Treatment of Autoimmune Hemolytic Anemia Karen E. King and Paul M. Ness The appropriate therapy of autoimmune hemolytic anemia (AIHA) is dependent on the correct diagnosis and classification of this family of hemolytic disorders. Although the majority of cases are warm AIHA, there are several distinct types of cold AIHA and a number of drug-induced etiologies of AIHA, which must be investigated to determine if stopping a drug will induce a remission. In warm AIHA, corticosteroids are standard, followed by consideration of splenectomy in recalcitrant cases. If steroids and splenectomy are insufficient, other forms of immunosuppressive therapy are typically initiated. In cold AIHA, keeping the patient warm in often sufficient, but therapy directed at an underlying lympholiferative disorder may be helpful. Brisk hemolysis, inadequate responses to therapy, and worsening anemia require transfusion therapy. Although the pretransfusion workup is made difficult by the presence of the autoantibody, transfusion services can usually provide blood safe for transfusion by excluding underlying alloantibodies. When transfusion is urgently required and compatible blood cannot be located, incompatible blood may be provided as a life-saving measure. Communication between the transfusion service and the hematologist is critical to assess the risks in these settings. Hemoglobinbased oxygen carriers may provide an important bridging therapy in the future. Requests for “least incompatible” blood do not enhance transfusion safety and often result in unnecessary delays. Semin Hematol 42:131-136 © 2005 Elsevier Inc. All rights reserved.

A

utoimmune hemolytic anemia (AIHA) can be classified into two major types on the basis of the in vivo and in vitro characteristics of the causative autoantibody: those associated with warm antibodies, reacting optimally at 37°C, and with cold antibodies, reacting optimally at 0 –5°C. The AIHAs associated with cold antibodies can be further subdivided into the more common cold agglutinin syndrome and the rare paroxysmal cold hemoglobinuria (PCH). Both the warm and cold types of AIHA can be idiopathic, or secondary to infection or diseases such as leukemia or systemic lupus erythematosus. The diagnosis of AIHA is usually made on the basis of both clinical findings and laboratory results of a peripheral blood smear, the direct antiglobulin test (DAT), eluate studies, and indirect antiglobulin tests. A peripheral blood smear reveals anisocytosis with microspherocytes. Polychromatophilia is often seen, indicating bone marrow compensation. In cold agglutinin syndrome, agglutination may be prominent on the peripheral smear. The DAT using anti-IgG and anti-C3 (C3d) will detect the presence of IgG and/or complement coating

Transfusion Medicine Division, Johns Hopkins Medical Institutions, Baltimore, MD. Address correspondence to Paul M. Ness, MD, Transfusion Medicine Division, Johns Hopkins Medical Institutions, Baltimore, MD 21287-6667.

0037-1963/05/$-see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1053/j.seminhematol.2005.04.003

the red blood cells (RBCs). Eluate studies, testing the autoantibody dissociated from the patient’s RBCs against a panel of RBCs, typically reveal a panagglutinin. In the indirect antiglobulin test, the patient’s serum is tested against untreated and enzyme-treated RBCs at 20°C and 37°C; tests are read for hemolysis and agglutination. Such a battery of assays will delineate whether the antibody is an agglutinating and/or sensitizing antibody and define its hemolytic potential. Results can be combined with the clinical and hematologic findings to classify the patient. Sometimes more specific testing must be performed, such as a cold agglutinin titer and thermal range for cold agglutinin syndrome, a Donath Landsteiner test for PCH, adsorption procedures to differentiate alloantibody from autoantibody, serum and eluate studies investigating the specificity of the autoantibody, and assessment for drug-associated antibodies. These results allow classification into warm AIHA, cold agglutinin syndrome, or PCH (Table 1). Occasionally, unusual patients may not be classifiable into one of the three types of AIHA. Some have mixed-type AIHA, a combination of warm and cold type AIHA1; they may have the classical serologic features of both: IgG and complement on RBCs and a high-titer IgM cold antibody together with an IgG warm autoantibody in the serum, or IgG and C3 on the RBCs with low-titer IgM cold autoagglutinins of high thermal 131

