An unusual case of autoimmune hemolytic anemia with reticulocytopenia, erythroid dysplasia, and an IgG2 autoanti-U

BACKGROUND: Autoantibodies with anti-U specificity, usually in combination with autoantibodies of other specificities, have occasionally been identified in association with autoimmune hemolytic anemia. A case of life-threatening autoimmune hemolytic anemia, characterized by several atypical features, including apparent intravascular hemolysis associated with an IgG2 anti-U, reticulocytopenia, and bone marrow dyserythropoiesis is described. CASE REPORT: A 36-year-old man with a severe case of acute-onset autoimmune hemolytic anemia was admitted to another hospital; he had a hematocrit of 15 percent, elevated bilirubin and lactate dehydrogenase, and positive direct and indirect antiglobulin tests. He received 7 units of incompatible red cells without improvement in hematocrit, and he was transferred to University Hospitals of Cleveland (OH). He was jaundiced and became syncopal in the sitting position. His serum was reddish pink; he had a hematocrit of 11.8 percent and a reticulocyte count of 2.5 percent. No spherocytes were observed in the peripheral blood smear. Shortly after admission, the hematocrit fell to 6.9 percent. He was given 3 units of “least-incompatible” red cells and was started on prednisone, with little improvement. An IgG2 autoanti-U was detected in his serum. Seven units of U- red cells were transfused over the next 4 days. The hematocrit improved to 23 percent and continued to rise without further transfusion. A bone marrow examination, initially revealing erythroid hyperplasia accompanied by dyserythropoiesis, became morphologically normal. Drug studies failed to show evidence of drug-related hemolysis. He remains well 2 years after discharge without evidence of recurrent hemolysis. CONCLUSION: Severe life-threatening autoimmune hemolytic anemia, in this instance induced by an autoanti-U, may be associated with IgG2 autoantibody and characterized by apparent intravascular hemolysis and bone marrow dyserythropoiesis. Early treatment with U- blood, in addition to steroids, may be beneficial.     

A "normal" individual with a positive direct antiglobulin test: case complicated by pregnancy and unusual autoantibody specificity

A "normal" individual with a positive, direct antiglobulin test is described. In common with many other "normal" persons in whom a similar finding has been made, there was no evidence of an increased rate of in vivo red blood cell destruction in this patient. The patient successfully completed a pregnancy during the time that detailed serologic studies on her autoantibodies were being performed. Although the maternal autoantibodies were demonstrable in both an eluate made from the red blood cells of her newborn infant, and in the cord serum, there was no reason to believe that the antibodies caused red blood cell destruction in the infant. The case was of further interest because of the specificities of some of the autoantibodies. Although the mother and child were both C-negative, eluates from their red blood cells contained what ostensibly appeared to be anti-C. Studies that showed that the antibody could be totally adsorbed with C-negative, as well as C-positive red blood cells, proved that this was another example of an autoantibody mimicking an alloantibody. Although this autoantibody appeared initially to have anti-C specificity it was eventually shown to be more closely related to anti-Hr or anti-Rh34, than to anti-C.     

Autoanti-K antibody mimicking an alloantibody

A patient with pancytopenia and hypercellular bone marrow but without evidence of hemolysis had a positive direct antiglobulin test due to a red blood cell autoantibody. Although the patient was K-negative, eluates made from his erythrocytes contained what seemed to be anti-K antibody. The latter could be adsorbed onto and eluted from both K- positive and K-negative cells. We concluded that this patient is another example of autoanti-K antibody mimicking an alloanti-K antibody.     

Clinical significance of serologic markers related to red blood cell autoantibodies production after red blood cell transfusion—severe autoimmune hemolytic anemia occurring after transfusion and alloimmunization: successful treatment with rituximab

