Haematological manifestations of COVID‐19: From cytopenia to coagulopathy

Abstract Emerging data from the management of patients with coronavirus disease 2019 (COVID‐19) suggests multi‐systemic involvement, including the hemopoietic system. The haematological manifestations of COVID‐19 include blood count anomalies notably lymphopenia and neutrophilia which are of prognostic significance. Hyperferritinemia and elevated lactate dehydrogenase have also been associated with increased mortality. Furthermore, there is considerable evidence of a distinct coagulopathy associated with COVID‐19 characterised by elevated D‐dimers and an increased risk of thrombotic events. This comprehensive review summarises the latest evidence from published studies and discusses the implications of the various haematological manifestations of COVID‐19 with a view to guiding clinical management and risk stratification in this rapidly evolving pandemic.


| 541
AGBUDUWE AnD BASU lower in severe cases of COVID-19. [14][15][16] In one study of haematological parameters in hospitalised COVID-19 patients in Singapore, the median nadir of the ALC was significantly lower in patients requiring admission to intensive care unit (ICU) (0.4 × 10 9 /L vs 1.2 × 10 9 /L) as was neutrophilia (11.6 × 10 9 /L vs 3.5 × 10 9 /L). 13 Another study from Wuhan revealed a similar pattern with more severe cases having higher neutrophil (4.3 × 10 9 /L vs 3.2 × 10 9 /L; P < .001) counts, lower lymphocyte counts (0.8 vs 1.0 ×10⁹/L ; P <0.001), higher neutrophil-to-lymphocyte ratio (5.5 vs 3.2; P < .001) as well as lower percentages of monocytes, eosinophils and basophils. 17 In addition to a significant reduction in both CD4+ and CD8+ T lymphocyte subsets in COVID-19 patients, severe cases had much lower CD8+ lymphocytes and a subsequent increase positively correlated with improved clinical outcomes. 18 Mild thrombocytopenia (100-150 × 10 9 /L) has been reported in up to 20%-36% of COVID-19 cases [11][12][13] however, severe thrombocytopenia (<50 × 10 9 /L) is unusual. In a case series of patients admitted to an intensive care unit in Wuhan, a platelet count of <100 × 10 9 /L was observed in only 5% of patients. 14 Similar findings were reported in a French cohort study with mild thrombocytopenia identified in about a quarter of COVID-19 patients on admission to hospital and this was independently predictive of the risk of admission to ICU, mechanical ventilation or death. 19 Furthermore, data from multiple studies suggest that anaemia is not a prominent feature of COVID-19 even in severe cases. 11,13,14 A case series of autoimmune haemolytic anaemia occurring in COVID-19 patients was recently published but about half of the patients in this study had an underlying haematological disorder known to be associated with autoimmune haemolysis. 20

| PERIPHER AL B LOOD FILM AND MORPHOLOG IC AL FE ATURE S
Examination of the peripheral blood smear remains a crucial part of the haematological assessment and can often aid the diagnosis of many conditions. Notable features reported in COVID-19 cases include an increased frequency of reactive and plasmacytoid lymphocytes, 13,21 significant left-shifted granulopoiesis with hypergranular, occasionally vacuolated neutrophils 22 and leucoerythroblastic features. 23 The presence of schistocytes or red cell fragments has not been reported.
Furthermore, from the experience of the authors, bone marrow hemophagocytosis can be a feature of severe COVID-19 and this has been observed in a small number of patients managed at New Cross Hospital, Wolverhampton, United Kingdom. The bone marrow aspirate from one of the patients is shown in Figure 1. This pathologic finding corroborates recent reports of cases of COVID-19 meeting the clinical criteria for secondary hemophagocytic lymphohistiocytosis (sHLH). 24,25

| OTHER B LOOD MARK ER S
Alongside other acute-phase markers such as C-reactive protein (CRP), procalcitonin and erythrocyte sedimentation rate, an elevated ferritin has been associated with increased mortality in COVID- 19. 17,26,27 Significant elevation of lactate dehydrogenase has also been observed in patients with severe disease. 13

| COVID -19 C Y TOK INE S TORM AND HAEMATOLOG IC AL HYPERINFL AMMATORY SYNDROME S
Observations from the clinical, biochemical and serological manifestations of COVID-19 strongly support an immunological basis for the severe manifestations of the disease. 27 The marked elevation of pro-inflammatory markers such as IL-1β, IL-2, IL-4, IL-6, IL-10, TNF-α and IFNγ frequently seen in severe COVID-19, 28-30 point to a state of disordered and exaggerated immune response to SARS-CoV-2 infection, often referred to as the "cy- Tocilizumab administration also resulted in clinical improvement in a patient with COVID-19 and multiple myeloma. 43 A recently published retrospective cohort study of patients treated with tocilizumab, the largest to date, observed a significant reduction in the risk of mechanical ventilation or death in tocilizumab-treated patients compared with controls. 38 Several randomised clinical trials of tocilizumab are ongoing and the results of these studies are eagerly awaited.