Treatment of autoimmune hemolytic anemia

132

Table 1 Classification and Typical Serologic Features of the Autoimmune Hemolytic Anemias

Direct antiglobulin test Immunoglobulin class Eluate Serum

Specificity

Warm Autoimmune Hemolytic Anemia

Cold Agglutinin Syndrome

Paroxysmal Cold Hemoglobinuria

IgG, IgG and C3 IgG (sometimes IgA) IgG IgG antibody reacting with most, or all, RBCs at the antihuman globulin phase Rh

C3 only IgM Nonreactive IgM agglutinating antibody, often with titers >1,000, reacting at 30°C in albumin I, i

C3 only IgG Nonreactive IgG biphasic hemolysin (Donath-Landsteiner antibody)

range in the serum. Another group of unusual patients with hemolytic anemia has all the clinical and hematologic features of warm AIHA present with a negative DAT and no detectable serum antibodies— so-called “Coombs negative” AIHA.2 More sensitive assays, such as the enzyme-linked antiglobulin test or the radiolabeled antiglobulin test, may be necessary to identify RBC-bound IgG in these cases.

Therapeutic Approaches to AIHA Corticosteroids For warm AIHA, corticosteroids are the initial therapy of choice and the mainstay of therapy (Table 2). Many hematologists use a standard approach in adults, consisting of an initial course of prednisone, 60 to 100 mg per day for 1 to 3 weeks. There should be some evidence of clinical improvement within several days to 1 week. Approximately 80% of patients have a good initial response to corticosteroid therapy.3 After stabilization of hematologic parameters or the elimination of the requirement for transfusion support, the dosage of steroids may be gradually tapered. If a relapse occurs, the dose is often increased to produce another clinical remission. Most clinicians consider a maintenance dose of prednisone of greater than 15 mg per day to represent a therapeutic failure. To reduce the side effects of chronic steroid administration, many clinicians will attempt to treat the steroid-dependent patient on an alternate-day schedule.

Table 2 Therapy for Autoimmune Hemolytic Anemia Warm Autoimmune Hemolytic Anemia Corticosteroids

Cold Agglutinin Syndrome

Avoidance of cold Splenectomy Cytotoxic drugs Immunosuppression Rituximab Danazol Plasma exchange IVIg High-dose cyclophosphamide therapy

Paroxysmal Cold Hemoglobinuria Supportive care

P

The basis of the clinical response to corticosteroids is probably multifactorial. Steroids have been shown to have an early effect on tissue macrophages, so that IgG- and C3-coated RBCs are not cleared as rapidly by the reticuloendothelial system (RES); this activity can be demonstrated within the early days of therapy.4 Another mechanism by which corticosteroids act in AIHA is to alter antibody avidity.5 Several weeks of therapy are usually required before a third mechanism, decrease in antibody production, is observed. Among adults, permanent remission of AIHA occurs in approximately 20% to 35% of patients.6 Consequently, additional therapy is generally required for the initial presentation or the relapses that are common in warm AIHA.

Splenectomy Approximately 50% of patients will have a complete initial response to splenectomy, although patients may continue to require low doses of prednisone (⬍15 mg/d) to maintain adequate hemoglobin levels.7 Late relapses are not uncommon, presumably due to enhanced antibody synthesis and increased hepatic sequestration after the spleen, the primary location of RES destruction, is removed.8 Patients who have had a splenectomy need to be educated about the risks of overwhelming postsplenectomy sepsis syndrome. This entity is often due to infections with encapsulated bacteria. This is a medical emergency since patients may rapidly progress from an apparent flu-like illness to bacteremic shock. The risk of overwhelming postsplenectomy sepsis syndrome has been quantitated as 3.2% with a mortality rate of 1.4%.9 The risks of both infection and mortality can be reduced by the use of pneumococcal and meningococcal vaccines. Prophylactic antibiotic regimens are controversial; however, many advocate the use of penicillin, or trimethoprim-sulfamethoxazole as an alternative. Febrile illnesses in these patients must be given prompt attention and antibiotics should be started rapidly.

