Hem 1-4

HEMATOLOGY BOARD REVIEW MANUAL
PUBLISHING STAFF
Immune Hemolytic Anemia
PRESIDENT, GROUP PUBLISHER
Series Editor and Contributing Author:
EXECUTIVE EDITOR
Richard S. Stein, MD, FACP
SENIOR EDITOR
Vanderbilt University Medical Center CONTRIBUTING EDITOR
ASSISTANT EDITOR
Contributing Author:
Anne T. Neff, MD
EDITORIAL ASSISTANT
Assistant Professor of Pathology and Medicine EXECUTIVE VICE PRESIDENT
Vanderbilt University Medical Center PRODUCTION DIRECTOR
PRODUCTION ASSOCIATES
Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
ADVERTISING/PROJECT MANAGER
Clinical Presentation. . . . . . . . . . . . . . . . . . . . . . . . . 2
Autoimmune Hemolytic Anemia . . . . . . . . . . . . . . . . 5
NOTE FROM THE PUBLISHER:
Management of Autoimmune Hemolytic Anemia . . . 7
This publication has been developed withoutinvolvement of or review by the American Drug-Related Autoimmune Hemolytic Anemia . . . . . 9
Paroxysmal Cold Hemoglobinuria . . . . . . . . . . . . . 10
Endorsed by the
Association for Hospital
Follow-Up of Case Patient . . . . . . . . . . . . . . . . . . . 11
Medical Education
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
The Association for Hospital Medical Educationendorses HOSPITAL PHYSICIAN for the pur- References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
pose of presenting the latest developments inmedical education as they affect residency pro-grams and clinical hospital practice.
Cover Illustration by Christine Schaar
Copyright 2001, Turner White Communications, Inc., 125 Strafford Avenue, Suite 220, Wayne, PA 19087-3391, www.turner-white.com. Allrights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means,mechanical, electronic, photocopying, recording, or otherwise, without the prior written permission of Turner White Communications, Inc.
The editors are solely responsible for selecting content. Although the editors take great care to ensure accuracy, Turner WhiteCommunications, Inc., will not be liable for any errors of omission or inaccuracies in this publication. Opinions expressed are those of theauthors and do not necessarily reflect those of Turner White Communications, Inc.
Hematology Volume 1, Part 4
HEMATOLOGY BOARD REVIEW MANUAL
Immune Hemolytic Anemia
Series Editor and Contributing Author:
Contributing Author:
Richard S. Stein, MD, FACP
Anne T. Neff, MD
Assistant Professor of Pathology and Medicine Vanderbilt University Medical Center Vanderbilt University Medical Center In discussing hemolytic anemias, this review will INTRODUCTION
focus specifically on autoimmune hemolytic anemia(AIHA), reviewing its definition, classifications, etiology, Hemolytic anemias are a diverse group of clinical dis- pathophysiology, clinical features, diagnosis, and man- orders characterized by decreased survival of erythro- agement. The case of a 34-year-old man with a hemo- cytes in the circulation. Because of their multiple causes, lytic anemia and spherocytosis will be presented to illus- hemolytic anemias are often difficult for hematologists, as well as internists and primary care physicians, to diag-nose. This clinical confusion can be lessened somewhatby following a logical, structured approach to diagnosis CLINICAL PRESENTATION
in patients suspected of having a hemolytic anemia.
First, it must be determined that hemolysis is actual- ly present and that the anemia is not caused instead by INITIAL EVALUATION OF CASE PATIENT
bleeding or bone marrow failure. The laboratory and A 34-year-old man is seen in the emergency depart- other tests most useful in determining whether or not ment reporting shortness of breath with exertion. He hemolysis exists are listed in Table 1. Next, the most
says that he has been unable to climb the stairs to his common causes of hemolysis (ie, acquired autoim- third-floor apartment during the past 2 weeks without mune hemolytic anemia, hypersplenism, congenital stopping several times; previously, he could climb to the hemoglobinopathies) should be ruled in or out, with third floor very easily. Medical history is unremarkable the specific cause identified, if possible. If a common except for arthroscopic knee surgery performed 2 years cause is not easily ascertained, a “checklist” approach to ago following a skiing injury. He takes no medicine other the diagnosis of hemolytic anemia can be pursued. The than an occasional nonsteroidal anti-inflammatory drug numerous specific causes of hemolysis for the most part to treat headaches and knee pain. Family history reveals fall into 1 of 4 general categories: (1) processes outside that his mother was treated for breast cancer 6 years ago; the erythrocyte, (2) alterations of the erythrocytic mem- his father and 2 siblings are in good health. The patient brane, (3) abnormalities in the hemoglobin molecule, does not smoke or use illicit drugs but does drink 2 to or (4) decreased levels of an enzyme in erythrocytes (Table 2).1 When evidence of one of the disorders that
Physical examination reveals a pale man in no dis- commonly cause hemolytic anemia cannot be found, tress at rest. Blood pressure is 140/76 mm Hg, pulse is this checklist is useful as a guide to other problems to 124 bpm, and respiratory rate is 22 breaths/min. He is afebrile. Nail beds and conjunctivae are pale; the sclerae Hospital Physician Board Review Manual
I m m u n e H e m o l y t i c A n e m i a are icteric. Auscultation of the heart reveals a grade II/VI Table 1. Tests for Determining the Presence of a
systolic ejection murmur at the apex; lungs are clear to auscultation. The liver and spleen are not palpable.