BACKGROUND: The association of autoantibody formation with blood transfusion was previously noted. Severe autoimmune hemolytic anemia (AIHA) diagnosed after red blood cell (RBC) transfusion determined us to undertake this study and investigate the incidence and clinical significance of autoantibodies occurring after transfusion by a retrospective review of blood bank and medical records.
STUDY DESIGN AND METHODS: We report a lymphoma patient who developed severe autohemolysis after blood transfusion and alloantibody production. The hemolysis was refractory to steroids and chemotherapy and ceased after rituximab. We also retrospectively assessed the blood bank records for a 2-year period to identify the patients who developed autoantibodies after blood transfusion and examined laboratory, clinical features, and outcome.
RESULTS: From January 2005 through December 2006, 375 direct antiglobulin tests (DATs) and 3409 indirect antiglobulin tests (IATs) were found to be positive. Thirty-eight patients with positive DATs and IATs had demonstrable RBC warm-type autoantibodies occurring after blood transfusion; 27 of them had also one or more alloantibodies. Clinical and laboratory signs of hemolysis were absent in all patients (except the case reported). In another 5 patients alloantibodies were retrieved from RBC eluate and serum without evidence of autoantibodies; therefore, a delayed serologic transfusion reaction was diagnosed.
CONCLUSION: RBC autoantibodies are quite commonly found after blood transfusion. Nevertheless, clinically significant AIHA is a rare but at times a life-threatening phenomenon. We describe a first case of successful treatment with rituximab of refractory posttransfusion AIHA. Rituximab must be further evaluated for this indication.     

Influence of immunohematological markers on severity of in vivo hemolysis in human warm autoimmune haemolytic anemia

Autoantibody production in autoimmune haemolytic anemia (AIHA) is the result of the loss of self-immunological tolerance of the host. Here we investigated the various immunohematological markers that may influence the severity of in vivo hemolysis in warm AIHA (WAIHA). Complete direct antiglobulin test (DAT) evaluation and immunohematological characterization were performed in 247 patients of WAIHA following departmental protocols. Clinical and laboratory details of patients were obtained from patient file. The median age of WAIHA patients was 47 years with a female preponderance. Lymphoproliferative diseases were the major underlying causes of secondary WAIHA. The mean haemoglobin (Hb) and reticulocyte count (Retic) were 6.43 gm/dL and 7.58% respectively. Single autoantibody bound to red cells was investigated in 151 patients. The main IgG subclass was IgG1. Multiple autoantibodies like IgG+ C, IgG+IgA and IgG+IgA+C were found in 87 (35.2%) patients. Free autoantibodies were observed in 112 patients with a median indirect antiglobulin test (IAT) reactivity of 2+. Derangement of haematological and biochemical values was statistically significant with increase in DAT reactivity, presence of multiple autoantibodies on red cells, coating of red cells by IgG3 or multiple IgG subclass, higher DAT dilution and increasing IAT reactivity. We conclude that several important but simple immunohematological parameters may influence the degree of in vivo hemolysis in WAIHA. Since a set of common haematological and biochemical test determines the severity of in vivo hemolysis therefore a comprehensive clinical and immunohematological evaluation is advisable for a correct diagnostic and therapeutic workup of WAIHA.     

Evidence suggesting the occurrence of C3-independent intravascular immune hemolysis. Reactive hemolysis in vivo

The authors present circumstantial evidence for the involvement of reactive hemolysis, i.e., C3-independent binding of the cytolytic C5b-9 complement complex to bystander red cells (RBC), in a case of intravascular immune hemolysis. Fresh serum obtained from a 6-year-old patient during the hemolytic episode, but not obtained thereafter, induced C5b-9-dependent hemolysis of human RBCs but the indirect C3 antiglobulin test remained negative. Particles (presumably RBC ghosts) isolated from the patient's plasma anticoagulated with EDTA at the peak of hemolysis were coated with C5b-9 complexes, whereas the direct antiglobulin test was strongly positive for IgA, only weakly positive for IgG, and negative for C3. Moreover, neither the autoantibodies isolated by elution (IgG plus IgA), nor free serum autoantibodies (IgA alone) activated complement in vitro. Additionally, serum samples collected later during the 12-month period of observation contained normal levels of C3, C4, C8, and C9, but markedly reduced levels of C7. These serums all produced strong reactive lysis in agarose plates, but not in test tubes. These results appear compatible with the working hypothesis that the intravascular hemolytic episode in this patient might have arisen through a local initiation of complement activation with subsequent C3-independent binding of C5b-9 to and hemolysis of bystander RBCs.     