Other Immunosuppressive Therapy Several immunosuppressive agents have been reported to be successful in the treatment of warm AIHA, but in a literature of case reports and small series of patients. The unfortunate paucity of randomized therapeutic trials makes it difficult to determine which immunosuppressive drugs and what doses should be employed for patients who have failed initial ther-

K.E. King and P.M. Ness apy with corticosteroids and splenectomy. Immunosuppressive regimens should be considered when conventional therapy has failed or cannot be tolerated as in (1) the patient who fails to respond to splenectomy or relapses after the spleen has been removed; (2) when splenectomy represents an unacceptable medical risk; and (3) for the patient who cannot tolerate corticosteroid therapy. Azathioprine, cyclophosphamide, and other single agents have been effective in warm AIHA.10 However, their adverse effects can be limiting, including gastrointestinal intolerance, hemorrhagic cystitis, and bone marrow suppression. Several case reports describe the successful use of cyclosporine in warm AIHA,11 but results overall have been mixed. Cyclosporine has nephrotoxicity and must be used with caution in patients with renal failure and kidney function should be closely monitored in all patients. Mycophenolate mofetil is now frequently used to prevent kidney allograft rejection and to treat other autoimmune conditions. A few case reports show efficacy of this drug in warm AIHA, and mycophenolate is worthy of further evaluation in clinical trials.12 High-dose cyclophosphamide (50 mg/kg/d for 4 days) followed by granulocyte colony-stimulating factor has been effective in patients with refractory AIHA and other chronic autoimmune diseases that have not responded to less aggressive forms of conventional therapy.13 Early studies have achieved complete remission in the majority of patients and partial remission in others, but long-term evaluation will be required to assess cyclophosphamide’s role in warm AIHA. Rituximab is a genetically engineered chimeric murine/ human monoclonal anti-CD20 antibody which targets B-cell precursors and mature B cells. Plasma cells do not carry the CD20 antigen. The reported success of rituximab has not been limited to warm AIHA secondary to B-cell neoplasms. Rituximab has induced complete remissions in some patients with idiopathic warm AIHA, including in refractory disease and in children.14 The typical dosing regimen of rituximab for this indication is 375 mg/m2, once a week for 2 to 4 weeks. One series of 15 patients showed continuous remission in 10 patients, relapse after remission in three cases, and failure to respond in two others. Warm AIHA is similar in many aspects to the more common autoimmune thrombocytopenic purpura (ITP), and there are increasing different immunosuppressive therapies that are being evaluated for ITP. It is likely that single agents or therapeutic regimens combining drugs that prove useful in refractory ITP later will be evaluated in AIHA, but the opportunities for large evidence-based clinical trials of these agents will be limited by the lower number of AIHA patients.

Additional Therapies for Warm AIHA Danazol, an attenuated androgen, has been successful in the treatment of some cases of warm AIHA.15 In this clinical setting, its mechanism of action is uncertain. Because of the experience for intravenous immunoglobulin (IVIg) in ITP, it was predicted to be rewarding in warm AIHA; however, the literature is discouraging: there

133 are case reports of success and failure. The results are probably incomplete and potentially misleading because of the lack of controlled trials, use of IVIg in conjunction with multiple other agents, and its application to the most refractory cases. Plasmapheresis has a temporizing measure and it is not a practical approach for long-term management of AIHA. Plasmapheresis is often less efficient in treating diseases caused by IgG antibodies such as warm AIHA than for syndromes due to IgM antibodies because of the extravascular distribution of IgG antibodies and their rapid reaccumulation in the blood after mechanical removal.

