Pulmonary Embolism in COVID-19 Patients: New Numbers, Little Consensus

New numbers showing pulmonary embolism (PE) was diagnosed in roughly one in five critically ill COVID-19 patients admitted to a single ICU highlight yet another hot spot in this pandemic. Is PE a distinguishing “signature” of this disease and, if so, how best can it be prevented and treated?

Hematologist Sophie Susen, MD, PhD (Institut Coeur-Poumon/CHU Lille, France), senior author of the report published April 27, 2020, in Circulation, said the first COVID-19 patient in France died from PE but that at that point she was unaware how much of her time would be spent dealing with COVID-19 patients. At her large, academic hospital, the first COVID-19 patient was admitted by the end of February. By mid-March, Susen, along with lead author Julien Poissy, MD, PhD (Hôpital Roger Salengro/CHU Lille), were aware that some patients were deteriorating rapidly with PE or localized thrombosis.  

“At the end of March, I wrote the first draft of the letter and we had 10 patients,” Susen told TCTMD. “Not all of them could have [computed tomography pulmonary angiography (CTPA)] so in some cases it was a suspicion of PE, but we were convinced that it was unusual.”

The final submitted paper ultimately included 107 consecutive patients with confirmed COVID-19 admitted to their ICU for pneumonia over a month-long period ending March 31, 2020. Of these, 20.6% had a CTPA-confirmed diagnosis of PE a median of 6 days after admission. At the time of diagnosis, 20 of the 22 patients diagnosed with PE were receiving prophylactic antithrombotic treatment with either unfractionated heparin or low-molecular-weight heparin.

In contrast, just 6.1% of patients hospitalized to the ICU during the same time last year developed PE during the admission, report investigators. The frequency of PE among COVID-19 patients in the ICU was also nearly three times higher than the frequency observed in patients admitted to the ICU for influenza in 2019.   

Susen said the decision to perform CTPA was based on a suspicion of PE upon admission and/or acute degradation of hemodynamic or respiratory status. She noted that some patients with COVID-19 may be too sick to go for CTPA and as such the frequency of PE may be even higher than 20%.

Victor Tapson, MD (Cedars-Sinai Medical Center, Los Angeles, CA), who wasn’t part of the study but is involved in the PE response team at his hospital, said he believes thrombotic complications could well be a hallmark feature of COVID-19. He pointed out to TCTMD that this is still controversial, with some physicians believing that the frequency of venous and arterial thrombotic complications might only be observed because of the increase in volume of very sick patients.

“My sense is that [thrombosis] could be a signature of this virus,” said Tapson. “It was seen in SARS, in MERS, and in H1N1, but those overall numbers were lower so maybe it wasn’t as noticeable.”

Behnood Bikdeli, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), who recently led a global collaboration highlighting the risks of thrombotic disease in COVID-19 patients, said the new French series provides yet more information about venous thromboembolism (VTE) in this population.

“This study raises the concern that the risks of pulmonary embolism COVID-19 might be above and beyond acute medical illness, that there might be something inherent to the disease contributing to thrombosis,” he told TCTMD. “To that extent, I think it’s a very important study, but we shouldn’t jump to conclusions. I think the authors were reasonably cautious in their interpretation.”

This study raises the concern that the risks of pulmonary embolism COVID-19 might be above and beyond acute medical illness, that there might be something inherent to the disease contributing to thrombosis. Behnood Bikdeli

What is not yet known is the true prevalence of VTE in COVID-19 patients, he said, noting that the few studies to date have shown a frequency ranging from 7% to 30%, with that range largely explained by variable testing, the severity of patient illness, and comorbidity profiles. “What we don’t know is the exact point estimate and what we need to do to reduce the risk,” said Bikdeli. “That’s still uncertain.” Larger studies from more hospitals will be needed to identify the true prevalence of VTE among ICU-admitted patients with COVID-19, and also among non-ICU patients, he added. “Is this study informative? Absolutely,” said Bikdeli. “Is it conclusive? No.”   

To TCTMD, Mahesh Madhavan, MD (NewYork-Presbyterian/Columbia University Irving Medical Center), who partnered with Bikdeli in their recent review, said the rate of PE appears comparable to what has been seen in COVID-19 patients requiring ICU level of care in other reports, despite the use of pharmacologic thromboprophylaxis in the majority of patients who developed PE. “The data [also] suggest that mechanisms contributing to thrombotic risk may be different between COVID-19 and other viral illnesses such as influenza,” he said.

Obesity May Be a Factor

In the French series, the median body mass index (BMI) of the COVID-19-positive patients admitted to the ICU was 30 kg/m², which is on the low end of the obese range. In general there is a growing appreciation that there may be a link between obesity and COVID-19 disease severity. In the current study, researchers hypothesize that obesity may be a contributing risk factor to the higher than expected frequency of PE in their patient cohort. Moreover, they note, a previous study showed there was a higher rate of obesity among individuals with SARS-CoV-2 who required mechanical ventilation.

