Drug-Induced Immune Hemolytic Anemia
GEORGE GARRATTY, F.I.M.L.T.’ LAWRENCE D. PETZ, M.D. San Francisco. California
From the Harkness Community Hospital and Medical Center, and the University of California Medical Center, San Francisco, California. This study was supported in part by Grant NS-09568 from the National Institutes of Health, by the Research Evaluation and Allocation Committee (Gilbert Fund) of the University of California, San Francisco, California, and by Eli Lilly & Co., Indianapolis, Indiana. Requests for reprints should be addressed to Mr. George Garratty. Manuscript accepted May 10, 1974. ’ Present address: Hematology-Immunology Research Unit, Room 229, Research Building, Institutes of Medical Sciences Pacific Medical Center, 2200 Webster Street. San Francisco, California 94115.
399
March 1975
Drug administration causes from 16 to 16 per cent of cases of acquired immune hemolytic anemia. The pathogenesis of erythrocyte sensitization by drug-related antibody with or without fixation of complement is variable, and there is a relationship between the responsible drug, the mechanism of red cell sensitization, clinical manifestations and laboratory methods of diagnosis. Drugs such as phenacetin and quinidine form a complex with the antidrug antibody, and the immune complex attaches to red cells usually fixing complement and causing acute intravascular hemolysis. Other drugs (e.g., penicillins), when given in large doses, coat normal red cells in vivo and in some patients a high titer IgG anti-drug antibody develops which reacts with the coated cells. Hemolytic anemia may develop with red cell destruction being primarily extravascular. Cephalosporins cause positive direct antiglobulin tests in a small percentage of pattents either by the same mechanism as penicillins or by modification of the red cell membrane leading to nonimmunologic absorption of serum proteins. Hemolytic anemia has been reported only rarely. A few drugs (notably alpha methyldopa) cause the development of autoimmune hemolytic anemia. Knowledge of clinical manifestations and laboratory aids to diagnosis is necessary to distinguish immunohematologic abnormalities caused by drugs from other causes. The administration of drugs may lead to the development of a wide variety of hematologic abnormalities including immune hemolytic anemia [l]. The clinical importance of drug-induced immune hemolytic anemia is emphasized by the fact that acquired immune hemolytic anemia among 16 per cent of our series of 200 patients [2] and 18 per cent of 79 patients of Dacie and Worlledge in 1967-1968 [3] were caused by drugs. In addition, immunohematologic abnormalities caused by drugs are a source of confusion in the blood transfusion laboratory where they may interfere with compatibility procedures. This review considers the various mechanisms leading to the development of red cell sensitization by drug-related antibodies, and the clinical and laboratory features of drug-induced immunohematologic abnormalities involving red cells. The first drug that was proved to cause a positive direct antiglobulin test (DAT) and immune hemolytic anemia was Fuadine (stibophen) which consists of 2 molecules of sodium catechol disulfonate linked by 1 atom of antimony. In 1954 [4] and 1956 [5], Harris described a patient with schistosomiasis treated with Fuadin in whom acute intravascular hemolysis developed. The patient’s
The American Journal of Medicine
Volume 58
DRUG-INDUCEDIMMUNE HEMOLYTIC ANEMIA-GARRAlTY,
serum contained a factor which agglutinated his own or normal red cells, or sensitized them to agglutination with antiglobulin serums. These reactions only took place in the presence of the drug. Since then many other drugs have been described which can cause a positive direct antiglobulin test and immune hemolytic anemia. Table I lists most of these drugs and pertinent references are cited [443]. Penicillin and alpha methyldopa (Aldome@)) are by far the most common drugs that cause immune hemolytic anemia. Methadone has also been reported as a cause of positive direct antiglobulin tests [44], although this was not confirmed by other workers [ 451. The pathogenesis of erythrocyte sensitization by drug-related antibody with or without fixation of complement is variable and is of particular importance since there is a relationship between the responsible drug, the mechanism of red cell sensitization, clinical manifestations and laboratory method of diagnosis (Table II). There seem to be four mechanisms that can cause a positive direct antiglobulin test or hemolytic anemia due to drugs. MECHANISMS OF THE POSITIVE DIRECT ANTIGLOBULIN ANEMIA DUE
TEST
AND
POSSIBLE
HEMOLYTIC
TO DRUGS
A. Immune-Complex Adsorption to Red Cells. In his studies on drug-induced purpura, Ackroyd [46] suggested that drugs may act as a hapten, combining loosely with the cell membrane and stimulating antibody production to the combined antigen. Schulman [47-491 proposed a different mechanism which is accepted by most investigators as the most probable explanation for the reactions seen with all the drugs mentioned in Table I except penicillin, cephalosporin, alpha methyldopa, L-dopa and possibly carbromal. Schulman extended the findings of Miescher and co-workers [50,51] by showing that drugs such as quinine, quinidine and stibophen have a far stronger affinity for their respective antibodies than for cell membranes. Such drugs, when present together with antibody in the patient’s plasma, will combine to form an immune complex and be adsorbed to the cell membrane, often activating complement in the process (Figure 1). It is unknown why this immune complex once formed sometimes causes red cell destruction only and other times platelet destruction only. The clinical and laboratory findings relating to immune abnormalities caused by this group of drugs follow: (1) The patient needs to take only a small quantity of the drug. (2) Acute intravascular hemolysis with hemoglobinemia and hemoglobinuria is the usual clinical presentation (16 of 19 reported cases [52]).
TABLE
PETZ
Drugs that Have Been Reported
I
Positive Direct Antiglobulin Anemia
Drug Fuadin@ (stibophen) Quinidine P-aminosalicylic acid (PAS) Quinine Phenacetin Penicillin insecticides (chlorinated hydrocarbons) Antihistamine (Antihistine@) Sulfonamides lsonicotinic acid hydrazide (isoniazid) Chlorpromazine Pyramido@ Dipyrone a-Methyldopa (Aldomet@:) L-phenylalanine mustard (melphalan) Cephalothin (Keflin@) Mefenamicacid (Ponstel@) Carbromal (Carbrital@)* Sulfonylurea derivatives Insulin L-Dopa Rifampicin * Carbromal antiglobulin
to Cause a Test and Hemolytic
Year First Described
Important References
1954 1956 1956 1958 1958 1959 1959
(4-W ]71 18,91 [lOI 18-111 [12-15)
1959 1960 1960
1171 118,191 IW’OI
1961 1961 1966 1966 1967
WI I221 I231
1967 1968 1970 1970 1970’ 1971 1972
1161
[24-271
WI [29-311 [32-34) [351 136,371 1381 (39-411 142,431
has been reported to cause positive tests but not hemolytic anemia so far.
direct
(3) Renal failure is frequent [6,8,16,21,23]. (4) The serum antidrug antibody is often IgM and capable of activating complement. (5) The direct antiglobulin test is positive, often due to the presence of complement components on the red cell surface, usually without detectable immunoglobulins. This may be explained by the fact that the immune complex does not bind very firmly to the red cells and may dissociate from the cells and be
DRUG
Y. COIUPLEX FORMATIOM
Figure 1. Mechanism of development of positive direct antiglobulin test caused by drugs such as phenacetin (see text). From Garratty [94].
March 1975
The American Journal of Medicine
Volume 58
399
DRUG-INDUCED IMMUNE HEMOLYTIC ANEMIA-GARRATTY,
TABLE
II
PET2
Correlation Between Mechanism of Red Cell Sensitization Drug-Induced lmmunohematologic Abnormalities
and Clinical and Laboratory
Features in
Serologic Evaluation
I
Immune complex formation (drug and antidrug antibody)
II
Drug adsorbed onto red cell membrane; reacts with high titer serum drug antibody
III
Membrane modification (nonimmunologic absorption of proteins)
IV Unknown
Clinical Findings
Prototype Drugs
Mechanism
Quinidine, phenacetin
Usually only complement components detected but IgG can be present
Drug + patient’s serum + RBC (especially enzyme-treated) hemolysis, agglutination, or sensitization; antibody frequently IgM and capable of fixing complement; RBC eluate often nonreactive
Penicillins, cephalosporins
Large doses of penicillin (10 million units or more daily); other manifestations of allergy not necessarily present; usually subacute extravascular hemolysis; penicillin one of most common causes of drug-induced immune hemolysis; rare case of immune hemolytic anemia caused by cephalosporins
Strongly positive (IgG) when hemolytic anemia present; rarely weaker complement sensitization is also present
Drug-coated RBC + serum + agglutination or sensitization (rarely hemolysis); high titer antibody associated with hemolytic anemia is always IgG; RBC eluate reacts only with antibiotic-coated RBC
Cephalosporins
No cases of cephalosporin-induced hemolytic anemia caused by nonimmunologic mechanisms
Positive with antiserums to a variety of serum proteins
Drug-coated RBC + serum + sensitization to antiglobulin serums in low titer (nonimmunologic protein absorption); RBC eluate nonreactive
a-methyldopa
Hemolysis in 0.8 percent of patients taking drug for at least 3 mo; gradual onset of hemolytic anemia; one of most common causes of drug-induced immune hemolysis
Strongly positive (IgG) when hemolytic anemia present; only rarely cells are sensitized with complement as well
Antibody sensitizes normal RBC without drug; antibody in serum and eluate identical to that found in warm antibody AIHA; no in vitro relationship to drug demonstrable
to react with other cells. This in turn may explain why such a small amount of drug complex can cause so much red cell destruction [53]. Furthermore, red cell sensitization by IgM antibodies is not readily detectable by the antiglobulin test [54]. Some reports of negative direct antiglobulin tests, particularly in the earlier literature, may be due to inadequate anticomplement properties of the antiglobulin serum [54]. (6) In vitro reactions (agglutination, lysis and/or sensitization to antiglobulin serums) are only ob-
March 1975
Antibody Identification
Small doses of drug; acute intravascular hemolysis usual; renal failure common; thrombocytopenia occasionally
free
400
Direct Antiglobulin Test
The American Journal of Medicine
served when patient’s serum, drug and red cells are all incubated together. B. Drug Adsorbed Onto Red Cells. In 1958, Ley et al. [ 121 first reported the presence of circulating penicillin antibody. In the course of routine blood bank testing, a serum was encountered which agglutinated the red cells of 25 group 0 members of an antibody identification panel. This panel had been stored in a preservative solution containing penicillin. The same panel with no penicillin present did not
Volume 58
DRUGlNDUCED
react with the serum. In the paper, a patient was described who had a positive direct antiglobulin test and hemolytic anemia after receiving penicillin. In 1966, Petz and Fudenberg [ 141 defined the role of penicillin in causing immune hemolytic anemia. Penicillin is one of the few drugs that binds firmly to proteins. Penicillin will combine with the protein on normal red cell membranes both in vivo and in vitro. Spath et al. [55] performed a detailed study of numerous variables affecting sensitization of red cells by penicillin in vitro (pH, drug concentration, temperature and duration of incubation, etc.). These workers defined optimal and very practical methods for preparing penicillin-treated red cells and for using such cells for penicillin antibody detection. The drug cannot be removed from red cells in vitro even by multiple washes in saline solution. This is in contrast to the drugs mentioned in section A, in which it is not possible to prepare drug-coated cells for antibody detection. Considerable experimental work has demonstrated that the immunogenicity of penicillin is due to its ability to react chemically with tissue proteins to form several different haptenic groups [56-651. The major haptenic determinant is the benzylpenicilloyl (BPO) group. It is not clear yet whether the BP0 group is formed by the direct reaction of penicillin with amino groups of protein, or through the intermediate formation of benzylpenicillenic acid. Penicillin can be detected on the red cell membranes of all patients receiving large doses of penicillin [66,67]. If a sensitive technic is used [55,68], over 90 per cent of unselected serums can be shown to contain penicillin (BPO) antibodies [68]. Most serums contain IgM antibodies alone (approximately 80 per cent); approximately 13 per cent contain IgG antibodies as well. These antibodies are usually easily neutralized by BP0 hapten. The antibodies associated with immune hemolytic anemia due to penicillin are IgG, and are not easily neutralized by BP0 hapten. The high percentage of penicillin antibodies in the normal population is probably due to the continual exposure to penicillin in our modern environment. In approximately 3 per cent of the patients who receive massive doses of intravenous penicillin, positive direct antiglobulin tests will develop [66,69]; and in a small percentage of these, hemolytic anemia will develop. The mechanism of the positive direct antiglobulin test and hemolytic anemia seems clear [14,67]. The drug is adsorbed to the red cells and the antipenicillin present in the patient’s plasma will react with the penicillin on the red cells (Figure 2). The quantity of penicillin (BPO) antibody sensitizing the red cell is limited by the number of BP0 haptenic
IMMUNE HEMOLYTIC ANEMIA-GARRATTY,
PET2
I? 6.C ANTIBODY
% Figure 2. antiglobulin ratty [94].
