Michael J. Day, BSc, BVMS(Hons), PhD, DSc, DECVP, FASM, FRCPath, FRCVS
Antibody and/or complement-mediated destruction of circulating red blood cells (RBCs) is known as immune-mediated haemolytic anaemia (IMHA). The immunological destruction of RBCs occurs by extravascular phagocytosis in the spleen or liver, or by intravascular osmotic lysis following the generation of terminal membrane attack complexes of the complement pathway. In some cases, the bone marrow erythroid precursors are targeted giving rise to more severe diseases termed non-regenerative IMHA (NRIMHA) and pure red cell aplasia (PRCA).
In primary IMHA no underlying cause for the presence of RBC antibody can be identified on thorough clinical and laboratory investigation. Primary IMHA may therefore be considered as a true idiopathic, autoimmune disease and the diagnostic label autoimmune haemolytic anaemia (AIHA) applied. AIHA and AITP may occur concurrently (Evans' syndrome) or may be part of a multisystemic autoimmune syndrome (e.g., systemic lupus erythematosus). AIHA may also occur in combination with autoimmune neutropenia and sometimes all three immune-mediated cytopenias may be concurrent. By contrast, in secondary IMHA, the presence of erythrocyte-associated antibody is secondary to an underlying cause - most commonly neoplastic or infectious disease.
In primary, idiopathic AIHA the specificity of the autoantibodies is for an autoantigenic component of the erythrocyte membrane. The target membrane antigens in canine AIHA include the anion exchange molecule (band 3), the cytoskeletal molecule spectrin and a series of membrane glycoproteins (glycophorins).
A high proportion of dogs with IMHA are believed to have true primary, idiopathic AIHA. Further evidence for the existence of primary AIHA in the dog comes from the strong breed and familial associations that are documented; particularly for cocker spaniels, old English sheepdogs and English springer spaniels. A strong predisposition in the Maltese terrier is reported from Victoria. A predisposition amongst Clumber Spaniels is widely recognized in North America. Our recent research has defined MHC gene associations with primary IMHA in these breeds and we are currently performing genome-wide association studies (GWAS) searching for genetic associations in the English cocker and springer spaniels. Primary IMHA is thought to be relatively less common than secondary disease in the cat. There is little evidence for breed-associated or familial feline AIHA.
Secondary Immune-Mediated Disease
IMHA may be secondary to:
1. Neoplasia (e.g., lymphoma, leukaemia, myeloproliferative disease in dogs and cats or haemangiosarcoma in the dog).
2. Chronic inflammatory disease (e.g., inflammatory bowel disease in dogs, abscesses, FIP infection, pancreatitis or chronic interstitial nephritis in cats).
3. Exposure to an infectious agent (e.g., viruses, arthropod-borne microparasites such as Babesia, Ehrlichia, Leishmania, Rickettsia and Cytauxzoon). There is a strong association between feline haemoplasma infection (M. haemofelis), haemolytic anaemia and Coombs positivity. Canine haemoplasmas (Mycoplasma haemocanis, 'Candidatus Mycoplasma haematoparvum') are not associated with IMHA.
4. Recent administration of a vaccine (polyvalent, modified-live, adjuvanted vaccines are generally incriminated).
5. Drug therapy, the best example of which involves administration of sulphonamide-trimethoprim to Doberman pinschers and dogs of a range of other breeds. In the cat, medical management of hyperthyroidism with methimazole is associated with IMHA, IMTP and serum ANA positivity.
In general terms, two clinical presentations for canine IMHA are recognized. The most common involves a relatively slow onset (days to weeks) of clinical signs including anorexia and increasing lethargy and exercise intolerance. Physical examination reveals pale mucous membranes, and in cases of concurrent thrombocytopenia, petechiae can be seen. Tachypnoea and tachycardia are present to compensate for anaemia. A systolic cardiac murmur can develop as a result of a decrease in blood viscosity. There may be evidence of hepatosplenomegaly and lymphadenomegaly. The alternative clinical presentation is an acute onset haemolytic crisis. These patients may be relatively normal one day, and collapsed the next. Such patients will more frequently have haemoglobinuria and icterus due to hypoxaemic hepatic necrosis combined with increased bilirubin turnover.
