Blood Reviews (2008) 22, 1731

www.elsevierhealth.com/journals/blre

REVIEW

Hemolytic anemia due to warm autoantibodies

Charles H. Packman
*

University of North Carolina School of Medicine, Chapel Hill, North Carolina Hematology-Oncology Section, Department of Internal Medicine and Blumenthal Cancer Center, Carolinas Medical Center, 1000 Blythe Boulevard, MMP-6, Charlotte, NC 28203, United States

KEYWORDS
Autoantibodies; Hemolytic anemia; Spherocytes; Reticulocytes; Direct antiglobulin test; Coombs test; Immunoglobulin; Complement

Summary The diagnosis of autoimmune hemolytic anemia (AHA) requires evidence of shortened red blood cell (RBC) survival mediated by autoantibodies directed against autologous RBCs. About 80 percent of patients with AHA have warm-reactive antibodies of the IgG isotype; the remainder exhibit cold-reactive autoantibodies. Typical patients exhibit anemia, reticulocytosis, spherocytes and polychromasia on the blood lm and a positive direct antiglobulin test (DAT). Increased indirect serum bilirubin, urinary urobilinogen and serum lactate dehydrogenase (LDH), and decreased serum haptoglobin are not required for the diagnosis, but are frequently present. Patients with AHA and no underlying associated disease are said to have primary or idiopathic AHA. AHA in patients with associated autoimmune disease and certain malignant or infectious diseases is classied as secondary. The etiology of AHA is unknown. Patients with symptomatic anemia require transfusion of RBCs. Prednisone and splenectomy may provide long term remission. Rituximab, intravenous immunoglobulin, immunosuppressive drugs and danazol have been effective in refractory cases and for patients who are poor candidates for surgery. c 2007 Elsevier Ltd. All rights reserved.


Introduction

Hemolytic icterus was familiar to clinicians by the start of the 20th century. Two forms, congenital (or familial) and acquired, were recognized but they could not be precisely distinguished and some authorities doubted the existence of acquired

* Tel.: +1 704 355 2884; fax: +1 704 446 5143. E-mail address: cpackman@carolinas.org

hemolytic icterus.1 Many patients with hemolytic icterus exhibited jaundice, splenomegaly, anemia, spherocytosis, reticulocytosis and increased osmotic fragility of red blood cells (RBCs). Sera from some patients with hemolytic icterus directly agglutinated normal or autologous RBCs in vitro. Sera from other patients caused hemolysis. The responsible serum factors were respectively termed direct (saline) agglutinins or hemolysins. Much later it was determined that most agglutinins were IgM antibodies and that hemolysins were

0268-960X/$ - see front matter c 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.blre.2007.08.001

18 also antibodies that required the presence of fresh serum complement to mediate lysis. However, most patients with hemolytic icterus exhibited neither direct agglutinins nor hemolysins and their infrequent association with hemolytic icterus offered no distinguishing diagnostic information. That was the general state of affairs for over 50 years,1 until the seminal paper of Coombs, Mourant, and Race2 in 1945 reported that RBCs coated in vitro with nonagglutinating Rh antibodies (later shown to be IgG isotype) could be agglutinated by rabbit antiserum to human c-globulin; this is now termed the indirect antiglobulin test (IAT). Initially unknown to Coombs and colleagues, the antiglobulin principle, i.e., that antiglobulin serum could cross-link antibodycoated particles to produce visible agglutination, had been previously described 37 years earlier in an obscure Germanlanguage journal by Moreschi.3 Credit was given to Dr. Moreschi in a subsequent paper4 and later by Dr. Coombs in other articles and lectures. Shortly thereafter, Boorman, et al.,5 and Loutit, et al., 6 reported that rabbit antiglobulin serum agglutinated RBCs taken directly from patients with suspected acquired hemolytic anemia, including those lacking saline agglutinins or hemolysins.

C.H. Packman This procedure is now termed the direct antiglobulin (Coombs) test (DAT). RBCs from patients with presumed congenital hemolytic anemia, which we now recognize as hereditary spherocytosis, did not agglutinate. The positive direct antiglobulin reaction in AHA is attributable to in vivo coating of the RBCs with immunoglobulins (mainly IgG) and/or complement proteins and it is the main distinction between hereditary spherocytosis (congenital hemolytic icterus) and AHA (acquired hemolytic icterus).

Demographics and classication of AHA

Warm antibody AHA is uncommon, with an annual incidence of about 1 per 75,00080,000 population.7 It has been diagnosed in people of all ages, but most patients are over age 40, and the peak incidence occurs around the seventh decade. This age distribution may be related to the increased frequency of lymphoproliferative malignancies in the elderly, resulting in an age-related increase in secondary AHA due to lymphoma. Most cases of primary AHA arise sporadically; familial cases are rare.810

Table 1
I. A. B. 1. 2. 3. 4. II. A. 1. 2. a. b. B. 1. 2. a. b. III. A. B. 1. IV. A. B. C.

CLASSIFICATION OF HEMOLYTIC ANEMIA MEDIATED BY ANTIBODIES.

Warm-autoantibody type Primary or idiopathic warm antibody AHA Secondary warm antibody AHA associated with: Lymphoproliferative disorders, e.g., Hodgkin disease, lymphoma Rheumatic/ autoimmune disorders, e.g., SLE, ulcerative colitis Certain nonlymphoid neoplasms, e.g., ovarian tumors Ingestion of certain drugs, e.g., a-methyldopa Cold-autoantibody type Mediated by cold agglutinins Idiopathic (primary) chronic cold-agglutinin disease (most of these exhibit evidence of monoclonal B-lymphoproliferation) Secondary cold-agglutinin hemolytic anemia associated with: Infections (e.g., Mycoplasma pneumoniae or infectious mononucleosis) Clinically-evident malignant B-cell lymphoproliferative disorder Mediated by cold hemolysins Idiopathic (primary) paroxysmal cold hemoglobinuria Secondary Donath-Landsteiner hemolytic anemia, usually associated with an acute viral syndrome in children Associated with congenital or tertiary syphilis in adults Mixed cold and warm autoantibodies Primary or idiopathic mixed AHA Secondary mixed AHA Associated with the rheumatic disorders, especially SLE Drug-immune hemolytic anemia Hapten or drug-adsorption mechanism Ternary (immune) complex mechanism True autoantibody mechanism

Hemolytic anemia due to warm autoantibodies AHA is commonly classied by the optimal temperature at which the autoantibodies bind to patient RBCs in vivo (Table 1). Warm antibody AHA comprises 8090 percent of cases of AHA. Hemolysis in these patients is mediated by warm-reactive autoantibodies,7,11,12 so called because they react optimally with human RBCs at 37 C. In most of the remaining cases hemolysis is mediated by cold-reactive autoantibodies which exhibit greater afnity for RBCs at temperatures below 37 C and cause RBC membrane injury by activating complement. The great majority of cold-reactive autoantibodies are cold agglutinins. Cold hemolysins are much less common. A small number of patients exhibit both cold-reactive and warm-reactive autoantibodies,1315 RBC destruction is generally more severe in such mixed cases. The distinction is important because the pathophysiology of RBC injury and the therapeutic approaches differ in warm- and cold- antibody AHA. AHA may also be classied by the presence or absence of pathophysiologically or etiologically associated underlying diseases (see Table 1). When no recognizable underlying disease is present, the AHA is termed primary or idiopathic. When AHA appears to be a manifestation or complication of an underlying disorder, the term secondary AHA is applied. AHA may be considered secondary 1) when conjunction of AHA and the underlying disease occurs more frequently than can be accounted for by chance alone; 2) when correction of the associated disease reverses the AHA; or 3) when AHA and the associated disease are related by evidence of immunologic aberration.7 Using strict classication criteria, primary warm-antibody AHA comprises about 50 percent of cases. Chronic lymphocytic leukemia (CLL) and lymphomas account for about half of secondary warm antibody AHA cases. Autoimmune diseases, particularly systemic lupus erythematosus (SLE), account for a considerable proportion of the remaining secondary warm antibody AHA cases. Of interest, most patients with apparently primary cold agglutinin disease have evidence for clonal B-lymphocyte proliferation.16 Many other patients with cold agglutin disease have clinically evident chronic lymphocytic leukemia or low grade lymphoma. Infectious mononucleosis and Mycoplasma pneumoniae are occasionally associated with cold antibody AHA. Many patients with mixed cold and warm autoantibodies have SLE.13,14 A positive DAT is not unusual in patients infected with the human immunodeciency virus (HIV), but hemolysis is relatively rare in these patients.1719 Other associated diseases are listed in Table 1.

