COVID-19 and venous thromboembolism: Known and unknown for imaging decisions

Abstract

As we continue to fight against the current coronavirus disease-2019 (COVID-19) pandemic, healthcare professionals across the globe are trying to answer questions surrounding how to best help patients with the up-to-date available science while awaiting the development of new therapies and mass vaccination. Since early in the pandemic, studies indicated a heightened risk of venous thromboembolism (VTE) in COVID-19 infected patients. There have been differing expert opinions about how to assess pretest probability of VTE in this patient population. This has been partly due to the high prevalence of respiratory failure in this patient population and the use of D-dimer as a prognostic test which is also frequently elevated in patients with COVID-19 in absence of VTE. Some experts have argued for an approach similar to usual care with testing if clinical suspicion is high enough. Some have argued for more routine screening at different points of care. Others have even suggested empiric therapeutic anti-coagulation in moderate to severely ill COVID-19 patients. In the following article, we review and summarize the most current literature in hopes of assisting clinicians in decision making and guidance for when to be concerned for VTE in COVID-19 patients. We also discuss research gaps and share pathways currently being used within our institution.

Key Words: COVID-19, Venous thromboembolism, Computed tomography scan, Anticoagulation, Point of care ultrasound

Core Tip: As we continue to fight against the current coronavirus disease-2019 (COVID-19) pandemic, healthcare professionals across the globe are trying to answer questions surrounding how to best help patients with the up-to-date available science while awaiting the development of new therapies and mass vaccination. Since early in the pandemic, studies indicated a heightened risk of venous thromboembolism (VTE) in COVID-19 infected patients. There have been differing expert opinions about how to assess pretest probability of VTE in this patient population. This has been partly due to the high prevalence of respiratory failure in this patient population and the use of D-dimer as a prognostic test which is also frequently elevated in patients with COVID-19 in absence of VTE. Some experts have argued for an approach similar to usual care with testing if clinical suspicion is high enough. Some have argued for more routine screening at different points of care. Others have even suggested empiric therapeutic anticoagulation in moderate to severely ill COVID-19 patients. In the following paper we review and summarize the most current literature in hopes of assisting clinicians in decision making and guidance for when to be concerned for VTE in COVID-19 patients. We also discuss research gaps and share pathways currently being used within our institution.

INTRODUCTION

At the time this paper was written, globally there are over 100 million patients who have tested positive for coronavirus disease-2019 (COVID-19) infection, with around 2.1 million patients having lost their lives due to this disease[1]. COVID-19 is caused by the novel coronavirus named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Though COVID-19 infections have a tendency to involve multiple organ systems, the respiratory system is primarily affected resulting in inflammatory infiltrates, and in severe cases leading to hypoxemia and respiratory failure. High risk of venous thromboembolism (VTE) in COVID-19 patients was recognized early on in the pandemic, with one study suggesting enoxaparin prophylaxis was associated with lower mortality[2]. However, despite thromboprophylaxis, the risk for VTE remains high[3]. Timely identification of deep vein thrombosis (DVT) and pulmonary embolism (PE) is critical in making clinical decisions regarding therapeutic anticoagulation. Computed tomography pulmonary angiography (CTPA) is considered the gold standard test for diagnosis of pulmonary artery clot. In patients presenting with COVID-19 infection, deciding when to screen or rule out pulmonary artery thromboembolism remains a challenge for physicians due to frequently fluctuating oxygenation requirements. Different approaches have been suggested and debated by experts including use of clinical decision-making tools, the use of D-dimer testing, universal CTPA or lower extremity ultrasound screening on admission to the hospital or at the time of admission to critical care units, and empiric higher than prophylactic anticoagulation. In the following review, we will explore current literature regarding clinical decision-making for imaging in the diagnosis of VTE in COVID-19 patients in the form of common clinical questions. We will also share our institution’s pathway for diagnosing VTE in this patient population.

WHAT IS THE RISK OF VTE IN COVID-19?

Since the beginning of the pandemic, studies have indicated increased risk of both venous and arterial thromboembolism in COVID-19 patients, including DVT, PE, ischemic stroke, myocardial infarction and peripheral arterial thromboembolism[4]. One study, which compared national databases of viral pneumonias, showed COVID-19 was associated with higher incidences of thrombotic complications compared to other viral pneumonias[5]. The reported frequency of pulmonary embolism in critically ill COVID-19 patients is approximately 20%-30%[6,7]. Evidence suggests that small vessel pulmonary thrombi are more common than large pulmonary vessel involve-ment in COVID-19[8,9]. Pulmonary embolism is a serious thrombotic complication of COVID-19 pneumonia, with mortality rates for patients with COVID-19 and PE estimated at approximately 45%[10]. With available literature it is clear that risk of VTE is very high in patients with COVID-19, especially those requiring intensive care during hospitalization. Patients with COVID-19 and VTE have high risk of mortality.