Petechiae, bruises, adenopathy, and edema are not pres- Laboratory measurements
ent, and there is no tenderness over the liver.
Results of laboratory testing show a hematocrit of 16%, a leukocyte count of 7.8 × 103/mm3 (with a normal differential), and a platelet count of 417 × 103/mm3.
Reticulocyte count is 9.3% of erythrocytes. Total serum bilirubin level is 7.2 mg/dL. Fractionation of the biliru-bin reveals a direct bilirubin level of 0.6 mg/dL, giving a calculated indirect bilirubin level of 6.6 mg/dL. Blood urea nitrogen level is within normal limits, as are serum Examination of the peripheral blood smear levels of electrolytes, creatinine, aspartate aminotrans-ferase (AST, SGOT), and alkaline phosphatase. Serumlactate dehydrogenase (LDH) level is 2900 U/L (nor-mal, 200 to 600 U/L). Serum haptoglobin level is with- often is more informative. The normal reticulocyte num- in normal limits. A peripheral blood smear reveals poly- ber is 60–90 × 103/mm3. In a patient with a hematocrit of 20% and an erythrocyte count of 2 × 106/mm3, an elevat-ed reticulocyte count of 3% represents a reticulocyte DETERMINING IF AN ANEMIA IS HEMOLYTIC
number of only 60 × 103/mm3. In cases of clinically sig-nificant hemolytic anemia, a reticulocyte number greater • What role do reticulocyte count, bilirubin level, and
than 150 × 103/mm3 is expected. The reticulocyte index other laboratory measurements play in determining
(which adjusts for the fact that reticulocytes spend more whether an anemia is hemolytic?
time in the circulation when there is marked hemolysis • How does a peripheral blood smear help in the
and a very high reticulocyte count) adds little to the deter- diagnosis of hemolytic anemia?
mination of whether anemia is or is not hemolytic.
The serum bilirubin level and bilirubin fractionation Diagnosis of any anemia involves examining a periph- value are helpful in establishing that hemolytic anemia eral blood smear and obtaining a reticulocyte count. In is present. Bilirubin is a breakdown product of erythro- general, when the reticulocyte count is elevated (ie, cytes, and its level will be elevated when these cells break greater than 1.5%), hemolytic anemia should be suspect- down in the bone marrow (as in the ineffective erythro- ed. However, an elevated reticulocyte count is not specific poiesis of myelodysplasia), in extravascular sites (as in for diagnosing hemolytic anemia but can occur, for exam- reabsorbed hematomas), or in the circulation (as in ple, after an episode of acute bleeding or after initial treat- hemolytic anemia). In all of these circumstances, the indirect or unconjugated bilirubin level is elevated. In deficiency. Even the rebound erythrocyte production fol- contrast, an elevated direct or conjugated bilirubin level lowing a normal menstrual period can be associated with occurs in liver disease as a result of the regurgitation of an elevated reticulocyte count in the range of 2% to 4%.
bilirubin into the circulation from injured hepatocytes Nevertheless, a reticulocyte count greater than 5% strong- (in cases of hepatitis) or from the bile ducts (in cases of ly suggests the presence of a hemolytic process.
An elevated reticulocyte count also is not totally sen- Unconjugated hyperbilirubinemia also can occur in sitive for diagnosing hemolytic anemia. Approximately Gilbert’s syndrome, a congenitally acquired nondisease 10% of patients with hereditary spherocytosis and up to state in which the activity of an enzyme responsible for 50% of patients with thalassemia minor might have conjugating bilirubin (ie, hepatic uridine diphosphate– reticulocyte counts within the normal range. The more glucuronosyltransferase) is decreased. Gilbert’s syn- severe the hemolytic anemia, however, the more likely drome is present in 5% of the population2 but is rarely it is that the reticulocyte count will be elevated; patients associated with bilirubin levels greater than 3.0 mg/dL.
with hemolytic anemia and reticulocyte counts within Because patients with Gilbert’s syndrome generally are normal limits generally have minimal anemia.
not anemic, differentiating hemolytic anemia from Although the reticulocyte count is reported as a per- Gilbert’s syndrome should not create a diagnostic centage of total erythrocytes, the reticulocyte number problem. However, patients with Gilbert’s syndrome Hematology Volume 1, Part 4
I m m u n e H e m o l y t i c A n e m i a Table 2. Specific Causes of Hemolytic Anemia*
Processes outside the erythrocyte
Lecithin–cholesterol acyltransferase deficiency† Abnormalities in the hemoglobin molecule
Hemoglobin E and other disorders associated with uneven production of α- and β-hemoglobin chains Decreased levels of an erythrocytic enzyme
Glucose-6-phosphate dehydrogenase deficiency (90% or more of all cases of enzyme-deficient hemolysis) Pyruvate kinase deficiency (most other cases of enzyme- Glutamylcysteine synthetase deficiency† Alterations of the erythrocytic membrane
*Categorized according to the site of the pathophysiologic process.