Mechanism of development of positive direct test caused by penicillins (see text). From Gar-
groups on the cell, the plasma concentration of BP0 specific antibodies and the avidity of the antibodies Complement is not usually involved in this reaction and intravascular hemolysis rarely occurs [ 70,7 11. The IgG sensitized red cells are probably removed extravascularly by the reticuloendothelial system in the same way as Rhesus (IgG) sensitized cells. There is no direct correlation between the presence of IgG and IgM penicillin hemagglutinating antibodies and allergic reactions. Most workers have found no correlation at all, but a few have found that high titer IgG antibodies occur more often in the allergic group [72,73]. Perhaps a better correlation will be found with IgE penicillin antibodies. The clinical and laboratory features of penicillininduced immune hemolytic anemia are quite constant and are as follows: (1) Hemolysis typically develops only in patients who are receiving very large doses of penicillin (at least 10 million units daily for a week or more). (2) Hemolytic anemia is usually less acute in onset than that caused by drugs of group A but may be lifethreatening if the etiology is unrecognized and penicillin administration is continued. (3) A high titer IgG penicillin antibody is present in the serum. (4) The direct antiglobulin test is strongly positive due to sensitization with IgG. Barely, complement components are detected as well, and complement activation may contribute to the immune hemolysis [70,71]. (5) Antibody eluted from the patient’s red cells will
react only against penicillin-treated red cells. (6) Cessation of penicillin therapy is followed by complete recovery, but hemolysis of decreasing severity may persist for several weeks.
March 1975
The American Journal of Medicine
Volume 59
401
DRUGlNDUCED IMMUNE HEMOLYTICANEMIA-GARRATTY. PET2
.
RETICULOCYTES
PENICILLIN
CAPA
TITER
8 6t
II/4 IO/5 12/3 O/I5 Figure 3. Clinical course of a patient with penicillin-induced immunohemolytic anemia. CAPA refers to circulating antipenicillin antibody. From Petz and Fudenberg [ 141.
(7) Other manifestations of penicillin allergy are not necessarily present. Figure 3 illustrates the course of a characteristic case of penicillin-induced hemolytic anemia. Other drugs known to bind firmly to red cells and cause positive direct antiglobulin tests by the same mechanism as penicillin are cephalothin [30,31] and possibly carbromal [35]. Cephalothin [29] has been described as a rare cause of immune hemolytic anemia but carbromal so far has not.
402
March 1975
The American Journal of Medicine
C. Modification of Red Cell Membrane by Drug Allowing Nonimmunologic Absorption of Protein. In 1967, Gralnick [30] and Molthan [31], with their coworkers, described positive direct antiglobulin tests following the administration of cephalothin (Keflinm). Three mechanisms for the positive direct antiglobulin test due to cephalothin have been proposed: (1) The cell membrane may be modified by the drug so that the cell now takes up proteins nonimmunologically [30,31,74] (Figure 4). Cephalothin-sensi-
Volume 58
DRUWNDUCED
tized cells incubated in normal plasma in vitro become coated with albumin, IgG, IgA, IgM, alpha, antitrypsin, alpha, macroglobulin, C3, C4 and fibrinogen [2,74]. (2) The drug may combine with the red cell membrane, as does penicillin, and being so closely related chemically to penicillin, cross-reacting penicillin antibodies will now combine with the drug [30,66,7476]. (3) The drug may combine with the cell membrane in a similar fashion to penicillin, and specific anticephalothin antibodies react with the drug on the cell [29,74,77]. Spath et al. [74] confirmed that all three of these mechanisms may lead to a positive direct antiglobulin test. Whether the positive direct antiglobulin test has resulted from nonimmunologic protein absorption or by fixation of antibody may be determined in an individual case as outlined in Table II. Other data confirm that cephalothin will modify the red cell membrane. Indeed, Sirchia, Ferrone and coworkers [78,79] have shown that normal red cells after treatment with cephalothin in vitro, resemble cells from a patient with paroxysmal nocturnal hemoglobinuria. In vitro studies showed a low acetylcholinesterase activity, reduced oxygen uptake in the presence of methylene blue and a positive Ham’s acid serum lysis test. They could show no enzymatic or metabolic abnormalities of red cells from patients with positive direct antiglobulin tests following cephalothin administration. Gralnick et al. [29] described two patients with immune hemolytic anemia following cephalothin administration, but this is a much rarer event than hemolytic anemia following penicillin therapy. Other cephalosporin drugs closely related to cephalothin such as cephalexin [80,81], cephaloridine [82,83] and cefazolin [81] have also been described as a cause of positive direct antiglobulin tests. Cephalothin, like penicillin, will bind firmly to red cells. These cells can be washed many times in vitro without the drug being removed. Cephalothin-treated cells can be used in a similar fashion to penicillintreated cells in the detection of cephalothin or penicillin cross-reacting antibodies [55]. Serums from unselected donors show IgM antibodies to cephalothin-treated cells in 88 per cent, IgG antibodies in 10 per cent, and both IgG and IgM antibodies in 15 per cent [74]. Spath et al. [74] were able to confirm and extend the findings of Gralnick [ 751, Nesmith 1761, Abraham [77] and co-workers in showing that penicillin antibodies may cross react with cephalothin-treated red cells: thus high titer penicillin antibodies may react with cephalothin-treated cells both in vivo and in vitro. In addition, specific anticephalothin antibodies
IMMUNE HEMOLYIK: ANEMIA-QARRArTY.
PETZ
MEM0f?AlVE MODIFCA77Oht
04
0
One of the mechanisms of development of a positive direct antiglobulin test caused by cephalosporins (see text). From Garratty (941.
Figure 4.
