Increased Blood Viscosity Tied to Mortality in Hospitalized COVID-19 Patients

More-viscous blood, which increases the propensity for clotting, is a sign that patients hospitalized for COVID-19 are more likely to do worse over the short term, according to data from a New York City health system.

The risk of in-hospital all-cause mortality increased along with rising estimates of blood viscosity, with patients with the highest estimates of high-shear and low-shear blood viscosity having 53% and 36% greater relative risks, respectively, compared with their counterparts with the least-viscous blood.

“If you have elevated blood viscosity, high shear or low shear, those individuals are in an accelerated pathway for death,” senior author Robert Rosenson, MD (Icahn School of Medicine at Mount Sinai, New York, NY), told TCTMD, noting that the relationship held true even after adjustment for many of the biomarkers that have been used to stratify risk in the setting of COVID-19.

“Those individuals should be appropriately hydrated to decrease the plasma viscosity, they should receive high-intensity glucocorticoids to decrease the inflammatory response, and they should be monitored much more closely because of the prognostic significance of an elevated blood viscosity,” he advised.

Ongoing studies are exploring whether additional interventions can improve the prognosis of patients with COVID-19 and high blood viscosity. In the meantime, Rosenson said that based on the results of a multiplatform trial published last year—from the REMAP-CAP, ACTIV-4a, and ATTACC investigators—therapeutic-dose heparin, which provided a benefit in noncritically ill patients, can be recommended.

The findings of the current study were published online today ahead of the July 26, 2022, issue of the Journal of the American College of Cardiology, with lead authors Daein Choi, MD, and Ori Waksman, MD (both Icahn School of Medicine at Mount Sinai).

Integrating Inflammatory and Thrombotic Markers

COVID-19 is characterized by the presence of diffuse microthrombi, which can lead to organ dysfunction, long-term disability, and death, but what the infection is doing to increase clotting is not clear, Rosenson explained. Prior research has examined markers like D-dimer and fibrinogen to try to understand which patients are most at risk for thrombotic events, he noted.

“There’s been a lot of disappointment regarding these biomarkers, and that’s because the inflammatory process also results in upregulation of the genes and the pathways that manufacture these acute-phase proteins like fibrinogen,” he said.

Blood viscosity is an integrative measure of all of these influences, and a prior study showed that directly assessed viscosity (using centipoise as the unit of measure) was higher in patients hospitalized for COVID-19 than in people without the infection and remained so for 2 months after discharge, Rosenson said. Direct measurement of viscosity can be challenging, however, and a method for estimating it—the Walburn-Schneck model, which incorporates routinely measured lab values like blood count, hematocrit, and globulin levels—has been developed.

Estimates using that method are what was done in the current study, which included 5,621 patients hospitalized for COVID-19 within the Mount Sinai Health System between February 27, 2020, and November 27, 2021.

This is definitely the first time that anyone has shown a prognostic role of estimated blood viscosity in COVID patients. Cheryl Maier

The researchers differentiated between high-shear and low-shear blood viscosity: the former applies to high-flow situations in medium and large arteries and has been associated with endothelial damage, whereas the latter applies to low-flow situations and is associated with the propensity for clotting. Every one-centipoise increase in high-shear and low-shear estimated blood viscosity was associated with a 36.0% and 7.0% increase, respectively, in in-hospital mortality (P < 0.001 for both).

Compared with patients in the lowest quartile of high-shear estimated blood viscosity, those in the highest had an elevated in-hospital mortality risk after adjustment for confounders (adjusted HR 1.53; 95% CI 1.27-1.84). The relationship tended to be stronger in Hispanics, patients with diabetes, and those without any comorbidities.

Similarly, patients with the highest estimated values for low-shear blood viscosity had a greater risk of in-hospital mortality compared with those with the lowest values (adjusted HR 1.36; 95% CI 1.14-1.64).

Other factors associated with a greater risk of mortality were levels of C-reactive protein and interleukin-6, although the relationship between high-shear blood viscosity and death remained significant even after accounting for these inflammatory markers.

“As new emerging antiviral agents suggest benefits in patients at high risk of progressing to severe illness, identifying high-risk populations in the earlier stage of the disease becomes crucial,” the investigators write. “From a translational perspective, the variables to calculate estimated blood viscosity (hematocrit, albumin, and total protein) are readily available to practitioners and are easily obtained from most admission labs, suggesting a possible use of estimated blood viscosity as an efficient and simple risk assessment of patients with COVID-19 to offer proper preventive therapy.”

Moreover, they write, “further studies investigating the impact of targeted reduction of whole blood viscosity are merited given the association between estimated blood viscosity and mortality.”

Potentially Major Implications for Patient Management

Commenting for TCTMD, Cheryl Maier, MD, PhD (Emory University School of Medicine, Atlanta, GA), said “this is definitely the first time that anyone has shown a prognostic role of estimated blood viscosity in COVID patients. And it’s also really helpful to the field that they were able to do this from a calculated measurement, because one of the major limitations in doing [a direct] assessment of viscosity is that most healthcare settings just don’t have the ability to do it.”

A key question is around the mechanism that links more-viscous blood with mortality in the setting of COVID-19, Maier said. A feature that is unique to COVID-19 is just how high fibrinogen levels rise in response to the acute infection, and this would directly increase blood viscosity, she noted. But the explanation for how that might ultimately lead to the multiorgan microvascular damage seen with COVID-19 is still missing.

The answer to that question, and whether blood viscosity is a marker of disease or is directly mediating the disease course, will influence the clinical implications of these findings, Maier indicated. “If it’s mediating the disease, it absolutely could have impact patient management, because we should have targeted therapies that decrease the viscosity rather than just some of these more-standard approaches that we’ve already been trying.”

A lot of these patients are given anticoagulants, but some still develop thrombotic events, Maier pointed out. “If we knew that the viscosity was directly causing those clots . . . then we could target the viscosity itself rather than just using typical anticoagulation.” She noted that despite the fact that anticoagulants are commonly called blood thinners, they don’t necessarily thin the blood and instead work by stopping the clotting cascade.

Maier’s group has conducted a small trial evaluating plasma exchange as a way to lower blood viscosity in the setting of COVID-19, though she said further research would be needed to establish that reducing viscosity will improve patient outcomes before implementing changes to practice.

In an accompanying editorial, Aldo Bonaventura, MD, PhD (Ospedale di Circolo e Fondazione Macchi, ASST Sette Laghi, Varese, Italy), and Nicola Potere, MD (G. d’Annunzio University of Chieti-Pescara, Chieti, Italy), also say that additional studies are needed.

“While still requiring adequate—both external and prospective—validation, estimated blood viscosity is likely to represent an attractive biomarker, as it was shown to be an early and robust predictor of mortality, and it is widely available and relatively inexpensive,” they write. “In light of the translational potential associated with [it], further investigation is eagerly warranted to confirm and expand the present findings, in order to advance estimated blood viscosity into the clinical scenario.”