| COVID -19-A SSO CIATED COAG U LO PATH Y (C AC)
Recent data emerging from the management of patients with COVID-19 suggests an increased thrombotic tendency. Approximately onethird of patients with COVID-19 had CT scan evidence of pulmonary embolism (PE) in a French study. 44 Notably, two-thirds of the patients without PE in this cohort also had elevated D-dimers with a higher cut off value of 2660 µg/L being more predictive of PE in this cohort. A Dutch study of COVID-19 patients admitted to ICU similarly identified a 31% incidence of thrombotic events including PE, deep vein thrombosis (DVT) and ischaemic strokes despite all patients having received prophylactic anticoagulation. 45 A retrospective study of COVID-19 patients admitted to ICU identified DVT in 25% with advanced age, lower lymphocyte counts and elevated D-dimers being significant risk factors. 46 The prognostic importance of D-dimer testing was further demonstrated in a prospective study of hospitalised COVID-19 patients which revealed that significantly higher D-dimer and prolonged prothrombin time (PT) were associated with a higher probability of mortality. 47 In a study investigating prolonged activated partial thromoplastin time (APTT) in patients with COVID-19, the lupus anticoagulant was detected in 91%. 48 In addition, markedly elevated Von Willebrand factor (VWF) levels (VWF:antigen-555%, VWF:activity-520%) and factor VIII (clotting activity of 369%) in addition to antiphospholipid antibodies were observed in a patient with severe COVID-19. 49 Taken together, these findings strongly support the existence of a syndrome of COVID-19-associated CAC characterised by derangements in clotting tests (PT and APTT), elevated D-dimer and an increased thrombotic tendency.
In a study comparing coagulation parameters in hospital-  Table S1 for the relevant study data). for overt disseminated intravascular coagulation (DIC) compared with 1 (0.6%) of survivors. 47 Therefore, the ISTH DIC scoring system which includes platelet count, fibrinogen, D-dimer and prothrombin time ( Table 2) is likely to be a useful prognostic tool for COVID-19associated Coagulopathy (CAC). The CAC appears to be distinct from DIC due to other causes in elevated fibrinogen, 50 modest prolongation of the APTT and the absence of schistocytes on the blood film. 51 Interestingly, despite the derangements in coagulation tests, abnormal bleeding is unusual. 52,53 Furthermore, much of the data for CAC in COVID-19 has been from hospitalised patients with the severe form of the disease. It is yet unclear but unlikely that patients with mild COVID-19 are at increased risk of thrombosis.

| D ISCUSS I ON
While still not fully understood, the immune response to SARS-CoV-2 is thought to be two-phased; an initial T-cell mediated adaptive response during the presymptomatic or non-severe stage which in a proportion of individuals is ineffective, leading to uncontrolled viral replication, triggering an exaggerated innate immune response which results in multi-organ failure in the most severe cases. 55,56 The dysregulated secretion of pro-inflammatory cytokines perpetuates hyperinflammation and is thought to be the basis of the cytokine storm in COVID-19. This two-phased model of the immune response would probably account for the lack of efficacy of some immunosuppressive therapies in non-severe cases of the disease. 37 Furthermore, ageing is generally associated with diminished adaptive immune function and an exaggerated innate immune response to pathogens, notably a shift towards myelopoiesis and neutrophil accumulation in animal models. 57,58 These age-related changes in the immune system may, at least partly, explain the increased mortality of COVID-19 among the elderly. A lot is still not known about the marked difference between immune responses to SARS-CoV-2 between individuals and it is likely that genetic and environmental factors also play a role. Of particular interest is the association between certain ABO blood group genotypes and the likelihood of severe COVID-19. In a recently published genomewide association study, Blood group A and a few other single nucleotide polymorphisms were associated with an increased risk of COVID-19-induced respiratory failure (Blood group O was apparently protective). 59 Interestingly, blood group A has also been shown to be associated with increased odds of thromoboembolic and cardiovascular events. 60 17,28,66 Of particular interest was the finding of reduced regulatory T cells (Treg) in severe cases. 28 The precise mechanism of the development of lymphopenia in COVID-19 is not known but lymphocytes have been shown to express ACE2 67 and lymphoid cell apoptosis may be a consequence of SARS-CoV-2 infection. 68 Trafficking of lymphocytes away from the peripheral blood to the lungs or other sites of infection may also play a role. 69 The neutrophilia observed in severe cases of COVID-19 is most likely a response to the cytokine storm which has been implicated in the most severe manifestations of the disease. Furthermore, significant elevations of acute-phase markers such as ferritin, CRP and procalcitonin have been associated with mortality from COVID-19 and these biomarkers are positively correlated with increased pro-inflammatory cytokines such as Interleukin-6 and TNF-α. 26,28 Therefore, serial monitoring of these markers may facilitate early therapeutic intervention with experimental immunomodulatory agents.

| RECOMMENDATI ON S AND CON CLUS IONS
In view of the reported high incidence of thrombotic events despite prophylactic anticoagulation in critically ill patients with COVID-19, 53,78,79 there is a strong argument for intensification of anticoagulation in this setting but the benefit of this strategy remains to be determined due to the lack of clinical trial data. While there is currently insufficient evidence to recommend therapeutic anticoagulation in all patients with COVID-19 in the absence of venous thromboembolism, an individualised risk-based approach to intensification of anticoagulation may be considered.
In summary, severe COVID-19 represents a state of immune dys-

ACK N OWLED G EM ENT
The authors wish to thank Dr Richard Whitmill for the bone marrow aspirate image included in this publication.

CO N FLI C T O F I NTE R E S T S
The authors declare no conflict of interests in relation to this article.