Treatment of Cold Autoimmune Hemolytic Anemias The mainstay of therapy for cold agglutinin syndrome is avoidance of the cold (Table 2). Acute hemolytic crises can be prevented if patients maintain a high temperature in their indoor environment and wear additional clothing outdoors. By diligent avoidance of the cold, many patients with cold agglutinin syndrome can be managed without transfusions; mild, compensated anemia may not require treatment for prolonged periods of time or only episodic transfusions. For more severe, partially compensated, idiopathic cold agglutinin syndrome, medical therapy is generally unsatisfactory, and corticosteroids are not efficacious in most. Splenectomy is not usually effective because the liver is the dominant site of sequestration of RBCs heavily sensitized with C3. Chlorambucil has been used with some success but the associated hematopoietic suppression can be limiting.16 In patients without underlying hematologic malignancy, rituximab has been encouraging, but long-term evaluations of more patients are needed. One study treated 27 patients with primary chronic cold agglutinin disease with 37 courses of rituximab and achieved complete remission in one course of therapy, partial remission in 19, and no remission in 17 courses of therapy.17 These results are consistent with the pathophysiology of cold agglutinins typically associated with abnormal lymphocytic clones detected by flow cytometry, even when bone marrow morphology fails to disclose a lymphoproliferative disease. As with warm AIHA, plasmapheresis may be a temporizing measure. Plasmapheresis is more likely to be successful in cold agglutinin disease, since the typical antibody is IgM, approximately 80% of IgM is in the intravascular space, and it can be removed efficiently by plasmapheresis. Special attention may be required to maintain the core body temperature at 37°C, as with use of an in-line blood warmer. PCH is an unusual form of cold AIHA that most commonly occurs in children following a viral infection; therapy is generally supportive. Although many children may receive corticosteroids, they typically are already improving as the therapy is initiated. The efficacy of corticosteroids for these patients has not been established and their use cannot be confidently recommended.

134 Table 3 Serologic Problems in AIHA ● Alloantibody stimulated by previous transfusion or pregnancy can be masked by autoantibody ● All cross-matches may be incompatible due to autoantibody ● Positive DAT causes phenotyping problems

Transfusion Therapy for AIHA Serologic Difficulties in AIHA Transfusion therapy can be complicated in AIHA due to serologic complexities (Table 3). Although AIHA may present with an ABO and/or Rh typing discrepancy, accurate blood group identification usually can be determined without specialized techniques. In an emergency or if ABO results are not clear, group O donor RBCs are employed. Since the panagglutinin in the patient’s serum typically reacts with all donor RBCs, cross-matching blood may be a difficult and timeconsuming process. Patients with AIHA often are candidates for chronic transfusion programs, and knowledge of the patient’s RBC phenotype, including Rh (C, E, c, e), Kell, Duffy, and Kidd blood group antigens, may be useful. By dissociating autoantibody from the patient’s RBCs, phenotyping procedures can be performed. Aliquots of the patient’s untransfused RBCs can be frozen for extended phenotyping later. The phenotype can be used to guide the exclusion of alloantibodies by indicating which antigen specificities the patient is at risk of developing. The most important problem in a patient with a history of previous pregnancies or multiple transfusions is detecting and identifying an alloantibody that may be hidden by the autoantibody.18 Sophisticated immunohematology laboratories use a combination of techniques including differential allogeneic adsorption and autoadsorption, to detect underlying alloantibody. Another difficult issue is transfusion of patients whose autoantibody demonstrates a specificity. Some cases of warm AIHA can be shown to have Rh specificity, most often auto anti-e, but is is unclear if e negative blood transfusions are advantageous and result in improved erythrocyte survival, as compared to antigen-positive blood.19 In cases of cold agglutinin syndrome, the autoantibody often demonstrates specificity against I or i antigens, and auto anti-P is commonly associated with PCH. When compatible units cannot be located, many clinicians request “least incompatible” blood with the hopes that additional safety will be provided. Such requests only serve to delay needed blood transfusions and offer no additional protection to the patient.20 Of critical importance is that the clinician and the transfusion medicine physician cooperate in assessing and managing the transfusion needs in AIHA when anemia is severe and progressing. Many patients with AIHA require chronic transfusion support, but they may present occasionally with urgent transfusion needs when a complete serologic evaluation can not be performed. A recent study by Shirey et al21 evaluated an algorithm of providing antigen-matched donor blood for pa-

Treatment of autoimmune hemolytic anemia tients with warm autoantibodies: by determining the patient’s red cell phenotype, extensive serologic evaluation to exclude underlying alloantibodies can be circumvented, allowing blood to be safely and expeditiously provided.

