Final Diagnosis -- Hemolytic Cold Agglutinin Syndrome


Hemolytic Cold Agglutinin Syndrome


Cold agglutinin autoantibodies (CAA) are usually of IgM-class and can be acute or chronic, benign or pathogenic. Specificity of CAA is most often for the Ii antigen system.1 Expression of the I and i antigens change with age. Newborns express i antigen on their red blood cells, and children and adults primarily express I antigen. Low titers (≤256) of CAA are common in the general population and have not been implicated in disease. However, higher titers (≥512) of CAA are more likely to cause hemolysis. Hemolysis is most likely if the CAA have a high thermal amplitude, that is if the CAA continue to cause agglutination at higher temperatures, for example temperatures above 30°C.1-2 The CAA can be a manifestation of a chronic, primary disease (e.g. Cold Agglutinin Disease), which often involves a clonal B cell abnormality, such as lymphoplasmablastic lymphoma.2 However, CAA can also be secondary to infection and cause acute and transient symptoms.

Cold Agglutinin Syndrome Associated with Infection

Respiratory infections can instigate Cold Agglutinin Syndrome (CAS). The two most commonly implicated infections are those of Mycoplasma pneumoniae and Epstein-Barr virus (EBV). Rarely, other bacterial or viral infections can precede CAS. M. pneumoniae infections often result in the formation of CAA. In fact, demonstration of CAA has historically been used as evidence of M. pneumoniae infection. Although CAA are common in Mycoplasma pneumonia, they are rarely clinically significant. Of all autoimmune hemolytic anemias, it is estimated that Mycoplasma infections cause only 8%.2 The CAA associated with Mycoplasma infection usually become detectable 2-3 weeks after the prodrome. When Mycoplasma-related CAA cause hemolytic anemia, they are classically high-titer (≥512) IgM with anti-I specificity.2 The CAA and anemia usually resolve in several weeks without intervention, but rare cases cause severe anemia.

EBV infections uncommonly result in the formation of CAA. Only about 1% of autoimmune hemolytic anemias have been associated with EBV-related CAA.2 The EBV-related CAA that cause hemolytic anemia are typically lower-titer (≤512) IgG or IgM with anti-i specificity, and any resulting anemia is mild.2

Pathophysiology of Cold Agglutinin Autoimmune Hemolytic Anemia

CAS is a result of antibodies directed against a patient's own red blood cells (RBCs). The targeted antigen is usually in the Ii antigen system. Most commonly, these CAA are IgM class. By definition, these IgM CAA have increased avidity at lower temperatures. In vivo, these lower temperatures are encountered in the body's extremities. Once bound, IgM's pentameric structure makes it an efficient activator of the complement cascade. IgM may dissociate from the RBC membrane when the cell returns to the body's trunk and rises to core body temperature, but the complement activation that began while the IgM was bound to the membrane will continue even after the IgM has dissociated. The complement cascade leads to intravascular hemolysis.

In contrast, warm autoimmune antibodies (WAA) causing hemolytic anemia are often IgG class. IgG-coated RBCs are typically removed from circulation by splenic macrophages, which results in extravascular hemolysis.

Blood Bank Laboratory Detection of Cold Agglutinin Antibodies

CAA can be evidenced in several ways in the blood bank laboratory. First, anticoagulated blood may seem to have clotted. These agglutinated RBCs often become fluidic upon warming the blood tube to 37°C. Second, the patient's plasma or serum may cause unexpected reactions when combined with reagent red blood cells. This can result in ABO discrepancies due to additional reverse type reactions and/or panagglutination when the plasma is tested against screening cells and panel cells. Third, the C3d complement component can be identified on the patient's RBC membranes by using the anti-C3d DAT reagent. When the C3d molecule is detected, the presence of IgM antibody is inferred because IgM efficiently activates the complement cascade.

When performing a DAT, often a DAT screening test is initially performed using a polyspecific (anti-C3d & anti-IgG) reagent. If this screening DAT is positive, as was the case in this patient presentation (Table 2), then two separate DATs are performed-one using only anti-C3d and the second using only anti-IgG. These two DATs help to distinguish between the presence of IgM antibodies and IgG antibodies (Table 3).

Table 3. Expected results of direct antiglobulin testing (DAT)

With CAA that interfere with blood bank testing, modified testing should be performed.3 Samples with severe CAA should be collected and maintained at 37°C until the plasma and RBCs are separated. Alternatively, the sample can be warmed to 37°C for 10-15 minutes and then separated. After separation, the patient's RBCs are washed with warm saline to remove residual plasma containing CAA. Handling the plasma and RBCs in this way is important in order to obtain the most accurate test results. The warmth discourages adsorption of the CAA onto the patient RBCs. Minimizing adsorption during sample processing allows an accurate titer of CAA to be determined, and it helps to prevent CAA from adsorbing to the patient's RBCs, which could cause false-positive results (e.g. agglutination during reverse typing).