Adapted from Lee RL. Hemolytic disorders: general considerations. In: Lee GR, Foerster J, Lukens J, et al, editors. Wintrobe’s clinical hematol-ogy. 10th ed. Baltimore: Williams and Wilkins; 1999:1112.
sometimes develop anemia from other causes, and their duction by the marrow cannot exceed 6 times the nor- mal level. When a steady state has been achieved, Although fractionation of the bilirubin is the stan- destruction of erythrocytes also is no greater than dard means of documenting whether an elevated level 6 times the normal rate, so the indirect bilirubin is involves either direct or indirect bilirubin, a helpful unlikely to be more than 6 times the normal rate. In the clue can come from examining the urine. Only direct presence of chronic stable hemolysis combined with hepatic disease, the total serum bilirubin level can be In chronic hemolytic anemia in which a steady state markedly elevated, but the indirect serum bilirubin has been achieved, the total serum bilirubin level is level is rarely greater than 6 mg/dL. If massive acute rarely greater than 6 mg/dL, because erythrocyte pro- hemolysis occurs beyond the ability of the bone marrow 4 Hospital Physician Board Review Manual
I m m u n e H e m o l y t i c A n e m i a to compensate, significant elevations of indirect serum their names suggest, elliptocytes can be seen in the bilirubin levels (up to 15 mg/dL) can be seen. Patients peripheral blood smears of patients with hereditary with such massive hemolysis most likely would have an elliptocytosis, and sickled cells can be seen in the unstable and rapidly falling hematocrit.
peripheral blood smears of patients with sickle cell ane- Measurement of the serum LDH level is a sensitive mia. Whereas hemoglobin C usually is associated with but highly nonspecific screening test for hemolytic crystals of abnormal hemoglobin, such crystals are gen- anemia. Erythrocytes contain the electrophoretically erally not seen unless the patient has already had a fast-moving fraction LDH-1, which is also present in splenectomy; however, target cells are usually present in myocardial muscle fibers and renal cortex cells. Con- hemoglobin C disease. Finally, marked microcytosis, sequently, the total serum LDH level and especially anisocytosis, and poikilocytosis generally are seen in the the LDH-1 fraction are not specific for hemolysis.
peripheral blood smears of patients with thalassemia However, a total serum LDH level can be helpful in major (beyond what is seen in cases of iron deficiency); ruling out hemolytic anemia. Given that the total only minimal microcytosis usually is observed in cases of serum LDH level might be elevated because of other thalassemia minor, although anisocytosis and poikilo- LDH fractions (eg, hepatic LDH), LDH fractionation might be more helpful in confirming hemolysis; how- When tests suggest that hemolysis, if present, is min- ever, this procedure is rarely performed.
imal, the patient should be carefully evaluated for Serum haptoglobin is an α -globulin that acts as a occult bleeding. Occult gastrointestinal bleeding often scavenger protein and binds any hemoglobin released is an intermittent problem and cannot be ruled out by into the blood from intravascular hemolysis. This bind- a single pair of stool specimens that are negative for ing is an important protective mechanism, because free occult blood. Occult bleeding can indicate the pres- hemoglobin can precipitate in the kidneys and cause ence of a life-threatening disorder; it would be disas- renal failure, as can occur in mismatched transfusion trous, for example, to miss diagnosing a curable neo- reactions. Serum haptoglobin is less affected when plasm by embarking on a search for a phantom hemolysis occurs at extravascular sites, specifically in the Unfortunately, the serum haptoglobin test is of lim- ited value for 2 reasons. First, serum haptoglobin is an AUTOIMMUNE HEMOLYTIC ANEMIA
acute phase reactant and thus will be elevated inpatients with myocardial infarction, cancer, infection,and similar conditions. As such, the serum haptoglobin DEFINITION AND ETIOLOGY
level in these patients, although potentially decreased AIHA occurs when patients produce autoantibodies by hemolysis, might still be within the normal range.
that bind to erythrocytes, leading to their destruction Secondly, up to 30% of a transfused unit of blood can and a resultant anemia. As such, AIHA represents a fail- hemolyze intravascularly within 24 hours of transfusion ure of self-tolerance. However, the specific mechanism and thus lower the serum haptoglobin level. Therefore, by which self-tolerance fails in AIHA is not known.
if the serum haptoglobin test is performed after a Although many cases of AIHA are idiopathic, some of patient has received a transfusion, its results are of lim- the conditions associated with AIHA are autoimmune disorders (Table 3); the latter classification includes sys-
Examination of a patient’s peripheral blood smear temic lupus erythematosis and other disorders of the can be helpful in diagnosing hemolysis because several immune system (eg, chronic lymphocytic leukemia, hemolytic disorders are associated with abnormal ery- Waldenström’s macroglobulinemia, and other lym- throcytic morphology. Although reticulocytes them- selves can be detected only by special supravital stains, The syndromes of AIHA are generally classified on the the presence of an elevated reticulocyte count is associ- basis of the relationship between antibody activity and ated with polychromatophilia (ie, large dark cells) on temperature. Warm active antibodies are generally IgG routine peripheral blood smears. Spherocytes can be an molecules, which may or may not fix complement and indication of either hereditary spherocytosis or AIHA.
have the greatest affinity for erythrocytes at body temper- Microangiopathic changes can occur in hemolytic ane- ature (ie, 37°C [98.6°F]). Cold active antibodies are gen- mias associated with thrombotic thrombocytopenic erally IgM molecules, which fix complement and have the purpura, dysfunctioning artificial heart valves, or renal greatest affinity for erythrocytes between 0°C (32°F) and diseases (eg, transplant rejection, scleroderma). As 4°C (39.2°F). Because such low body temperatures are Hematology Volume 1, Part 4
I m m u n e H e m o l y t i c A n e m i a Table 3. Conditions Associated with Autoimmune
curs most readily in the sinuses of the spleen, where ery- throcytes are separated (relatively speaking) from plas-ma, leaving the antibody-coated erythrocyte more like- Warm antibody–mediated hemolysis
For patients with cold antibody–mediated hemoly- sis, complement on the cells leads to destruction any-where in the circulation that complement is cleared.