Issues Complicating Transfusion Therapy Blood transfusion may be required when AIHA presents with fulminant hemolysis or for chronic hemolysis that becomes symptomatic before a response occurs to primary therapy. The presence of RBC autoantibody coating the patient’s RBCs or appearing as a panagglutinin in the serum contributes to the general risks of a blood transfusion (allergic and febrile nonhemolytic transfusion reactions and transfusion-transmitted infectious diseases). The autoantibody can make it difficult to detect coexisting RBC alloantibodies, which can cause severe hemolytic transfusion reactions; in addition, the RBC autoantibody itself can increase destruction of donor RBCs. Despite these concerns, transfusion may be required in AIHA and it can be life-saving (Table 4). Conley et al documented a series of patients with AIHA and reticulocytopenia despite intense erythroid hyperplasia in the bone marrow22: in all five cases, transfusions were required as a life-sustaining measure for profound anemia in patients transferred from other hospitals, where transfusions had been withheld because of serologic incompatibilities. Reticulocytopenia persisted from 4 to 160 days before the onset of adequate erythropoiesis or control of the hemolytic process with corticosteroids or splenectomy. This experience emphasizes that transfusion should not be withheld in AIHA, despite old dogma to the contrary, and in many cases transfusion requirements should be regarded as a medical emergency.22 The transfusion management of patients with AIHA requires careful communication between the clinician and the transfusion service. It is important to know the capabilities of the transfusion service for performing specialized immunohematologic procedures or where these procedures may be available on a referral basis. It is also critical to assess the history of pregnancies and previous transfusions with great care; it is unlikely, for example, that a previously untransfused male will have masked alloantibodies.

Table 4 Transfusion Therapy in AIHA ● If patient is clinically stable and responding to therapy, transfusions may not be required. ● Reticulocytopenia indicates high likelihood of early need for transfusion support. ● Incompatible blood may be required; “least incompatible” blood is not useful and may cause significant delay in the provision of urgently needed RBCs. ● Neurologic signs such as confusion suggest transfusion is urgently required. ● Leukocyte reduced donor RBCs are recommended to avoid febrile nonhemolytic transfusion reactions that may mimic a hemolytic reaction.

