Title: Impact of COVID‐19 on red blood cell rheology
Abstract: Coronavirus disease 2019 (COVID-19) is an infectious disease caused by severe acute respiratory coronavirus 2 (SARS-CoV-2). Among hospitalized patients, many have developed typical manifestations of sepsis-like cold extremities, weak peripheral pulses or severe metabolic acidosis, indicating microcirculation dysfunction.1 Microvascular dysfunction has been reported in sepsis2, 3 and could be partly caused by red blood cell (RBC) rheological alterations such as decreased RBC deformability (RBCD),4 increased RBC aggregation5 and abnormal RBC morphology/physiology.6 Whether COVID-19 patients also have impaired RBC rheology is unknown. The aim of this preliminary study was to compare the haemorheological profile (RBCD, RBC aggregation, blood viscosity) between COVID-19 patients, sepsis patients and healthy controls. Seven COVID-19 patients (COVID) hospitalized in Edouard-Herriot Hospital (Hospices Civils de Lyon, Lyon, France) were included. COVID-19 diagnosis was made by specific real-time polymerase chain reaction (RT-PCR; n = 5) and/or by typical chest computerized tomography (n = 7). Seven patients admitted for non-COVID-related sepsis (SEPSIS) and seven healthy controls (CONTROL) were also included. SEPSIS patients were identified by the quick Sequential Organ Failure Assessment (quickSOFA) score according to Sepsis-3 definition.7 Clinical reports were reviewed by a physician to collect general and clinical characteristics (Table I). The study was approved by the local ethics committee (No. 20-108) and the Hospices Civils de Lyon biological resource centre (CRB HCL, BB-0033-00046). Haematological parameters were determined using a haematology analyser (XN-9000; Sysmex Corporation, Kobe, Japan) for COVID and SEPSIS patients. The haemorheological parameters (RBCD, RBC aggregation and blood viscosity) were measured by ektacytometry, Myrenne aggregometer and cone/plate viscometer respectively, as previously described.8 The Hemox-Analyzer (TCS, Medical Products Division, Southampton, PA, USA) was used to plot the oxygen haemoglobin dissociation curves and determine the P50 values (the oxygen tension at which 50% of Hb is saturated with oxygen), with adjustment of pH at 7.4 and temperature at 37°C.9 Gender distribution was not different between the three groups (Table I). The age of the SEPSIS group was significantly higher than the CONTROL group (P < 0.05). No other differences in age between groups were detected. In the COVID group, none of the patients were smokers, but every patient had co-morbidities such as obesity, high blood pressure, diabetes, metabolic syndrome, cardiac, hepatic or renal failure. COVID patients were globally severe with marked lung alterations, as evidenced by the extent of parenchymal lesions (>50% in 43% of cases; data not shown) and a transfer to intensive care unit (ICU) for three of them (none in the SEPSIS group). No patient in any group died during the study. White blood cell counts and C-reactive protein were significantly lower in the COVID than in the SEPSIS group (P < 0.05). The other common biological parameters did not differ between these two groups. There were no differences in P50 values between groups. RBCD was significantly lower in the COVID group compared to the SEPSIS group at low shear stresses (i.e., from 0.53 to 3 Pa; P < 0.05). In addition, RBCD from COVID patients was significantly lower than the CONTROL group for shear stresses, ranging from 1.69 to 30 Pa (P < 0.05). RBCD was also significantly reduced in the SEPSIS compared to the CONTROL group at 30 Pa (P < 0.05). Maximum RBCD was reduced in both COVID and SEPSIS groups compared to the CONTROL group (P < 0.05). Blood viscosity was not significantly different between the three groups. No difference in fibrinogen levels was shown between COVID and SEPSIS patients. Despite no data available in the CONTROL group, the concentrations were above the normal range (reference values of our laboratory: 2–4 g/l). RBC aggregation measured at stasis was significantly higher in the COVID group compared to the two other groups, while RBC aggregation measured at a low shear rate was significantly higher in both the COVID and SEPSIS groups compared to the CONTROL group. Our results showed that in COVID-19 patients, RBC aggregation is increased both at stasis and at a low shear rate, compared to CONTROL individuals, while RBC aggregation in SEPSIS patients was higher than healthy individuals at low shear rate only. RBC aggregation at a low shear rate, in comparison to a static condition, increases the chance of RBCs to make contact and aggregate, which would better reflect in-vivo conditions. It is possible that the high fibrinogen levels observed in both COVID and SEPSIS patients, relative to the 2–4 g/l reference value used in our laboratory, was a contributing factor to increased RBC aggregation,10 but further work is needed to test the contribution of RBC aggregability (i.e., cellular factors). Clot stability has been shown to be affected by the RBC rheological properties.11 Indeed, enhanced RBC aggregation could affect clot structure and increase their resistance. The lack of peripheral blood smears in the present study did not allow us to clearly differentiate RBC aggregation from RBC agglutination, which may also be a factor in COVID-19 pathophysiology. The present study also showed a decrease in RBCD in the COVID group compared to the CONTROL group at shear stresses ranging from 0.53 to 30 Pa. Changes in RBC membrane lipids composition and RBC membrane protein fragmentation could explain the RBCD reduction in COVID-19 patients.12 RBCD at 30 Pa and maximum RBCD were also different between the CONTROL and SEPSIS groups, suggesting that RBC from sepsis patients would have increased internal viscosity and/or a loss in the surface/volume ratio, but not impaired membrane elasticity, in contrast to COVID-19 patients.13 Despite the differences in RBC rheological properties, no difference in blood viscosity was observed between the three groups, possibly because haematocrit tended to be decreased in the COVID and SEPSIS groups. The lack of difference in measured P50 between COVID-19 and healthy individuals contrasts with a recent study where P50 was calculated from blood gas analyzer.14 Our results suggest that SARS-CoV-2 does not affect Hb oxygen affinity directly, but COVID-19 patients may have increased Hb oxygen affinity due to hypoxia/metabolic adaptations.14 In conclusion, this preliminary study shows that RBC rheological properties are impaired in COVID-19 patients. This could increase the risks for thromboembolic events, as well as affect microvascular blood flow. Nevertheless, the sample size of this study was limited and further studies are needed. The authors thank the Biological Resource Centre of Hospices Civils de Lyon. Céline Renoux, Romain Fort, Elie Nader, Philippe Joly, Emeric Stauffer, Mélanie Robert, Agnès Cibie, Alexandra Gauthier and Philippe Connes designed the research study. Romain Fort, Céline Renoux and Sandrine Girard included patients or controls. Camille Boisson, Elie Nader and Philippe Joly performed the research. Céline Renoux, Romain Fort, Elie Nader, Philippe Joly and Philippe Connes analysed the data. Céline Renoux, Romain Fort, Elie Nader and Philippe Connes wrote the paper. All the authors revised the paper critically and approved the final version. The authors declare no conflicts of interest.