Rare case of warm IgM autoimmune hemolytic anemia

Abstract

Cold agglutinins are IgM‐mediated autoimmune hemolytic processes that most often cause destruction of red blood cells at colder temperatures. Here we describe a case of a patient with history of living donor renal transplantation that developed an atypical autoantibody causing profound anemia and hemolysis which, although IgM in nature, reacted best at warmer temperatures and was caused by a lymphoproliferative disorder. The investigations and serological findings here raise awareness to less common forms of autoimmune hemolytic disorders and describe specialized tests that can help uncover these antibodies. This case also highlights a rare presentation of post‐transplant lymphoproliferative disorder manifesting clinically as hemolytic anemia.

Abbreviations

+ve

positive result

DAT

direct antiglobulin test

DTT

dithiothreitol

IAT

indirect antiglobulin test

IS

immediate spin

RESt

rabbit erythrocyte stroma

−ve

negative result

ZZAP

dithiothreitol and papain

CKD

chronic kidney disease

IgM

immunoglobulin M

WAIHA

warm autoimmune hemolytic anemia

CAD

cold agglutinin disease

1. INTRODUCTION

Warm IgM autoimmune hemolytic anemia is a rare phenomenon where typically cold‐reactive IgM autoantibody causes red cell hemolysis exclusively at warmer temperatures. ¹ This contrasts with cold agglutinin disease (CAD), where most IgM‐related autoantibodies cause hemolysis at colder temperatures. ² Both warm and cold‐related IgM‐mediated immune hemolysis may occur in the context of malignancy [where the latter is then referred to as cold agglutinin syndrome (CAS)], though warm IgM autoimmune hemolytic anemia often presents with a more aggressive clinical course. ³ , ⁴ , ⁵ , ⁶ These antibodies represent a diagnostic challenge, which may be aided by thermal amplitude testing, among other investigations. ⁷

Here we describe the case of a patient with a history of living donor kidney transplant on chronic immunosuppressive therapy who became progressively anemic and was found to have life‐threatening, atypical IgM autoantibody‐associated hemolysis. Serological investigations involving both the local blood bank and national immunohematology reference laboratory (NIRL) were able to demonstrate that the hemolytic antibody was a warm‐reacting IgM autoantibody in the setting of posttransplant lymphoproliferative disease.

2. CASE PRESENTATION

A 19‐year‐old male with a history of living donor renal transplantation was admitted with anemia (hemoglobin 59 g/L). Past medical history was significant for premature delivery (33 weeks' gestation) and chronic kidney disease (CKD) secondary to urosepsis from posterior urethral valves. He underwent bilateral nephrectomy, bladder augmentation, and living renal transplant from his mother at age 13. Chronic immunosuppression consisted of azathioprine 100 mg PO daily, tacrolimus 8.5 mg PO daily, and prednisone 5 mg PO daily; he was also taking ferrous fumarate 300 mg PO daily. Social history included daily cannabis use (smoking).

Upon presentation, he described 2 months of progressive fatigue, unintentional 10‐pound weight loss, and darkened urine. Prior to symptom onset, he received Evusheld (tixagevimab/cilgavimab) for COVID‐19 prophylaxis. Examination showed generalized pallor and diffuse reticular rash affecting the arms, legs, and torso. Small palpable lymph nodes were present in the anterior cervical zone. General cardiovascular, respiratory, and abdominal examinations were unremarkable, and there were no signs of joint stiffness or swelling.

Historical hemoglobin values ranged between 100 and 120 g/L, which fell to 80–90 g/L 6 months before presentation. This correlated with rising creatinine (from 120 to 140 μmol/L to 204 μmol/L). A renal biopsy showed nonspecific fibrosis. One month prior to admission, complete blood count (CBC) showed hemoglobin 92 g/L, WBC 8.3 × 10E9/L, PLTs 215 × 10E9/L without abnormalities on peripheral blood smear. Two weeks later, smear showed red cell agglutination at room temperature, slight macrocytosis, and intermittent eosinophilia. Pertinent lab values during the initial clinic visit included Hb 59 g/L, MCV 93 fL, WBC 10 × 10E9/L, eosinophils 0.9 (chronically elevated for at least 2 years), and platelets 177 × 10E9/L.