19 Many drugs also set off immune injury to RBCs (Table 1). Drug-mediated immune injury to RBCs is often referred to as drug-immune hemolytic anemia to distinguish these forms of RBC injury from de novo AHA. These are not generally thought of or classied as autoimmune hemolytic anemias, but some drugs, e.g., alphamethyldopa and udarabine are capable of provoking true autoantibodies which react with RBCs in the absence of the inciting drug. The other types of drug-immune injury require the simultaneous presence of drug and antibody to cause RBC injury. However, even some of these so-called drug-dependent antibodies exhibit specicity for a particular RBC antigen and can injure only RBCs that express the antigen, suggesting that auto-reactivity plays a role in the genesis of these antibodies.2024 The incidence of drugimmune hemolytic anemia is estimated at 1 per million population, about 88% of which are due to the second and third generation cephalosporins, cefotetan and ceftriaxone.25

Mechanisms of RBC destruction by warm autoantibodies

In warm-antibody AHA, the autoantibodies are predominantly IgG globulins which possess relatively high afnity for human RBCs at body temperature (37 C). As a result, most of the autoantibody is bound to the patients circulating RBCs with only a amall amount circulating free in plasma. In patients with primary AHA, erythrocyte autoantibodies are the only recognizable immunologic aberration. The autoantibodies in a given patient are usually specic for a single RBC membrane protein (see below). This narrow spectrum of autoreactivity suggests that the development of primary AHA may not be due to a general defect in immune regulation but more likely arises through an aberrant immune response either to a self-antigen or to an unknown immunogen that mimics a selfantigen. On the other hand, in patients with AHA secondary to lymphoma or SLE for example, the autoantibodies probably do arise through an underlying defect in immune regulation. AHA occurs more frequently in patients with low-grade lymphoma or CLL treated with udarabine26 or 2-chlorodeoxyadenosine (cladribine).27 These drugs profoundly suppress T-lymphocytes which may exacerbate the preexisting tendency of these patients to form autoantibodies. Alpha-methyldopa induces warm-reacting IgG anti-RBC autoantibodies in about 10% of otherwise

20 normal individuals. The autoantibodies are indistinguishable from autoantibodies arising in unprovoked AHA (see below), exhibiting similar Rh-related serologic28 and immunochemical29 features. These drug-associated autoantibodies disappear when the drug is discontinued. This suggests that many immunologically normal individuals possess the latent potential to form this type of anti-RBC autoantibody, but whatever mechanisms create such autoantibodies do not necessarily also create a sustained autoimmune state. RBC destruction in warm antibody AHA is mediated by the autoantibodies. Several lines of evidence support this idea: 1) In patients with warm-antibody AHA, transfused RBCs lacking the antigen targeted by the patients autoantibodies generally survive normally, in contrast to the shortened survival of the patients own RBCs.11,30,31 2) Hemolytic disease of the newborn results from passage of IgG anti-RBC autoantibodies from a mother with AHA to the fetus, via the placenta.32 3) In individual patients, an inverse relationship exists between the quantity of RBC-bound IgG antibody and RBC survival in the patient.3338 Destruction of the autoantibody-coated RBCs is mediated primarily by sequestration and phagocytosis in macrophages of the Billroth cords in the spleen and to some extent in Kupffer cells of the liver.30,33,34,3640 The macrophages express surface receptors for the Fc region of IgG, particularly for IgG1 and IgG341,42 and for opsonic fragments of complement proteins C3 and C4.4345 When present together on the RBC surface, IgG and complement fragments act synergistically to enhance trapping of the RBC and phagocytosis.36,37,4448 Most RBC sequestration in warm-antibody AHA occurs in the spleen,30,3739 but trapping in the liver occurs in the presence of large quantities of RBC-bound IgG33,34,40 or the concurrent presence of complement proteins on the RBCs.33,36,37 A trapped RBC may be partially or completely ingested by a macrophage. Partial phagocytosis is most common and leads to formation of spherocytes. After a portion of adherent RBC membrane is internalized by the macrophage, the remainder of the RBC escapes back into the circulation. Since membrane is lost in excess of volume, the escaped RBC assumes a spherical shape, a sphere being the geometric shape with the lowest ratio of surface area to volume.4951 Spherical RBCs, more rigid and less deformable than normal RBCs, are subject to further fragmentation and destruction in future passages through the spleen. Spherocytosis is consistent and diagnostically important in AHA.52 The degree of spherocytosis correlates with the severity of hemolysis.11

C.H. Packman Although many warm autoantibodies x complement, intravascular complement-mediated hemolysis and hemoglobinuria are unusual in warmantibody AHA. The failure of complement-coated RBCs to be hemolyzed by the terminal complement cascade (C5C9) has been attributed in part to the ability of complement regulatory proteins in plasma and on the RBC surface to abort xation of terminal complement components.53

Clinical presentation of warm antibody AHA

History and physical ndings
Patients with warm-antibody AHA generally present with symptoms of anemia, but occasionally jaundice brings the patient to the physician. The onset of symptoms is often slow and insidious, perhaps over several months, but some patients experience rapid onset of symptoms of severe anemia and jaundice over a few days. Dark urine is uncommon because the bilirubin elevation is mostly of the indirect type and hemolysis is usually not rapid enough to cause hemoglobinuria. In patients with secondary AHA, the underlying disease often dominates the hemolytic anemia and its associated features. The physical examination is often normal in patients with only mild anemia, particularly if there is no underlying disese. Mild to moderate splenomegaly is common in patients with more severe hemolytic anemia. In severe cases with acute onset, ndings may include fever, pallor, jaundice, hepatosplenomegaly, tachypnea, tachycardia, angina, or heart failure.

Laboratory features
Patients with warm antibody AHA are anemic. The degree of anemia may be mild or life-threatening. Some patients with a near-normal hematocrit have only reticulocytosis as a clue to the diagnosis. Occasional patients have leucopenia, neutropenia or thrombocytopenia,11,54 although a normal platelet count and mild leukocytosis and neutrophilia are the rule. The constellation of AHA and thrombocytopenia is termed Evans syndrome.55,56 The blood lm contains important diagnostic hallmarks of AHA. Polychromasia is the morphologic correlate of reticulocytosis, an indicator of increased prevalence of newly formed RBCs in the blood. The presence of spherocytes is constant in patients with moderate or severe AHA (Fig. 1). Fragmented and nucleated RBCs, and occasionally

Hemolytic anemia due to warm autoantibodies

21 bilirubin is increased. Serum haptoglobin levels are low and LDH levels are elevated. Hemoglobinuria is only rarely encountered in patients with warm antibody AHA; its presence should lead one to consider other forms of hemolytic anemia such as cold agglutinin disease, paroxysmal cold hemoglobinuria, paroxysmal nocturnal hemoglobinuria or drug-immune hemolytic anemia.