Quantitatively, this means that hepatic clearance is the Neoantigen formation (quinidine/stibophen) type dominant mode of clearance of erythrocytes in cold antibody–mediated hemolysis. Because warm antibod- ies also can fix complement, warm antibodies can be associated with hepatic destruction of erythrocytes as DIAGNOSTIC LABORATORY TESTING
• What tests best establish the diagnosis of autoim-
mune hemolytic anemia?
Cold antibody–mediated hemolysis
AIHA should be suspected in any patient with acquired hemolytic anemia, especially when the peripheral blood smear shows spherocytosis (suggest- ing warm antibody–mediated hemolysis) or rouleaux formation (suggesting cold antibody–mediated hemol- ysis). However, even in the absence of those findings on peripheral blood smears, specific testing for possi-ble AIHA is part of the evaluation of any patient with hemolytic anemia. The direct antiglobulin test (DAT), also known as the direct Coombs’ test, is designed to detect immunoglobulin and/or complement on the surface of erythrocytes.3 In this test, a patient’s ery- throcytes are first collected in EDTA to limit in vitro adherence of complement to the erythrocytes. Cellsare washed to eliminate any nonspecifically bound pro- teins and then mixed with antiglobulin serum ob-tained from rabbits. Either polyspecific antiglobulinserum or monospecific anti-IgG or -C3 antiserum is incompatible with life, the biologic effect of cold antibod- used. The cell-antiserum mixture is then centrifuged ies depends on how much activity is present at colder tem- peratures (eg, at temperatures of 22°C [71.6°F] to 30°C Although positive results on this test can give further [86°F], which actually can occur in the periphery on weight to the diagnosis of AIHA, the standard DAT is not extremely sensitive and requires 150 to 200 IgG mol- One reason to distinguish warm antibody–mediated ecules per cell to give positive results. Because autoim- from cold antibody–mediated hemolysis is that the mune hemolysis can occur with fewer molecules per mechanism of destruction of erythrocytes differs in cell, more sensitive techniques might be necessary when these 2 types of hemolysis, thus necessitating different AIHA is suspected and the standard DAT produces neg- treatments. For patients with warm antibody–mediated ative results.4 Unfortunately, these more sensitive tests hemolysis, the antigen-binding fragment (Fab) of the also can produce negative results, in which case the antibody attaches to antigens on the membrane of ery- diagnosis is established by exclusion of other causes and throcytes; cells are destroyed when the crystallizable by a response to immunotherapy. The mechanism of fragment (Fc) of the IgG molecule attaches to cells in DAT-negative immune hemolysis is not known, but such the monocyte-macrophage system. This scenario oc- cases might be mediated by low-titer antibodies that are 6 Hospital Physician Board Review Manual
I m m u n e H e m o l y t i c A n e m i a lost in the wash phase of the DAT. When the DAT pro- experience postinfectious cold antibody–mediated duces positive results, the antibody can be eluted off ery- AIHA after mycoplasmal infections or infectious mono- throcytes with acid or xylene, and the specificity of the antibody can be evaluated by means of the indirectantiglobulin test.
Whereas the DAT uses antiserum and patient cells, MANAGEMENT OF AUTOIMMUNE HEMOLYTIC
the indirect antiglobulin test uses the patient’s serum and a panel of erythrocytes to detect agglutination. Thistest is positive in approximately three fourths of patientswho have AIHA, and its use has shown that autoanti- WARM ANTIBODY–MEDIATED HEMOLYSIS
bodies can react against a number of erythrocytic anti- Therapy of warm antibody–mediated hemolysis gens. Failure of autoantibodies to react against very rare depends on the severity of the disease. Positive results Rh-null cells initially suggested that all autoantibodies on a DAT, if associated with a hematocrit that is within react against a basic component of the Rh system. How- normal limits and a slightly elevated reticulocyte count, ever, although the Rh antigen is the most common tar- do not warrant corticosteroid therapy. Treatment may get antigen for warm antibody–mediated AIHA, other be limited to folate replacement and close observation antigens (such as LW antigens, glycophorins A, B, C, to detect progressive anemia. Additionally, an evalua- and D, and rarely Kidd or Kell group antigens) also can tion seeking the underlying causes of the hemolysis is CLINICAL FEATURES
Corticosteroid Therapy
The clinical picture in warm antibody–mediated When the degree of hemolysis is such that anemia is AIHA, which represents the majority of cases of AIHA,6 present, corticosteroids are the initial therapy of can range from a minimal increase in hemolysis of no choice. The usual starting dose for prednisone is clinical significance to fatal fulminant hemolysis. Slow 1 mg/kg body weight per day, but higher doses can be onset of weakness, fatigue, and exertional dyspnea is used in the face of massive hemolysis. A response to common. In elderly patients unable to tolerate ane- corticosteroid therapy usually occurs within 4 to 7 days, mia, angina or even bowel infarction can occur. Even if and a slow increase in hemoglobin level of 2 to 3 g/dL an underlying lymphoproliferative disorder is not pres- per week is expected. Once the hemoglobin level ent, splenomegaly commonly is observed, and lym- reaches 10 g/dL, a slow tapering of the prednisone phadenopathy occasionally can be seen. The presence dose can begin. The generally recommended ap- of adenopathy and splenomegaly in association with proach is to taper to a dose of 0.5 mg/kg over a period AIHA must lead to a consideration of chronic lympho- of 4 to 6 weeks. In a 70-kg (154-lb) patient, this taper- cytic leukemia or lymphoma. Also, because many cases ing would represent a decrease from 70 mg/day to of AIHA are drug related, a careful history of recent 35 mg/day. Once the latter level of daily prednisone is drug use (including prescription, over-the-counter, achieved, very slow tapering of the daily dose is recom- mended. The usual aim is to decrease the daily dose to In cases of cold antibody–mediated AIHA, the 5 to 10 mg over a period of 3 to 4 months. Whether clinical picture is generally that of chronic hemolytic long-term maintenance at that level is necessary to pre- anemia but with exacerbations occurring during the vent relapse has not been established by clinical trials.