A hemolytic workup was conducted, whose results are summarized in Table 1. In the context of the rapid drop in hemoglobin and peripheral blood findings, this was felt to show possible compatibility with hemolysis. Studies for antinuclear antibodies (ANA), rheumatoid factor, serum protein electrophoresis, HIV, hepatitis B and C, parvovirus and mycoplasma serology, cryoglobulins, flow cytometry for paroxysmal nocturnal hemoglobinuria, BK virus, CMV and HIV PCR were all negative.

TABLE 1.

Summary of hemolytic workup on presentation.

Test	Result
Haptoglobin	Undetectable
LDH	208 IU/L (local ref. range 120–250 IU/L)
Unconjugated Bilirubin	20 μmol/L (local ref. range <19 μmol/L)
Reticulocytes	33 × 10E9/L
Serum free hemoglobin	Positive
Total hemoglobin	59 g/L (140–180 g/L)

Historical ABO typing (10 years prior) was Group A, RhD‐positive, with no known red cell antibodies or history of transfusion. Due to agglutination on peripheral blood smear and biochemistry suggesting hemolysis, additional pre‐transfusion investigations were ordered. Repeat ABO typing showed unexpected additional reactivity on forward typing (agglutination at immediate spin (IS) using anti‐B antisera). RhD forward typing control also returned positive. When repeating using the pre‐warming technique at 37°C, the result was unchanged.

At presentation, RBC antibody screening was and extended antibody panel using solid‐phase testing was panreactive. Polyspecific direct antiglobulin testing (DAT) could not be interpreted due to positive albumin control. Similar to ABO typing, pre‐warming the sample at 37°C did not resolve the reactive control. Cold agglutinin screening at 4°C was negative using a 1:64 dilution. Thermal amplitude testing was positive at 22°C, 30°C, and 36°C but negative at 4°C (using 1:4 dilution), suggestive of an unusual IgM antibody active at warmer temperatures.

ABO and RhD testing were repeated at the Canadian National Immunohematology Reference Laboratory (NIRL) following treatment of patient red cells with dithiothreitol (DTT). Results were consistent with historic typing (Group A, RhD‐positive). DAT using DTT‐treated RBCs was positive for C3d, with a negative eluate using Gel indirect antiglobulin testing (IAT) with group O reagent cells and with Group A and B red cells, ruling out anti‐H. Lack of reactivity against group O RBCs suggested against an anti‐H.

Antibody screening at NIRL was panreactive at 37°C and IAT, while negative at IS (Gel and SIAT). Autocontrol was also positive at 37°C and IAT only. Following DTT treatment, the patient's plasma was negative in all phases of testing against reagent panel RBCs. Rabbit erythrocyte stroma (RESt) adsorption‐treated plasma failed to remove reactivity at IAT. Cold agglutinin panel (group testing at 4°C was panreactive and nonspecific). Alloadsorption using an Rh phenotype‐matched adsorbing cell was done to avoid removing potential antibodies to Rh and remained reactive at IS. Repeat adsorption using Rh phenotype‐matched ZZAP‐treated adsorbing cells (resulted in no reactivity at IS, 37°C, or IAT). Incubation of reagent cells with ZZAP at 37 degrees destroys the Kell, Kidd, and Duffy antigens as this reagent includes DTT and the enzyme Papain. DTT‐treated patient plasma and a control plasma containing anti‐P1 IgM antibody both showed no reactivity with reagent panel cells at IAT phase (Table 2).

TABLE 2.

Summary of serological testing results.