exist, and these can sometimes be shown to be present in eluates from red cells demonstrating a positive direct antiglobulin test due to cephalothin [29,74]. In the original reports, from 40 to 75 per cent of patients who received cephalothin had positive direct antiglobulin tests [30,31], but in further studies recently, Spath, Garratty and Petz [55,74] found positive tests in only 3 per cent. These latter workers postulated that the differences between their findings and those of earlier workers may be due to the following reasons: (1) Dosage of drug: The dosage and duration of therapy are likely to influence the incidence of positive direct antiglobulin tests. The patients studied by Spath et al. were unselected patients in a community hospital, where the mean daily cephalothin dosage was 6.3 g and the mean duration of therapy 5.5 days. The patients with a positive direct antiglobulin test received 7 g for 10 days. The earlier workers reported a higher average drug dosage (e.g., patients with positive direct antiglobulin tests repeated by Gralnick et al. [30] receiving from 8 to 14 g of cephalothin daily). (2) Patient selection: There is evidence to suggest that the percentage of positive direct antiglobulin tests due to cephalothin are higher in azotemic patients. Gralnick et al. [30] suggested a relationship between hypoalbuminemia and positive direct antiglobulin tests following cephalothin administration. (3) Antiglobulin technic: Cephalothin-treated red cells absorb proteins nonspecifically. Since such cells are coated by numerous proteins, the number of positive direct antiglobulin tests obtained will be proportional to the spectrum of antibodies in the antiglobulin serums used. For instance, cephalothintreated cells take up albumin very efficiently. On screening commercial antiglobulin serums and antiserums prepared in our own laboratory [2], we found
March 1975
The Am&can
Journal of Modlclne
Volume 59
403
DRUG-INDUCED IMMUNE HEMOLYTIC ANEMIA-GARRATTY,
PET2
that most commercial antiglobulin serums, as sold for blood banking purposes, contained very little or no antialbumin. One or two contained a strongly reacting antialbumin. If the latter antiglobulin serums were used, a higher percentage of positive direct antiglobulin tests would be obtained. D. Red Cell Autoantibodies Induced by Drugs: Unknown Mechanism. In 1966, Carstairs, Worlledge and co-workers [26,27] described the first positive direct antiglobulin tests and autoimmune hemolytic anemia due to the drug alpha methyldopa (Aldomet). This reaction was different from any of the other types described in sections A, B or C, as in patients receiving this drug antibodies were made that acted directly with normal red cells, even in the absence of the drug. Indeed, the antibodies reacted similarly to those seen in the idiopathic autoimmune hemolytic anemias, in that they often showed some Rhesus blood group specificity. In contrast, of course, the antibodies in patients receiving the drugs mentioned in sections A, B and C were directed against the drug itself and not intrinsic red cell antigens. Immune mechanisms described in sections A, B and C nevertheless cause red cells to be affected as if antibody had been specifically directed against them (these mechanisms of sensitization are often referred to as “innocent bystander” mechanisms [84]). The precise role of alpha methyldopa in the pathogenesis of red cell autosensitization is at present unclear although several theories have been suggested: (1) Adsorption of drug or drug-protein complex to red cells: Wurzel and Silverman [85] showed that normal washed red cells and red cells in plasma would adsorb alpha methyldopa and that if normal whole blood was incubated in a high concentration (e.g., 0.5 mg/ml) of drug for several days, in vitro, then a positive antiglobulin test was obtained. Gottlieb and Wurzel [86] recently demonstrated that alpha methyldopa or one of its metabolites could be covalently bound to gamma globulin when drug and plasma were incubated together for 2 days or more. If red cells were then added, a positive antiglobulin test occurred. No Rhesus specificity was noted. Other workers [2,27,87] have not been able to demonstrate a positive antiglobulin test after incubation of red cells in alpha methyldopa or its metabolites. Furthermore LoBuglio and Jandl [87] demonstrated that alpha methyldopa had very little affinity for plasma proteins or red cell membranes, although it should be noted that the incubation period employed was shorter than that used by Gottlieb and Wurzel [86]. (2) Altered autoantigens: Worlledge et al. [27,88] have postulated that the drug or one of its metabolites may combine with the red cell membrane or enter the red cell and in some way alter the Rhesus antigens so that the normal immune system no long-
494
March 1975
The American Journal of Medicine
er recognizes them as “self.” The incorporation may occur at the normoblast or reticulocyte stage, thus explaining the delay in development of a positive direct antiglobulin test. (3) Forbidden clones: Dameshek [89] suggested that alpha methyldopa may result in errors or aberrations in the proliferation of normal lymphocytes, thus producing clones of abnormal but nevertheless immunologically competent cells. These cells, although functional, would be “intolerant” of, or fail to recognize as “self,” normal antigens, especially those of the ubiquitous red cell. However, antibodies in patients receiving alpha methyldopa are “polyclonal” (i.e., kappa and lambda light chains and distinct Hchain types) [24]. This finding necessitates for clonal interpretation the proposition that at least three different clones of immunocytes lost the ability to recognize the Rhesus locus of the red cell itself, which seems unlikely. Two other drugs have recently been described as acting in a similar fashion to alpha methyldopa. One is a closely related drug L-dopa, which was described as causing positiveadirect antiglobulin tests in about 6 to 9 per cent of the patients receiving the drug [39,40]. Recently, this drug has been reported as a cause of hemolytic anemia [41]. The other drug, mefenamic acid, is unrelated to alpha methyldopa and has been described as causing positive direct antiglobulin tests and autoimmune hemolytic anemia in three separate reports of five patients [32-341. The clinical and laboratory characteristics of alpha methyldopa-induced abnormalities are as follows: (1) Direct antiglobulin test is positive in 15 per cent of patients receiving alpha methyldopa (reports vary from 10 to 36 per cent). Incidence seems to vary between racial groups, being highest in Caucasians, lower in Chinese and almost absent in Blacks
WI. (2) The direct antiglobulin test usually becomes positive after 3 to 6 months of treatment. (It is interesting that this delay is not shortened when administration of the drug is resumed in a patient who previously had a positive test.) (3) The development of the positive direct antiglobulin test appears to be dose-dependent. About three times as many patients (36 per cent) have positive tests when taking more than 2 g of the drug daily than patients taking less than 1 g daily (11 per cent); 19 per cent had a positive test with 1 to 2 g daily. (4) The positive direct antiglobulin test is due to red cell sensitization with IgG, and all patients with hemolytic anemia have a very strongly positive direct antiglobulin test. Rarely, red cells are weakly sensitized with complement as well [ 21. (5) Various reports indicate that from 0 to 5 per
Volume 59
DRUG-INDUCEDIMMUNE HEMOLYTIC ANEMIA-GARRATTY. PET2
cent of patients receiving alpha methyldopa have hemolytic anemia. The cumulative evidence is 0.8 per cent [52]. (6) Worlledge et al. [27] commented that the autoantibodies associated with alpha methyldopa commonly show Rhesus specificity. However, such specificity is based on subtle differences in titer and/or cross absorption and elution technics [24,91-931 using cells of varying genotypes, rather than the clear patterns one sees with Rhesus alloantibodies. Worlledge herself states in a later article [52] that these autoantibodies are similar to those seen in association with idiopathic warm autoimmune hemolytic anemia. This has been our experience, too, in our own series of 200 cases of immune hemolytic anemias studied since 1988. In 20 cases of immune hemolytic anemia associated with alpha methyldopa, the serology was indistinguishable from idiopathic warm autoimmune hemolytic anemia. The specificity of the autoantibody is often associated with the Rh locus, but the definition of this specificity is depen-
dent on the extent of the technical investigation and the availability of rare red cell phenotypes (e.g., Rh,,rr. LW negative, U negative), without which the autoantibody often appears nonspecific. Vos et al. [93] discuss some of these variables in a recent publication. We believe differences in opinion in the literature [86] on the blood group specificity of these autoantibodies can be attributed directly to these differences in technic. (7) The positive direct antiglobulin reaction gradually becomes negative once the administration of alpha methyldopa is stopped. This may take from 1 month to 2 years. Fortunately, the hematologic values usually improve strikingly within the first week or two. Alpha methyldopa is one of the most common causes of positive direct antiglobulin tests due to drugs. Worlledge [52] has reported that the total of alpha methyldopa-induced autoimmune hemolytic anemias exceeds the total of all other drug-induced immune hemolytic anemias so far described.
REFERENCES 1. 2. 3. 4. 5. 6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
16.
Huguley CM, Lea JW, Butts JA: Adverse hematologic reactions to drugs. Proor Hemato15: 105. 1966. Garratty G, Pe& LD: Unpublished observations, 1974. Dacie JV. Worlledge SM: Auto-immune hemolytic anemias. Progr Hemato16: 82, 1989. Harris JW: Studies on the mechanism of a drug-induced hemolytic anemia. J Lab Clin Med 44: 809, 1954. Harris JW: Studies on the mechanism of a drug-induced hemolytic anemia. J Lab Clin Med 47: 760, 1958. De Torregrosa MVV, Rodriguez Rosada AL, Montiila E: Hemolytic anemia secondary to stibophen therapy. JAMA 186: 598, 1983. Freedman AL, Tarr PS, Brody EA: Hemolytic anemia due to quinidine: observations on its mechanism. Am J Med 20: 806, 1956. MacGibbon BH, Loughridge LW, Hourihane DO’B, Boyd DW: Autoimmune haemolytic anaemia with acute renal failure due to phenacetin and paminosalicylic acid. Lancet 1: 7, 1960. Van Loghem JJ: Autoimmune haemolytic anaemia with renal failure due to phenacetin and PAS. Lancet 1: 434, 1960. Muirhead EE. Halden ER, Groves M: Drug-dependent Coombs’ (antiglobulin) test and anemia. Arch Intern Med 101: 87, 1958. Dausset J, Contu L: A case of haemolytic anaemia due to phenacetin allergy. VOX Sang 9: 599, 1964. Ley AB, Harris JP, Brinkley M, Liles B, Jack JA. Cahan A: Circulating antibodies directed against penicillin. Science 127: 1118, 1958. Clayton EM, Aftshuler J, Bove JR: Penicillin antibody as a cause of positive direct antiglobulin tests. Am J Clin Patho144: 648, 1965. Petz LD, Fudenberg HH: Coombs-positive hemolytic anemia caused by penicillin administration. N Engl J Med 274: 171, 1966. Swanson MA, Chanmougan D, Schwartz RS: Immunohemolytic anemia due to antipenicillin antibodies. N Engf J Med 274: 178, 1966. Muirhead EE, Groves M, Guy R. Halden ER, Bass RK: Ac-
17.
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
March 1975
quired hemolytic anemia, exposures to insecticides and positive Coombs’ test dependent on insecticide preparations. VOX Sang 4: 277, 1959. Mahssenet R, Dreyfus B, Antoine B: Anemic hemolytique immunologique due a I’antistine. Proceedings of the 7th Congress International Society of Haematology, Rome, 1958, vol 2. New York, Grune & Stratton, 1960. p 614. Fukase M, Nakano H. Konda 5, Tajima H, Takeuchi S, Tsunematsu T, Nishida R. Kanagawa K, Sugano A. Watanabe K, Suzuka H, lnagaki M: Two cases of drug-allergy followed by acquired autoimmune hemolytic anemia. Acta Haematol Jap 23: 70, 1960. Tajima l-l: Clinical studies on hemolytic anemia. I. Autoimmune hemolytic anemia. Acta Haematol Jap 23: 188, 1960. Robinson MG, Foadi M: Hemolytic anemia with positive Coombs’ test. Association with isoniazid therapy. JAMA 208: 658, 1969. Lindberg LG, Norden A: Severe hemolytic reaction to chlorpromazine. Acta Med Stand 170: 195, 1961. Bernasconi C, Bedartia G, Pollini G, Sartori S: Studio del meccanismo di emolisi in un case di anemia emolitica acquista da piramidone. Haematologica (Pavia) 46: 897, 1981. Lay WH: Drug-induced haemolytic reaction due to antibodies against the erythrocyte/dipyrone complex. VOX Sang 11: 801, 1966. Bakemeier RF. Leddy JP: Erythrocyte autoantibody associated with alpha-methyldopa: heterogeneity of structure and specificity. Blood 32: 1, 1968. Carstairs KC, Breckenridge A, Dollery CT, Worlledge SM: Incidence of a positive direct Coombs’ test in patients on a-methyldopa. Lancet 2: 133. 1966. Carstairs K. Worlledge S, Dollery CT, Breckenridge A: Methyldopa and haemolytic anaemia. Lancet 1: 201, 1966. Worlledge SM, Carstairs KC, Dacie JV: Autoimmune haemolytic anaemia associated with a-methyldopa therapy. Lancet 2: 135, 1966. Eyster ME: Melphalan (Alkeran) erythrocyte agglutinin and
The American Journal of Medlclne
Volume 58
405
ORUGINDUCEO
29.
30.
31.
32. 33.
34.
35.
36.
37.
36. 39.
40. 41.
42.
43. 44.
45.
46. 47.
46.
49.
50.
51.
52. 53.
406
IMMUNE HEMOLYTIC ANEMIA-GARRATTY.