Three examples of Rh haemolytic disease of the newborn with a negative direct antiglobulin test

SUMMARY. Typically the serological diagnosis of alloimmune haemolytic disease of the newborn (HDN) includes a positive direct antiglobulin test on the infant's red cells, and the presence of an IgG red cell alloantibody in both maternal and cord sera. HDN with a negative direct antiglobulin test has been reported with anti-A and anti-B, but not with other red-cell alloantibodies.
In this report we describe four examples of HDN in infants whose red cells had a negative direct antiglobulin test. The first case was diagnosed retrospectively when the infant was admitted to hospital aged 3 weeks with severe anaemia and cardiac failure, and subsequently died. Maternal and infant sera were both shown to contain anti-C: however, the direct antiglobulin test on the infant's red cells was negative. Approximately 1 year later the mother of this infant gave birth to triplets: soon after birth one of the triplets required an exchange transfusion, one had hyperbili-rubinaemia, and the third was unaffected. Anti-C and anti-e were detectable in the maternal serum at this time. The most probable Rh genotypes of the two affected infants were R₁R₂ (CDe/cDE), while the Rh genotype of the unaffected infant was R₂R₂ (cDE/cDE). Anti-c was implicated as causing HDN in a fourth infant (from a different family) who was a hydropic stillborn. The direct antiglobulin test on fetal blood was negative and other causes of non-immune hydrops were excluded. These four infants provide evidence that the direct antiglobulin test may be negative in some severely affected and even fatal cases of HDN.     

Immune hemolytic anemia induced by 6-mercaptopurine

BACKGROUND: Myelosuppression is the main hematotoxic effect of 6-mercaptopurine (6-MP), which is an antimetabolite chemotherapy drug. Immune hemolytic anemia associated with this drug has not been previously reported. CASE REPORT: A 67-year-old man with chronic myelomonocytic leukemia presented with anemia 2 weeks after 6-MP therapy had been initiated. Additional tests provided laboratory evidence of hemolysis. When treatment was stopped, the patient's condition and laboratory results showed a progressive improvement. RESULTS: The direct antiglobulin test was positive for IgG. The eluate and the serum were not reactive with panel red cells but reacted with 6-MP-treated red cells, while the normal serum pool was unreactive. The direct antiglobulin test was no longer positive by 20 days after the cessation of 6-MP therapy. CONCLUSION: This drug, 6-MP, should be added to the list of drugs that have been reported to cause immune hemolytic anemia by means of the so-called hapten mechanism.     

Detection of complement (c3d) coated cells in a newborn due to maternal anti-i

A case of a positive direct antiglobulin test on a newborn's cord red blood cells caused by maternal anti-i is reported. The direct antiglobulin test was moderately positive using polyspecific antiglobulin reagent and monospecific anti-C3d, but negative using monospecific anti-IgG. The infant showed no clinical signs of hemolytic disease of the newborn.     

Hemolytic disease of the newborn due to the Scianna antibody, anti-Sc2

BACKGROUND: Alloantibodies to the low-frequency antigen Scianna 2 (Sc2) are uncommon and not previously reported to cause immune hemolysis. CASE REPORT: A group B, Rh-negative infant born to a group B, Rh- positive mother had a 2+ direct antiglobulin test, as well as modest hyperbilirubinemia and a hematocrit of 45 percent. Ongoing immune hemolysis led to a hematocrit of 17.3 percent on Day 20 of life, and the infant required hospitalization and red cell transfusions. The routine maternal antibody screen was negative, but anti-Sc2 was detected during work-up for a low-frequency red cell antigen, and the father's red cells typed as Sc:1,2. CONCLUSION: Anti-Sc2 can cause clinically significant hemolytic disease of the newborn. Although the antibody is uncommon, its frequency and hemolytic potential may be underappreciated, in part because investigations often are not carried out in the infant whose red cells are ABO-incompatible with maternal blood.     

二例自身免疫性溶血性贫血患儿的输血前检测及疗效评价

目的：探讨自身免疫性溶血性贫血患者的临床输血前检测对血液选择及输注疗效的重要性。方法：对2例自身免疫性溶血性贫血患儿进行血型鉴定、不规则抗体筛查、交叉配血,并评估输血疗效。结果：2例标本的血型鉴定反定型中,均表现为O型红细胞凝集,抗体筛查均为阳性。根据抗体鉴定反应格局,一例标本含类抗-C抗体,另一例标本抗体性质不明。予输注适宜洗涤红细胞悬液后,患者病情得到改善,输血疗效评价为有效。结论：自身免疫性溶血性贫血患者体内不规则抗体比较复杂,应根据抗体性质尽量选择无相关抗原的血液输注。如自身抗体显示Rh血型等相关特异性时,应选择缺乏此抗原的红细胞悬液输注;如自身抗体性质不明,则应选择多份ABO血型相同的血液做配合性试验,采用患者血清与献血者红细胞反应最弱的血液;若患者直接抗球蛋白试验阳性,特别是抗C3d阳性,可对献血者血液进行洗涤处理后予以输注,避免献血者血液中残留的补体成分引起患者溶血。     