winter. Fulminant hemolysis is rare, and the hemoglo- Many physicians simply stop administering prednisone bin level is generally greater than 7 g/dL. Acrocyanosis after a slow taper to a daily dose of 5 mg and then fol- can occur from agglutination of cells in the hands, nose, feet, or ears. A mottled appearance of the limbs, The response rate to prednisone is excellent, with known as livedo reticularis, also may be present. As more than 80% of patients responding within a week occurs in warm antibody–mediated AIHA, splen- and almost 90% of patients showing some improve- omegaly also can be present. However, whenever more ment within 2 weeks of receiving the drug.7 Prednisone than minimal splenomegaly is present, lymphoma decreases hemolysis by 2 mechanisms. Its most immedi- and infectious mononucleosis–related cold antibody– ate effect is to decrease clearance of erythrocytes by the mediated hemolysis must be considered as possible monocyte-macrophage system. Animal studies have causes. Whereas cold antibody–mediated AIHA is most shown that prednisone is most effective in this regard commonly a disease of the elderly, young patients often when erythrocytes are covered with IgG alone. When Hematology Volume 1, Part 4
I m m u n e H e m o l y t i c A n e m i a both IgG and complement are present, prednisone is among patients with warm antibody–mediated AIHA less effective, and prednisone is least effective when only who are refractory to corticosteroids and splenectomy.7 complement is present on the erythrocytes.8 This latter No standard dose has been established, but it seems rea- fact is consistent with the limited efficacy of prednisone sonable to initiate cyclophosphamide at a dose of 1.5 to in cold antibody–mediated hemolysis. A second docu- 2.0 mg/kg per day or azathioprine at a dose of 2 mg/kg mented mechanism of action of prednisone involves per day, with or without prednisone (1 mg/kg per day).
decreased production of autoantibody. As the response When prednisone is included, its dose is usually tapered to prednisone occurs, results of the DAT generally over 3 months; the immunosuppressive drug is contin- remain positive, although the strength of the reaction ued for 6 months prior to dose reduction.10,11 decreases (eg, from a result reported as “+++” to one Immunoglobulin has been administered intra- venously to patients with AIHA, but the results of this Although prednisone produces responses in the treatment have been far less encouraging in cases of majority of patients with AIHA, clinical improvement is AIHA than they are in cases of immune thrombocy- not equivalent to a complete recovery. Slightly more topenia.7 Plasmapheresis has been attempted on occa- than half of the patients who respond to prednisone sion, but this approach has met with limited success, therapy will relapse, even if the corticosteroid is tapered given the fact that most IgG is extravascular. Anecdotal gradually. It is estimated that prednisone alone will pro- success has also been reported with use of danazol and duce permanent responses in only a third of cases.9 For patients who are primarily refractory to prednisone orwho require frequent retreatment with the drug, thera- Transfusions
py with splenectomy or cytotoxic drugs is indicated.
Transfusion therapy in patients with AIHA charac- terized by warm antibody–mediated hemolysis presents Splenectomy
a major clinical problem because of difficulties in cross- Splenectomy is indicated when patients require more matching and so should be avoided in most cases. How- than 20 mg/day of prednisone, when patients suffer ever, in patients with critically low levels of hemoglobin, from severe adverse effects at the dose required to main- transfusions can be used as a life-saving treatment to tain a response, or when frequent relapses require re- temporize the patient until a response to cortico- treatment at a high dose (ie, near 1 mg/kg per day).6,10 steroids or splenectomy occurs. In patients with warm Splenectomy theoretically would be effective because the autoantibodies, ABO and Rh typing generally presents spleen is both a major site of destruction of erythrocytes little difficulty. If there is a question about the accuracy in warm antibody–mediated hemolysis and a major site of typing, the antibody can be eluted from the cells and of autoantibody production. As is the case with pred- nisone, splenectomy is of very limited value in patients The problem in arranging transfusions for patients who have cold antibody–mediated hemolysis. with warm autoantibodies is that antibody screening is Although the response rate to splenectomy varies difficult to perform; crossmatching presents similar diffi- considerably in published studies, a rate of approxi- culties. Free autoantibody may react with all cells, ren- mately 75% seems consistent with the overall data.7 dering all crossmatches incompatible. When this circum- Despite the frequent relapses seen after splenectomy, it stance occurs, the detection of alloantibodies induced by is estimated that a third to a half of patients have their prior transfusions or pregnancies can be difficult. The disease permanently controlled by splenectomy. Ad- best that a blood bank often can to do in this situation is ditionally, many patients who require high doses of to define the “least incompatible” units. If transfused, prednisone before splenectomy can be maintained on such units must be given slowly with close observation for lower, safer doses after splenectomy.