K.E. King and P.M. Ness

135

Clinical Issues With Transfusion

Transfusion in Cold Agglutinin Syndrome

Occasionally a transfusion may be required before serologic evaluation is completed; even after thorough serologic evaluation, the optimal blood for transfusion may be incompatible. The clinician must understand that in some circumstances, serologically incompatible blood is safe for transfusion and can be expected to have in vivo survival comparable to the patient’s own RBCs. Reluctance to transfuse patients due to serologic incompatibility or an incomplete evaluation can be devastating. Patients with severe anemia may appear to be hemodynamically stable, but they have life-threatening anemia and should be transfused immediately regardless of the serologic evaluation or compatibility. The onset of confusion in a patient with worsening anemia is a particular clinical indicator of an imminent transfusion requirement (Table 4). While it is difficult to generalize as to the appropriate transfusion trigger for a patient with severe AIHA, factors such as the rate of onset, presence or absence of accompanying hypovolemia, and, most important, the underlying health status and cardiorespiratory reserve must be considered. If the heart and lungs are healthy and there is no significant hypoperfusion, good tissue oxygenation can be maintained at subnormal hemoglobin levels. Experiments in acute, normovolemic anemia in rats,23 dog,24 and baboon25 have focused on the whole-body oxygen extraction ratio (ER): the heart is the major organ at risk. With progressive hemodilution, healthy animals with normal coronary vasculature maintained normal levels of oxygen consumption through a moderate increased cardiac output, coronary blood flow, and in the ER up to a ratio of 50%. As the hematocrit fell below 10%, however, oxygen consumption declined, and the animals were no longer able to increase the ER sufficiently to compensate for the low oxygen blood tension. An ER of 50% represents the critical point at which the myocardium converts from aerobic to anaerobic metabolism, as reflected in net lactate production. At this point metabolic acidosis ensues, resulting in hemodynamic instability. Some illustrative studies compared the response to acute normovolemic anemia in healthy dogs to the response in dogs with a critical coronary stenosis; both groups converted to anaerobic metabolism and developed congestive heart failure at an ER greater than 50%. Dogs with a critical coronary stenosis experienced an ER greater than 50% at a hematocrit of 17%, in contrast to the healthy dogs who experienced an ER greater than 50% at a hematocrit of 8.6%.26 Thus transfusions are indicated at a higher hematocrit in patients with AIHA who are elderly and likely to have underlying, undiagnosed cardiovascular disease. When the decision has been reached to transfuse, several additional steps can enhance patient safety. In some cases, transfusions of small aliquots of blood may be sufficient to provide relief of symptoms while avoiding the complications of fluid overload. Leukocyte-reduced RBCs are recommended to avoid febrile, nonhemolytic transfusion reactions, in which the initial presentation may mimic a hemolytic reaction with more severe consequences.

Similar to warm AIHA, cold agglutinin syndrome may have complex serology that complicates transfusion management. ABO discrepancies can make it difficult to determine the ABO group. Washing the patient’s RBCs with warm (37°C) normal saline to remove the IgM autoantibody can facilitate determination of an accurate ABO group. If the ABO typing is not clear from the serologic testing, group O RBCs can be administered. If transfusion is required for patients with cold agglutinin syndrome, blood warmers are often suggested despite limited data to support this recommendation.

Alternatives to Red Blood Cell Transfusion In recent years, numerous biotechnology products have been developed as alternatives to blood transfusion and progress in the development of blood substitutes as artificial oxygen carriers has accelerated. Although there have been substantial efforts to develop chemicals such as perfluorocarbons into clinically useful blood substitutes, hemoglobin based oxygen carriers (HBOC) appear more likely to provide an alternative to blood transfusion. Currently, the most promising HBOC preparations consist of extracted hemoglobin from lysed RBCs recovered from outdated human or bovine blood. These materials are chemically manipulated by polymerization or binding to macromolecules to increase in vivo survival and to decrease the potential for adverse reactions due to vasoconstriction. Case reports have demonstrated their successful use in AIHA and sickle cell anemia for patients refusing blood therapy.27 These products will have an important role in transfusion practice in the future. For patients with AIHA without available compatible blood, HBOC may become a lifesaving therapeutic bridge until compatible blood can be found.27

References 1. Shulman I, Branch DR, Nelson JM, Thompson JC, Saxena S, Petz LD: Autoimmune hemolytic anemia with both cold and warm autoantibodies. JAMA 253:1746-1748, 1985 2. Gilliland BC, Baxter E, Evans RS: Red cell antibodies in acquired hemolytic anemia with negative antiglobulin tests. N Engl J Med 285:252256, 1971 3. Allgood JW, Chaplin H: Idiopathic acquired autoimmune hemolytic anemia: A review of forty-seven cases treated from 1955 through 1965. Am J Med 43:254-273, 1967 4. Frank MM, Schreiber AD, Atkinson JP, Jaffe CJ: Pathophysiology of immune hemolytic anemia. Ann Intern Med 87:210-222, 1977 5. 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 50:734-743, 1971 6. Zupanska B, Sylwestrowicz T, Pawelski S: The results of prolonged treatment of autoimmune haemolytic anaemia. Haematologia 4:425433, 1981 7. Petz LD: Treatment of autoimmune hemolytic anemias. Curr Opin Hematol 8:411-416, 2001 8. Pirofsky B: Autoimmunization and the Autoimmune Hemolytic Anemias. Baltimore, MD, Williams & Wilkins, 1969 9. Bisharat N, Omari H, Lari I, Raz R: Risk of infection and death among post- splenectomy patients. J Infect 43:182-186, 2001 10. Worlledge S: Immune haemolytic anemias, in Hardesty RM, Weatherall DJ (eds): Blood and Its Disorders. Oxford, UK: Blackwell, 1974, pp 714-762