Pulmonary (and other organ) thromboembolism associated with disseminated intravascular coagulation (DIC) is a recognized complication of canine IMHA and considered an important risk factor for mortality. The pathogenesis of this effect is poorly understood although changes in coagulation profiles and platelet hypercoagulability are widely recognized. A recent study has shown no evidence for autoantibodies against vascular endothelium as are often observed in human thromboembolic processes and the 'lupus anticoagulant' is rarely identified in dogs with IMHA.
The same spectrum of clinical disease may occur in the cat.
Initial clinical and historical assessments are supported by haematological analysis and bone marrow assessment in non-regenerative forms of disease. Key features of the haemogram consistent with IMHA include:
Strongly regenerative, severe haemolytic anaemia (except in cases of NRIMHA/PRCA or acute onset disease)
Autoagglutination of erythrocytes macro- or microscopically. Macroscopic agglutination should be further investigated by the 'in-saline agglutination test.'
Spherocytosis (in dogs)
Left shift neutrophilia related to centrilobular hepatocyte degeneration and necrosis
Concurrent thrombocytopenia in Evans' syndrome or neutropenia in concurrent IMNP
Haemostatic abnormalities in dogs with IMHA may underlie the predisposition to development of pulmonary thromboembolism. The acute phase proteins (APPs) C-reactive protein (CRP) and 1 acid glycoprotein (AAG) are elevated at the time of presentation with IMHA.
The identification of erythrocyte-bound antibody and/or complement is based on the Coombs or direct antiglobulin test. The Coombs test involves the incubation of a washed RBC suspension with antisera that will cause gross agglutination. The polyvalent Coombs reagent detects all immunoreactants, whilst specific antisera (to IgG, IgM and complement C3) may be used in parallel. Testing with a panel of reagents is more likely to give a positive diagnosis than testing with polyvalent reagent alone. Moreover, there are distinct differences in the pattern of test reactivity between dogs with primary and secondary disease. Dogs with primary disease are more likely to have IgG antibody and less likely to have IgM antibody compared with dogs having secondary IMHA.
Although other ELISA-based tests for IMHA have been developed, they are more complicated than is required for routine clinical diagnosis. In some countries the osmotic fragility test is favoured as an adjunct diagnostic test. Flow cytometric detection of erythrocyte-bound antibody has greater sensitivity, but less specificity than the Coombs test and depends on the availability of a flow cytometer.
For those cases in which IMHA is secondary to a defined underlying cause, appropriate therapy for the primary disease must be administered. Adjunct therapy for the anaemic state may also be considered, particularly transfusion of whole blood or packed erythrocytes.
Adjunct administration of heparin or ultra-low dose aspirin is routinely practiced as a means of preventing the potential secondary thromboembolic complications of IMHA.
The specific therapy for these disorders involves administration of immunosuppressive doses of glucocorticoids, with major effect of downregulating Fc receptor expression by phagocytic cells. A range of other drugs (e.g., danazol, azathioprine, cyclophosphamide, ciclosporin, mycophenolate mofetil and liposomal-encapsulated clodronate) has been co-administered with glucocorticoids to reduce the likelihood of glucocorticoid side effects ('steroid sparing') or have additional effects on the lymphoid cells that initiate the disease or the macrophages responsible for extravascular haemolysis. Despite the widespread use of such protocols, there is no clear consensus as to which (if any) combination is most efficacious, and which individual cases are the most likely candidates for combined therapy. In fact, some studies suggest that the use of drugs such as cyclophosphamide or ciclosporin have no added value (ciclosporin), or result in a less successful clinical outcome than use of glucocorticoid alone (cyclophosphamide). At present, the majority of authors favour azathioprine when an adjunct to glucocorticoid is required. Blockade of Fc receptors by administration of human gammaglobulin has proven useful, but the expense and availability of this product generally precludes its widespread application. The management of feline IMHA also involves immunosuppressive glucocorticoids, remembering that the use of azathioprine in cats is generally to be avoided. All cats should be tested for haemoplasma infection and if positive only an appropriate course of antimicrobial therapy (e.g., doxycycline) is required.
IMHA is a severe disease and in a referral setting several studies have now shown that approximately 50% of dogs will die during initial hospitalization. IMHA with evidence of intravascular haemolysis (hyperbilirubinaemia), autoagglutination, targeting of bone marrow precursors, thrombocytopenia or development of pulmonary thromboembolism carries a more guarded prognosis. Risk factors for thromboembolic complications include thrombocytopenia and hyperbilirubinaemia. Those animals that recover from the initial disease episode are at risk for relapse, often months or years later.