PET2
hemolytic anemia. Ann Intern Med 66: 573, 1967. Gralnick HR, McGlnnlss MH, Elton W, McCurdy P: Hemolytlc anemia associated with cephalothin. JAMA 217: 1143, 1971. Gralnick HR, Wright LD, McGinniss MH: Coombs’ positive reactions associated wlth sodium cephalothin therapy. JAMA 199: 725, 1967. Molthan L, Reidenberg MM, Eichman MF: Positive direct Coombs’ tests due to cephalothin. N Engl J Med 277: 123, 1967. Farid NR. Johnson RJ, Low WT: Haemolytic reaction to mefenamic acid. Lancet 2: 382, 197 1. Robertson JH, Kennedy CC, Hill CM: Haemolytic anaemia associated with mefenamic acid. Ir J Med Sci 140: 226, 1971. Scott GL, Myles AB. Bacon PA: Autoimmune haemolytic anaemia and mefenamic acid therapy. Br Med J 3: 534. 1968. Stefanini M, Johnson NL: Positive antihuman globulin test in patients receiving carbromal. Am J Med Sci 259: 49, 1970. Logue GL, Boyd AE, Rosse WF: Chlorpropamide-induced immune hemolytic anemia. N Engl J Med 283: 900, 1970. Bird GWG, Eeles GH, Litchfield JA. Rahman M, Wingham J: Haemolytic anaemia with antibodies to tolbutamide and phenacetin. Br Med J 1: 728, 1972. Faulk WP, Tomsovic EJ, Fudenberg HH: Insulin resistance in juvenile diabetes mellitus. Am J Med 49: 133, 1970. Henry RE, Golberg LS, Sturgeon F, Ansel RD: Serologic abnormalities associated with Ldopa therapy. VOX Sang 20: 306, 1971. Joseph C: Occurrence of positive Coombs’ test in patients treated with levodopa. N Engl J Med 286: 140, 1972. Territo MC, Peters RW, Tanaka K: Autoimmune hemolytic anemia due to levodopa therapy. JAMA 226: 1347, 1973. Poole G, Stradling P, Worlledge S: Potentially serious side effects of highdose twice-weekly rifampicin. Br Med J 3: 343,197l. Lakshminarayan S. Sahn SA, Hudson LD: Massive haemoJysiscaused by rifampicin. Br Med J 2: 282, 1973. Sivamurthy S, Frankfurt E, Levine ME: Positive antiglobulin tests in patients maintained on methadone. Transfusion 13: 418, 1973. Sherwood GK, McGinniss MH. Katon RN. DuPont RL, Webster JB: Negative direct Coombs’ tests in narcotic addicts receiving maintenance doses of methadone. Blood 40: 902, 1972. Ackroyd JF: Allergic purpura, including purpura due to foods, drugs and infections. Am J Med 14: 605, 1953. Shulman NR: Mechanism of blood cell damage by adsorption of antigen-antibody complexes. Immunopathology, 3rd International Symposium, La Jolla, 1963, p 338. Shulman NR: Mechanism of blood cell destruction in individuals sensitized to foreign antigens. Trans Assoc Am Physicians 76: 72, 1963. Shulman NR: A mechanism of cell destruction in individuals sensitized to foreign antigens and its implications in autoimmunity. Ann Intern Med 60: 506, 1964. Miescher P, Cooper N: The fixation of soluble antigen-antibody-complexes upon thrombocytes. VOX Sang 5: 138, 1960. Miescher P, Straessle R: Experimentelle Studien uber den Mechanismus der Thrombocyten-Schtidigung durch Antigen-Antikorper-Reaktionen. VOX Sang 1: 83, 1956. Worlledge SM: Immune drug-induced haemolytic anaemias. Semin Hematol6: 18 1, 1969. Croft JD Jr, Swisher SN Jr, Gilliland BC, Bakemeier RF, Leddy JP, Weed RA: Coombs’-test positivity induced by drugs. Ann Intern Med 68: 176, 1968.
March 1975
The American Journal of Medlclne
Volume 58
54.
Garratty 0, Petz LD: An evaluatlon of commercial antiglobulin sera with particular reference to their antlcomplement properties. Transfusion 11: 79, 1971. 55. Spath P, Garratty G, Petz LD: Studies on the immune response to penicillin and cephalothin in humans. I. Optimal conditions for titration of hemagglutinating penicillin and cephalothin antibodies. J lmmunol 107: 854, 197 1. 56. De Week AL: Studies on penicillin hypersensitivity. I. The specificity of rabbit “anti-penicillin” antibodies. Int Arch Allergy Appl lmmunol 21: 20, 1962. 57. Levine BB: Studies on the mechanism of the formation of the penicillin antigen. I. Delayed allergic cross-reactions among penicillin G and its degradation products. J Exp Med 112: 1131, 1960. 58. Levine BB, Ovary Z: Studies on the mechanism of the formation of the penicillin antigen. Ill. The N-(Dalpha-benzyl-penicilloyl) group as an antigenic determinant responsible for hypersensitivity to penicillin G. J Exp Med 114: 875. 1961. 59. Parker CW, De Week AL, Kern M, Eisen HN: The preparation and some properties of penicillenic acid derivatives relevant to penicillin hypersensitivity. J Exp Med 115: 803, 1962. 60. Parker CW, Shapiro J, Kern M, Eisen HN: Hypersensitivity to penicillenic acid derivatives in human beings with penicillin allergy. J Exp Med 115: 821, 1962. 81. De Week AL: Newer developments in penicillin immunochemistry. Int Arch Allergy Appl lmmunol 22: 245, 1963. 62. Levine 88. Price VH: Studies on the immunological mechanisms of penicillin allergy. II. Antigenic specificities of atlergic wheal-and-flare skin responses in patients with histories of penicillin allergy. Immunology 7: 542, 1964. 63. Levine 88. Redmond AP: Minor haptenic determinant-specific reagins of penicillin hypersensitivity in man. Int Arch Allergy Appl lmmuno135: 445, 1969. 64. Siegel BB, Levine BB: Antigenic specificities of skin-sensitizing antibodies in the sera from patients with immediate septemic allergic reactions to penicillin. J Allergy 35: 488. 1964. 65. Thiel JA, Mitchell S, Parker CW: The specificity of hemagglutination reactions in human and experimental penicillin hypersensitivity. J Allergy 35: 399, 1964. 66. Abraham GN, Petz LD, Fudenberg HH: Immunohaematological cross-allergeniclty between penicillin and cephalothin in humans. Clin Exp Immunol’3: 343, 1968. 67. Levine B, Redmond A: lmmunochemical mechanisms of penicillin induced Coombs’ positive and hemolytic anemia in man. Int Arch Allergy Appl lmmunol 31: 594, 1967. 68. Levine BB, Fellner MJ, Levytska V: Benzylpenicilloyl specific serum antibodies to penicillin in man. J lmmunol 96: 707, 1966. 69. Petz LD: Immunologic reactions of humans to cephalosporins. Postgrad Med J 47 (suppl): 64, 197 1. 70. Kerr RO, Cardamone J, Dalmasso AP, Kaplan ME: Two mechanisms of erythrocyte destruction in penicillin-induced hemolytic anemia. N Engl J Med 287: 1322, 1972. 71. Reis CA, Garratty G, Petz LD, Fudenberg HH: Massive intravascular hemolysis in penecillin-induced immune hemolytic anemia. JAMA (in press). 72. Levine BB, Redmond AP, Fellner MJ, Voss HE, Levytska V: Penicillin allergy and the heterogeneous immune responses of man to benzylpenicillin. J Clin Invest 45: 1895. 1966. 73. De Week AL, Blum G: Recent clinical and immunological aspects of penicillin allergy. Int Arch Allergy Appl Immunol 27: 221, 1965. 74. Spath P, Garratty G. Petz LO: Studies on the immune response to penicillin and cephalothin in humans. II. Immunohematologic reactions to cephalothin administration. J lmmunol 107: 860, 1971.
DRUGINDUCED
75. 76. 77.
78.
79.
80.
81. 82.
83.
84. 85.
Gralnick HR, McGinniss MH: Immune crossreactivity of penicillin and cephalothin. Nature (Lond) 216: 1026, 1967. Nesmith LW, Davis JW: Hemolytic anemia caused by penicillin. JAMA 203: 27, 1968. Abraham GN. Petz LD, Fudenberg HH: Cephalothin hypersensitivity associated with anti-cephalothin antibodies. Int Arch Allergy Appl lmmunol34: 65, 1968. Sirchia G, Merculiali F, Ferrone S: Cephalothin-treated normal red cells: a new type of PNH-like cells. Experientia 24: 495, 1968. Ferrone S. Zanella A, Schlamogna M: Red cell metabolism in positive direct Coombs test after cephalothin therapy. Experientia 27: 194, 1971. Kosakai N, Miyakawa C: Fundamental studies on the positive Coombs’ tests due to cephalosporins. Postgrad Med J 46 (suppl): 107, 1970. Kuwahara S, Mine Y, Nishiia M: lmmunogenicity of Cefazoline. Antimicrot Chemother 10: 374. 1970. Fass R, Perkins R, Saslow S: Positive direct Coombs’ tests associated with cephaloridine therapy. JAMA 213: 121, 1970. York P, Landes R, Seay L: Coombs’ posttive reactions associated with cephaloridine therapy. JAMA 206: 1086, 1968. Dameshek W: Autoimmunity: theoretical aspects. Ann NY Acad Sci 124: 6, 1965. Wurzel HA, Silverman JL: The effects of alpha-methyl-3,4-
86.
87. 88.
89.
90. 91.
92.
93.
94.
March 1975
IMMUNE HEMDLYTIC ANEMIA-GARRATTY.
PET2
dihydroxy-L-phenylalanine (methykfopa, Aldomet) on erythrocytes. Transfusion 8: 84, 1968. Gottlier AJ, Wurzel HA: Protein-quinone interaction: in vitro induction of indirect antiilobulin reactions with methyldopa. Blood 43: 85, 1974. LoBuglio AF, Jandl JH: The nature of the alpha-methyldopa red-cell antibody. N Errol J Med 276: 658. 1967. Worlledge SM: Autoantibody formation associated with methyldopa (Akfomet) therapy. Br J Haematol 16: 5, 1969. Dameshek W: Alpha-methyldopa red-cell antibody: crossreaction or forbidden clones? N Engl J Med 276: 1382. 1967. Worlledge SM: Immune drug-induced hemolytic anemias. Semin Hematol 10: 327, 1973. Vos GH, Petz LD, Fudenberg HH: Specificity of acquired haemolytic anaemia autoantibodies and their serological characteristics. Br J Haematol 19: 57, 1970. Vos GH, Petz LD, Fudenberg HH: Specificity and immunoglobulin characteristics of autoantibodies in acquired hemolytic anemia. J lmmunol 106: 1172, 1971. Vos GH, Petz LD, Garratty G, Fudenberg HH: Autoantibodies in acquired hemolytic anemia with special reference to the LW system. Blood 42: 445. 1973. Garratty G: Drug-related problems. A Seminar on Problems Encountered in Pre-Transfusion Tests, Washington, D.C., American Association of Blood Banks, 1972, p 33.