WHAT IS THE MECHANISM BEHIND VTE IN COVID-19 PATIENTS?

The pathway for clot formation in acutely ill patients is Virchow’s Triad, which includes the predisposing factors of venous stasis, hypercoagulability and endothelial damage. All critically ill patients, despite underlying etiology, usually face a combination of the above factors and are therefore considered high risk for VTE. Post-mortem studies have raised significant concerns regarding microvascular thrombosis, as well as macrovascular involvement in COVID-19 patients[11,12]. Data suggests SARS-CoV-2 can infect pulmonary endothelial cells, triggering a cascade of local immune response involving leukocyte activation, complement deposition and platelet aggregation[13]. In a small study of 25 patients with COVID-19 who were admitted to the intensive care unit, screening bilateral lower extremity venous ultrasounds between days 5 and 10 of admission showed an overall incidence of proximal DVT of 24%, indicating lower extremity thrombosis is also a major contributor for pulmonary embolism in COVID-19 patients[14]. Apart from known factors that put critically ill patients at high risk of VTE, direct injury to the endothelium by the virus and strong local immune response seems to play a large role, especially in small pulmonary vessel in-situ thromboses’ in patients with COVID-19. Several studies also reported other hemostatic abnormalities in COVID-19 patients, including positive antiphospholipid antibodies, abnormal platelet function and abnormal coagulation parameters that likely add a complex interplay further increasing risk of thromboembolism[15].

WHAT IS THE ROLE OF CLINICAL PROBABILITY SCORES IN DIRECTING EVALUATION FOR VTE IN COVID-19 PATIENTS?

Clinical probability scores have been shown to assist in determining pretest probability of pulmonary embolism with more accuracy than clinician gestalt[16,17]. The Well’s criteria are the most popular and commonly used tool internationally used to aid in clinical decision making for diagnosis of VTE. Studies on the use of clinical probability scoring in COVID-19 patients is thus far very limited. One study indicated a Wells score > 2 had a higher correlation for VTE on imaging in critically ill COVID-19 patients[18]. Despite the limited evidence, use of a clinical prediction scoring tool should be considered in conjunction with clinical judgement when defining pretest probability of VTE in COVID-19 patients.

WHAT IS THE ROLE OF D-DIMER IN DECISION MAKING FOR EVALUATION OF VTE IN COVID-19 PATIENTS?

D-dimer is a soluble fibrin degradation product resulting from fibrinolysis of thrombi. It is frequently elevated in acute VTE, but is non-specific, being frequently elevated in many other non-thrombotic conditions including pregnancy, cancer and inflam-mation[19,20]. Cochrane review suggests D-dimer sensitivity ranging from 80%-100% and specificity from 23%-63% in prediction for VTE[21]. Due to lack of specificity and high false positive results, D-dimer is a good test to rule out VTE in low pretest probability patients if D-dimer results are normal, but should not be used to establish diagnosis of VTTE when levels are elevated[22].

In multiple studies D-dimer has shown to be frequently elevated in COVID-19 positive patients in the absence of VTE. Studies indicate up to 40%-50% of patients with COVID-19 will have elevated D-dimer during hospitalization[23,24]. In one study, admission D-dimer was found to be the same in those patients who were found to have VTE during hospitalization vs those without evidence of VTE[25]. Many retrospective studies suggest significantly higher D-dimer levels in patients with confirmed pulmonary embolism on CTPA vs patients without pulmonary embolism on CTPA[26,27]. Though data clearly indicates higher D-dimer values are associated with higher probability for pulmonary embolism on CTPA, there is ongoing debate about serial D-dimer testing and the cut-off value for D-dimer at which imaging to evaluate for VTE should be performed. Based on available studies, we think absolute D-dimer levels and changes over time should be taken into account in decision making on when to obtain imaging to evaluate for VTE despite the absence of significant clinical suspicion, but exact cut off values or percentage of change from initial D-dimer at which imaging should be performed remains controversial.

SHOULD ALL PATIENTS PRESENTING TO THE EMERGENCY DEPARTMENT WITH SUSPECTED COVID-19 INFECTION UNDERGO LOWER EXTREMITY VENOUS ULTRASOUND OR CTPA FOR EVALUATION OF VTE?

Currently, CTPA is considered the gold standard for diagnosis of pulmonary embolism and venous duplex ultrasound is considered standard for diagnosis of DVT. Pulmonary embolism is seen as a hypodense filling defect on CTPA (Figure 1). A normal CTPA effectively rules out pulmonary embolism with high negative predictive values of around 99%[28]. However, like any other testing modality, CTPA carries risks including exposure to ionizing radiation and use of intravenous iodinated contrast, placing the patient at risk for renal toxicity especially in patients with existing kidney disease and hypersensitivity reactions. Overtreatment of clinically insignificant pulmonary emboli also comes with significant risk given the need for therapeutic anticoagulation medications and resulting risk of bleeding[16].