136 11. Emilia G, Messora C, Longo G, Bertesi M: Long-term salvage treatment by cyclosporine in refractory autoimmune haematological disorders. Br J Haematol 93:341-344, 1996 12. Howard J, Hoffbrand AV, Prentice HG, Mehta A: Mycophenolate mofetil for the treatment of refractory autoimmune haemolytic anaemia and auto-immune thrombocytopenia purpura. Br J Haematol 117:712715, 2002 13. Moyo VM, Smith D, Brodsky I, Crilley P, Jones RJ, Brodsky RA: Highdose cyclophosphamide for refractory autoimmune hemolytic anemia. Blood 100:704-706, 2002 14. Zecca M, Nobili B, Ramenghi U, Perrotta S, Amendola G, Rosito P, et al: Rituximab for the treatment of refractory autoimmune hemolytic anemia in children. Blood 101:3857-3861, 2003 15. Pignon JM, Poirson E, Rochant H: Danazol in autoimmune haemolytic anaemia. Br J Haematol 83:343-345, 1993 16. Worlledge SM, Brain MC, Cooper AC, Hobbs JR, Dacie JV: Immunosuppressive drugs in the treatment of autoimmune haemolytic anaemia. Proc R Soc Med 61:1312-1315, 1968 17. Berentsen S, Ulvestad E, Gjertsen BT, Hjorth-Hansen H, Langholm R, Knutsen H, et al: Rituximab for primary chronic cold agglutinin disease: A prospective study of 37 courses of therapy in 27 patients. Blood 103:2925-2928, 2004 18. Branch DR, Petz LD: Detecting alloantibodies in patients with autoantibodies. Transfusion 39:6-10, 1999

Treatment of autoimmune hemolytic anemia 19. Petz LD, Garratty G: Immune Hemolytic Anemias (ed 2). New York, NY, Churchill Livingstone, 2004 20. Petz LD: “Least incompatible” units for transfusion in autoimmune hemolytic anemia: Should we eliminate this meaningless term? A commentary for clinicians and transfusion medicine professionals. Transfusion 43:1503-1507, 2003 21. Shirey RS, Boyd JS, Parwani AV, Tanz WS, Ness PM, King KE: Prophylactically antigen-matched donor blood for patients with warm autoantibodies: An algorithm for transfusion management. Transfusion 42: 1435-1441, 2002 22. Conley CL, Lippman SM, Ness PM, Petz LD, Branch DR, Gallagher MT: Autoimmune hemolytic anemia with reticulocytopenia and erythroid marrow. N Engl J Med 306:281-286, 1982 23. Adams RP, Dielman LA, Cain SM: A critical value for O2 transport in the rat. J Appl Physiol 53:660-664, 1982 24. Cain SM: Oxygen delivery and uptake in dogs during anemic and hypoxic hypoxia. J Appl Physiol 42:228-234, 1977 25. Wilkerson DK, Rosen AL, Gould SA, Sehgal LR, Sehgal HL, Moss GS: Oxygen extraction ratio: A valid indicator of myocardial metabolism in anemia. J Surg Res 42:629-634, 1987 26. Wilkerson DK, Rosen AL, Sehgal LR, Gould SA, Sehgal HL, Moss GS: Limits of cardiac compensation in anemic baboons. Surgery 103:665670, 1988 27. Buetens OW, Ness PM: Red blood cell transfusion in autoimmune hemolytic anemia. Curr Opin Hematol 10:429-433, 2003