Evans syndrome in the SARS-CoV-2 era: “springing up like mushrooms”

Osti and colleagues describe an elderly patient with Evans syndrome (ES), a rare entity characterised by an heterogeneous association and sequence of autoimmune cytopenias, namely autoimmune haemolytic anaemia (AIHA), immune thrombocytopenia (ITP) and, rarely, autoimmune neutropenia. The patient had been previously diagnosed with warm type AIHA concomitant with ITP and, 5 months later, experienced an AIHA relapse with evidence of cold reactive (and complement-activating) autoantibodies upon severe pneumonia during coronavirus disease 19 (COVID-19)¹. In the last 2 years severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiological agent of COVID-19, has become one of the major triggers for the development of secondary autoimmune diseases, including ES, and their reactivation^2–4. During the course of ES, several relapses and complications may occur, particularly infections and thrombosis. The latter resulted in a mortality rate of about 2.4 per 100 person-years in a recent analysis⁵. In this context SARS-CoV-2 infection and vaccination may play a detrimental role by increasing the rate of ES exacerbations requiring immunosuppressive treatment, thus increasing the burden of immune impairment and medicalisation of these patients⁶.

The case presented by Osti and colleagues highlights several aspects regarding SARS-CoV-2-associated ES, including the pathogenesis of secondary ES, the reliability of diagnostic techniques (particularly the direct antiglobulin test, DAT) during septic state, the changing epidemiology, and specific therapeutic considerations.

From a pathogenic point of view, molecular mimicry among SARS-CoV-2 antigens and platelet, red cell and neutrophil epitopes seems to be the prominent mechanism⁷; additionally, the hyper-inflammation triggered by the virus also results in cross-activation of several arms of the immune system (humoral and cellular immunity, complement, coagulation cascade, etc.) against self-antigens in an “innocent bystander” fashion. Furthermore, complement deposition as well as the septic state may favour the exposure of phosphatidylserine as a result of the oxidative stress against erythrocyte and platelet membranes, thus causing blood cell consumption^8,9. These mechanisms are found in various infections as well, and may trigger autoimmunity for host predisposition, i.e. in the presence of primary immunodeficiencies, other autoimmune conditions, underlying lymphoproliferative disorders, etc. “Mutatis mutandis”, it has recently been shown that about 10% of patients with critical COVID-19 harbour neutralising autoantibodies to type I interferons, whilst infected individuals with asymptomatic or mild disease lack these autoantibodies. These autoantibodies are probably pre-existing and have an immunological impact early in the course of COVID-19¹⁰.

Diagnosis of ES mainly relies on the DAT positivity for AIHA and, for ITP and autoimmune neutropenia, on the exclusion of other causes of thrombocytopenia and neutropenia, with or without autoantibody positivity⁵. According to isotype and thermal amplitude of the autoantibody, AIHA is generally classified into warm AIHA (with DAT positive for IgG, or IgG plus complement at low titre and thermal range close to 37°C), cold agglutinin disease (CAD, usually due to IgM autoantibodies reactive at low temperatures, i.e. 4°C to 34°C, DAT positive for complement, and autoagglutination at 20°C), mixed (with DAT positive for IgG plus complement at high titre and autoagglutination at 20°C) and atypical forms (IgA or warm reactive IgM autoantibodies, and DAT-negative cases)¹¹. However, DAT findings are not always straightforward since there may be a continuity among the various thermal AIHA forms. Indeed, in the case described by Osti and colleagues, complement activation was detected even during the first episode, so that a mixed AIHA might be suspected. It is generally accepted that up to 50% of warm AIHA cases can also activate complement, resulting in DAT positivity for IgG plus complement at low titres. These cases are generally more severe and relapse more frequently compared to those positive for IgG only¹². Cases of mixed AIHA, which are rarer, are characterized by the coexistence of a warm IgG and a cold IgM with distinct antigen specificity. However, classifications may be artificial in clinical practice, since the various tests (Ig class, titre, thermal amplitude, and specificity) are not always concordant and positivities may vary over time and in clinical setting (host immunity, previous treatments, presence of triggers). Actually, in the case reported, the high-dose steroid treatment given for the first AIHA episode might have shut down the “warm” part of a mixed AIHA. Furthermore, it can be hypothesised that the B-cell population may release autoantibodies of different Ig classes, particularly in the case of polyclonal activation as observed during infection. This may also occur during primary IgM immune responses which may be immature and less specific. Finally, the presence of a complement-activating trigger, such as SARS-CoV-2, may favour complement deposition on red blood cells and further interfere with DAT findings. It should be mentioned that in AIHA in the context of ES the underlying immunological alteration is thought to be much more profound than in isolated autoimmune cytopenias⁵. In fact, in a large series of AIHA, the occurrence of ITP was linked to higher relapse and mortality rates¹². In this context, it can be hypothesised that the activation of the immune system against multiple antigens may have also triggered the shift from a warm to a cold autoantibody-mediated AIHA. Collectively, these considerations suggest that the DAT should be performed at each relapse, evaluated and supported by an expert immunohaematologist, also considering sepsis and previous therapies.