The level of CAA can be determined by titering the plasma in serial twofold dilutions with saline and subsequently incubating the solutions with type O adult RBCs at 4°C for 1-2 hours before examining for agglutination.3 The reciprocal of the most dilute solution that yields agglutination is referred to as the "titer" of the CAA. For example, if a mixture of 1 part plasma and 511 parts saline causes agglutination, but a mixture of 1 part plasma and 1023 parts saline does not cause agglutination, then the titer would be determined to be "512."

False positive results due to CAA can usually be avoided if conditions are carefully maintained at 37°C throughout testing. However, it may be necessary to use cold-adsorption studies to remove the CAA from the patient plasma if the CAA are interfering with alloantibody testing at the anti-human globulin phase. The cold-adsorbed plasma can be tested against reagent screening cells to determine if an alloantibody is present. If the screening cells agglutinate with un-adsorbed plasma and do not agglutinate with cold-adsorbed plasma, then it is determined that the patient has CAA but no alloantibodies.

Additionally, specificity of the CAA can be determined by testing patient plasma against type O blood from adults and newborns. Adult RBCs are I antigen positive, and infant RBCs are i antigen positive. If patient plasma causes preferential agglutination of adult RBCs, then the CAA are determined to have anti-I specificity. If the plasma reacts preferentially with cord blood, then the CAA are determined to have anti-i specificity.

Clinical Conclusion

This patient's cold agglutinin antibodies (CAA) were detected by the blood bank. Reaction of the CAA with reagent screening cells initially suggested the presence of alloantibodies (Table 2). Testing against a panel of RBCs resulted in panagglutination, including the autocontrol. The positive autocontrol prompted DAT testing, which demonstrated C3d on the patient's RBCs (Table 2). The patient plasma was tested against type O cord RBCs and type O adult RBCs at 4°C; the titers were 128 and ≥512, respectively. Tube testing the patient plasma against reagent screening cells demonstrated agglutination at 4°C, room temperature, and 30°C; but no agglutination occurred at 37°C. A cold autoadsorption was performed, and testing the cold-adsorbed plasma against the reagent screening cells resulted in no agglutination. These tests determined that the patient demonstrated IgM CAA with anti-I specificity at a titer of ≥512 and a thermal amplitude ≥30°C. The patient had no detectable underlying alloantibodies to red blood cell antigens.

The CAA were determined to be caused by cold agglutinin syndrome (CAS). Her CAS was caused by her immune response to a Mycoplasma pneumonia. Production of CAA during Mycoplasma pneumonia is common, but rarely do they cause the marked hemolysis that occurred in this patient. This patient reported seeing blood in her urine, but the patient was actually seeing hemoglobinuria due to intravascular hemolysis. She also reported blood in her stool, but the occult blood test was negative (Table 1). Initial laboratory findings that suggested hemolysis include decreased haptoglobin, elevated LDH, and gross hemolysis in the samples (Table 1). The patient's large quantity of free hemoglobin was causing injury to her kidneys, which is evidenced by a high serum creatinine level (Table 1). Her serum creatinine continued to rise after admission. The patient's plasma free hemoglobin was checked on Day 3 and was found to be 394mg/dL (normal <6 mg/dL).

This patient's underlying pneumonia was treated with antibiotics. The patient's anemia was treated with transfusions of warmed packed red blood cell units that were fully cross-match compatible with the patient's cold-adsorbed plasma. She was also advised to remain covered with blankets in an attempt to keep her entire body's temperature, including her extremities, above the thermal amplitude of the CAA. Plasmapheresis was commenced on Day 3 in an effort to alleviate the kidneys' plasma free hemoglobin burden.

In spite of plasmapheresis, the patient's serum creatinine levels remained elevated until Day 5 of admission. Thereafter, the serum creatinine levels trended downward. The patient's anemia was corrected after receiving red blood cell transfusions. Repeat testing revealed a negative DAT on Day 10. Plasmapheresis was stopped after Day 11. The patient was discharged home on Day 20 with instructions to follow up in clinic.


  1. Sokol RJ, Hewitt S, Stamps BK. Autoimmune haemolysis: an 18-year study of 865 cases referred to a regional transfusion centre. Br Med J (Clin Res Ed). Jun 20 1981;282(6281):2023-2027.
  2. Berentsen S, Tjonnfjord GE. Diagnosis and treatment of cold agglutinin mediated autoimmune hemolytic anemia. Blood Rev. May 2012;26(3):107-115.
  3. Roback JD, Grossman BJ, Harris T, Hillyer CD, ed Technical Manual. 17th ed: AABB; 2011.

Contributed by Daniel D. Rhoads, MD, Sarah Harm, MD, Lirong Qu, MD, PhD

Case IndexCME Case StudiesFeedbackHome