At their center, all patients received thromboprophylaxis based on critical care guidelines from the American College of Chest Physicians and American Society of Hematology, but Susen said they initially did not adjust treatment based on BMI. At the time, there were no reports from China, the country with the largest experience dealing with COVID-19, about the frequency or risk of PE in this population. However, given the high rate of PE at their center and the emerging data about the thrombotic complications associated with COVID-19, they now administer a weight-adjusted dose of antithrombotic therapy and give therapeutic anticoagulation in patients at high risk for VTE, such as those with obesity, acute respiratory distress syndrome (ARDS), and evidence of coagulation abnormalities.  

“We always increase the dose in obese patients,” said Susen. “In our ARDS patients, and in patients in which a diagnosis [of PE] is more than expected, we give a therapeutic dose of heparin. We don’t wait for the CTPA because in a short time frame, the patient can become more severe and you can’t always confirm with CTPA.”

Bikdeli said he isn’t entirely convinced of obesity’s causative role in COVID-19 severity, noting that in the two control groups—the influenza patients admitted to ICU in 2019 and the patients admitted to the ICU during the same time period last year—the BMI ranges overlapped with that of the COVID-19 patients. Weight-based prophylactic anticoagulation is still a debated issue, even outside of the COVID-19 setting.

The critical care guidelines raised the issue of obesity as an open question requiring further research. While there isn’t enough hard evidence to support dose adjustment in obesity for PE prophylaxis, many hospitals do adopt dose-adjusted protocols, mainly based on observations linking drug concentrations and surrogate markers, said Bikdeli.  

What Dose of Anticoagulation Is Best?

A recent review published in Blood by Jean Marie Connors, MD (Brigham and Women’s Hospital, Boston, MA), and Jerrold Levy, MD (Duke University School of Medicine, Durham, NC), suggest physicians should provide therapeutic anticoagulation in COVID-19 patients with confirmed VTE or presumed PE. The reviewers also recommend an escalated dose of VTE prophylaxis in COVID-19 patients admitted to the ICU and those with ARDS. In the international review drafted by Bikdeli and Madhavan, the group highlighted the lack of randomized evidence to support therapeutic anticoagulation in COVID-19, but noted that some members of the collaboration do use it in their practice.

Tapson acknowledged the absence of evidence supporting therapeutic anticoagulation in COVID-19 patients at high risk for VTE and said while there must be concerns about potential bleeding, the pendulum has swung a bit over recent weeks where many physicians are now concerned with thrombotic complications. He also emphasized the need for more randomized, controlled evidence before there can be widespread recommendations for more-aggressive prophylactic anticoagulation. Morbidly obese patients with COVID-19 and perhaps active cancer patients could be exceptions where an increase in the usual prophylactic dose might be considered with careful risk assessment. While he believes there is a prothrombotic component to COVID-19, he is reluctant to act on it in all situations.

We’re going to figure this out, though, that’s for sure. Unfortunately, there seem to be enough patients [with COVID-19] to get the studies done. Victor Tapson

“We need more data, and there are ongoing studies looking at therapeutic versus prophylactic anticoagulation in these patients, or higher-dose prophylactic anticoagulation, which we really need,” he said. “We can’t keep relying on a few patients. . . . There is so much to learn, but honestly there is no real consensus for how to handle this.”

If a critically ill COVID-19 patient comes in with a very, very high D-dimer level and markedly elevated inflammatory markers, some colleagues will consider higher anticoagulation doses, particularly if the D-dimer levels continue to trend higher, he said.

“Some of my New York colleagues feel this way,” continued Tapson. “Some who have seen substantial numbers of COVID-19 patients feel there is an increased risk of clotting over and above what they would otherwise be seeing. I think this is logical, but we don’t have randomized controlled data to support it. I like the approach that some are taking by considering higher doses in very carefully selected cases with scrutiny of the bleeding risk, but not uniformly increasing doses. But some colleagues are against any changes without strong clinical trial data.”

And despite the need for concrete evidence, Tapson said he understands that “when you’re a clinician taking care of patients, you have to think through this stuff, and sometimes you have to make decisions when you don’t have a strong evidence base. You have to use your insight and clinical experience. We’re going to figure this out, though, that’s for sure. Unfortunately, there seem to be enough patients [with COVID-19] to get the studies done.”

To TCTMD, Bikdeli said that his experiences to date working in ICUs have shown a fair amount of variability in anticoagulation dosing among prescribing physicians.

“Some clinicians have a lower threshold for high-dose blood thinners, some are restrictive, and some land in the middle,” he said. “Nobody knows what the correct answer is right now. With our international collaboration, we provided consensus guidance, but that was a consensus in the absence of high-quality data. That’s why I think there is such great and urgent need for randomized trials.” At present, there are at least six registered randomized trials comparing different anticoagulation strategies for patients with COVID-19, he added.

In their series, Susen also noted they observed a low rate of deep vein thrombosis in COVID-19 patients, suggesting a localized thrombosis rather than PE. However, their findings need to be confirmed because a systematic check for deep vein thrombosis is difficult to perform in these patients and wasn’t always possible when the hospital was hit with a wave of COVID-19 cases, she said.