Figure 1 Computed tomography angiography images. A and B: Chest axial views; C and D: Chest coronal views. Computed tomography angiography chest axial and coronal views soft tissue window shows large hypodense filling defects in distal right and left main pulmonary arteries and all segmental pulmonary branches in both lungs (orange arrows) in a 58-year male patient with shortness of breath. Findings are compatible with bilateral extensive pulmonary embolisms.

High resolution CT of lungs can identify ground glass opacities with significant accuracy. In the early part of pandemic there was concern regarding accuracy of reverse transcriptase- polymerase chain reaction (RT-PCR) which led to studies suggesting lung CT scan as a more sensitive modality for diagnosing COVID-19 pneumonia[29]. Though RT-PCR remains the gold standard for confirmation of COVID-19 diagnosis, it has been advised to consider CT chest as a primary modality for diagnosis, especially if RT-PCR availability is limited or there is a delay in testing results[30]. Many institutions use CT modalities over chest x-ray due to the poor sensitivity of chest x-ray in diagnosing COVID-19 pneumonia. Questions naturally arise if one should consider CTPA as a triage test due to the potential added value of evaluating for pulmonary embolism in addition to imaging the lung parenchyma, especially in patients presenting to the emergency department.

Studies evaluating the role of CTPA as a triage and universal evaluation strategy in emergency departments are limited. In a single center retrospective study in the United Kingdom, 48 patients with COVID-19 like symptoms, but without clinical concern for PE, were screened with non-contrast CT. All patients who had findings concerning for COVID-19 or RT-PCR confirmed COVID-19 underwent CTPA. Overall, there was only one positive CTPA (2%) for pulmonary embolism[31]. On the other side in one retrospective study, emergency department clinicians referred COVID-19 patients for CTPA based on clinical suspicion for PE with results showing detection of PE on CTPA in 18% of patients[32]. Of note, data from early in the pandemic in France suggested that PE’s in COVID-19 positive patients typically occurred around day 6 of infection (median)[33]. Currently there are no studies to our knowledge evaluating venous ultrasound as mandatory screening in emergency room in patients with COVID-19.

Though there are many studies indicating high prevalence of VTE in COVID-19 patients, most studies were performed involving patients in the intensive care setting. Data suggests that critically ill patients are at high risk for VTE despite primary cause of that illness[34,35].

Due to the risk associated with intravenous contrast exposure, the role of CTPA as universal testing for all emergency room or hospitalized patients with COVID-19 is not advisable. Similarly routine lower extremity ultrasound of all COVID-19 patients in emergency department does not have cost benefits and is also associated with risks, especially in regards to detecting small distal DVTs with unknown clinical significance.

A high index of clinical suspicion especially in patients with significant hypoxemia along with the use of clinical probability scores and D-dimer should be the driving factors in determining when to obtain CTPA or lower extremity ultrasound in COVID-19 patients.

SHOULD ALL PATIENTS ADMITTED TO THE HOSPITAL OR CRITICAL CARE UNITS WITH COVID-19 BE SCREENED FOR VTE?

Hospitalization due to medical illness is associated with increased risk of VTE. Critically ill patients have an even higher risk of VTE despite underlying diagnosis, with many critically ill patients unable to express their symptoms. Physical diagnosis can be challenging and usually not very high yield in diagnosis of DVT[36].

In Prophylaxis for Thromboembolism in Critical Care Trial study pre-COVID 3764 intensive care unit (ICU) patients were randomized to receive either prophylactic low molecular weight heparin, dalteparin or unfractionated heparin. Patients underwent mandatory twice weekly lower extremity ultrasound. The overall VTE rate in the study was 9.1% and DVT rate was 5.5%[37]. In a study published recently involving medical-surgical critically ill patients, twice weekly surveillance with lower extremity ultrasonography lead to 9.6% rate of DVT and was associated with higher detection of DVT compared to non-surveillance standard care group and a lower 90 d mortality (adjusted HR: 0.75, 95%CI: 0.57 to 0.98)[38].

To date, there are not many studies involving systematic screening ultrasound for detection of DVT in COVID-19 patients. In a study involving 26 critically ill patients with COVID-19, when surveillance ultrasonography was mandated, DVT rate was close to 50%. In this study, all patients were mechanically ventilated and about 90% of patients were on vasopressor therapy[39]. Based on limited available data, some institutes and expert groups recommend screening lower extremity ultrasound for patients with COVID-19 who need ICU level care[40,41]. Factors such as size of the hospital, as well as location and local treatment cultures can play a role in which patients are cared for in ICU settings. Smaller hospitals may treat patients on high oxygen or noninvasive mechanical ventilation in intensive care units. Flow management can lead to patients spending some time in intensive care beds due to lack of availability of beds on medical wards. These factors should all be considered when making surveillance imaging decisions.