The direct antiglobulin test

The diagnosis of AHA depends on the demonstration of immunoglobulin and/or complement proteins bound to patient RBCs, usually by the DAT or Direct Coombs Test (Fig. 2, Panel A). A broad-spectrum antiglobulin reagent which contains antibodies directed against human immunoglobulin and complement proteins is used to screen the patients RBCs. If agglutination is noted, monospecic antisera with antibodies to IgG and complement respectively are used to dene the pattern of RBC sensitization. Monospecic antisera to IgM or IgA may also be used in selected cases. A positive DAT indicates that autoantibody and/or complement proteins are present on circulating RBCs in vivo. There are three possible major patterns of direct antiglobulin reaction in AHA: 1) RBCs coated with IgG alone, 2) RBCs coated with IgG plus complement components, and 3) RBCs coated with complement components alone.11,6163 Positive antiglobulin reactions with anti-IgA or anti-IgM are encountered less commonly, often in association with bound IgG and/or complement.6469 The diagnostic signicance of each of these major patterns is summarized in Table 2 and discussed below. Circulating in the blood of patients with warm antibody AHA, most of the autoantibody is bound to the patients RBCs and is detected by the DAT. Unbound autoantibody may also be detected in the plasma of these patients by means of the indirect antiglobulin test (IAT, Fig. 2, Panel B). In the IAT, the patients serum or plasma is rst incubated with normal donor RBCs. These donor RBCs, now coated with patient antibody, are then washed and tested for agglutination by antiglobulin serum. In the patient, RBC-bound autoantibody and free plasma autoantibody exist in a reversible dynamic equilibrium.70,71 The amount of autoantibody in plasma depends upon the amount of antibody present and the afnity of the antibody for RBC antigens. In general, the more autoantibody present and the lower the binding afnity, the more likely a patient will exhibit plasma autoantibody. Patients with a positive IAT due to a warm-reactive

Figure 1 Panel A. Camera lucida drawing of spherocytes from a patient with hemolytic icterus and from a normal subject (Vanlair and Masius, 1871, cited in reference 1). Panel B. Microphotograph of blood smear made from a patient with autoimmune hemolytic anemia showing spherocytes (arrows).

erythrophagocytosis by monocytes may be seen in severe cases. Agglutination of RBCs is uncommon in warm antibody AHA, but it is frequently seen in cases mediated by cold agglutinins. Typically the reticulocyte count is elevated, but early in the presentation, as many as one-third of patients have transient reticulocytopenia in spite of normal or hyperplastic erythroid marrow.5760 More persistent reticulocytopenia is seen in patients with marrow function compromised by an underlying disease, parvovirus infection, toxic chemicals, or nutritional deciency. Such patients may develop severe anemia rapidly and early transfusion may be indicated. Although normally not necessary in the evaluation of AHA, marrow examination usually reveals erythroid hyperplasia and also may provide evidence of an underlying lymphoproliferative disorder. When present, unconjugated (indirect) hyperbilirubinemia is highly suggestive of hemolytic anemia. Total bilirubin is rarely increased above 5 mg/dl, and the conjugated (direct) fraction usually constitutes less than 15 percent of the total. Urinary urobilinogen is increased, but bilirubin is not detected in the urine unless serum conjugated

22

C.H. Packman

Figure 2 Panel A. Direct antiglobulin test. IgG-coated RBCs taken from a patient with warm antibody autoimmune hemolytic anemia are washed and suspended in buffer. Antiglobulin reagent containing antibody to IgG is added and the cells sedimented in a centrifuge. Visible agglutination occurs when anti-IgG crosslinks the IgG-coated RBCs, conrming that the cells are coated with IgG. RBCs coated with complement proteins may be similarly agglutinated with an antiglobulin reagent containing anti-complement antibody. Panel B. Indirect antiglobulin test. Serum from a patient with circulating free IgG antibody is incubated with RBCs from a potential donor. The cells are washed, resuspended in buffer, and further incubated with antiglobulin reagent containing antibody to IgG. The cells are sedimented in a centrifuge. Visible agglutination indicates that IgG antibody from the patient serum has bound to the RBCs.

Table 2 IgG alone

Major Reaction Patterns of the Direct Antiglobulin Test and Differential Diagnosis of Immune Injury. Differential Diagnosis 1) Warm antibody autoimmune hemolytic anemia 2) Drug-immune hemolytic anemia of the Hapten-drug adsorption type or autoantibody type 1) Warm antibody autoimmune hemolytic anemia with sub-threshold IgG deposition 2) Cold agglutinin disease 3) Paroxysmal cold hemoglobinuria 4) Drug-immune hemolytic anemia, Ternary complex type 1) Warm antibody autoimmune hemolytic anemia 2) Drug-immune hemolytic anemia: autoantibody type (rare)

Direct Antiglobulin Test Result

Complement alone

IgG plus complement

Hemolytic anemia due to warm autoantibodies autoantibody must also have a positive DAT. A patient with antibody in the plasma (positive IAT) but not on RBCs (negative DAT) does not have an autoimmune process, but more likely an alloantibody caused by prior transfusion or fetal to maternal transfer of RBCs during pregnancy. The intensity of the direct antiglobulin reaction correlates well with the number of IgG molecules bound per RBC as determined by sensitive quantitative techniques. A trace-positive antiglobulin reaction with anti-IgG or complement reagents detects 300400 molecules of IgG per cell and 60115 molecules of C3 per cell, respectively.72,73 More sensitive quantitation of RBC-bound IgG has allowed the identication of a subset of AHA patients who have the usual features of warm-antibody AHA but a negative DAT using standard reagents.7274 In many such patients, the RBCs are coated with IgG autoantibody in amounts below the sensitivity threshold for a positive DAT (subthreshold IgG). Specialized methods such as antiIgG consumption assays, automated enhanced agglutination techniques, enzyme-linked immunoassays and radioimmunoassays are able to detect small quantities of cell-bound IgG molecules which have been identied as warm-reacting anti-RBC autoantibodies.72 Such DAT-negative AHA patients generally have mild hemolysis and usually respond favorably to glucocorticoid therapy. Using the same methods, subthreshold IgG has also been detected in a patients who exhibit the complement alone pattern of direct antiglobulin reaction in the absence of drug sensitivity or cold agglutinins. In these cases, the subthreshold quantities of bound IgG antibodies were capable of xing clinically detectable quantities of C3 to the cell membrane.72 Among warm-antibody AHA patients, the strength of the antiglobulin reaction (IgG molecules per RBC) and the rate of RBC destruction correlate poorly. This is related in part to the subclass of IgG autoantibody. IgG1, the most commonly encountered subclass, and IgG3 autoantibodies appear to be more effective in decreasing RBC life span than do those of the IgG2 or IgG4 subclasses.64,75,76 The greater afnity of macrophage Fc receptors for IgG1 and IgG341,42 and the higher complement-xing activity of IgG1 and IgG3 antibodies relative to that of IgG2 or IgG4 antibodies53 partly explains this difference. Autoantibodies from patients with AHA typically bind to all the common types of human RBCs used in test panels by blood banks but these antibodies are not nonspecic. Rather, the autoantibodies from a given patient bind specically to one or more antigens which are common to almost all human RBCs, so called public antigens. About half

23 of all warm autoantibodies are specic for epitopes on Rh proteins.7,11,7780 These autoantibodies react with all human RBCs except those with the rare Rhnull phenotype which completely lack expression of the Rh complex. Some patients have anti-Rh autoantibodies with anti-e, anti-E, or anti-c (or, more rarely, anti-D) specicity; these patients nearly always have other autoantibodies reactive with all human RBCs, except Rhnull. These autoantibodies are designated collectively as Rh related.29,80 The major target of the Rh-related autoantibodies is a 32- to 34-kDa nonglycosylated polypeptide lacking on Rhnull RBCs.29,81 Other patients with warm-antibody AHA have IgG autoantibodies that are fully reactive with Rhnull RBCs.7,11,7780 Careful studies using RBCs of appropriate antigen-decient phenotype has allowed identication of autoantibody specicity for public blood group antigens outside the Rh system including anti-Wrb,77 anti-Ena,82 anti-LW,83 anti-U,84 anti-Ge,69,85 anti-Sc1,86 or antibodies to Kell blood group antigens.87 This group of autoantibodies is designated nonRh related.29,80 Autoantibodies with non-Rh serologic specicity react with the band 3 anion transporter29,88 or with both band 3 and glycophorin A.29 The latter autoantibodies may react with an epitope formed through the interaction of these two proteins on the RBC membrane.89