Changes in peripheral blood CD5 (Bla) B-cell populations and autoantibodies following blood transfusion

BACKGROUND: CD5 B cells and the natural autoantibodies they produce play a role in antigen presentation, tolerance induction, and maintenance of an idiotypic immune network. The effects of transfusion on autoantibodies and peripheral blood CD5 B cells were studied. STUDY DESIGN AND METHODS: Eight previously transfused patients with sickle cell anemia and five patients who underwent orthopedic surgical procedures with transfusion were enrolled in the study. Patients in both groups received 1 to 2 units of allogeneic packed red cells. Ten untransfused healthy adults and five patients who underwent orthopedic surgery without transfusion were enrolled as controls. Peripheral blood CD5 B cells, serum levels of IgM, antinuclear antibodies, rheumatoid factor, and anticardiolipin IgM were quantitated either at the beginning of the study (baseline sample), before transfusion, or before surgery and either at 1-, 2-, 4-, 6-, and 8-week intervals after transfusion, after surgery, or after the baseline sample was obtained. RESULTS: IgM levels and the absolute number of B cells that coexpressed CD5 rose to twice pretransfusion levels in six of eight transfused sickle cell anemia patients and in four of five transfused orthopedic surgery patients. No comparable increases in CD5 B cells were noted in untransfused controls. Preexisting rheumatoid factor and antinuclear antibody levels increased in four of five transfused orthopedic surgery patients. One sickle cell anemia patient developed anti-Fya despite receiving Fya-negative blood. Increasing titers of anti-Fya paralleled the increases in IgM and CD5 B cells after transfusion. One patient who developed a positive direct antiglobulin test after transfusion had large increases in serum anticardiolipin IgM. Anticardiolipin IgM was subsequently eluted from direct antiglobulin test-positive red cells obtained after transfusion. Antibodies with anti-Fya-like activity and anticardiolipin IgM were produced in vitro by CD5 B cells and not by conventional CD5-negative B cells. CONCLUSION: An association was found between transfusion-induced increases in CD5 B cells and increased autoantibody production. These data may have implications for immunologic intervention to prevent the induction of red cell antibodies and other changes in the immune system caused by exposure to foreign antigens via blood transfusion.     

Maternal ABO-mismatched blood for intrauterine transfusion of severe hemolytic disease of the newborn due to anti-Rh17

BACKGROUND: Clinically significant antibodies to high-incident antigens present a challenge in hemolytic disease of the newborn. Antigen-negative blood may be difficult to obtain for intrauterine transfusion (IUT). In these instances, maternal blood is de facto compatible regardless of an ABO mismatch. CASE REPORT: A group B/D-- woman with a history of hemolytic disease of the newborn due to anti-Rh17 (titer 256) presented to the obstetrical clinic at 12 weeks gestation for management of her third pregnancy. She consented to donate blood for possible IUT. STUDY DESIGN AND METHODS: Washed maternal packed cells were suspended in saline to 75 percent Hct and irradiated before transfusion. The fetus was transfused via the intrahepatic vein. RESULTS: Ultrasound examination at 19 weeks indicated a hydropic fetus. The fetal blood group was O Rh+, direct antiglobulin test 4+, and hemoglobin 22 g per L. A total of 368 mL of maternal blood was transfused during seven procedures. Labor was induced at 38 weeks, and a 2560-g male infant was delivered by Caesarian-section due to fetal distress. The infant grouped as B Rh+, direct antiglobulin test negative. No group O red blood cells were detected. The hemoglobin level was 143 g per L rising to 209 g per L at discharge 3 days later. The indirect bilirubin was 55 micromol/L and remained stable during the hospital stay. Phototherapy was discontinued after 1 day, and the infant was discharged without an exchange or top-up transfusion. CONCLUSIONS: Maternal ABO-mismatched blood is an alternate source for IUT in instances when antigen-compatible allogenic blood is unavailable.     