signs of intravascular hemolysis (eg, development of backpain and/or hemoglobinuria). However, in most cases, Immunosuppressive Therapy
transfused cells will be destroyed at a rate neither faster For patients not receiving or responding to cortico- nor slower than that at which native erythrocytes are steroids who are not surgical candidates or for whom splenectomy fails as a treatment, immunosuppressivepharmacologic therapy is indicated. Azathioprine and COLD ANTIBODY–MEDIATED HEMOLYSIS
cyclophosphamide are the most commonly used drugs.
For many patients with cold antibody–mediated No data from controlled clinical trials are available, but hemolysis, also known as cold agglutinin disease, the reviews suggest a response rate of approximately 50% disease process is a chronic illness rather than a severe 8 Hospital Physician Board Review Manual
I m m u n e H e m o l y t i c A n e m i a acute episode. As such, adequate therapy may require of α-methyldopa–related AIHA, therapy identical to that no more than having the patient avoid temperatures at employed in drug-independent AIHA might be re- which the antibody shows activity. In patients with more severe disease, more aggressive therapy is indicated.
As mentioned previously, prednisone therapy and DRUG ADSORPTION
splenectomy play no role in the usual management of In the drug adsorption–type of hemolysis, a drug cold antibody–mediated hemolysis. Instead, standard binds to erythrocytic membranes, so the antibody is therapy is aimed at decreasing antibody production by directed against the drug, not the erythrocytes. The means of immunosuppression.6 Cyclophosphamide and prototype for this phenomenon is penicillin-induced chlorambucil are the drugs most commonly used for this hemolysis, which can occur when penicillin is given at purpose. Chlorambucil is often administered at a dose of doses greater than 10 million units per day. Because 2 to 4 mg/day and cyclophosphamide at a dose of 100 to such doses are rarely used in clinical practice, the phe- 150 mg/day. Pulse therapy with higher levels of drugs nomenon is largely of academic interest. However, given for 4 days every 2 to 3 weeks can also be employed. when such high doses are administered, approximately Because the antibody in cold antibody–mediated 3% of patients will develop antipenicillin antibodies.7 In hemolysis is IgM, which has an intravascular distribution, contrast to true autoimmune disease, eluates from plasmapheresis is of greater theoretical value than in these patients do not react against normal cells because warm antibody–mediated hemolysis. Unfortunately, that the antigen against which the antibody is directed is theoretical value does not translate into significant clinical penicillin. It is worth noting that this phenomenon was value, perhaps because plasmapheresis does nothing to first described in a transfusion recipient whose serum decrease antibody production. Accordingly, plasma- reacted with all erythrocytes that had been stored in pheresis should not be used alone but instead should be penicillin.12 Because the antibody involved is an IgG combined with immunosuppressive therapy. Obviously, if antibody, other types of penicillin sensitivity (eg, plasmapheresis is employed, great care must be taken not urticaria, anaphylaxis) are usually not observed. The to lower the temperature of the blood to the point where drug adsorption–type of hemolysis can also occur with cephalosporins, tetracycline, tolbutamide, and semisyn- Transfusion therapy also is not generally needed for patients with cold antibody–mediated hemolysis. How-ever, when transfusions are needed, all testing must be NEOANTIGEN FORMATION
performed at 37°C (98.6°F) to minimize the effects of A second mechanism of drug-induced hemolysis is the cold antibodies (or agglutinins) and allow for the neoantigen formation, previously known by the less detection of alloantibodies. Whereas most cold anti- accurate terms of “immune complex phenomenon” bodies are directed against the I antigen, locating i and “innocent-bystander mechanism.” Former theories cells is not practical because these cells are extremely held that, in this type of hemolysis, drugs formed a com- rare. Instead the “least incompatible” units should be plex with an antidrug antibody, which then attached to given through a blood warmer designed to keep the an erythrocyte that was not a target of the antibody but was instead an innocent bystander. However, recentstudies have shown that the antibodies react against acombined drug-erythrocyte complex.13 As in the peni- DRUG-RELATED AUTOIMMUNE HEMOLYTIC
cillin mechanism of hemolysis, the drug or its metabo- lites are required for antibody binding to occur. Unlikepenicillin, the drugs involved in this type of reaction Although AIHA most often is an idiopathic process, bind only very loosely to erythrocytes. However, only a drugs also can cause this type of anemia. Three main small amount of the drug is necessary for hemolysis to mechanisms of drug-induced immune hemolysis have occur. This type of hemolysis is mediated by comple- been described: (1) drug adsorption, (2) neoantigen ment and can be quite massive, leading to hemoglobin- formation, and (3) true autoimmune disease. Moreover, there are other drugs that cause nonspecific binding ofproteins (eg, immunoglobulins) to erythrocytes but do TRUE AUTOIMMUNE DISEASE
not cause hemolysis. Treatment of drug-related AIHA A third type of drug-related autoimmune hemolysis consists of recognizing the possible responsible drug is the α-methyldopa type of hemolysis. In this type (Table 4)7 and discontinuing its administration. In cases
of hemolysis, antibodies bind to erythrocytes in a Hematology Volume 1, Part 4
I m m u n e H e m o l y t i c A n e m i a Table 4. Commonly Used Drugs That Can Cause Autoimmune Hemolysis
Adapted with permission from Thomas AT. Autoimmune hemolytic anemias. In: Lee GR, Foerster J, Lukens J, et al, editors. Wintrobe’s clinical hem-atology. 10th ed. Baltimore: Williams and Wilkins; 1999:1251.
