Pulmonary Embolism in Patients With COVID-19

We report a case series of patients with coronavirus disease 2019 (COVID-19) with pulmonary embolism (PE) in our institution. Lille University Hospital is the tertiary care center for the North of France, the second greatest French region in population density (189 people per 1 km²), also considered a metabolic area with high number of overweight patients. The study was approved by the institutional data protection authority of Lille University Hospital.

Among the 107 first consecutive patients with confirmed COVID-19 admitted to the intensive care unit (ICU) for pneumonia from February 27 to March 31, we noticed an unexpectedly high number of PEs during their stay in the ICU: 22 (20.6%) at the time of analysis (April 9), within a median time from ICU admission of 6 days (range, 1–18 days). To determine whether this represents an increase in the expected incidence of PE over a similar time interval, we analyzed the files of 196 patients hospitalized in our ICU during the same time interval in 2019. Despite a similar severity score on admittance to the ICU, the frequency of PE in our COVID-19 series was twice as high as the frequency we found in this control period (20.6% versus 6.1%; absolute increased risk, 14.4% [95% CI, 6.1–22.8]). It was also twice as high as the 7.5% frequency of PE in the 40 patients with influenza admitted to the ICU between January 1 and December 30, 2019 (3 PEs; absolute increased risk, 13.1% [95% CI, 1.9–24.3]). A qualitative description of the main characteristics of the patients with PE in the different periods is given in the Table.

Taking into account the ICU duration at time of analysis, we estimated the cumulative incidence of PE using the Kalbfleisch and Prentice method by taking into account death (n=15) and discharged alive (n=48) as competing events. The 22 patients still hospitalized in the ICU without PE at the time of analysis (median ICU length of stay, 15 days; range, 10–30 days) were treated as censored observations. At day 15 of ICU admission, the cumulative incidence of PE in patients with COVID-19 in the ICU was estimated to be 20.4% (95% CI, 13.1–28.7). In terms of the main data at ICU admission (using the univariable Fine and Gray model to estimate subhazard ratios of PE), D-dimers (estimate subhazard ratio per log-SD increase, 1.81 [95% CI, 1.03–3.16]), plasma factor VIII activity (estimate subhazard ratio per log-SD increase, 1.73 [95% CI, 1.10–2.72]), and von Willebrand factor antigen (estimate subhazard ratio per log-SD increase, 1.69 [95% CI, 1.12–2.56]) values seem to be associated with a greater PE risk.

At the time of PE diagnosis, 20 of 22 patients were receiving prophylactic antithrombotic treatment (unfractionated heparin or low-molecular-weight heparin) according to the current guidelines in critically ill patients.^1,2 One patient with a history of deep venous thrombosis was receiving fluindione with an international normalized ratio in the therapeutic range, and 1 patient was receiving therapeutic unfractionated heparin because of atrial fibrillation. The criteria for deciding to perform computed tomography pulmonary angiography (CTPA) were suspicion of PE on admission and acute degradation of hemodynamic or respiratory status. All CTPAs were performed with multibar computed tomography with no difference in the injection protocol regardless of whether the CTPA was performed for PE diagnosis. The number of CTPAs was higher in patients with COVID-19 than in patients hospitalized in the ICU during the same time period in 2019. This historical control group reflects the global practice in our ICU. Because only 34% of patients from this group have respiratory failure requiring CTPA (Table), a potential bias of an increased detection of PE in patients with COVID-19 could have been generated. That is why we compared patients with COVID-19 and patients with influenza admitted to the ICU for respiratory failure in 2019. Even if the number of CTPAs performed in patients with influenza was higher than in patients with COVID-19, fewer PEs were identified, reinforcing the increased risk of PE in patients with COVID-19. The low number of associated deep venous thromboses in patients with COVID-19 may suggest that they have pulmonary thrombosis rather than embolism.

PE frequency has not yet been reported in the different series of patients with COVID-19. All our patients received thromboprophylaxis according to the current recommendations for critically ill medical patients. However, we suspect that the high obesity prevalence in our patient group contributes to the increased PE frequency.³ Because of the lack of specific studies in this population, the recommendations do not mention adaptation of the prophylaxis regimen in overweight patients or a need to monitor anti–factor Xa concentrations. Heparin could have benefic effects in COVID-19 through different mechanisms. However, the effective dose and monitoring are discussed, in particular in very high-risk patients and those with high body mass index.⁴ Indeed, during the H1N1 flu pandemic, some centers reported an increased thrombotic risk in patients with severe acute respiratory distress syndrome and suggested the use of higher doses of heparin.⁵

In conclusion, there is an urgent need for replication on a much larger scale of our data on PE frequency in COVID-19 infection in patients in the ICU. Failure to identify and accurately manage this risk could worsen the prognosis of patients with COVID-19.

Acknowledgments

The authors wish to thank the Lille COVID Research Network (LICORNE) and the i-site for their support during the COVID-19 pandemic.

Footnotes