SHOULD CTPA BE PERFORMED TO EVALUATE FOR PULMONARY THROMBOEMBOLISM IF OXYGEN REQUIREMENTS SEEM OUT OF PROPORTION TO LUNG INFILTRATES OR OXYGEN REQUIREMENTS ARE INCREASING IN COVID-19 PATIENTS WITH STABLE LUNG INFILTRATES?

This is one area where there is agreement amongst most experts and professional societies. The European Society of Radiology, European Society of Thoracic Imaging and European Society of Cardiology suggested that CTPA should be performed to evaluate for pulmonary embolism in COVID-19 patients with limited extent of disease on non-contrast imaging and significant supplemental oxygen needs[42,43].

SHOULD WE GIVE THERAPEUTIC ANTICOAGULATION TO ALL MODERATE TO SEVERELY ILL PATIENTS WITH COVID-19 WITHOUT IMAGING CONFIRMATION OF VTE?

Given the prevalence of VTE in COVID-19 patients, many physicians and professional societies have contemplated the role of empiric therapeutic anticoagulation for all hospitalized COVID-19 patients. Currently, randomized data is lacking to support the use of empiric therapeutic anticoagulation even amongst critically ill patients. One recent randomized controlled trial comparing therapeutic and prophylactic enoxaparin showed therapeutic enoxaparin improved gas exchange and need for mechanical ventilation in severe COVID-19 patients[44], Many institutions have created alternate guidelines supporting the use of “ intermediate” or full therapeutic anticoagulation[6]. Of course, the use of higher intensity anticoagulation comes with its own set of risks, with several small retrospective studies showing major bleeding events and even fatalities associated with its use[45]. Current guidelines recommend prophylactic dose anticoagulation for hospitalized adults with COVID-19[46]. In addition, guidelines recommend empiric treatment of suspected PE if imaging is expected to take > 4 h or for DVT if imaging is expected to take > 24 h[47]. Currently optimal dosing of intermediate anticoagulation with goal of pharmacoprophylaxis in COVID-19 patients remains unknown[48]. Randomized controlled trials are underway to answer these questions. Results of these trials will help clarify more precise use of anticoagulation strategy in near future[49]. At this point, at our institution, we do not recommend universal intermediate or therapeutic anticoagulation for all patients with COVID-19. We suggest universal pharmacologic prophylactic anticoagulation (if bleeding risk is acceptable) and maintaining a high index of clinical suspicion to help in early diagnosis of VTE events and escalation to appropriate therapeutic dosing when indicated.

WHAT IS THE ROLE OF POINT OF CARE ULTRASOUND IMAGING IN EVALUATION OF VTE IN COVID-19?

Point of care ultrasound (POCUS) in care of all patients is rapidly evolving. Currently, training in POCUS is variable across different medical institutions. Availability of good quality ultrasound machines for point of care use is an additional challenge. Evaluation of the lower extremity deep veins with POCUS for evaluation of deep venous thrombosis is reasonable if the provider has the skills for acquisition and interpretation of images. At our institute, bedside clinicians are not trained and do not use venous ultrasound for thrombosis evaluation. Although periodic screening for deep venous thromboses in medical patients was performed in previous VTE prophylaxis efficacy trials, it has not been studied as an intervention, and, therefore, cannot be recommended[49-55]. For providers with this set of skills, however, lower extremity venous POCUS can be considered in critically ill patients with COVID-19. For patients with moderate to severe COVID-19 with hemodynamic worsening or sudden instability, POCUS use is recommended for rapid evaluation of cor pulmonale[41].

OUR APPROACH

Due to the coronavirus pandemic, a disaster preparedness group of health system experts came together to form the COVID Clinical Content Group. A new webpage on COVID care was created on the health systems website to assist providers with current evidence and local expert guidance. Evidence on risk of thrombosis and management is routinely evaluated panel of experts which include hospitalists, intensivist, vascular medicine specialist, hematologist and anticoagulation pharmacist. Consensus recommendations are posted on this webpage, and periodic educational webinars are hosted. Our current algorithm is described below (Figure 2).

Figure 2 Diagnosis and management of venous thromboembolism in hospitalized patients with coronavirus disease-2019. COVID-19: Coronavirus disease-2019; DVT: Deep vein thrombosis; PE: Pulmonary embolism; VTE: Venous thromboembolism.

CONCLUSION

VTE remains a concerning complication in patients with COVID-19 infections. Currently, there remain many unanswered questions related to imaging and anticoagulation strategies. Maintaining a high index of suspicion and use of imaging for early diagnosis of VTE without universal screening appears to be the most logical method in managing this issue until further research can be completed and validated.