Coronavirus disease 2019 induces multi‐lineage, morphologic changes in peripheral blood cells

Abstract The clinical course of coronavirus disease 2019 (COVID‐19) varies from mild symptoms to acute respiratory distress syndrome, hyperinflammation, and coagulation disorder. The hematopoietic system plays a critical role in the observed hyperinflammation, particularly in severely ill patients. We conducted a prospective diagnostic study performing a blood differential analyzing morphologic changes in peripheral blood of COVID‐19 patients. COVID‐19 associated morphologic changes were defined in a training cohort and subsequently validated in a second cohort (n = 45). Morphologic aberrations were further analyzed by electron microscopy (EM) and flow cytometry of lymphocytes was performed. We included 45 COVID‐19 patients in our study (median age 58 years; 82% on intensive care unit). The blood differential showed a specific pattern of pronounced multi‐lineage aberrations in lymphocytes (80%) and monocytes (91%) of patients. Overall, 84%, 98%, and 98% exhibited aberrations in granulopoiesis, erythropoiesis, and thrombopoiesis, respectively. Electron microscopy revealed the ultrastructural equivalents of the observed changes and confirmed the multi‐lineage aberrations already seen by light microscopy. The morphologic pattern caused by COVID‐19 is characteristic and underlines the serious perturbation of the hematopoietic system. We defined a hematologic COVID‐19 pattern to facilitate further independent diagnostic analysis and to investigate the impact on the hematologic system during the clinical course of COVID‐19 patients.

well as ameliorating morbidity and mortality in severely ill patients by drugs limiting viral replication or modulating dysbalanced immune response [3,4]. In terms of smart containment strategies widespread testing with high sensitivity and specificity is a prerequisite. However, the "gold standard" test by real-time polymerase chain reaction (RT-PCR) takes several hours in time and requires substantial human und laboratory resources. Moreover, the percentage of false negative pharyngeal samples is quite high partly due to the focused selection of a single compartment for routine diagnostics [5]. Additional laboratory findings supporting diagnosis and enabling follow up diagnostics to evaluate the course of the disease of COVID-19 would be helpful, especially when they are everywhere and rapidly available and cost effective [6].
Yet, no systematic evaluation of morphologic changes in hematopoietic cells in COVID-19 patients has been performed. Such observed changes in peripheral blood cells could in principal represent a specific SARS-CoV-2-mediated impact on hematopoiesis. This assumption is supported by hematologic findings during other viral diseases (ie, mononucleosis, dengue fever, etc), which help clinicians to secure a correct diagnosis [19,20].
Our prospective study aimed to provide a comprehensive description of all morphologic aberrations in blood cells associated with COVID-19. We demonstrate that a systematic search for this hematologic pattern can provide a straightforward and easy-to-perform additional diagnostic tool. Moreover, COVID-19-induced hematologic changes might provide hints to what extent the hematopoietic system contributes to the pathophysiology of this new viral disease.

Patients
In in contrast, the term "reactive/atypical lymphocyte" was applied to describe lymphocytes with a benign/malignant etiology.

Electron microscopy
For electron microscopy analyses, leukocytes and mononuclear cells were isolated from selected whole blood samples by density gradient centrifugation on Ficoll-Paque. The cell pellet was fixed with buffered glutaraldehyde (Karnovsky fixative), and then enclosed in low melting agarose. The embedding process was performed in standardized, automated manner by use of the LYNX microscopy tissue processor (Reichert-Jung, Wetzlar, Germany). Semi-thin-sections and ultra-thin sections (80 nm) were cut using the Reichert Ultracut S Microtome (Leica-Reichert, Wetzlar, Germany). Ultra-thin-sections were contrasted with aqueous 2% uranyl-acetate and 2% lead-citrate solution for 10 min each. Electron-microscopic analysis was performed using the EFTEM LEO 912AB electron-microscope (Zeiss, Oberkochen, Germany).

Immunophenotyping of circulating lymphocyte subsets
A detailed description of multiparametric immunophenotyping is given in the Supporting Information.