The American Journal of Medicine
Volume 58
497
GEORGE GARRATTY, F.I.M.L.T.’ LAWRENCE D. PETZ, M.D. San Francisco. California
From the Harkness Community Hospital and Medical Center, and the University of California Medical Center, San Francisco, California. This study was supported in part by Grant NS-09568 from the National Institutes of Health, by the Research Evaluation and Allocation Committee (Gilbert Fund) of the University of California, San Francisco, California, and by Eli Lilly & Co., Indianapolis, Indiana. Requests for reprints should be addressed to Mr. George Garratty. Manuscript accepted May 10, 1974. ’ Present address: Hematology-Immunology Research Unit, Room 229, Research Building, Institutes of Medical Sciences Pacific Medical Center, 2200 Webster Street. San Francisco, California 94115.
399
March 1975
Drug administration causes from 16 to 16 per cent of cases of acquired immune hemolytic anemia. The pathogenesis of erythrocyte sensitization by drug-related antibody with or without fixation of complement is variable, and there is a relationship between the responsible drug, the mechanism of red cell sensitization, clinical manifestations and laboratory methods of diagnosis. Drugs such as phenacetin and quinidine form a complex with the antidrug antibody, and the immune complex attaches to red cells usually fixing complement and causing acute intravascular hemolysis. Other drugs (e.g., penicillins), when given in large doses, coat normal red cells in vivo and in some patients a high titer IgG anti-drug antibody develops which reacts with the coated cells. Hemolytic anemia may develop with red cell destruction being primarily extravascular. Cephalosporins cause positive direct antiglobulin tests in a small percentage of pattents either by the same mechanism as penicillins or by modification of the red cell membrane leading to nonimmunologic absorption of serum proteins. Hemolytic anemia has been reported only rarely. A few drugs (notably alpha methyldopa) cause the development of autoimmune hemolytic anemia. Knowledge of clinical manifestations and laboratory aids to diagnosis is necessary to distinguish immunohematologic abnormalities caused by drugs from other causes. The administration of drugs may lead to the development of a wide variety of hematologic abnormalities including immune hemolytic anemia [l]. The clinical importance of drug-induced immune hemolytic anemia is emphasized by the fact that acquired immune hemolytic anemia among 16 per cent of our series of 200 patients [2] and 18 per cent of 79 patients of Dacie and Worlledge in 1967-1968 [3] were caused by drugs. In addition, immunohematologic abnormalities caused by drugs are a source of confusion in the blood transfusion laboratory where they may interfere with compatibility procedures. This review considers the various mechanisms leading to the development of red cell sensitization by drug-related antibodies, and the clinical and laboratory features of drug-induced immunohematologic abnormalities involving red cells. The first drug that was proved to cause a positive direct antiglobulin test (DAT) and immune hemolytic anemia was Fuadine (stibophen) which consists of 2 molecules of sodium catechol disulfonate linked by 1 atom of antimony. In 1954 [4] and 1956 [5], Harris described a patient with schistosomiasis treated with Fuadin in whom acute intravascular hemolysis developed. The patient’s
The American Journal of Medicine
Volume 58
DRUG-INDUCEDIMMUNE HEMOLYTIC ANEMIA-GARRAlTY,
serum contained a factor which agglutinated his own or normal red cells, or sensitized them to agglutination with antiglobulin serums. These reactions only took place in the presence of the drug. Since then many other drugs have been described which can cause a positive direct antiglobulin test and immune hemolytic anemia. Table I lists most of these drugs and pertinent references are cited [443]. Penicillin and alpha methyldopa (Aldome@)) are by far the most common drugs that cause immune hemolytic anemia. Methadone has also been reported as a cause of positive direct antiglobulin tests [44], although this was not confirmed by other workers [ 451. The pathogenesis of erythrocyte sensitization by drug-related antibody with or without fixation of complement is variable and is of particular importance since there is a relationship between the responsible drug, the mechanism of red cell sensitization, clinical manifestations and laboratory method of diagnosis (Table II). There seem to be four mechanisms that can cause a positive direct antiglobulin test or hemolytic anemia due to drugs. MECHANISMS OF THE POSITIVE DIRECT ANTIGLOBULIN ANEMIA DUE
TEST
AND
POSSIBLE
HEMOLYTIC
TO DRUGS
A. Immune-Complex Adsorption to Red Cells. In his studies on drug-induced purpura, Ackroyd [46] suggested that drugs may act as a hapten, combining loosely with the cell membrane and stimulating antibody production to the combined antigen. Schulman [47-491 proposed a different mechanism which is accepted by most investigators as the most probable explanation for the reactions seen with all the drugs mentioned in Table I except penicillin, cephalosporin, alpha methyldopa, L-dopa and possibly carbromal. Schulman extended the findings of Miescher and co-workers [50,51] by showing that drugs such as quinine, quinidine and stibophen have a far stronger affinity for their respective antibodies than for cell membranes. Such drugs, when present together with antibody in the patient’s plasma, will combine to form an immune complex and be adsorbed to the cell membrane, often activating complement in the process (Figure 1). It is unknown why this immune complex once formed sometimes causes red cell destruction only and other times platelet destruction only. The clinical and laboratory findings relating to immune abnormalities caused by this group of drugs follow: (1) The patient needs to take only a small quantity of the drug. (2) Acute intravascular hemolysis with hemoglobinemia and hemoglobinuria is the usual clinical presentation (16 of 19 reported cases [52]).
TABLE
PETZ
Drugs that Have Been Reported
I
Positive Direct Antiglobulin Anemia
Drug Fuadin@ (stibophen) Quinidine P-aminosalicylic acid (PAS) Quinine Phenacetin Penicillin insecticides (chlorinated hydrocarbons) Antihistamine (Antihistine@) Sulfonamides lsonicotinic acid hydrazide (isoniazid) Chlorpromazine Pyramido@ Dipyrone a-Methyldopa (Aldomet@:) L-phenylalanine mustard (melphalan) Cephalothin (Keflin@) Mefenamicacid (Ponstel@) Carbromal (Carbrital@)* Sulfonylurea derivatives Insulin L-Dopa Rifampicin * Carbromal antiglobulin
to Cause a Test and Hemolytic
Year First Described
Important References
1954 1956 1956 1958 1958 1959 1959
(4-W ]71 18,91 [lOI 18-111 [12-15)
1959 1960 1960
1171 118,191 IW’OI
1961 1961 1966 1966 1967
WI I221 I231
1967 1968 1970 1970 1970’ 1971 1972
1161
[24-271
WI [29-311 [32-34) [351 136,371 1381 (39-411 142,431
has been reported to cause positive tests but not hemolytic anemia so far.
direct
(3) Renal failure is frequent [6,8,16,21,23]. (4) The serum antidrug antibody is often IgM and capable of activating complement. (5) The direct antiglobulin test is positive, often due to the presence of complement components on the red cell surface, usually without detectable immunoglobulins. This may be explained by the fact that the immune complex does not bind very firmly to the red cells and may dissociate from the cells and be
DRUG
Y. COIUPLEX FORMATIOM
Figure 1. Mechanism of development of positive direct antiglobulin test caused by drugs such as phenacetin (see text). From Garratty [94].
March 1975
The American Journal of Medicine
Volume 58
399
DRUG-INDUCED IMMUNE HEMOLYTIC ANEMIA-GARRATTY,
TABLE
II
PET2
Correlation Between Mechanism of Red Cell Sensitization Drug-Induced lmmunohematologic Abnormalities
and Clinical and Laboratory
Features in
Serologic Evaluation
I
Immune complex formation (drug and antidrug antibody)
II
Drug adsorbed onto red cell membrane; reacts with high titer serum drug antibody
III
Membrane modification (nonimmunologic absorption of proteins)
IV Unknown
Clinical Findings
Prototype Drugs
Mechanism
Quinidine, phenacetin
Usually only complement components detected but IgG can be present
Drug + patient’s serum + RBC (especially enzyme-treated) hemolysis, agglutination, or sensitization; antibody frequently IgM and capable of fixing complement; RBC eluate often nonreactive
Penicillins, cephalosporins
Large doses of penicillin (10 million units or more daily); other manifestations of allergy not necessarily present; usually subacute extravascular hemolysis; penicillin one of most common causes of drug-induced immune hemolysis; rare case of immune hemolytic anemia caused by cephalosporins
Strongly positive (IgG) when hemolytic anemia present; rarely weaker complement sensitization is also present
Drug-coated RBC + serum + agglutination or sensitization (rarely hemolysis); high titer antibody associated with hemolytic anemia is always IgG; RBC eluate reacts only with antibiotic-coated RBC
Cephalosporins
No cases of cephalosporin-induced hemolytic anemia caused by nonimmunologic mechanisms
Positive with antiserums to a variety of serum proteins
Drug-coated RBC + serum + sensitization to antiglobulin serums in low titer (nonimmunologic protein absorption); RBC eluate nonreactive
a-methyldopa
Hemolysis in 0.8 percent of patients taking drug for at least 3 mo; gradual onset of hemolytic anemia; one of most common causes of drug-induced immune hemolysis
Strongly positive (IgG) when hemolytic anemia present; only rarely cells are sensitized with complement as well
Antibody sensitizes normal RBC without drug; antibody in serum and eluate identical to that found in warm antibody AIHA; no in vitro relationship to drug demonstrable
to react with other cells. This in turn may explain why such a small amount of drug complex can cause so much red cell destruction [53]. Furthermore, red cell sensitization by IgM antibodies is not readily detectable by the antiglobulin test [54]. Some reports of negative direct antiglobulin tests, particularly in the earlier literature, may be due to inadequate anticomplement properties of the antiglobulin serum [54]. (6) In vitro reactions (agglutination, lysis and/or sensitization to antiglobulin serums) are only ob-
March 1975
Antibody Identification
Small doses of drug; acute intravascular hemolysis usual; renal failure common; thrombocytopenia occasionally
free
400
Direct Antiglobulin Test
The American Journal of Medicine
served when patient’s serum, drug and red cells are all incubated together. B. Drug Adsorbed Onto Red Cells. In 1958, Ley et al. [ 121 first reported the presence of circulating penicillin antibody. In the course of routine blood bank testing, a serum was encountered which agglutinated the red cells of 25 group 0 members of an antibody identification panel. This panel had been stored in a preservative solution containing penicillin. The same panel with no penicillin present did not
Volume 58
DRUGlNDUCED
react with the serum. In the paper, a patient was described who had a positive direct antiglobulin test and hemolytic anemia after receiving penicillin. In 1966, Petz and Fudenberg [ 141 defined the role of penicillin in causing immune hemolytic anemia. Penicillin is one of the few drugs that binds firmly to proteins. Penicillin will combine with the protein on normal red cell membranes both in vivo and in vitro. Spath et al. [55] performed a detailed study of numerous variables affecting sensitization of red cells by penicillin in vitro (pH, drug concentration, temperature and duration of incubation, etc.). These workers defined optimal and very practical methods for preparing penicillin-treated red cells and for using such cells for penicillin antibody detection. The drug cannot be removed from red cells in vitro even by multiple washes in saline solution. This is in contrast to the drugs mentioned in section A, in which it is not possible to prepare drug-coated cells for antibody detection. Considerable experimental work has demonstrated that the immunogenicity of penicillin is due to its ability to react chemically with tissue proteins to form several different haptenic groups [56-651. The major haptenic determinant is the benzylpenicilloyl (BPO) group. It is not clear yet whether the BP0 group is formed by the direct reaction of penicillin with amino groups of protein, or through the intermediate formation of benzylpenicillenic acid. Penicillin can be detected on the red cell membranes of all patients receiving large doses of penicillin [66,67]. If a sensitive technic is used [55,68], over 90 per cent of unselected serums can be shown to contain penicillin (BPO) antibodies [68]. Most serums contain IgM antibodies alone (approximately 80 per cent); approximately 13 per cent contain IgG antibodies as well. These antibodies are usually easily neutralized by BP0 hapten. The antibodies associated with immune hemolytic anemia due to penicillin are IgG, and are not easily neutralized by BP0 hapten. The high percentage of penicillin antibodies in the normal population is probably due to the continual exposure to penicillin in our modern environment. In approximately 3 per cent of the patients who receive massive doses of intravenous penicillin, positive direct antiglobulin tests will develop [66,69]; and in a small percentage of these, hemolytic anemia will develop. The mechanism of the positive direct antiglobulin test and hemolytic anemia seems clear [14,67]. The drug is adsorbed to the red cells and the antipenicillin present in the patient’s plasma will react with the penicillin on the red cells (Figure 2). The quantity of penicillin (BPO) antibody sensitizing the red cell is limited by the number of BP0 haptenic
IMMUNE HEMOLYTIC ANEMIA-GARRATTY,
PET2
I? 6.C ANTIBODY
% Figure 2. antiglobulin ratty [94].