Concerning epidemiology, although ES is more frequently observed during infancy, the fact that SARS-CoV-2 infections occur mainly in adulthood, explains the higher frequency of COVID-19-induced ES in adults and elderly patients¹³. Moreover, while only one-fifth of cases of primary AIHA are accompanied by ITP, most SARS-CoV-2-induced cases are preceded by or concomitant with ITP^1,2. Furthermore, in this context, the prevalence of cold complement-activating autoantibodies is much higher (up to 50%) than in primary ES, suggesting a more profound immune dysregulation encompassing the aforementioned mechanisms of molecular mimicry and complement activation^1,12. In the patient described by Osti et al., a small B-cell clone was noted in the bone marrow, representing an additional risk factor for the development and recrudescence of AIHA. In CAD, a B-cell clone, distinct from that of other non-Hodgkin’s lymphomas, is almost invariably present, and an IgM monoclonal gammopathy is often present. B-cell clones are rarer in warm cases and, as in the case described, may not produce the autoantibody that is in fact polyclonal. Bone marrow evaluation is therefore usually advised in CAD patients at diagnosis and reserved for relapsed/refractory cases in warm AIHA, unless suggested by clinical suspicion (increased b2 microglobulin levels, palpable lymph nodes and spleen, reticulocytopenia, etc.)¹¹. Given the high rate of secondary cases, the diagnosis of adulthood ES represents an additional possible indication to perform upfront bone marrow assessment⁵. Overall, small lymphoid clonalities should be considered with caution, particularly in a septic context, since they are not necessarily the source of the autoantibody and may either grow or disappear at re-evaluation. However, they allow the exclusion of underlying lymphoproliferative syndromes (by computed tomography scan, genetic testing for specific non-Hodgkin’s lymphoma mutations, etc.) which may require a different treatment approach.