Test	Hospital investigations	National immunohematology reference laboratory investigations
ABO & Rh typing	Additional reactivity w/Anti‐B; Rh‐control +ve	Pt RBCs treated w/DTT Result: Group A, Rh +ve
DAT & Eluate	Albumin control +ve (test uninterpretable)	After DTT tx of RBCs Result: DAT‐C3d +ve, eluate −ve (Gel IAT)
Cold Agglutinin Screen	Negative: 1:64 dilution w/4°C	Not performed
Thermal Amplitude Testing; cord + adult RBCs	Positive: 37°C, 30°C, 22°C Negative 4°C, 1:4 dilution	Not performed
Antibody Screen	Solid phase: Panreactive	37C and IAT: Panreactive IS −ve; autocontrol (all phases) +ve DTT Treated Pt. Plasma: −ve
Antibody Panel	Sold Phase: Panreactive	Gel & SIAT: Panreactive (37C +ve; IS −ve) RESt adsorption: 37C +ve; IS −ve Cold agg: panreactive; no spec. to (I, i, Pr)
Alloadsorption (w/ZZAP)	Not performed	ZZAP treated Rh‐matched RBCS “Artificial” auto adsorption showed no signs of alloantibodies

Abbreviations: +ve, positive result; DAT, direct antiglobulin test; DTT, dithiothreitol; IAT, indirect antiglobulin test; IS, immediate spin; RESt, rabbit erythrocyte stroma; −ve, negative result; ZZAP, dithiothreitol and papain.

Given the lack of prior transfusions, pattern of serological test results and interference in DAT testing with positive controls, the findings were consistent with an IgM autoantibody. Unusually, this IgM antibody was nonreactive at colder temperatures (4°C) and IS with increasing reactivity at 37°. The conclusion was that this represented a rare, warm‐reactive IgM autoantibody.

During admission, the markedly reduced hemoglobin (59 g/L) and progressive fatigue prompted transfusion with 2 units of RBCs, resulting in a post‐transfusion Hb of 87 g/L, which fell to 67 g/L the next day. Aranesp 60 μg/weekly and folate 5 mg PO daily were started at this time. The nadir hemoglobin during hospitalization was 51 g/L. Three consecutive daily 1:1 plasma volume exchanges using albumin as replacement fluid achieved recovery in Hb >120 g/L. Peak LDH pre‐exchange was 272 IU/L, which normalized to 169 IU/L afterward. Unconjugated bilirubin fell from 44 mmol/L to 24 mmol/L (in the context of being post‐plasma exchange). Reticulocyte counts could not be assessed because of extensive sample agglutination at the height of illness. Bone marrow aspirate and biopsy were done to evaluate for malignancy. The findings showed mild megaloblastoid dyserythropoiesis, querying stress erythropoiesis secondary to hemolysis; other cell lines were unremarkable, with no evidence of disease.

CT scans revealed enlarged (15 × 13 × 36 mm) right retroperitoneal and mesenteric lymph nodes. Retrocaval node biopsy showed a polymorphic B‐cell post‐transplant lymphoproliferative disorder. EBV PCR was negative on whole blood and lymph node biopsies. Azathioprine was discontinued, and treatment was initiated with pulse steroids (solumedrol 40 mg daily × 5 days and then baseline prednisone) and rituximab 375 mg/m² weekly for four treatments; 26 days following rituximab therapy the patient became transfusion independent (Figure 1).

FIGURE 1.

Timeline of interventions and corresponding Hb/LDH measurements. [Color figure can be viewed at wileyonlinelibrary.com]

3. DISCUSSION

Autoimmune hemolytic anemia is frequently categorized as either warm or cold, based on the thermal characteristics of the culprit autoantibody. Warm autoimmune hemolytic anemia (WAIHA) is typically associated with immunoglobulin G (IgG) autoantibodies reacting optimally at 37°C, whereas cold agglutinin disease (CAD) is caused by immunoglobulin M (IgM) autoantibodies reacting optimally at 4°C but with a thermal amplitude often reaching 28°C–30°C. ¹ It is unusual for cold agglutinin disease to present with a preference for warmer temperatures or lack reactivity at colder temperatures such as in this case. ² Thermal amplitude testing is performed at varying temperatures to demonstrate the presence of optimal temperature‐dependent reactivity. Outcomes for patients with warm‐reactive IgM‐mediated disease include more severe anemia, hemolysis, and higher mortality than is typically seen with warm autoimmune hemolytic anemia. ¹

In classical cold agglutinin disease, cold autoantibodies are associated with intravascular hemolysis due to the ability of IgM antibodies to fix complement and lyse RBCs. This may also interfere with common serological transfusion medicine testing, leading to positive controls and unexpected additional reactivity on ABO typing. A monospecific DAT for C3d is typically positive, and antibody panels are panreactive without specificity. Dithiothreitol (DTT) is a reagent capable of disrupting disulfide bonds contained within IgM pentamers, removing their reactivity and enabling distinction from IgG‐related disease. ⁸ DTT treatment of RBCs resolved the ABO discrepancy and confirmed that the bound antibody was IgM.