Mechanism of development of positive direct test caused by penicillins (see text). From Gar-
groups on the cell, the plasma concentration of BP0 specific antibodies and the avidity of the antibodies Complement is not usually involved in this reaction and intravascular hemolysis rarely occurs [ 70,7 11. The IgG sensitized red cells are probably removed extravascularly by the reticuloendothelial system in the same way as Rhesus (IgG) sensitized cells. There is no direct correlation between the presence of IgG and IgM penicillin hemagglutinating antibodies and allergic reactions. Most workers have found no correlation at all, but a few have found that high titer IgG antibodies occur more often in the allergic group [72,73]. Perhaps a better correlation will be found with IgE penicillin antibodies. The clinical and laboratory features of penicillininduced immune hemolytic anemia are quite constant and are as follows: (1) Hemolysis typically develops only in patients who are receiving very large doses of penicillin (at least 10 million units daily for a week or more). (2) Hemolytic anemia is usually less acute in onset than that caused by drugs of group A but may be lifethreatening if the etiology is unrecognized and penicillin administration is continued. (3) A high titer IgG penicillin antibody is present in the serum. (4) The direct antiglobulin test is strongly positive due to sensitization with IgG. Barely, complement components are detected as well, and complement activation may contribute to the immune hemolysis [70,71]. (5) Antibody eluted from the patient’s red cells will
react only against penicillin-treated red cells. (6) Cessation of penicillin therapy is followed by complete recovery, but hemolysis of decreasing severity may persist for several weeks.
March 1975
The American Journal of Medicine
Volume 59
401
DRUGlNDUCED IMMUNE HEMOLYTICANEMIA-GARRATTY. PET2
.
RETICULOCYTES
PENICILLIN
CAPA
TITER
8 6t
II/4 IO/5 12/3 O/I5 Figure 3. Clinical course of a patient with penicillin-induced immunohemolytic anemia. CAPA refers to circulating antipenicillin antibody. From Petz and Fudenberg [ 141.
(7) Other manifestations of penicillin allergy are not necessarily present. Figure 3 illustrates the course of a characteristic case of penicillin-induced hemolytic anemia. Other drugs known to bind firmly to red cells and cause positive direct antiglobulin tests by the same mechanism as penicillin are cephalothin [30,31] and possibly carbromal [35]. Cephalothin [29] has been described as a rare cause of immune hemolytic anemia but carbromal so far has not.
402
March 1975
The American Journal of Medicine
C. Modification of Red Cell Membrane by Drug Allowing Nonimmunologic Absorption of Protein. In 1967, Gralnick [30] and Molthan [31], with their coworkers, described positive direct antiglobulin tests following the administration of cephalothin (Keflinm). Three mechanisms for the positive direct antiglobulin test due to cephalothin have been proposed: (1) The cell membrane may be modified by the drug so that the cell now takes up proteins nonimmunologically [30,31,74] (Figure 4). Cephalothin-sensi-
Volume 58
DRUWNDUCED
tized cells incubated in normal plasma in vitro become coated with albumin, IgG, IgA, IgM, alpha, antitrypsin, alpha, macroglobulin, C3, C4 and fibrinogen [2,74]. (2) The drug may combine with the red cell membrane, as does penicillin, and being so closely related chemically to penicillin, cross-reacting penicillin antibodies will now combine with the drug [30,66,7476]. (3) The drug may combine with the cell membrane in a similar fashion to penicillin, and specific anticephalothin antibodies react with the drug on the cell [29,74,77]. Spath et al. [74] confirmed that all three of these mechanisms may lead to a positive direct antiglobulin test. Whether the positive direct antiglobulin test has resulted from nonimmunologic protein absorption or by fixation of antibody may be determined in an individual case as outlined in Table II. Other data confirm that cephalothin will modify the red cell membrane. Indeed, Sirchia, Ferrone and coworkers [78,79] have shown that normal red cells after treatment with cephalothin in vitro, resemble cells from a patient with paroxysmal nocturnal hemoglobinuria. In vitro studies showed a low acetylcholinesterase activity, reduced oxygen uptake in the presence of methylene blue and a positive Ham’s acid serum lysis test. They could show no enzymatic or metabolic abnormalities of red cells from patients with positive direct antiglobulin tests following cephalothin administration. Gralnick et al. [29] described two patients with immune hemolytic anemia following cephalothin administration, but this is a much rarer event than hemolytic anemia following penicillin therapy. Other cephalosporin drugs closely related to cephalothin such as cephalexin [80,81], cephaloridine [82,83] and cefazolin [81] have also been described as a cause of positive direct antiglobulin tests. Cephalothin, like penicillin, will bind firmly to red cells. These cells can be washed many times in vitro without the drug being removed. Cephalothin-treated cells can be used in a similar fashion to penicillintreated cells in the detection of cephalothin or penicillin cross-reacting antibodies [55]. Serums from unselected donors show IgM antibodies to cephalothin-treated cells in 88 per cent, IgG antibodies in 10 per cent, and both IgG and IgM antibodies in 15 per cent [74]. Spath et al. [74] were able to confirm and extend the findings of Gralnick [ 751, Nesmith 1761, Abraham [77] and co-workers in showing that penicillin antibodies may cross react with cephalothin-treated red cells: thus high titer penicillin antibodies may react with cephalothin-treated cells both in vivo and in vitro. In addition, specific anticephalothin antibodies
IMMUNE HEMOLYIK: ANEMIA-QARRArTY.
PETZ
MEM0f?AlVE MODIFCA77Oht
04
0
One of the mechanisms of development of a positive direct antiglobulin test caused by cephalosporins (see text). From Garratty (941.
Figure 4.