Diagnosis and differential diagnosis

The differential diagnosis of warm antibody AHA includes the other causes of spherocytic anemias. The most common of these and the one that resembles warm antibody AHA most closely is hereditary spherocytosis (HS). HS is often rst diagnosed in adulthood and patients generally exhibit splenomegaly on physical examination, in common with AHA. The family history is helpful in distinguishing the two, as patients with HS often have affected family members, whereas patients with AHA rarely do. The DAT is always negative in HS. Zieves syndrome, clostridial sepsis, and the hemolytic anemia that precedes Wilsons disease are other diseases characterized by spherocytic anemia. In patients with hemolytic anemia and a positive DAT, serologic characterization of the autoantibody will usually distinguish warm-antibody AHA from cold-reacting autoantibody syndromes, e.g., manifested by a high titer cold agglutinin with complement alone pattern in the DAT, or the presence of a cold hemolysin. In patients who have been transfused recently, a positive DAT reaction may reect the binding of a newly formed alloantibody to donor

24 RBCs in the patients circulation, indicative of a delayed transfusion reaction which can resemble an autoimmune process. Characterization of the antibody will determine whether it is directed against transfused or autologous RBCs. A patients medication list should be compared to published lists of drugs known to cause drugimmune hemolytic anemia90 Conrmation of drug-immune hemolysis requires demonstration that 1) patient antibody binds only to RBCs previously coated with drug in vitro (hapten/drug adsorption mechanism) or 2) patient antibody supports deposition of complement on RBCs only when drug, patient serum containing antibody, and a source of fresh complement are simultaneously present in the reaction mixture with RBCs (ternary complex mechanism). The diagnosis is further bolstered when discontinuation of the drug is followed in a few days by cessation of hemolysis. In other acquired types of hemolytic anemia, spherocytes are not prominent on the blood lm and the DAT is negative, making a diagnosis of AHA unlikely. Dark urine (hemoglobinuria) is seen frequently in patients with paroxysmal nocturnal hemoglobinuria (PNH) and in patients with cold-antibody hemolytic syndromes but rarely in patients with warm-antibody AHA. In PNH, the acidied serum test and the sucrose hemolysis test are usually positive and decreased levels of CD55 and CD59 on blood cells are frequently detected by ow analysis; in AHA, these tests are characteristically negative. Microangiopathic hemolytic disorders such as thrombotic thrombocytopenic purpura and hemolytic uremic syndrome are characterized by marked RBC fragmentation and minimal spherocytosis and can be distinguished from AHA by examining the blood lm. Thrombocytopenia, a frequent accompaniment of microangiopathic hemolytic anemia is uncommon in either warm- or cold-antibody AHA. A special type of autoimmune hemolysis may develop in recent recipients of allogeneic blood or marrow stem cell transplants or solid organ transplants.91 In the case of stem cell transplants, antibodies are produced by the stem cell graft against RBCs also produced by the stem cell graft, i.e., both antibodies and RBCs are of donor origin. In the case of solid organ transplants, the recipients own lymphocytes produce autoantibody against recipient RBCs. In both situations, it is thought that the autoimmunity arises from immunosuppressive therapy which causes delayed reconstition or dysfunction of T-cell immunity, leading to development of antibodies to RBCs autologous to the offended immune system.
Pathogenesis of antibody formation

C.H. Packman
Immunosuppressive therapy causes T-lymphocyte dysfunction/ dysregulation leading to autoantibody production Same Alloantibody production by donor Blymphocytes against recipient RBCs Persistent isoantibody in recipient plasma causes transient hemolysis of donor RBCs Mixed hematopoietic chimerism in which persistent recipient B lymphocytes make isoantibody against RBCs derived from the HSC graft RBCs made by HSC graft Recipient RBCs Recipeint RBCs Donor RBCs Persistent recipient iso-anti-A or B in plasma Blood Group O Blood Group A or B Recipient B lymphocytes Donor B-lymphocytes Donor B lymphocytes Any Blood Group A or B Any Blood Group O Blood Group A or B Blood Group O Any Persistent recipient B-lymphocytes RBCs made by HSC graft
HSC, hematopoietic stem cells.

Types of Immune RBC Destruction in Organ Transplant Recipients.

Donor Blood Group Donor Organ

Recipient Blood Group

Origin of antibodies

Target of antibodies

Any

Solid organ HSC or solid organ HSC

Table 3

HSC

HSC

Hemolytic anemia due to warm autoantibodies An alloimmune hemolytic anemia that imitates warm antibody AHA can develop in recipients of kidney, liver, or marrow transplants by several mechanisms. The most common setting is an organ from a blood group O donor transplanted into a blood group A recipient. In these cases, passenger B lymphocytes in the donated organ or marrow form alloantibodies against recipient RBCs.9296 In a different setting, a blood group O recipient of marrow from a blood group A or B donor may develop a transiently positive DAT and hemolysis of RBCs made by the marrow graft, due to temporary persistence of previously synthesized host anti-A or anti-B.97 Finally, group O marrow transplant recipients of a group A or B marrow graft may develop mixed hematopoietic chimerism; the recipients surviving B lymphocytes make alloantibodies against group A or B RBCs of donor origin made by the newly engrafted marrow.97 Since autoantibodies against the major blood group antigens A and B are extremely rare, in these settings, the nding of a positive DAT due to anti-A or anti-B is diagnostic of an alloimmune hemolytic process. The features of these types of transplant-related immune hemolysis are summarized in Table 3.

25 are more likely than the autoantibody to cause a severe hemolytic transfusion reaction. Patients with a history of pregnancy or prior transfusion are more likely to harbor alloantibodies.80,98100 Patients who have neither been pregnant nor exposed to blood are unlikely to have an alloantibody. It is important not to delay transfusion in a severely anemic symptomatic patient and it may be necessary to select the least incompatible blood unit for transfusion before the workup can be completed. Consultation between the clinician and the blood bank physician is useful. Once selected, the packed RBCs should be administered slowly. During the transfusion, the patient should be monitored for signs of a hemolytic transfusion reaction. The transfused cells may be destroyed as fast as the patients own cells or perhaps even faster. However, the increased oxygen-carrying capacity provided by the transfused cells may be sufcient to maintain the patient during the acute interval required for other modes of therapy to become effective. Oxygen therapy is also helpful by increasing dissolved oxygen in the blood.

Glucocorticoids

Management
Transfusion
The clinical severity of warm antibody AHA is related to the degree of the anemia and its rapidity of onset. Patients whose anemia has developed slowly are relatively asymptomatic and do not generally require RBC transfusions. On the other hand, patients who are symptomatic at rest or with minimal exertion, or those who have symptomatic cardiovascular disease should receive transfusions to maintain the hematocrit at a clinically acceptable level until other measures to stop the hemolysis can take effect. Transfusion of RBCs in AHA is complicated by the difculty of crossmatching and by the short survival of the transfused RBCs. Serocompatible donor blood is generally found only in those rare cases when the autoantibody is specic for a dened blood group antigen; most autoantibodies react with all potential donor RBCs. Thus, one must choose donor RBCs that are least incompatible with the patients serum in cross-match testing. It is also important to detect coexisting alloantibody in the patients serum, which can be difcult and time consuming. Alloantibodies

The effect of glucocorticoids on hemolysis in warm-antibody AHA is mediated by at least 2 mechanisms, i.e., decresed sequestration of RBCs in the spleen and decreased autoantibody synthesis. Glucocorticoids appear to suppress sequestration of RBCs by splenic macrophages.37,38,47,101 A decreased number of Fc c receptors is observed on blood monocytes of AHA patients during glucocorticoid therapy.102 This mechanism may explain the early improvement in hemolysis, often seen within one to three days after glucocorticoid therapy is started, well before any change in strength of the DAT is observed. Decreased strength of the DAT and IAT may be seen later in patients who have responded to glucocorticoids, suggesting that a decrease in synthesis of anti-RBC autoantibodies contributes to improved RBC survival as well.35,70 Treatment is usually initiated with oral prednisone, 60100 mg daily, but patients with very rapid hemolysis may require intravenous methylprednisolone, 100200 mg daily, often given in divided doses, sometimes for as long as 1014 days. Once the hematocrit begins to increase, the prednisone dose is decreased in a step-wise fashion to approximately 30 mg/day over about two weeks. As the hemolysis continues to slow, as indicated by