Patient characteristics
The patient cohort consisted of 45 hospitalized patients suffering from  Figure S1).

Blood differential
The first observations of distinct morphologic changes in blood smears of COVID-19 patients led to the analysis of an initial (training) cohort (n = 15 COVID-19 patients). In this initial group of patients, marked morphologic anomalies in all three hematologic lineages were found ( Figure 1). In particular, "aberrant lymphocytes" differing from reactive lymphocytes that are commonly seen in other viral diseases such as dengue fever and infectious mononucleosis, as well as "aberrant" monocytes were observed ( Figure 1). Moreover, marked aberrations in granulo-, erythro-, and thrombopoiesis were also seen (detailed composition of aberrations is shown in Table S1).
In order to further evaluate whether these morphologic changes ( Figure 1; Table S1) are associated with COVID-19, we performed a prospective analysis in a larger cohort of COVID-19 patients using an "experimental" blood differential in which all suspected COVID-19 associated aberrations (Table S1) were quantified separately. The results are demonstrated in Table 2. Median differential counts for "aberrant" lymphocytes and "aberrant" monocytes were 1% and 3%, respectively. Pronounced aberrations in all lineages were observed.

Electron microscopy
To assess the ultrastructural equivalents of the distinctive aberrations seen by light microscopy, electron microscopy on peripheral blood TA B L E 2 Light microscopic blood differential blood smears of COVID-19 patients

Quantitative alterations of circulating lymphocyte subpopulations
In addition to the microscopic leukocyte differential, a quantitative analysis of lymphocytes, regarding their major subpopulations, was performed. Relative T cells counts (as % of leucocytes) were significantly lower in the COVID-19 group than in the healthy control group.
There was no difference between the two groups in respect of absolute T cell and B cell counts. Interestingly, there was a significantly lower absolute NK cell count in the COVID-19 group as compared to the healthy control group (Table S2). In addition to the quantification of circulating lymphocyte subpopulations, cell surface expression of all major antigens (ie, CD3, CD4, CD5, CD8) on T cells and their subsets were determined by flow cytometry. We found significant higher expression of CD3, CD5, and CD8 on the CD8 + T cell subpopulation in the COVID-19 group, as compared to the healthy control group ( Figure S4).

DISCUSSION
Our morphologic observations reveal that COVID-19 is associated with hematologic multi-lineage changes of all cellular components of the peripheral blood, that is, lymphocytes, monocytes, granulocytes, erythrocytes, and platelets. This is consistent with previous, sporadic reports describing particular aspects like abnormal lymphocytes, plasmocytoid cells, monocytes, granulocytic, or leukoerythroblastic reactions in COVID-19 patients [15][16][17][18]. Interestingly, while other virus infections (eg, Epstein-Barr virus or Dengue fever) are known to induce morphologic changes in lymphocytes (termed "reactive lymphocytes" or "virocytes"), these "classical" reactive lymphocytes were found only infrequently in COVID-19 patients in our study cohort [19,20]. hyperinflammation/cytokine release syndrome [22]. "Aberrant" cells such as Pelger-Huët neutrophils or monocytes are very likely to be non-/dysfunctional leading to a high susceptibility to secondary bacterial or fungal infections representing an additional serious threat to the life of patients, especially for the critically ill [23,24].
In addition to the immune dysregulation, a further characteristic clinical feature of COVID-19 is the high incidence of thromboembolism [25]. A perturbation of the plasmatic coagulation as well as TA B L E 3 Frequency of morphologic changes in blood smears of COVID-19 patients

All patients and samples
Still SARS-CoV-2 positive patients and samples  [26,27]. The high incidence of giant platelets (independent of absolute platelet counts) found in this study might also contribute to hypercoagulability in COVID-19. Giant platelets are also a common morphologic feature of MPNs, another group of diseases associated with a high risk for thromboembolism.
Our flow cytometry analyses showed distinct immunophenotypic alterations of T-lymphocyte subpopulations in the COVID-19 group consistent with recent reports [14,28,29]. Moreover, we found signif-