exist, and these can sometimes be shown to be present in eluates from red cells demonstrating a positive direct antiglobulin test due to cephalothin [29,74]. In the original reports, from 40 to 75 per cent of patients who received cephalothin had positive direct antiglobulin tests [30,31], but in further studies recently, Spath, Garratty and Petz [55,74] found positive tests in only 3 per cent. These latter workers postulated that the differences between their findings and those of earlier workers may be due to the following reasons: (1) Dosage of drug: The dosage and duration of therapy are likely to influence the incidence of positive direct antiglobulin tests. The patients studied by Spath et al. were unselected patients in a community hospital, where the mean daily cephalothin dosage was 6.3 g and the mean duration of therapy 5.5 days. The patients with a positive direct antiglobulin test received 7 g for 10 days. The earlier workers reported a higher average drug dosage (e.g., patients with positive direct antiglobulin tests repeated by Gralnick et al. [30] receiving from 8 to 14 g of cephalothin daily). (2) Patient selection: There is evidence to suggest that the percentage of positive direct antiglobulin tests due to cephalothin are higher in azotemic patients. Gralnick et al. [30] suggested a relationship between hypoalbuminemia and positive direct antiglobulin tests following cephalothin administration. (3) Antiglobulin technic: Cephalothin-treated red cells absorb proteins nonspecifically. Since such cells are coated by numerous proteins, the number of positive direct antiglobulin tests obtained will be proportional to the spectrum of antibodies in the antiglobulin serums used. For instance, cephalothintreated cells take up albumin very efficiently. On screening commercial antiglobulin serums and antiserums prepared in our own laboratory [2], we found
March 1975
The Am&can
Journal of Modlclne
Volume 59
403
DRUG-INDUCED IMMUNE HEMOLYTIC ANEMIA-GARRATTY,
PET2
that most commercial antiglobulin serums, as sold for blood banking purposes, contained very little or no antialbumin. One or two contained a strongly reacting antialbumin. If the latter antiglobulin serums were used, a higher percentage of positive direct antiglobulin tests would be obtained. D. Red Cell Autoantibodies Induced by Drugs: Unknown Mechanism. In 1966, Carstairs, Worlledge and co-workers [26,27] described the first positive direct antiglobulin tests and autoimmune hemolytic anemia due to the drug alpha methyldopa (Aldomet). This reaction was different from any of the other types described in sections A, B or C, as in patients receiving this drug antibodies were made that acted directly with normal red cells, even in the absence of the drug. Indeed, the antibodies reacted similarly to those seen in the idiopathic autoimmune hemolytic anemias, in that they often showed some Rhesus blood group specificity. In contrast, of course, the antibodies in patients receiving the drugs mentioned in sections A, B and C were directed against the drug itself and not intrinsic red cell antigens. Immune mechanisms described in sections A, B and C nevertheless cause red cells to be affected as if antibody had been specifically directed against them (these mechanisms of sensitization are often referred to as “innocent bystander” mechanisms [84]). The precise role of alpha methyldopa in the pathogenesis of red cell autosensitization is at present unclear although several theories have been suggested: (1) Adsorption of drug or drug-protein complex to red cells: Wurzel and Silverman [85] showed that normal washed red cells and red cells in plasma would adsorb alpha methyldopa and that if normal whole blood was incubated in a high concentration (e.g., 0.5 mg/ml) of drug for several days, in vitro, then a positive antiglobulin test was obtained. Gottlieb and Wurzel [86] recently demonstrated that alpha methyldopa or one of its metabolites could be covalently bound to gamma globulin when drug and plasma were incubated together for 2 days or more. If red cells were then added, a positive antiglobulin test occurred. No Rhesus specificity was noted. Other workers [2,27,87] have not been able to demonstrate a positive antiglobulin test after incubation of red cells in alpha methyldopa or its metabolites. Furthermore LoBuglio and Jandl [87] demonstrated that alpha methyldopa had very little affinity for plasma proteins or red cell membranes, although it should be noted that the incubation period employed was shorter than that used by Gottlieb and Wurzel [86]. (2) Altered autoantigens: Worlledge et al. [27,88] have postulated that the drug or one of its metabolites may combine with the red cell membrane or enter the red cell and in some way alter the Rhesus antigens so that the normal immune system no long-
494
March 1975
The American Journal of Medicine
er recognizes them as “self.” The incorporation may occur at the normoblast or reticulocyte stage, thus explaining the delay in development of a positive direct antiglobulin test. (3) Forbidden clones: Dameshek [89] suggested that alpha methyldopa may result in errors or aberrations in the proliferation of normal lymphocytes, thus producing clones of abnormal but nevertheless immunologically competent cells. These cells, although functional, would be “intolerant” of, or fail to recognize as “self,” normal antigens, especially those of the ubiquitous red cell. However, antibodies in patients receiving alpha methyldopa are “polyclonal” (i.e., kappa and lambda light chains and distinct Hchain types) [24]. This finding necessitates for clonal interpretation the proposition that at least three different clones of immunocytes lost the ability to recognize the Rhesus locus of the red cell itself, which seems unlikely. Two other drugs have recently been described as acting in a similar fashion to alpha methyldopa. One is a closely related drug L-dopa, which was described as causing positiveadirect antiglobulin tests in about 6 to 9 per cent of the patients receiving the drug [39,40]. Recently, this drug has been reported as a cause of hemolytic anemia [41]. The other drug, mefenamic acid, is unrelated to alpha methyldopa and has been described as causing positive direct antiglobulin tests and autoimmune hemolytic anemia in three separate reports of five patients [32-341. The clinical and laboratory characteristics of alpha methyldopa-induced abnormalities are as follows: (1) Direct antiglobulin test is positive in 15 per cent of patients receiving alpha methyldopa (reports vary from 10 to 36 per cent). Incidence seems to vary between racial groups, being highest in Caucasians, lower in Chinese and almost absent in Blacks
WI. (2) The direct antiglobulin test usually becomes positive after 3 to 6 months of treatment. (It is interesting that this delay is not shortened when administration of the drug is resumed in a patient who previously had a positive test.) (3) The development of the positive direct antiglobulin test appears to be dose-dependent. About three times as many patients (36 per cent) have positive tests when taking more than 2 g of the drug daily than patients taking less than 1 g daily (11 per cent); 19 per cent had a positive test with 1 to 2 g daily. (4) The positive direct antiglobulin test is due to red cell sensitization with IgG, and all patients with hemolytic anemia have a very strongly positive direct antiglobulin test. Rarely, red cells are weakly sensitized with complement as well [ 21. (5) Various reports indicate that from 0 to 5 per
Volume 59
DRUG-INDUCEDIMMUNE HEMOLYTIC ANEMIA-GARRATTY. PET2
cent of patients receiving alpha methyldopa have hemolytic anemia. The cumulative evidence is 0.8 per cent [52]. (6) Worlledge et al. [27] commented that the autoantibodies associated with alpha methyldopa commonly show Rhesus specificity. However, such specificity is based on subtle differences in titer and/or cross absorption and elution technics [24,91-931 using cells of varying genotypes, rather than the clear patterns one sees with Rhesus alloantibodies. Worlledge herself states in a later article [52] that these autoantibodies are similar to those seen in association with idiopathic warm autoimmune hemolytic anemia. This has been our experience, too, in our own series of 200 cases of immune hemolytic anemias studied since 1988. In 20 cases of immune hemolytic anemia associated with alpha methyldopa, the serology was indistinguishable from idiopathic warm autoimmune hemolytic anemia. The specificity of the autoantibody is often associated with the Rh locus, but the definition of this specificity is depen-
dent on the extent of the technical investigation and the availability of rare red cell phenotypes (e.g., Rh,,rr. LW negative, U negative), without which the autoantibody often appears nonspecific. Vos et al. [93] discuss some of these variables in a recent publication. We believe differences in opinion in the literature [86] on the blood group specificity of these autoantibodies can be attributed directly to these differences in technic. (7) The positive direct antiglobulin reaction gradually becomes negative once the administration of alpha methyldopa is stopped. This may take from 1 month to 2 years. Fortunately, the hematologic values usually improve strikingly within the first week or two. Alpha methyldopa is one of the most common causes of positive direct antiglobulin tests due to drugs. Worlledge [52] has reported that the total of alpha methyldopa-induced autoimmune hemolytic anemias exceeds the total of all other drug-induced immune hemolytic anemias so far described.
REFERENCES 1. 2. 3. 4. 5. 6.
7.
8.
9.
10.
11. 12.
13.
14.
15.
16.
Huguley CM, Lea JW, Butts JA: Adverse hematologic reactions to drugs. Proor Hemato15: 105. 1966. Garratty G, Pe& LD: Unpublished observations, 1974. Dacie JV. Worlledge SM: Auto-immune hemolytic anemias. Progr Hemato16: 82, 1989. Harris JW: Studies on the mechanism of a drug-induced hemolytic anemia. J Lab Clin Med 44: 809, 1954. Harris JW: Studies on the mechanism of a drug-induced hemolytic anemia. J Lab Clin Med 47: 760, 1958. De Torregrosa MVV, Rodriguez Rosada AL, Montiila E: Hemolytic anemia secondary to stibophen therapy. JAMA 186: 598, 1983. Freedman AL, Tarr PS, Brody EA: Hemolytic anemia due to quinidine: observations on its mechanism. Am J Med 20: 806, 1956. MacGibbon BH, Loughridge LW, Hourihane DO’B, Boyd DW: Autoimmune haemolytic anaemia with acute renal failure due to phenacetin and paminosalicylic acid. Lancet 1: 7, 1960. Van Loghem JJ: Autoimmune haemolytic anaemia with renal failure due to phenacetin and PAS. Lancet 1: 434, 1960. Muirhead EE. Halden ER, Groves M: Drug-dependent Coombs’ (antiglobulin) test and anemia. Arch Intern Med 101: 87, 1958. Dausset J, Contu L: A case of haemolytic anaemia due to phenacetin allergy. VOX Sang 9: 599, 1964. Ley AB, Harris JP, Brinkley M, Liles B, Jack JA. Cahan A: Circulating antibodies directed against penicillin. Science 127: 1118, 1958. Clayton EM, Aftshuler J, Bove JR: Penicillin antibody as a cause of positive direct antiglobulin tests. Am J Clin Patho144: 648, 1965. Petz LD, Fudenberg HH: Coombs-positive hemolytic anemia caused by penicillin administration. N Engl J Med 274: 171, 1966. Swanson MA, Chanmougan D, Schwartz RS: Immunohemolytic anemia due to antipenicillin antibodies. N Engf J Med 274: 178, 1966. Muirhead EE, Groves M, Guy R. Halden ER, Bass RK: Ac-
17.
18.
19.
20.
21. 22.
23.
24.
25.
26.
27.
28.
March 1975
quired hemolytic anemia, exposures to insecticides and positive Coombs’ test dependent on insecticide preparations. VOX Sang 4: 277, 1959. Mahssenet R, Dreyfus B, Antoine B: Anemic hemolytique immunologique due a I’antistine. Proceedings of the 7th Congress International Society of Haematology, Rome, 1958, vol 2. New York, Grune & Stratton, 1960. p 614. Fukase M, Nakano H. Konda 5, Tajima H, Takeuchi S, Tsunematsu T, Nishida R. Kanagawa K, Sugano A. Watanabe K, Suzuka H, lnagaki M: Two cases of drug-allergy followed by acquired autoimmune hemolytic anemia. Acta Haematol Jap 23: 70, 1960. Tajima l-l: Clinical studies on hemolytic anemia. I. Autoimmune hemolytic anemia. Acta Haematol Jap 23: 188, 1960. Robinson MG, Foadi M: Hemolytic anemia with positive Coombs’ test. Association with isoniazid therapy. JAMA 208: 658, 1969. Lindberg LG, Norden A: Severe hemolytic reaction to chlorpromazine. Acta Med Stand 170: 195, 1961. Bernasconi C, Bedartia G, Pollini G, Sartori S: Studio del meccanismo di emolisi in un case di anemia emolitica acquista da piramidone. Haematologica (Pavia) 46: 897, 1981. Lay WH: Drug-induced haemolytic reaction due to antibodies against the erythrocyte/dipyrone complex. VOX Sang 11: 801, 1966. Bakemeier RF. Leddy JP: Erythrocyte autoantibody associated with alpha-methyldopa: heterogeneity of structure and specificity. Blood 32: 1, 1968. Carstairs KC, Breckenridge A, Dollery CT, Worlledge SM: Incidence of a positive direct Coombs’ test in patients on a-methyldopa. Lancet 2: 133. 1966. Carstairs K. Worlledge S, Dollery CT, Breckenridge A: Methyldopa and haemolytic anaemia. Lancet 1: 201, 1966. Worlledge SM, Carstairs KC, Dacie JV: Autoimmune haemolytic anaemia associated with a-methyldopa therapy. Lancet 2: 135, 1966. Eyster ME: Melphalan (Alkeran) erythrocyte agglutinin and
The American Journal of Medlclne
Volume 58
405
ORUGINDUCEO
29.
30.
31.
32. 33.
34.
35.
36.
37.
36. 39.
40. 41.
42.
43. 44.
45.
46. 47.
46.
49.
50.
51.
52. 53.
406
IMMUNE HEMOLYTIC ANEMIA-GARRATTY.