26 increasing hematocrit and decreasing reticulocyte count, the prednisone dose is decreased at a rate of 5 mg/day every week, to a dose of 1520 mg/ day. Prednisone should be continued at that dose for 23 months after the hemolysis has subsided. Thereafter the patient may be weaned from the drug over 12 months using a daily, early morning dose schedule. Alternatively, when a dose of 2030 mg daily is reached with persistent remission, it is reasonable to begin changing to an every-other-day schedule by decreasing the prednisone dose on alternate days over two to three months. Once the alternate day dose is eliminated, the prednisone may be further weaned over 23 months. The advantage of an alternate-day schedule is fewer glucocorticoid side effects. The switch to an alternate day prednisone schedule should not be attempted before the hematocrit has achieved stability on daily prednisone, 2030 mg. It is also helpful to follow the DAT periodically; therapy should not be discontinued while the DAT remains positive, even if the hematocrit and reticulocyte count are normal. Increasing hemolysis during weaning is managed by increasing the prednisone dose to the previous level at which hemolysis was controlled, followed by another, more gradual attempt at weaning. Glucocorticoids decrease the rate of hemolysis in about two-thirds of patients.7,11,61,103,104 About 20 percent of patients achieve a complete response manifest by cessation of hemolysis and about 10 percent have no response. Relapses are common after prednisone is discontinued, even for patients who achieve a complete response. Therefore, long-term followup is important for at least several years. Recurrent hemolysis may require repeat prednisone therapy, splenectomy, or immunosuppression.

C.H. Packman show no change in the DAT following splenectomy. Thus, the benecial effect of splenectomy seems to be related to several factors interacting in complex fashion.106 The results of splenectomy are variable. It has not been possible to predict which patients will respond using measurements of 51Cr-RBC sequestration in the spleen.11,103,107 Thus the patients clinical data dictate who should receive splenectomy and when. In general, it is reasonable to continue glucocorticoids for 12 months while waiting for a maximal response, but about onethird of patients require prednisone in doses greater than 15 mg/day to maintain an acceptable hemoglobin concentration. These patients are candidates for splenectomy. Splenectomy should be done sooner if the patients condition deteriorates, if the anemia is very severe or if the patient exhibits no response at all within 3 weeks of starting prednisone. Approximately two-thirds of AHA patients will have signicant slowing or cessation of hemolysis following splenectomy103,106,108,109 but recurrent hemolysis is frequent. Patients with persistent or recurrent hemolysis often respond to further prednisone therapy, often at a lower dose than they required prior to splenectomy.11,61,103 Alternate-day therapy is preferred to daily therapy if hemolysis can be controlled. Splenectomized subjects have an increased risk for overwhelming sepsis due to encapsulated organisms.110 The lifetime risk may be as high as 5%, greater in the rst two years after splenectomy.111 Vaccinations for pneumococcus, Haemophilus inuenzae type b and meningococcal serogroup C are recommended, preferably two weeks prior to surgery or alternatively 2 weeks after, with revaccination recommended for pneumococcus every 56 years.111,112 The use of prophylactic phenoxymethyl penicillin or erythromycin for adult patients is controversial; a recent British guidline suggests prophylaxis for life in these patients.111 Because the risk for fulminant sepsis and death is high, it is important for patients to start empiric antibiotics at the onset of any febrile illness and to seek medical attention immediately.

Splenectomy
Splenectomy removes the primary site of RBC trapping. When hemolysis persists after splenectomy, RBC destruction occurs primarily in the liver, mediated by hepatic Kupffer cells.35,37,40 High levels of RBC-bound IgG are required to maintain a given rate of hemolysis post-splenectomy, as much as 6-10 times higher than in nonsplenectomized subjects.35,37 The amount of RBC-bound autoantibody decreases in some patients following splenectomy,11,103,105 as evidenced by a decrease in the strength of the DAT, suggesting that one of the effects of splenectomy is to reduce the rate of autoantibody production. However, many patients

Rituximab
Rituximab, a monoclonal antibody directed against the CD20 antigen expressed on B-lymphocytes, has been used successfully in refractory AHA and in patients who are medically unsuited for splenectomy. It is presumed to eliminate

Hemolytic anemia due to warm autoantibodies B-lymphocytes that make autoantibodies to RBCs. The dose of rituximab, 375 mg/m2 weekly for 24 weeks, is similar to that used to treat B-cell lymphoma. In the largest series to date,113 13 of 15 children with warm antibody AHA responded to rituximab with cessation or signicant slowing of hemolysis. Over 30 other case reports support the use of rituximab in adults, although the actual response rate is not known, since patients who do not respond are unlikely to be reported.

27 cystitis in patients taking daily cyclophosphamide. Cyclosporine A has been used successfully in a few patients with refractory AHA.116 The dose is 510 mg per kg daily given in divided dose twice a day. Protracted therapy may be necessary to prevent recurrent AHA. Mycophenylate mofetil mediated a complete or partial response in several patients with AHA117,118; doses range from 5002000 mg per day given twice daily. Both agents are associated with signicant toxicity and their application should be carefully monitored by a physician experienced in their use.

Immunosuppressive drugs
Most patients with warm-antibody AHA respond sufciently well to prednisone and/or splenectomy that they are not candidates for immunosuppressive therapy. Because of the potential risks to the patient, immunosuppressive therapy should be considered mainly for patients refractory to prednisone, splenectomy and rituximab, or for those patients resistant to these medical therapies and unt for surgery.114 Cytotoxic drugs such as cyclophosphamide, 6-mercaptopurine, azathioprine, or 6-thioguanine have been used to treat refractory cases of AHA in an attempt to suppress autoantibody synthesis. Whether that rationale explains the effect of these drugs is not known, but benecial responses have been observed.7,114 In one small series, high dose cyclophosphamide was given at 50 mg/kg ideal body weight/day for 4 consecutive days.115 Of 9 patients, 8 of whom had warm autoantibodies, all became transfusion independent and 6 of the 9 experienced complete remission. All patients had prolonged severe cytopenias requiring G-CSF support and hospitalization for a median of 21 days. This very high dose of cyclophosphamide should be given only by physicians and institutions experienced in the management of profound cytopenias. Many patients with refractory AHA will not tolerate prolonged cytopenias. For these patients, oral cyclophosphamide 100 mg per day or azathioprine 100150 mg per day may be given. Since the response may be slow in coming, it is reasonable to continue treatment for up to 6 months before conceding defeat. Because both drugs suppress hematopoiesis, the complete blood count and reticulocyte count should be watched carefully during therapy. Both agents have been associated with risk of subsequent neoplasia, and cyclophosphamide may cause hemorrhagic cystitis. Maintenance of dilute urine by increasing oral uid intake may help prevent hemorrhagic

Miscellaneous treatments
Plasmapheresis has been successfully used in patients with warm-antibody AHA in a few reported cases.119,120 A case series reported short-term responses to high-dose intravenous c-globulin in 29 of 73 of patients with refractory warm antibody AHA.121 Responses to danazol, a nonvirilizing androgen, have been observed in patients with AHA.122,123 Danazol may eliminate the need for splenectomy when combined with prednisone in rst-line therapy and may allow for a shorter duration of prednisone therapy.123 In patients with ulcerative colitis and refractory AHA cessation of hemolysis has been noted following colectomy.124 Similarly, AHA associated with an ovarian dermoid cyst has resolved following removal of the cyst.125

Followup, complications and outlook

Patients with warm antibody AHA must be followed and managed indenitely as the course is capricious, characterized by relapses and remissions. Furthermore, patients with apparently primary AHA may later develop an underlying disorder such as lymphoma.126 In general, the outlook for secondary AHA parallels the course of the underlying disease. Pulmonary emboli, infection, and cardiovascular collapse are causes of death and deep-vein thrombosis and splenic infarcts are common complications during active hemolysis103,127 particularly in patients with antiphospholipid antibodies.128 Prophylactic anticoagulation may help prevent thrombotic complications in patients with AHA, active hemolysis and antiphospholipid antibodies129 or other risk factors for venous thromboembolism.