drug-independent manner. In essence, the process cephalosporins.7 Cephalosporins, of course, also can results from drug administration but is identical in produce hemolysis by means of either the drug absorp- every other respect to idiopathic autoimmune hemolyt- ic anemia. Antibodies that are eluted off the erythrocytein these cases will bind in vitro to erythrocytes of pa-tients who have never received α-methyldopa and PAROXYSMAL COLD HEMOGLOBINURIA
are indistinguishable from antibodies seen in idiopath-ic AIHA. As many as a third of patients receiving A few words about paroxysmal cold hemoglobinuria α-methyldopa will eventually have positive results on a (PCH) also are appropriate. PCH can be distinguished DAT, but fewer than 1% of patients receiving the drug from cold antibody–mediated hemolysis by both clinical actually develop hemolysis.14 The specific mechanism and immunologic characteristics. Originally described by by which α-methyldopa alters the immune system and Donath and Landsteiner in 1904,15 the condition is char- acterized by sudden onset of fever, back or leg pain, andhemoglobinuria after exposure to the cold. Cold expo- NONHEMOLYTIC DRUGS
sure may be brief, and symptoms begin within minutes to Although not producing hemolysis, a final drug- hours. Urine is characteristically dark red to black but related immune mechanism worth noting involves clears in color over the course of a few hours. The syn- nonspecific attachment of proteins to erythrocytic drome is more common in children but also can occur membranes. These proteins lead to agglutination of erythrocytes and to positive results on a DAT but do not In contrast to the IgM antibody in cold antibody– produce destruction of erythrocytes. Such a situation mediated hemolysis, the antibody in PCH is an IgG that occurs in approximately 3% of all patients receiving binds to erythrocytes in the cold and fixes complement.
10 Hospital Physician Board Review Manual
I m m u n e H e m o l y t i c A n e m i a When the erythrocytes are warmed, hemolysis occurs.
consistent with the diagnosis. Although the serum hap- The Donath-Landsteiner test for this biphasic hemolysin toglobin level generally is low in cases of intravascular involves incubating erythrocytes in the patient’s serum at hemolysis, the destruction of erythrocytes in AIHA (as 0°C (32°F) to 4°C (39.2°F) and then warming the cells to well as in other hemolytic anemias) may be primarily 37°C (98.6°F) to produce lysis. The antibody is specific splenic; as previously mentioned, the serum haptoglo- for the P antigen. The generally accepted theory is that bin level is less affected when hemolysis occurs at such an antigen associated with a microorganism leads to pro- extravascular sites. Because the patient’s elevated serum duction of the IgG antibody, which cross-reacts with the bilirubin level is indirect, his urine would be straw col- P blood-group system. Because IgG does not remain on ored (normal) rather than brown (indicating the pres- the cell once complement is fixed, the DAT will have pos- itive results if nonspecific antiglobulin antiserum or anti- To confirm the diagnosis of AIHA, a DAT is per- complement antiserum is used, but not if specific anti- formed, with results reported as positive (“++++”).
Specific tests reveal that IgG is present on the erythro- In the past, PCH was associated with advanced or con- cytes (results also are reported as “++++”) but that com- genital syphilis. Patients with PCH still should be investi- plement is absent. The patient is treated with pred- gated for occult syphilis, but this association rarely is nisone (80 mg/day) and initially responds, with the hematocrit increasing to 42% as the reticulocyte count The anemia associated with PCH can be severe decreases to 3.4%. After the prednisone dose is slowly because of massive intravascular hemolysis. A hemoglo- tapered to 20 mg/day, the hematocrit falls to 27%. The bin level of 5 g/dL is not uncommon. Although the dose of prednisone is increased, leading to an increase reticulocyte count is elevated as recovery occurs, it may in the hematocrit to 41%. However, despite slow taper- be misleadingly low if patients are seen at the onset of a ing of the prednisone dose, the hematocrit again falls (to 26%) when the dose reaches 20 mg/day. At this Treatment of PCH consists of keeping patients warm point, the corticosteroid dose is again increased to raise and giving them transfusions, as necessary. Although the hematocrit above 30%, and a splenectomy is per- the antibody is directed against the P antigen, there is formed. This time, the patient is successfully tapered off no need to search for rare pp units of blood; antibody prednisone without experiencing a relapse of the will not fix to P antigen–positive cells unless the patient FOLLOW-UP OF CASE PATIENT
The survival of erythrocytes in the circulation is decreased in patients with hemolytic anemia. Because • What type of anemia is most likely in the case pa-
hemolytic anemia is uncommon and because its causes are diverse, its presentation can cause confusion for • Does the presence of spherocytes in the case pa-
clinicians. The first step in evaluating a patient suspect- tient’s peripheral blood smear necessarily indicate a
ed of having hemolytic anemia is to document that diagnosis of hereditary spherocytosis?
