have found no such pattern. Rather, there is widespread discordance between the incidence of PE and DVT, which has been  replicated by multiple prior studies.  Belmahos and Alam were the first team to try to explain this discordance. In their review, only 15% of PE patients had an identifiable DVT.  They proposed that pulmonary embolism may not originate from the peripheral veins as commonly believed, but instead may occur de novo in the lungs, coining the concept of  de novo PE. They argued that the systemic effects of trauma-related adrenergia and inflammation could lead to  pulmonary endothelial damage, creating localized thrombosis in the lungs. Knudsen et al. expanded on this work in their review of 3,738 post-traumatic PE patients  creating a proposed pathophysiologic model for PE formation. Based on the observation that de novo PE was more common following chest injury, they thought  that endothelial damage in the lungs could lead to localized thrombosis and de novo PE formation.  Furthermore, they suspected this process was spurred by the systemic effects of shock and coagulopathy endemic to trauma patients.  In contrast, their view of DVT formation was more of a product of stasis and venous injury.  Since these landmark papers, multiple authors have tried to explain the phenomenon of de novo pulmonary embolism, identifying several notable risk factors for de novo PE formation,  including young age, trauma, especially localized trauma, and pulmonary inflammatory conditions, such as COPD.  Considering the liver is the largest repository of clotting factors and inflammatory mediators  in the body, the fact that many forms of liver dysfunction often lead to pulmonary dysfunction,  and the direct anatomic relationship between the liver and lung vasculature, we hypothesized that liver injury could play a significant role in the pathophysiology of de novo PE  formation. To test our hypothesis, we collected all blunt and penetrating traumas in the Pennsylvania Trauma Outcomes Study Database.  The PTOS database contains data submitted by accredited Pennsylvania trauma centers throughout the state. Both blunt and penetrating traumas were included.  The minimum age for the database was 18 years old. We then compared all non-liver injury patients to liver injury patients.  Our primary outcome was the incidence of de novo PE, which was defined as PE in the absence of DVT. PE and DVT were defined based on the PTOS and National Trauma Data Standard.  The diagnosis of DVT required documentation in the patient's medical record, which may have been confirmed by venogram, ultrasound, or CT scan.  The patient must have been treated with anticoagulation, vena cava filter placement, or vena cava clipping.  The diagnosis of PE required a high probability VQ scan result, positive pulmonary arteriogram, positive CT angiogram, or a diagnosis of PE documented in the medical record.  Notably, sub-segmental PE was excluded. We performed several additional comparisons as well. We compared lung injury patients to non-lung injury patients and de novo PE patients to  non-de novo PE patients. T-testing and chi-squared were used to compare continuous and categorical variables between each study group.  Statistical analyses were two-sided, and P less than 0.05 was considered statistically significant.  Finally, we performed a regression analysis to find independent predictors of de novo PE formation. Factors with a P value less than 0.2 were included in the regression analysis.  On our initial analysis, both lung injury and liver injury resulted in a significantly higher rate of DVT, PE, and de novo PE formation.  On regression analysis, injury severity and liver injury were the strongest predictors of de novo PE formation, with odds ratios of 3.27 and 1.67 respectively.  Other predictors included lung injury and male gender.  We made several conclusions based on our findings. De novo PE is a unique phenomenon from DVT and PE, owing to unique predictive factors.  Furthermore, our data showed that injury severity and liver injury were the strongest predictors  of de novo PE formation, suggesting liver trauma may be a central orchestrator of the de novo PE process.  Liver injury and lung injury are independently predictive factors for de novo PE formation. Moreover, male gender is another notable risk factor.  Based on our findings, we proposed several theoretical ideas for the pathophysiology of de novo PE formation that take into account local, regional, and systemic factors.  From a local perspective, it's possible that localized trauma and pulmonary inflammatory conditions lead to lung endothelial damage that spur the phenomenon.  Lung injury did have a higher rate of de novo PE in our study and was independently predictive of de novo PE on regression analysis.  However, as noted, injury severity and liver injury were stronger predictors overall. We suggest, as Knudsen et al. did in their original work, that chest injury may require  a backdrop of inflammation and coagulopathy that liver injury and polytrauma may provide. From a regional perspective, the anatomic relationship between the liver and the lung,  namely that they are directly connected via the hepatic vein and IVC, may explain the liver's involvement.  It's possible that trauma to the liver causes localized endothelial damage and clot formation that propagates through the relatively unimpeded route to the lungs.  Finally, from a systemic perspective, there may be inflammatory factors specific to liver trauma that lead to thrombus in the lungs.  The medical literature has described multiple associations between liver injury and pulmonary inflammation, suggesting a unique link between the organ systems.  Chronic liver diseases have been implicated in wide-ranging inflammatory effects in the lungs, including COPD, pneumonia, asthma, emphysema, and hepatopulmonary syndrome.  Animal models inducing liver injury have shown increased blood concentrations of lung-specific inflammatory and oxidative mediators.  Some of these mediators are presented on the slide. Independent of the liver, shock on its own may lead to endothelial damage in the lungs as well.  Animal models have shown associations between polytrauma and reperfusion lung injury. Human studies have demonstrated an association between shock and several known lung-specific  inflammatory mediators. We should also note that liver injury is notorious for causing profuse bleeding and shock on its own.  Thus, liver trauma may indirectly contribute to these shock-related inflammatory reactions that target the lungs.  The association of male gender with de novo PE formation may point to different biochemical and hormonal coagulability profiles of male trauma patients related to women.  However, we should note that male gender may also be a confounding factor, as young men are much more likely to suffer from severe polytrauma.  These local, regional, and systemic theories are not mutually exclusive. Perhaps the most elegant model considers all of these factors.  This is a proposed model created by our team for de novo PE formation. As mentioned, local factors like lung injury and pulmonary inflammatory conditions may  lead to endothelial damage. External thromboembolism from liver trauma could also explain the phenomenon.  Finally, systemic inflammatory and oxidative as well as coagulopathic effects of shock  generally and the liver specifically may uniquely target the lung, leading to endothelial damage and de novo PE formation.  This is the first study to propose liver injury as a central agent in de novo pulmonary embolism formation.  We propose several novel theories about the origins of de novo pulmonary embolism that involve local, regional, and systemic factors.  Moreover, we introduce a unified model that places the liver as a central orchestrator of the process.  There is certainly a significant need to perform experimental prospective studies, as well as 3D imaging and autopsy-based studies to test these hypotheses.  In future studies, we plan to apply our methodology to a larger national database. From a theoretical perspective, our findings add to the physiologic model of localized  inflammation that has come to explain the phenomenon of de novo pulmonary embolism. From a clinical standpoint, identifying liver trauma as a causative factor in PE formation  may assist in screening trauma patients who may be high risk for de novo PE.  Perhaps more importantly, by understanding the pathophysiology of the phenomenon of de novo PE, we can more accurately formulate prophylactic measures and effective treatment plans.  We think better understanding of the pathophysiology of de novo PE can help us diagnose the phenomenon and tailor specific treatment strategies.  Namely, by identifying specific inflammatory mediators in liver to lung and shock-related  inflammatory pathways, we may be able to better prevent and better manage de novo PE formation.  Thank you very much.