28

C.H. Packman
14. Shulman IA, Branch DR, Nelson JM, Thompson JC, Saxena S, Petz LD. Autoimmune hemolytic anemias with both cold and warm autoantibodies. JAMA 1985;253:17468. 15. Kajii E, Miura Y, Ikemoto S. Characterization of autoantibodies in mixed-type autoimmune hemolytic anemia. Vox Sang 1991;60:4552. 16. Berentsen S, Bo K, Shammas F, Myking A, Ulvestad E. Chronic cold agglutinin disease of the idiopathic type is a premalignant or low-grade malignant lymphoproliferative disease. APMIS 1997;105:35462. 17. Telen MJ, Roberts KB, Bartlett JA. HIV-associated autoimmune hemolytic anemia: Report of a case and review of the literature. J Acquir Immune Dec Syndr 1990;3:9337. 18. Rapoport AP, Rowe JM, McMican A. Life-threatening autoimmune hemolytic anemia in patient with acquired immune deciency syndrome. Transfusion 1988;28: 1901. 19. Saif M. HIV Associated Autoimmune Hemolytic Anemia: An Update. AIDS Patient Care 2001;15:21724. 20. Habibi B, Basty R, Chodez S, Prunat A. Thiopental-related immune hemolytic anemia and renal failure. N Engl J Med 1985;312:3535. 21. Habibi B. Drug-induced red blood cell autoantibodies codeveloped with drug-specic antibodies causing a hemolytic anaemia. Br J Haematol 1985;61:13943. 22. Sosler SD, Behzad O, Garratty G, Lee CL, Postanay N, Khomo O. Acute hemolytic anemia associated with a chlorpropamide-induced apparent auto-anti-Jka. Transfusion 1984;24:2069. 23. Salama A, Mueller-Eckhardt C. Rh blood group-specic antibodies in immune hemolytic anemia induced by nomifensine. Blood 1986;68:12858. 24. Salama A, Mueller-Eckhardt C. On the mechanisms of sensitization and attachment of antibodies to RBC in druginduced immune hemolytic anemia. Blood 1987;69: 100610. 25. Arndt P, Garratty G. Cross-Reactivity of cefotetan and ceftriaxone antibodies, associated with hemolytic anemia, with other cephalosporins and penicillin. Am J Clin Pathol 2002;118:2562002. 26. Weiss R, Freiman J, Kweder S, Diehl LF, Byrd JC. Hemolytic anemia after udarabine therapy for chronic lymphocytic leukemia. J Clin Oncol 1998;16:18859. 27. Chasty RC, Myint H, Oscier DG, Orchard JA, Bussutil DP, Hamon MD. Autoimmune haemolysis in patients with B-CLL treated with chlorodeoxyadenosine (CDA). Leuk Lymphoma 1998;29:3918. 28. Worlledge SM. Immune drug-induced haemolytic anaemias. Semin Hematol 1969;6:181200. 29. Leddy JP, Falany JL, Kissel GE, Passodor ST, Rosenfeld SI. Erythrocyte membrane proteins reactive with human (warm-reacting) anti-red cell autoantibodies. J Clin Invest 1993;91:167280. 30. Mollison PL. Measurement of survival and destruction of red cells in haemolytic syndromes. Br Med Bull 1959;15: 5967. 31. Crowley LV, Bouroncle BA. Studies on the specicity of autoantibodies in acquired hemolytic anemia. Blood 1956;11:7007. 32. Chaplin H, Cohen R, Bloomberg G, Kaplan HJ, Moore JA, Dorner I. Pregnancy and idiopathic autoimmune haemolytic anaemia: A prospective study during 6 months gestation and 3 months post-partum. Br J Haematol 1973;24:21929. 33. Mollison PL, Crome P, Hughes-Jones NC, Rochna E. Rate of removal from the circulation of red cells sensitized with

Practice points
 The direct antiglobulin test distinguishes immune-mediated RBC destruction from nonimmune types of hemolysis.  In the differential diagnosis of immune RBC destruction, consider medications. It is helpful to consult published lists of drugs known to cause drug-immune hemolytic anemia.  Transfusion of RBCs may be life sustaining until other measures take effect. Do not withhold transfusion from a patient with symptomatic AHA.  Prednisone is the mainstay of therapy, but patients who cannot be weaned may require splenectomy.  Rituximab and cyclophosphamide may be effective in patients who are either unt for or whose hemolysis is refractory to splenectomy.

References
1. Packman C. Historical Review: The Spherocytic Haemolytic Anaemias. Br J Haematol 2003;112:88899. 2. Coombs RRA, Mourant AE, Race EE. Detection of weak and incomplete Rh agglutinins: A new test. Lancet ii 1945;246:156. 3. Moreschi C. Neue tatsachen uber blutkorperchenagglutination. Zentralbl Bacteriol Parasitenkd Infektkr Originale 1908;46:4951. 4. Coombs RRA, Mourant AE, Race RR. A new test for the detection of weak and incomplete Rh agglutinins. British Journal of Experimental Pathology 1945;26:25566. 5. Boorman KE, Dodd BE, Loutit JF. Haemolytic icterus (acholuric jaundice), congenital and acquired. Lancet 1946;1:8124. 6. Loutit JF, Mollison PL. Haemolytic icterus (acholuric jaundice), congenital and acquired. J Pathol Bacteriol 1946;58:71128. 7. Petz LD, Garratty G. Acquired Immune Hemolytic Anemias. London: Churchill Livingstone; 1980. 8. Pirofsky B. Hereditary aspects of autoimmune hemolytic anemia: A retrospective analysis. Vox Sang 1968;14: 33447. 9. Dobbs CE. Familial auto-immune hemolytic anemia. Arch Intern Med 1965;116:2736. 10. Cordova MS, Baez-Villasenor J, Mendez JJ, Campos E. Acquired hemolytic anemia with positive antiglobulin (Coombs test) in mother and daughter. Arch Intern Med 1966;117:6925. 11. Dacie JV. The Haemolytic Anaemias. 3d ed. The Autoimmune Haemolytic Anaemias, vol. 3. New York: Churchill Livingstone; 1992. 12. Sokol RJ, Hewitt S, Stamps BK. Autoimmune haemolysis: An 18 year study of 865 cases referred to a regional transfusion centre. Br Med J 1981;282:20237. 13. Sokol RJ, Hewitt S, Stamps BK. Autoimmune haemolysis: Mixed warm and cold antibody type. Acta Haematol 1983;69:26674.

Hemolytic anemia due to warm autoantibodies

different amounts of antibody. Br J Haematol 1965;11: 46170. Mollison PL, Hughes-Jones NC. Clearance of Rh-positive red cells by low concentration of Rh antibody. Immunology 1967;12:6373. Rosse WF. Quantitative immunology of immune hemolytic anemia: II. The relationship of cell-bound antibody to hemolysis and the effect of treatment. J Clin Invest 1971;50:73443. Schreiber AD, Frank MM. Role of antibody and complement in the immune clearance and destruction of erythrocytes: I. In vivo effects of IgG and IgM complement-xing sites. J Clin Invest 1972;51:57582. Atkinson JP, Schreiber AD, Frank MM. Effects of corticosteroids and splenectomy on the immune clearance and destruction of erythrocytes. J Clin Invest 1973;52: 150917. Atkinson JP, Frank MM. Complement independent clearance of IgG sensitized erythrocytes: Inhibition by cortisone. Blood 1974;44:62937. Jandl JH, Richardson-Jones A, Castle WB. The destruction of red cells by antibodies in man: I. Observations on the sequestration and lysis of red cells altered by immune mechanisms. J Clin Invest 1957;36:142859. Jandl JH, Kaplan ME. The destruction of red cells by antibodies in man: III. Quantitative factors inuencing the pattern of hemolysis in vivo. J Clin Invest 1960;39: 114556. Anderson CL, Looney RJ. Human leukocyte IgG Fc receptors. Immunol Today 1986;7:264. Ravetch JV, Kinet J-P. Fc receptors. Annu Rev Immunol 1991;9:45792. Gigli I, Nelson RA. Complement-dependent immune phagocytosis: I. Requirements of C1, C4, C2, C3. Exp Cell Res 1968;51:4567. Lay WF, Nussenzweig V. Receptors for complement on leukocytes. J Exp Med 1968;128:9911009. Ross GD. Opsonization and membrane complement receptors. In: Ross GD, editor. Immunobiology of the Complement System. Orlando: Academic Press; 1986. p. 87. Fischer JT, Petz LD, Garratty G, Cooper NR. Correlations between quantitative assay of red cell bound C3, serologic reactions, and hemolytic anemia. Blood 1974;44: 35973. Schreiber AD, Parsons J, McDermott P, Cooper RA. Effect of corticosteroids on the human monocyte IgG and complement receptors. J Clin Invest 1975;56:118997. Ehlenberger AG, Nussenzweig V. The role of membrane receptors for C3b and C3d in phagocytosis. J Exp Med 1977;145:35771. Abramson N, LoBuglio AF, Jandl JH, Cotran RS. The interaction between human monocytes and red cells: Binding characteristics. J Exp Med 1970;132:1191206. LoBuglio AF, Cotran RS, Jandl JH. Red cells coated with immunoglobulin G: Binding and sphering by mononuclear cells in man. Science 1967;158:15825. Rosse WF, de Boiseury A, Bessis M. The interaction of phagocytic cells and red cells modied by immune reactions: Comparison of antibody and complement coated red cells. Blood Cells 1975;1:34558. Dameshek W, Schwartz SO. Acute hemolytic anemia (acquired hemolytic icterus, acute type). Medicine 1940;19:231327. Leddy JP, Rosenfeld SI. Role of complement in hemolytic anemia and thrombocytopenia. In: Ross GD, editor. Immunobiology of the Complement System. Orlando: Academic Press; 1986. p. 213.