Oxaliplatin-induced immune pancytopenia

BACKGROUND: Oxaliplatin, a third-generation platinum compound, has been implicated in isolated cases of immune hemolytic anemia and/or immune thrombocytopenia. The first case of severe immune pancytopenia related to oxaliplatin is described. PATIENT AND METHODS: A 79-year-old woman with colorectal cancer was initially treated with 5-fluorouracil and she later received oxaliplatin and leucovorin every 2 to 4 weeks. During the 15th and 17th cycles of chemotherapy she developed thrombocytopenia, hemolysis, and neutropenia. No problems occurred during the 16th cycle without oxaliplatin. Serologic testing including detection of drug-dependent antibodies and autoantibodies was performed with standard techniques. RESULTS: Serologic findings included a positive immunoglobulin G direct antiglobulin test; nonreactive red blood cell (RBC) eluates; platelet (PLT)-bound antibodies to glycophorin (GP) IIb-IIIa, GPIb-IX, and GPIa-IIa; and oxaliplatin-dependent antibodies to PLTs, RBCs, and neutrophils. CONCLUSION: Oxaliplatin may lead to the production of ddabs to RBCs, PLTs, and neutrophils. Thus the risk of immune cytopenias should always be considered in patients treated with oxaliplatin.     

与免疫相关的血细胞减少患者骨髓造血细胞自身抗体的研究

目的　了解与免疫相关的血细胞减少患者骨髓造血细胞结合的自身抗体的类型、分布、数量及临床意义 ,考察骨髓单个核细胞直接抗人球蛋白试验 (BMMNC Coombs)的敏感性。方法　对 32例临床疑诊为免疫性全血细胞减少的患者进行BMMNC Coombs试验和流式细胞术 (FACS)双标法检测骨髓造血干 祖细胞、有核红细胞、粒细胞结合的自身抗体的种类及结合率。结果　FACS双标法检出的骨髓造血细胞自身抗体的阳性率为 90 .6 %,BMMNC Coombs试验检出的阳性率为 5 0 .0 %,前者显著高于后者 ( χ2 =12 .6 5 ,P <0 .0 5 )。FACS检测自身抗体阳性的 2 9例患者中IgG型占 6 .9%,IgM型占13 .8%,IgG +IgA型占 3 .4%,IgG +IgM型占 31.0 %,IgG +IgM +IgA型占 44 .8%;2 9例中含IgG型 2 5例 ,占 86 .2 %,含IgM型 2 6例 ,占 89.7%,含IgA型 14例 ,占 48.3 %。IgG型血红蛋白减少得最严重 ,含IgM型的患者可有血管内溶血的实验室发现 ,IgM和IgM +IgG型组治疗的起效时间明显短于其他组。91.3 %的患者可检出造血干 祖细胞上有自身抗体 ,而且多表现为全血细胞减少 ;约 5 0 .0 %的患者红、粒系细胞上有自身抗体 ,13例BMMNC Coombs试验阴性而FACS检测阳性的患者中有 11例在造血干细胞上有自身抗体。与有核红细胞和造血干 祖细胞结合的三种抗体     

Fetal hemolytic anemia and intrauterine death caused by anti-M immunization

BACKGROUND: Antibodies with anti-M specificity are detected in 10 percent of pregnant women with a positive antibody screen, but anti-M is only rarely associated with hemolytic anemia in the fetus. STUDY DESIGN AND METHODS: This study reports on three pregnancies in one family that all resulted in severe fetal anemia. The first fetus died in utero with hydrops fetalis during the 20th gestational week and the second child was delivered after 28 weeks of gestation with hydrops fetalis and a hemoglobin level of 16 g per L whereas the third affected child was treated with intrauterine red cell (RBC) transfusions before delivery at 28 weeks of gestation. RESULTS: The direct antiglobulin test was negative but anti-M in a low titer was detected through the three pregnancies, and its clinical relevance, which initially was uncertain, was confirmed by pronounced in vivo hemolysis in maternal blood of chromate ((51)Cr)-labeled M+ RBCs and normal survival of (51)Cr labeled M- RBCs. CONCLUSION: It is concluded that anti-M immunization in a few cases may cause severe fetal hemolytic anemia and intrauterine death. It remains to be elucidated why a normally clinically insignificant antibody is this aggressive in a small proportion of cases. Because the condition is treatable, anti-M must be considered as a possible cause of fetal anemia and intrauterine death.     

Studies of antibodies in the sera of patients who have made red cell autoantibodies