PET2
hemolytic anemia. Ann Intern Med 66: 573, 1967. Gralnick HR, McGlnnlss MH, Elton W, McCurdy P: Hemolytlc anemia associated with cephalothin. JAMA 217: 1143, 1971. Gralnick HR, Wright LD, McGinniss MH: Coombs’ positive reactions associated wlth sodium cephalothin therapy. JAMA 199: 725, 1967. Molthan L, Reidenberg MM, Eichman MF: Positive direct Coombs’ tests due to cephalothin. N Engl J Med 277: 123, 1967. Farid NR. Johnson RJ, Low WT: Haemolytic reaction to mefenamic acid. Lancet 2: 382, 197 1. Robertson JH, Kennedy CC, Hill CM: Haemolytic anaemia associated with mefenamic acid. Ir J Med Sci 140: 226, 1971. Scott GL, Myles AB. Bacon PA: Autoimmune haemolytic anaemia and mefenamic acid therapy. Br Med J 3: 534. 1968. Stefanini M, Johnson NL: Positive antihuman globulin test in patients receiving carbromal. Am J Med Sci 259: 49, 1970. Logue GL, Boyd AE, Rosse WF: Chlorpropamide-induced immune hemolytic anemia. N Engl J Med 283: 900, 1970. Bird GWG, Eeles GH, Litchfield JA. Rahman M, Wingham J: Haemolytic anaemia with antibodies to tolbutamide and phenacetin. Br Med J 1: 728, 1972. Faulk WP, Tomsovic EJ, Fudenberg HH: Insulin resistance in juvenile diabetes mellitus. Am J Med 49: 133, 1970. Henry RE, Golberg LS, Sturgeon F, Ansel RD: Serologic abnormalities associated with Ldopa therapy. VOX Sang 20: 306, 1971. Joseph C: Occurrence of positive Coombs’ test in patients treated with levodopa. N Engl J Med 286: 140, 1972. Territo MC, Peters RW, Tanaka K: Autoimmune hemolytic anemia due to levodopa therapy. JAMA 226: 1347, 1973. Poole G, Stradling P, Worlledge S: Potentially serious side effects of highdose twice-weekly rifampicin. Br Med J 3: 343,197l. Lakshminarayan S. Sahn SA, Hudson LD: Massive haemoJysiscaused by rifampicin. Br Med J 2: 282, 1973. Sivamurthy S, Frankfurt E, Levine ME: Positive antiglobulin tests in patients maintained on methadone. Transfusion 13: 418, 1973. Sherwood GK, McGinniss MH. Katon RN. DuPont RL, Webster JB: Negative direct Coombs’ tests in narcotic addicts receiving maintenance doses of methadone. Blood 40: 902, 1972. Ackroyd JF: Allergic purpura, including purpura due to foods, drugs and infections. Am J Med 14: 605, 1953. Shulman NR: Mechanism of blood cell damage by adsorption of antigen-antibody complexes. Immunopathology, 3rd International Symposium, La Jolla, 1963, p 338. Shulman NR: Mechanism of blood cell destruction in individuals sensitized to foreign antigens. Trans Assoc Am Physicians 76: 72, 1963. Shulman NR: A mechanism of cell destruction in individuals sensitized to foreign antigens and its implications in autoimmunity. Ann Intern Med 60: 506, 1964. Miescher P, Cooper N: The fixation of soluble antigen-antibody-complexes upon thrombocytes. VOX Sang 5: 138, 1960. Miescher P, Straessle R: Experimentelle Studien uber den Mechanismus der Thrombocyten-Schtidigung durch Antigen-Antikorper-Reaktionen. VOX Sang 1: 83, 1956. Worlledge SM: Immune drug-induced haemolytic anaemias. Semin Hematol6: 18 1, 1969. Croft JD Jr, Swisher SN Jr, Gilliland BC, Bakemeier RF, Leddy JP, Weed RA: Coombs’-test positivity induced by drugs. Ann Intern Med 68: 176, 1968.
March 1975
The American Journal of Medlclne
Volume 58
54.
Garratty 0, Petz LD: An evaluatlon of commercial antiglobulin sera with particular reference to their antlcomplement properties. Transfusion 11: 79, 1971. 55. Spath P, Garratty G, Petz LD: Studies on the immune response to penicillin and cephalothin in humans. I. Optimal conditions for titration of hemagglutinating penicillin and cephalothin antibodies. J lmmunol 107: 854, 197 1. 56. De Week AL: Studies on penicillin hypersensitivity. I. The specificity of rabbit “anti-penicillin” antibodies. Int Arch Allergy Appl lmmunol 21: 20, 1962. 57. Levine BB: Studies on the mechanism of the formation of the penicillin antigen. I. Delayed allergic cross-reactions among penicillin G and its degradation products. J Exp Med 112: 1131, 1960. 58. Levine BB, Ovary Z: Studies on the mechanism of the formation of the penicillin antigen. Ill. The N-(Dalpha-benzyl-penicilloyl) group as an antigenic determinant responsible for hypersensitivity to penicillin G. J Exp Med 114: 875. 1961. 59. Parker CW, De Week AL, Kern M, Eisen HN: The preparation and some properties of penicillenic acid derivatives relevant to penicillin hypersensitivity. J Exp Med 115: 803, 1962. 60. Parker CW, Shapiro J, Kern M, Eisen HN: Hypersensitivity to penicillenic acid derivatives in human beings with penicillin allergy. J Exp Med 115: 821, 1962. 81. De Week AL: Newer developments in penicillin immunochemistry. Int Arch Allergy Appl lmmunol 22: 245, 1963. 62. Levine 88. Price VH: Studies on the immunological mechanisms of penicillin allergy. II. Antigenic specificities of atlergic wheal-and-flare skin responses in patients with histories of penicillin allergy. Immunology 7: 542, 1964. 63. Levine 88. Redmond AP: Minor haptenic determinant-specific reagins of penicillin hypersensitivity in man. Int Arch Allergy Appl lmmuno135: 445, 1969. 64. Siegel BB, Levine BB: Antigenic specificities of skin-sensitizing antibodies in the sera from patients with immediate septemic allergic reactions to penicillin. J Allergy 35: 488. 1964. 65. Thiel JA, Mitchell S, Parker CW: The specificity of hemagglutination reactions in human and experimental penicillin hypersensitivity. J Allergy 35: 399, 1964. 66. Abraham GN, Petz LD, Fudenberg HH: Immunohaematological cross-allergeniclty between penicillin and cephalothin in humans. Clin Exp Immunol’3: 343, 1968. 67. Levine B, Redmond A: lmmunochemical mechanisms of penicillin induced Coombs’ positive and hemolytic anemia in man. Int Arch Allergy Appl lmmunol 31: 594, 1967. 68. Levine BB, Fellner MJ, Levytska V: Benzylpenicilloyl specific serum antibodies to penicillin in man. J lmmunol 96: 707, 1966. 69. Petz LD: Immunologic reactions of humans to cephalosporins. Postgrad Med J 47 (suppl): 64, 197 1. 70. Kerr RO, Cardamone J, Dalmasso AP, Kaplan ME: Two mechanisms of erythrocyte destruction in penicillin-induced hemolytic anemia. N Engl J Med 287: 1322, 1972. 71. Reis CA, Garratty G, Petz LD, Fudenberg HH: Massive intravascular hemolysis in penecillin-induced immune hemolytic anemia. JAMA (in press). 72. Levine BB, Redmond AP, Fellner MJ, Voss HE, Levytska V: Penicillin allergy and the heterogeneous immune responses of man to benzylpenicillin. J Clin Invest 45: 1895. 1966. 73. De Week AL, Blum G: Recent clinical and immunological aspects of penicillin allergy. Int Arch Allergy Appl Immunol 27: 221, 1965. 74. Spath P, Garratty G. Petz LO: Studies on the immune response to penicillin and cephalothin in humans. II. Immunohematologic reactions to cephalothin administration. J lmmunol 107: 860, 1971.
DRUGINDUCED
75. 76. 77.
78.
79.
80.
81. 82.
83.
84. 85.
Gralnick HR, McGinniss MH: Immune crossreactivity of penicillin and cephalothin. Nature (Lond) 216: 1026, 1967. Nesmith LW, Davis JW: Hemolytic anemia caused by penicillin. JAMA 203: 27, 1968. Abraham GN. Petz LD, Fudenberg HH: Cephalothin hypersensitivity associated with anti-cephalothin antibodies. Int Arch Allergy Appl lmmunol34: 65, 1968. Sirchia G, Merculiali F, Ferrone S: Cephalothin-treated normal red cells: a new type of PNH-like cells. Experientia 24: 495, 1968. Ferrone S. Zanella A, Schlamogna M: Red cell metabolism in positive direct Coombs test after cephalothin therapy. Experientia 27: 194, 1971. Kosakai N, Miyakawa C: Fundamental studies on the positive Coombs’ tests due to cephalosporins. Postgrad Med J 46 (suppl): 107, 1970. Kuwahara S, Mine Y, Nishiia M: lmmunogenicity of Cefazoline. Antimicrot Chemother 10: 374. 1970. Fass R, Perkins R, Saslow S: Positive direct Coombs’ tests associated with cephaloridine therapy. JAMA 213: 121, 1970. York P, Landes R, Seay L: Coombs’ posttive reactions associated with cephaloridine therapy. JAMA 206: 1086, 1968. Dameshek W: Autoimmunity: theoretical aspects. Ann NY Acad Sci 124: 6, 1965. Wurzel HA, Silverman JL: The effects of alpha-methyl-3,4-
86.
87. 88.
89.
90. 91.
92.
93.
94.
March 1975
IMMUNE HEMDLYTIC ANEMIA-GARRATTY.
PET2
dihydroxy-L-phenylalanine (methykfopa, Aldomet) on erythrocytes. Transfusion 8: 84, 1968. Gottlier AJ, Wurzel HA: Protein-quinone interaction: in vitro induction of indirect antiilobulin reactions with methyldopa. Blood 43: 85, 1974. LoBuglio AF, Jandl JH: The nature of the alpha-methyldopa red-cell antibody. N Errol J Med 276: 658. 1967. Worlledge SM: Autoantibody formation associated with methyldopa (Akfomet) therapy. Br J Haematol 16: 5, 1969. Dameshek W: Alpha-methyldopa red-cell antibody: crossreaction or forbidden clones? N Engl J Med 276: 1382. 1967. Worlledge SM: Immune drug-induced hemolytic anemias. Semin Hematol 10: 327, 1973. Vos GH, Petz LD, Fudenberg HH: Specificity of acquired haemolytic anaemia autoantibodies and their serological characteristics. Br J Haematol 19: 57, 1970. Vos GH, Petz LD, Fudenberg HH: Specificity and immunoglobulin characteristics of autoantibodies in acquired hemolytic anemia. J lmmunol 106: 1172, 1971. Vos GH, Petz LD, Garratty G, Fudenberg HH: Autoantibodies in acquired hemolytic anemia with special reference to the LW system. Blood 42: 445. 1973. Garratty G: Drug-related problems. A Seminar on Problems Encountered in Pre-Transfusion Tests, Washington, D.C., American Association of Blood Banks, 1972, p 33.
The American Journal of Medicine
Volume 58
497