In general, warm AIHA is effectively treated with steroids (prednisone 1 mg/kg day or equivalent), while first-line treatment for CAD, that it is responsive only to high doses ofsteroids, is rituximab. A literature review of SARS-CoV-2-associated AIHA and ES showed that the combination of steroids and intravenous immunoglobulins was the preferred choice, irrespectively of the type of AIHA. Rituximab was administered in only one case with cold reactive autoantibodies^1,13. The use of high-dose steroids in CAD is a common practice, either as a bridge therapy until DAT becomes available, or in very severe CAD for its faster response. Furthermore, immunosuppression related to rituximab may potentially induce severe COVID-19 manifestations, similarly to what described in inflammatory rheumatic diseases¹⁴. Beyond steroids, further supportive treatments, including transfusions, recombinant erythropoietin, and intravenous immunoglobulins should be considered¹⁵. Transfusions of both red blood cells and platelets are to be considered if clinically indicated (i.e. haemoglobin <7 g/dL and platelets <10–20×10⁹/L or higher in the case of comorbidities increasing oxygen requirements or bleeding risk), although with the caveat of alloimmunisation that requires attention and interaction with the transfusion centre. Plasma exchange may be considered in critically ill patients as a last option treatment. The use of convalescent plasma from donors who have recovered from COVID-19 has not proven beneficial and furthermore, in patients with autoimmune cytopenias, may be detrimental for the presence of pro-coagulant microparticles, anti-ADAMTS13, anti-MDA5 or anti-interferon autoantibodies¹⁶. Recombinant erythropoietin (epoetin alpha 40,000 IU/week) is suggested for patients with reticulocyte count or endogenous erythropoietin levels inadequate to the severity of anaemia and is reported effective in up to 70% of cases. Intravenous immunoglobulins (0.4 g/kg for 5 days), although generally less effective in AIHA than in ITP, reduce the reticuloendothelial uptake of blood cells, may exert an anti-idiotypic effect favouring the clearance of the autoantibodies, and also enhance host defences¹⁵. Finally, in the case described by Osti et al., hepatic infarctions were documented, pointing to the high risk of thrombosis in ES patients. Anticoagulant prophylaxis with low molecular weight heparin, provided that the patient has a safe platelet count, is recommended in this setting, particularly in the presence of intravascular haemolysis (lactate dehydrogenase ≥1.5 times the upper limit) and infection⁵. Several concerns have been raised regarding the use of SARS-CoV-2 vaccines in patients with autoimmune cytopenias, including the possibility of an inadequate response under immunosuppression and the risk of exacerbations of autoimmune cytopenias. Regarding the former, there are recent reports that haematological patients treated with monoclonal antibodies (such as rituximab, the anti-CD38 daratumumab, and the anti-B-cell early maturation antigen belantamab) do not mature a humoral response to vaccines¹⁷. Additionally, it has been shown that steroids reduce both humoral and T-cell responses to COVID-19 vaccine in haematological diseases, further complicating the selection of patients¹⁷. TherearereportsintheliteratureofuptosevencasesofAIHA (5 warm and 2 cold) and 25 episodes of ITP developing/ reactivating after vaccination against SARS-CoV-2^18–21, irrespective of the type and dose of vaccine administered. All patients have been successfully managed with steroids, adjustment of ongoing treatment, or supportive therapy. Additionally, two large epidemiological studies estimated that the incidence of ITP after vaccination was lower than that expected for primary ITP in the general population (0.8 vs 3.3 cases per 100,000 adults per year)^22,23. Overall, it is suggested that the time between vaccination and rituximab treatment in ES patients should be as long as possible, that the vaccine should be administered during treatment with the lowest steroid dose, and that the patients should be monitored closely for reactivations of autoimmune cytopenias since these are manageable if recognised promptly.

In conclusion, ES cases in the SARS-CoV-2 era are “springing up like mushrooms” and the diagnosis is challenged by the several confounders linked to the septic state and to the typical hyper-inflammation and complement activation (Figure 1). If promptly recognised, ES is generally manageable with steroids and supportive care, including transfusions, recombinant erythropoietin, intravenous immunoglobulins, and anticoagulant prophylaxis. Rituximab is generally deferred given the possible detrimental impact on the course of infections and because it impairs the response to COVID-19 vaccine in the months subsequent to its use.

Figure 1.

Clinical features and management of Evans syndrome upon SARS-CoV-2 infection

EV: extravascular; IV: intravascular; MAC: membrane attack complex. AIHA: autoimmune haemolytic anaemia; wAIHA warm type AIHA; CAD cold agglutinin disease; C: complement; PS: phosphatidylserine; DAT direct anti-globulin test; IVIG intravenous immunoglobulins; eEPO: endogenous erythropoietin; rEPO: recombinant erythropoietin; ROS: reactive oxygen species.