Rabbit erythrocyte stroma (RESt) absorbs common cold‐associated autoantibodies (anti‐I or anti‐HI) from patient plasma ⁹ and is useful in resolving serological discrepancies with IgM. Its ineffectiveness in this case helped confirm this was an atypical cold autoantibody. ZZAP is a product combining DTT and proteolytic enzymes (ficin or papain) which cleaves IgG and C3d from red cell membranes to allow investigation despite autoantibodies being present. However, it will also destroy multiple RBC antigens including M, N, Fy^a, Fy^b, and K. ¹⁰ The reference laboratory involved in this case used ZZAP to remove autoantibody and enable alloadsorption and rule out underlying alloantibodies toward these various antigens by removing their expression on adsorbing cells. Autoadsorption was not possible due to the recent transfusion.

In a published study of 49 cases of IgM warm AIHA identified over 25 years, authors reported that 78% (n = 38) of patients' RBCs showed spontaneous agglutination, with the majority (27 of 30 tested) resolving after DTT treatment. ¹ Autoagglutination at 37°C was frequently present, and DAT was commonly positive for C3d and/or other proteins in 44/49 patients, while IgM and IgA were present with/without other proteins in 14 and 3 patients, respectively. Authors noted that eluates from patient RBCs were usually reactive. Among 35 patients with thermal amplitude and titers performed, 24 patients' antibodies had significant reactivity at 37°C or 30°C. Among 40 patients with 4°C titers performed, all had detectable reactivity, with 83% (n = 33) having titers of 32 or less; one third of cases had antibody specificity. ¹

Warm IgM‐mediated autoimmune hemolytic anemia is often associated with an aggressive course in reported case studies. ³ , ⁴ , ⁵ , ⁶ Similar to other autoantibody‐mediated hemolysis, it may be associated with autoimmune disease or malignancy. Due to its rarity, evidence‐based therapy still eludes this condition. Immunosuppression using steroids and other agents, eculizumab, and plasma exchange (particularly effective for removal of IgM antibodies due to more intravascular distribution) can be considered. Treatment should also be directed toward any underlying etiology if targeted therapy is available.

Our patient demonstrated similar findings to case summaries: spontaneous RBC agglutination, 37°C autoagglutination, positive DAT for C3d, responsiveness to DTT treatment, and thermal activity at 37°C. However, the eluate performed at the reference laboratory was negative for this patient. Also unique was the lack of reactivity at 4°C (standard temperature for our center's cold agglutinin screens) but observed reactivity at higher temperatures. Due to their unusual nature, local transfusion services may not have the necessary reagents or capability to investigate these antibodies. Therefore, the availability of an immunohematology reference laboratory may be useful for diagnostic purposes.

4. CONCLUSION

This is a rare case of a warm‐reacting IgM autoantibody mediating a potentially life‐threatening hemolytic anemia. Plasma exchange effectively temporized IgM levels while allowing time for immunosuppressive therapy to take effect. Specialized serological testing performed (DTT treatment, RESt, and ZZAP) are helpful techniques to determine antibody etiology. Clinical acumen requires identification of abnormal findings that may prompt more advanced serological investigation at their center or larger reference laboratories.

CONFLICT OF INTEREST STATEMENT

The authors have disclosed no conflicts of interest.

Tordon B, Alhomsi N, Iqbal A, Gangji A, Clarke G, Ning S. Rare case of warm IgM autoimmune hemolytic anemia. Transfusion. 2025;65(9):1756–1760. 10.1111/trf.18353