Video Summary

The speaker, Ryan Cohen, presents a study on the origins of de novo pulmonary embolism (PE) and its connection to liver injury. Traditional understanding of PE formation suggests that thrombi in peripheral veins can lead to PE. However, studies have shown a widespread discordance between the incidence of PE and deep vein thrombosis (DVT). Based on this, the speaker investigates the concept of de novo PE, where PE may originate in the lungs instead of peripheral veins. The study examines data from trauma patients and finds that liver injury and injury severity are strong predictors of de novo PE formation. The speaker proposes theoretical models that involve local, regional, and systemic factors in the pathophysiology of de novo PE. The findings have implications for screening, prevention, and treatment of de novo PE.

Asset Subtitle

Pulmonary, Research, 2022

Asset Caption

INTRODUCTION: Over the last decade the concept of de novo pulmonary embolism (DNPE), or pulmonary embolism (PE) independent of a deep vein thrombosis (DVT), has been realized as a unique phenomenon, especially in trauma patients. Although the pathophysiology of DNPE remains unclear, most surmise that it stems from local inflammatory changes in the lungs. Considering the anatomic relationship between the liver and the pulmonary vasculature, we hypothesize that liver injury may play an important role in DNPE formation. The purpose of our study was to compare the relationship between hepatic and lung injury to DNPE. METHODS: A retrospective chart review from January 2010 to December 2019 identified all blunt and penetrating trauma patients in the Pennsylvania Trauma Outcome Study (PTOS) database. Patients were divided and compared based on whether or not they suffered a liver or lung injury. Our primary outcome was the incidence of DNPE, defined as the incidence of PE without DVT. DVT and PE were diagnosed based on requirements set by PTOS and the National Trauma Data Standard. In a separate analysis, multivariate regression was used to identify independent risk factors for DNPE formation. RESULTS: In total, we identified 381,676 patients. Compared to non-liver injury patients, liver-injury patients had a higher incidence of DNPE (1.3% vs. 0.3%, P < .05). Compared to non-lung injury patients, lung-injury patients had a higher incidence of DNPE (1.0% vs 0.3%, P < .05). Compared to non-DNPE patients, DNPE patients had a higher incidence of liver injury (8.9% vs. 2.4%, P < .05) and lung injury (18.5% vs. 6.4%, P < .05). On multivariate analysis, Injury Severity Score > 25 (OR 3.3), liver injury (OR 1.7), lung injury (OR 1.6), and male gender (OR 1.2), were all independent predictors of DNPE. CONCLUSIONS: Our data suggests that both liver injury and lung injury are independently associated with DNPE formation. We propose a novel theory, expanding on the lung-centered hypothesis for the pathophysiology of DNPE. Considering the direct anatomic relationship of the liver to the pulmonary vasculature, the local hypercoagulable, inflammatory, and thromboembolic effects of liver trauma may also promote DNPE formation. Our findings could have widespread implications on the diagnosis, prevention, and management of PE.

Meta Tag

Content Type Presentation

Knowledge Area Pulmonary

Knowledge Area Research

Knowledge Level Advanced

Membership Level Select

Tag Pulmonary Embolism

Tag Outcomes Research

Year 2022

Keywords

de novo pulmonary embolism

liver injury

thrombi

trauma patients

pathophysiology

	Society of Critical Care Medicine 500 Midway Drive Mount Prospect, IL 60056 USA Phone: +1 847 827-6888 Fax: +1 847 439-7226 Email: support@sccm.org	Contact Us About SCCM Newsroom Advertising & Sponsorship DONATE	MySCCM LearnICU Patients & Families Surviving Sepsis Campaign Critical Care Societies Collaborative	GET OUR NEWSLETTER
© Society of Critical Care Medicine. All rights reserved. \| Privacy Statement \| Terms & Conditions The Society of Critical Care Medicine, SCCM, and Critical Care Congress are registered trademarks of the Society of Critical Care Medicine.