hemolysis is present by determining the reticulocyte • Is the case patient’s normal serum haptoglobin level
count and the serum indirect bilirubin, LDH, and hap- inconsistent with a diagnosis of autoimmune hemo-
toglobin levels. Secondly, the specific cause of the lytic anemia?
hemolysis should be determined. If the hemolytic ane- • Would the case patient’s urine be expected to con-
mia is clearly acquired, one can start by evaluating the tain increased levels of bilirubin?
patient for the most common causes of hemolysis,namely acquired AIHA and hypersplenism.
Based on the clinical presentation and laboratory AIHA comes in 2 varieties. In warm antibody– findings, a diagnosis of hemolytic anemia with sphero- mediated AIHA, the antibody is usually an IgG, which cytosis is made in the case patient. AIHA is suspected.
may or may not fix complement. Destruction of ery- Spherocytes are not specific for hereditary spherocyto- throcytes occurs primarily in the spleen, and standard sis but can indicate the presence of AIHA, especially in therapy is prednisone administration. Splenectomy is the absence of a positive family history. The serum hap- reserved for corticosteroid-refractory cases. In cold toglobin level, although in the normal range, also is antibody–mediated AIHA, the antibody is an IgM that Hematology Volume 1, Part 4
I m m u n e H e m o l y t i c A n e m i a fixes complement, leading to destruction of erythrocytes Association of Blood Banks; 1991:33–72. in the liver and elsewhere in the monocyte-macrophage 6. Petz LD, Garratty G. Acquired immune hemolytic ane- system. Corticosteroids and splenectomy generally play mias. New York: Churchill Livingstone; 1980.
no role in the management of cold antibody–mediated 7. Thomas AT. Autoimmune hemolytic anemias. In: Lee AIHA. In fact, this type of AIHA may be so mild as to not GR, Foerster J, Lukens J, et al, editors. Wintrobe’s clinical require therapy; treatment with chemotherapy or plas- hematology. 10th ed. Baltimore: Williams and Wilkins; mapheresis is also possible. Appreciation of the underly- ing pathophysiology of AIHA may lead to a more rational 8. NIH conference. Pathophysiology of immune hemolytic approach to its diagnosis and treatment. anemia. Ann Intern Med 1977;87:210–22.
9. Allgood JW, Chaplin H Jr. Idiopathic acquired autoim- mune hemolytic anemia. A review of forty-seven cases treat- REFERENCES
ed from 1955 through 1965. Am J Med 1967;43:254–73.
10. Pirofsky B, Bardana EJ Jr. Autoimmune hemolytic anemia.
1. Lee RL. Hemolytic disorders: general considerations. In: II. Therapeutic aspects. Ser Haematol 1974;7:376–85.
Lee GR, Foerster J, Lukens J, et al, editors. Wintrobe’sclinical hematology. 10th ed. Baltimore: Williams and 11. Zupanska B, Sylwestrowicz T, Pawelski S. The results of pro- longed treatment of autoimmune haemolytic anaemia.
2. Deiss A. Destruction of erythocytes. In: Lee GR, Foerster Haematologia (Budap) 1981;14:425–33.
J, Lukens J, et al, editors. Wintrobe’s clinical hematology.
12. Ley AB, Harris JP, Brinkley M, et al. Circulating antibod- 10th ed. Baltimore: Williams and Wilkins; 1999:267–99.
ies directed against penicillin. Science 1958;127:1118–9.
3. Coombs RR, Mourant AE, Race RR. A new test for the 13. Salama A, Mueller-Eckhardt C. On the mechanisms of detection of weak and “incomplete” Rh agglutinins. Br J sensitization and attachment of antibodies to RBC in drug-induced immune hemolytic anemia. Blood 1987; 4. Burkhart P, Rosenfield RE, Hsu TC, et al. Instrumental PVP-augmented antiglobulin tests. I. Detection of allo- 14. Carstairs KC, Breckenridge A, Dollery CT, Worlledge SM.
geneic antibodies coating otherwise normal erythrocytes.
Incidence of a positive direct Coombs test in patients on alpha-methyldopa. Lancet 1966;2:133–5.
5. Garratty G. Target antigens for red-cell bound autoanti- 15. Donath J, Landsteiner K. Ueber paroxysmal haemoglo- bodies. In: Nance SJ, editor. Clinical and basic science binurie. [Article in German.] Munch Med Wochenschr aspects of immunohematology. Arlington (VA): American Copyright 2001 by Turner White Communications Inc., Wayne, PA. All rights reserved.
12 Hospital Physician Board Review Manual

Source: http://hemaonco.tmu.edu.tw/teachingbank/Immunehemolyticanemia.pdf