29
54. Evans RS, Duane RT. Acquired hemolytic anemia: I. The relation of erythrocyte antibody production to activity of the disease. II. The signicance of thrombocytopenia and leukopenia. Blood 1949;4:1196213. 55. Evans RS, Takahashi K, Duane RT, Payner R, Liu C. Primary thrombocytopenic purpura and acquired hemolytic anemia: Evidence for a common etiology. Arch Intern Med 1951;87:4865. 56. Pegels JG, Helmerhorst FM, vanLeeuwen EF, van de Plasvan Dalen C, Engelfriet CP, von dem Borne AE. The Evans syndrome: Characterization of the responsible autoantibodies. Br J Haematol 1982;51:44550. 57. Liesveld JL, Rowe JM, Lichtman MA. Variability of the erythropoietic response in autoimmune hemolytic anemia: Analysis of 109 cases. Blood 1987;69:8206. 58. Hegde UM, Gordon-Smith EC, Worlledge SM. Reticulocytopenia and absence of red cell autoantibodies in immune haemolytic anaemia. Br Med J 1977;2:14447. 59. Conley CL, Lippman SM, Ness P. Autoimmune hemolytic anemia with reticulocytopenia: A medical emergency. JAMA 1980;244:168890. 60. Greenberg J, Curtis-Cohen M, Gill FM, Cohen A. Prolonged reticulocytopenia in autoimmune hemolytic anemia of childhood. J Pediatr 1980;97:7846. 61. Eyster ME, Jenkins Jr DE. Erythrocyte coating substances in patients with positive direct antiglobulin reactions: Correlation of cG globulin and complement coating with underlying diseases, overt hemolysis and response to therapy. Am J Med 1969;46:36071. 62. Leddy JP. Immunological aspects of red cell injury in man. Semin Hematol 1966;l 3:4873. 63. Engelfriet CP, von dem Borne AEG Kr, Vander Giessen D, Beckers D, van Loghem JJ. Autoimmune haemolytic anaemias: I. Serological studies with pure antiimmunoglobulin reagents. Clin Exp Immunol 1968;3: 60514. 64. Engelfriet CP, von dem Borne AEG Kr, Beckers D, van Loghem JJ. Autoimmune haemolytic anaemia: Serological and immunochemical characteristics of the autoantibodies: Mechanisms of cell destruction. Ser Haematol 1974;7: 32847. 65. Suzuki S, Amano T, Mitsunaga M, Yagyu F, Ofuji T. Autoimmune hemolytic anemia associated with IgA autoantibody. Clin Immunol Immunopathol 1981;21: 24756. 66. Wolf CF, Wolf DJ, Peterson P, Brandstetter RD, Hansen DE. Autoimmune hemolytic anemia with predominance of IgA autoantibody. Transfusion 1982;22:23840. 67. Szymanski IO, Teno R, Rybak ME. Hemolytic anemia due to a mixture of low-titer IgG lambda and IgM lambda agglutinins reacting optimally at 22 C. Vox Sang 1986;51: 1126. 68. Reusser P, Osterwalder B, Burri H, Speck B. Autoimmune hemolytic anemia associated with IgA: Diagnostic and therapeutic aspects in a case with long-term follow-up. Acta Haematol 1987;77:536. 69. Go ttsche B, Salama A, Mueller-Eckhardt C. Autoimmune hemolytic anemia associated with an IgA autoanti-Gerbich. Vox Sang 1990;58:2114. 70. Evans RS, Bingham M, Boehni P. Autoimmune hemolytic disease: Antibody dissociation and activity. Arch Intern Med 1961;108:33852. 71. Evans RS, Bingham M, Turner E. Autoimmune hemolytic disease: Observations of serological reactions and disease activity. Ann NY Acad Sci 1965;124:42240. 72. Gilliland BC, Leddy JP, Vaughan JH. The detection of cellbound antibody on complement-coated human red cells. J Clin Invest 1970;49:898906.

34.

35.

36.

37.

38.

39.

40.

41. 42. 43.

44. 45.

46.

47.

48.

49.

50.

51.

52.

53.

30
73. Gilliland BC, Baxter E, Evans RS. Red cell antibodies in acquired hemolytic anemia with negative antiglobulin serum tests. N Engl J Med 1971;285:2526. 74. Gilliland BC. Coombs-negative immune hemolytic anemia. Semin Hematol 1976;13:26775. 75. Sokol RJ, Hewitt S, Booker DJ, Bailey A. Erythrocyte autoantibodies, subclasses of IgG and autoimmune haemolysis. Autoimmunity 1990;6:99104. 76. von dem Borne AE Kr, Beckers D, van der Meulen W, Engelfriet CP. IgG4 autoantibodies against erythrocytes, without increased hemolysis: A case report. Br J Haematol 1977;37:13744. 77. Issitt PD, Pavone BG, Goldnger D, Zwikder H, Issit CH, Tessel JA. Anti-Wrb and other autoantibodies responsible for positive direct antiglobulin test in 150 individuals. Br J Haematol 1976;34:518. 78. Weiner W, Vos GH. Serology of acquired hemolytic anemia. Blood 1963;22:60613. 79. Vos GH, Petz L, Funenberg HH. Specicity of acquired haemolytic anaemia autoantibodies and their serological characteristics. Br J Haematol 1970;19:5766. 80. Leddy JP, Peterson P, Yeaw MA, Bakemeier RF. Patterns of serologic specicity of human cG erythrocyte autoantibodies. J Immunol 1970;105:67786. 81. Barker RN, Casswell KM, Reid ME, Sokol RJ, Elson CJ. Identication of autoantigens in autoimmune haemolytic anaemia by a non-radioisotope immunoprecipitation method. Br J Haematol 1992;82:12632. 82. Bell CA, Zwicker H. Further studies on the relationship of anti-Ena and anti-Wrb in warm autoimmune hemolytic anemia. Transfusion 1978;18:5725. 83. Celano MJ, Levine P. Anti-LW specicity in autoimmune acquired hemolytic anemia. Transfusion 1967;7:2658. 84. Marsh WL, Reid ME, Scott EP. Autoantibodies of U blood group specicity in autoimmune haemolytic anaemia. Br J Haematol 1972;22:6259. 85. Shulman IA, Vengelen-Tyler V, Thompson JC, Nelson JM, Chen DC. Autoanti-Ge associated with severe autoimmune hemolytic anemia. Vox Sang 1990;59:2324. 86. Owen I, Chowdhury V, Reid ME, Poole J, Marsh JC, Hows JM. Autoimmune hemolytic anemia associated with antiSc1. Transfusion 1992;32:1736. 87. Marsh WL, Oyen R, Alicea E, Linter M, Horton S. Autoimmune hemolytic anemia and the Kell blood groups. Am J Hematol 1979;7:15562. 88. Victoria EJ, Pierce SW, Branks MJ, Masouredis SP. IgG red blood cell autoantibodies in autoimmune hemolytic anemia bind to epitopes on red blood cell membrane band 3 glycoprotein. J Lab Clin Med 1990;115: 7488. 89. Telen MJ, Chasis JA. Relationship of the human erythrocyte Wrb antigen to an interaction between glycophorin A and band 3. Blood 1990;76:8428. 90. Lichtman MA, Beutler E, Kipps T, Seligsohn U, Kaushansky K. Prchal. In: Williams Hematology. New York: McGraw Hill; 2005. 91. Sokol R, Stamps R, Booker D, Scott FM, Ladlaw ST, Vandenberghe EA. Posttransplant immune-mediated hemolysis. Transfusion 2002;42:198204. 92. Lundgren G, Asaba H, Bergstro m J, Groth CG, Magnusson G, Moller E. Fulminating anti-A autoimmune hemolysis with anuria in a renal transplant recipient: A therapeutic role of plasma exchange. Clin Nephrol 1981;16:2114. 93. Ramsey G, Nusbacher J, Starzl TE, Lindsay GD. Isohemagglutinins of graft origin after ABO-unmatched liver transplantation. N Engl J Med 1984;311:116770.