BACKGROUND: In patients in whom autoantibodies of broad specificity (panagglutinins) are present in the serum, adsorption studies are often necessary to identify alloantibodies that are simultaneously present. STUDY DESIGN AND METHODS: Samples from 138 patients in whom the direct antiglobulin test was positive and antibody was present in the serum were studied. When antibody identification studies before or after initial adsorption suggested the presence of an alloantibody, additional alloadsorptions were performed. RESULTS: Among the samples from 138 patients, 71 contained only panagglutinating autoantibody, and another 19 contained either autoantibodies or alloantibodies that were not accompanied by panagglutinins. The remaining 48 samples contained both panagglutinins and a total of 62 antibodies that appeared to be alloimmune in nature. Alloadsorption with antigen-negative red cells showed that 29 (47%) of the apparent alloantibodies were in fact partially adsorbed autoantibodies that mimicked alloantibodies by their reactions. CONCLUSION: Initial autoadsorption often left unadsorbed alloantibodies and autoantibodies with mimicking specificities. Initial alloadsorption more often left only true alloantibodies unadsorbed. From the screening tests, it appeared that 43 percent of the 138 patients were alloimmunized. Recognition of the mimicking nature of the partially adsorbed autoantibodies found that the real incidence of alloimmunization in the patients was 23 percent. Recognition of this phenomenon considerably simplifies the selection of blood for transfusion to these patients.     

Removing IgG antibodies from intact red cells: comparison of acid and EDTA, heat, and chloroquine elution methods

BACKGROUND: To accurately phenotype red cell from patients with a positive direct antiglobulin test (DAT), nonlytic elution procedures were assessed for their ability to dissociate IgG from antibody-coated red cells without altering red cell antigen expression. STUDY DESIGN AND METHODS: Antibodies coating red cells that were sensitized in vivo (warm-reactive autoantibodies: 8 patients) or in vitro (42 alloantibodies) were eluted by using glycine-HCl and EDTA (acid/ EDTA), heat (56 degrees C, 10 min), or chloroquine method. RESULTS: Acid/EDTA elution gave the best results, reducing DAT positivity to microscopic levels or rendering the DAT negative in 48 of 50 instances, whereas 4 samples remained resistant to heat elution and 24 to chloroquine. Standard DAT agglutination scores demonstrated that both acid/EDTA and heat elution were superior to the chloroquine method (p < 0.0001). With the gel low-ionic-strength saline indirect antiglobulin test, acid/ EDTA was superior to heat (p < 0.001). Overall, acid/ EDTA elution dissociated more antibodies than heat (p < 0.0001), especially for Kell system (K, k, Kpa, Kpb) alloantibodies. Common red cell antigens, other than Kell system antigens, were unaffected by acid/EDTA elution. In contrast, the expression of most blood group antigens was diminished after heat elution. However, it was possible to type red cell antigens by using gel low-ionic-strength saline indirect antiglobulin tests or tube agglutination methods. CONCLUSION: Although heat elution may be used on a limited basis, the acid/EDTA method appears to be the procedure of choice for typing red cell coated with warm-reactive IgG alloantibodies or autoantibodies.     

pH-dependent anti-U in autoimmune hemolytic anemia

Auto anti-U of unusual serologic characteristics is described in a Caucasian patient with autoimmune hemolytic anemia associated with chronic lymphocytic leukemia. The direct antiglobulin test was strongly positive (3+) with broad spectrum and anticomplement reagents. Auto anti-U was demonstrated only when the serum was acidified to pH 6.5 and anticomplement reagent was used in the antiglobulin test. Multiple transfusions of U positive red blood cells had clinically decreased survival manifested by jaundice and return of hemoglobin to pretransfusion levels within days. The autoantibody reacted weakly with U positive cord red blood cells interpreted as mosaicism of the U antigen possibly on a developmental basis.     

Acquired hemolytic anemia due to "auto"-anti-A or "auto"-anti-B induced by group O homograft in renal transplant recipients

Three patients developed severe but self-limited hemolytic anemia within 2 weeks of renal transplantation. All three had received kidneys from cadaver donors who were blood group O. Two of the recipients were blood group B while the third was blood group A. There was no pretransplant preparation of the donors or the recipients. Preoperative crossmatch and antibody screen were negative; however, subsequent to the hemolytic episodes, group-specific blood was incompatible and the patients were transfused with group O crossmatch-compatible blood. Blood bank serological tests showed a positive direct antiglobulin test (DAT), and anti-A and anti-B were eluted from group A and B patients, respectively. There was no evidence of hemolysis despite the positive DAT at 37 days following transplantation in two of the three patients who were maintained on cyclosporine immunosuppression. Retrospective analysis of renal transplant records showed that these "autoantibodies" appeared in three of the four renal transplant recipients who were on an immunosuppressive regimen of cyclosporine , with or without prednisone, but not in the 21 recipients who received radiotherapy to the donor kidney in addition to cyclosporine or azathioprine (p = less than 0.001). The possible pathogenetic mechanism for "autoantibody" formation by donor kidney and the role of immunosuppressive agents are discussed.