C.H. Packman
94. Mangal AK, Growe GH, Sinclair M, Stillwell GF, Reeve SC, Naiman SC. Acquired hemolytic anemia due to auto-anti-a or auto-anti-b induced by group O homograft in renal transplant recipients. Transfusion 1984;24:2015. 95. Hazlehurst GR, Brenner MK, Wimperis JZ, Knowles Sm, Prentice HG. Haemolysis after T-cell depleted bone marrow transplantation involving minor ABO incompatibility. Scand J Haematol 1986;37:13. 96. Solheim BG, Albrechtsen D, Egeland T, Flatmark A, Fauchald P, Froysaker T. Auto-antibodies against erythrocytes in transplant patients produced by donor lymphocytes. Transplant Proc 1987;6:45201. 97. Sniecinski IJ, Oien L, Petz LD, Blume KG. Immunohematologic consequences of major ABO-mismatched bone marrow transplantation. Transplantation 1988;45:5304. 98. Issitt PD. Autoimmune hemolytic anemia and cold hemagglutinin disease: Clinical disease and laboratory ndings. Prog Clin Pathol 1978;7:13763. 99. Wallhermfechtel MA, Pohl BA, Chaplin H. Alloimmunization in patients with warm autoantibodies. A retrospective study employing three donor alloabsorptions to aid in antibody detection. Transfusion 1984;24:4825. 100. Branch DR, Petz LD. Detecting alloantibodies in patients with autoantibodies. Transfusion 1999;39:610. 101. Greendyke RM, Bradley EB, Swisher SN. Studies of the effects of administration of ACTH and adrenal corticosteroids on erythrophagocytosis. J Clin Invest 1965;44: 74653. 102. Fries LF, Brickman CM, Frank MM. Monocyte receptors for the Fc portion of IgG increase in number in autoimmune hemolytic anemia and other hemolytic states and are decreased by glucocorticoid therapy. J Immunol 1983;131:12405. 103. Allgood JW, Chaplin HJ. Idiopathic acquired autoimmune hemolytic anemia: A review of forty-seven cases treated from 1955 to 1965. Am J Med 1967;43:25473. 104. Meyer O, Stahl D, Beckhove P, Huhn D, Salama A. Pulsed high-dose dexamethasone in chronic autoimmune haemolytic anaemia of warm type. Br J Haemotal 1997;98: 8602. 105. Habibi B, Homberg JC, Schaison G, Salmon C. Autoimmune hemolytic anemia in children: A review of 80 cases. Am J Med 1974;56:619. 106. Christensen BE. The pattern of erythrocyte sequestration in immunohaemolysis: Effects of prednisone treatment and splenectomy. Scand J Haematol 1973;10:1209. 107. Parker AC, MacPherson AIS, Richmond J. Value of radiochromium investigation in autoimmune haemolytic anemia. Br Med J 1977;1:2089. 108. Bowdler AJ. The role of the spleen and splenectomy in autoimmune hemolytic disease. Semin Hematol 1976;13: 33548. 109. Schwartz SI, Bernard RP, Adams JT, Bauman AW. Splenectomy for hematologic disorders. Arch Surg 1970;101: 33847. 110. Eichner ER. Splenic function: Normal, too much and too little. Am J Med 1979;66:311320. 111. Newland A, Provan D. Preventing severe infection after splenectomy. Brit Med J 2005;331:207. 112. Centers for Disease Control and Prevention: Recommended adult immunization schedule-United States, 2003-2004. MMWR 2003; 52:965969. 113. Zecca M, Nobili B, Ramenghi U, Perrotta S, Amendola G, Rosito P. Rituximab for the treatment of refractory autoimmune hemolytic anemia in children. Blood. 2003;101:385761.

Hemolytic anemia due to warm autoantibodies

114. Murphy S, LoBuglio AF. Drug therapy of autoimmune hemolytic anemia. Semin Hematol 1976;13:32334. 115. Moyo VM, Smith D, Brodsky I, Crilley P, Jones RJ, Brodsky RA. High-dose cyclophosphamide for refractory autoimmune hemolytic anemia. Blood 2002;100:7046. 116. Emilia C, Messora C, Longo G, Bertesi M. Long-term salvage treatment by cyclosporine in refractory autoimmune haematological disorders. Br J Haematol 1996;93:3414. 117. Howard J, Hoffbrand AV, Prentice HG, Mehta A. Mycophenolate mofetil for the treatment of refractory autoimmune haemolytic anaemia and auto-immune thrombocytopenic purpura. Br J Haematol 2002;117:712715,. 118. Fibich C, Schober C, Schmoll H-J. Treatment of refractory autoimmune hemolytic anemias with mycophenolatemofetil (Cellcept) [abstract}. Br J Haematol 1998;102:39. 119. Shumak KH, Rock GA. Therapeutic plasma exchange. N Engl J Med 1984;310:76271. 120. Council Report: Current status of therapeutic plasmapheresis and related techniques. JAMA 1985; 253:819825. 121. Flores G, Cunningham-Rundles C, Newland AC, Bussel JB. Efcacy of intravenous immunoglobulin in the treatment of autoimmune hemolytic anemia: Results in 73 patients. Am J Hematol 1993;44:23742. 122. Ahn YS, Harrington WJ, Mylvaganam R, Ayub j, Pall LM. Danazol therapy for autoimmune hemolytic anemia. Ann Intern Med 1985;102:298301.

31
123. Pignon J-M, Poirson E, Rochant H. Danazol in autoimmune haemolytic anaemia. Br J Haematol 1993;83: 3435. 124. Giannadaki E, Potamianos S, Roussomoustakaki M, Kyriakou D, Fragkiadakis SN, Manousos ON. Autoimmune hemolytic anemia and positive Coombs test associated with ulcerative colitis. Am J Gastroenterol. 1997;92:18724. 125. Cobo F, Pereira A, Nomdedeu B, Gallart T, Ordi J, Torne A. Ovarian dermoid cyst- associated autoimmune hemolytic anemia. Am J Clin Pathol. 1996;105:56771. 126. Sallah S, Wan J, Hanrahan L. Future development of lymphoproliferative disorders in patients with autoimmune hemolytic anemia. Clin Cancer Res 2001;7: 7914. 127. Dausset J, Colombani J. The serology and the prognosis of 128 cases of autoimmune hemolytic anemia. Blood 1959;14:1280301. 128. Pullarkat V, Ngo M, Iqbal S, Espina B, Liebman HA. Detection of lupus anticoagulant identies patients with autoimmune haemolytic anaemia at increased risk of venous thromboemoblism. Brit J Haematol. 2002;118: 11669. 129. Hendrick AM. Auto-immune haemolytic anaemia a highrisk disorder for thromboembolism? Hematology 2003;8: 536.

Available online at www.sciencedirect.com