Ask the Hematologist: Management Approach to Primary Cold Agglutinin Disease

The Question

What is your management approach to patients with primary cold agglutinin disease?

My Response

Primary chronic cold agglutinin disease (CAD) is an autoimmune hemolytic anemia mediated by cold agglutinins (CA), without any obvious underlying disease such as aggressive lymphoma, other overt malignancies, or specific infections.^1,2  CA are able to agglutinate red blood cells (RBC) at an optimum temperature of 3°C to 4°C, but are also active at higher temperatures, depending on the thermal amplitude.¹  Most CA in CAD are IgMκ antibodies and have specificity for the surface carbohydrate antigen termed I, whereas IgG, IgA, or λ phenotype are found in less than 10 percent of cases.² 

Classification

CAD is a well-defined clinicopathologic entity, as described below, and should be called a disease, not syndrome.¹  The term “cold agglutinin syndrome” (CAS) should be used for secondary CA-mediated hemolytic anemia occasionally complicating other specific clinical diseases, in particular Mycoplasma or Epstein Barr virus infection, or aggressive lymphomas.

Clonality and Histopathology

Monoclonal serum immunoglobulin (usually IgMκ) can be detected in 90 percent of patients with CAD, provided correct handling of samples and the use of immunofixation even when capillary or agarose electrophoresis does not show any obvious spike.²  By flow cytometry, the cellular κ/λ ratio in bone marrow aspirate is usually greater than 3.5.² 

Our understanding of the immunohistologic and cellular basis of the clonal immunoglobulin production has greatly improved during the last decade. Two large series have confirmed signs of a clonal lymphoproliferative bone marrow disorder in most patients.^2,3  The individual hematologic and histologic diagnoses, however, showed a striking heterogeneity within each series. In one study, lymphoplasmacytic lymphoma (LPL) was the most frequent finding, whereas marginal zone lymphoma (MZL) and unclassified clonal B-cell lymphoproliferation were also reported frequently. The explanation for this perceived heterogeneity has probably been revealed by a recent, comprehensive study of 54 patients with CAD.⁴  The bone marrow findings in these patients were consistent with a surprisingly homogeneous disorder, termed by the authors “primary CA-associated lymphoproliferative disease”, which was distinct from LPL, MZL, and other previously recognized entities. The MYD88 L265P mutation, found in most cases of LPL, could not be detected in 15 samples from CAD patients analyzed for this mutation.^4,5 

Mechanisms of Complement-Mediated Hemolysis

Following CA-binding to the I antigen at the RBC surface, the antigen-antibody complex binds complement protein complex C1 and thereby triggers the classical complement pathway.⁶  The Figure shows the details of this process, resulting in predominantly extravascular, C3b-mediated hemolysis.^6,7  Intravascular hemolysis mediated by the C5b6789 complex has also been shown to occur to some extent, at least in some patients and situations.^6,8 

Clinical Features and Diagnostic Workup

CAD is defined by chronic hemolysis, positive direct antiglobulin test (DAT), monospecific DAT positive for C3d, and CA titer ≥ 64 (much higher in most cases).^1,2  DAT is usually negative for IgG but may be weakly positive.² 

Approximately 90 percent of patients have cold-induced acrocyanosis and/or Raynaud phenomena.²  By definition, all patients have hemolysis, but occasional patients are not anemic because the hemolysis is fully compensated. Median hemoglobin (Hb) level has been estimated at 8.9 g/dL.²  In cold climates, a majority of individuals experience seasonal variations with worsening of anemia and circulatory symptoms during winter or at low ambient temperatures.⁹  At least 70 percent of patients have experienced exacerbations precipitated by febrile infection, major surgery, or major trauma.^1,2,10  Fifteen percent or more have hemoglobinuria.³  In a retrospective study, approximately 50 percent of patients were transfusion-dependent at some time during the course of the disease.² 

A focused history and clinical examination is, therefore, an essential part of the diagnostic workup. Quite often, “the history tells you the diagnosis.” Laboratory tests should include full blood counts, blood smear analysis, assessment of hemolysis (absolute reticulocyte count, lactate dehydrogenase, bilirubin, and haptoglobin), polyspecific and monospecific DAT, and CA titer. For differential diagnostic considerations, cryoglobulin assessment should be included. Thermal amplitude determination may be of interest but is time-consuming and often not necessary in routine clinical practice. Results of serum electrophoresis with immunofixation, immunoglobulin class quantification, complement protein (C3 and C4) levels, and bone marrow examination (biopsy and flow cytometry) are not part of the disease definition, but should always be obtained.^1-4 

Of critical importance, serum for CA titration, electrophoresis, and immunoglobulin assessments must be obtained from blood specimens that have been kept at 37°C to 38°C from the time of sampling until serum has been removed from the clot.¹ 

Management

The mainstay of nonpharmacologic management is warm clothing and avoidance of cold. Most patients, however, have discovered this measure before they see the hematologist. Transfusions can be given safely provided that specific precautions are observed; these are comprehensively described elsewhere.¹  Splenectomy is inefficient because most of the extravascular hemolysis takes place in the liver.^1-3  Because almost all IgM is located intravascularly, plasmapheresis is considered an efficient “first-aid” in acute situations or before surgery requiring hypothermia.¹  However, such remissions are short-lived.

Not all patients require drug therapy. In our opinion, however, pharmacologic treatment is indicated more frequently than traditionally advised in the literature and should be offered to patients with symptom-producing anemia or disabling circulatory symptoms. Recommendations to avoid drug therapy have often been based on poor efficacy (until the last 10 to 15 years), combined with an underestimation of the patients’ clinical problems.^1,2  In two retrospective studies from Norway and the US, respectively, 70 to 80 percent of the patients had received pharmacologic therapy.^2,3  Corticosteroids are inefficient, and unacceptably high maintenance doses are usually required to maintain remission in the few responders.^1-3  Corticosteroids should, therefore, not be used to treat CAD.

Two prospective trials of rituximab monotherapy (375 mg/m² weekly for four weeks) showed response rates of about 50 percent according to strict criteria.^11,12  Complete responses were rare. The median response duration was approximately one year. A retrospective, “real life” study confirmed these findings.²  Based on these results and its very favorable safety profile, rituximab monotherapy is now often recommended as first-line treatment.¹³ 

The safety and efficacy of combination therapy with fludarabine and rituximab were studied in 29 patients in a prospective trial.¹⁴  The participants received rituximab 375 mg/m² on days 1, 29, 57, and 85; and fludarabine 40 mg/m² orally on days 1 to 5, 29 to 34, 57 to 61, and 85 to 89. The same response criteria were used as in the rituximab monotherapy trial. Twenty-two patients (76%) responded, with six (21%) achieving complete response and 16 (55%) achieving partial response. Median time to response was four months. An impressive estimated median response duration of more than 66 months was achieved. Short-term hematologic toxicity was significant, however, with 12 patients (41%) experiencing grade 3 or 4 toxicity. Furthermore, although not directly observed in this study, the possibility of long-term toxicity may be a concern, particularly in younger patients.

In conclusion, patients with CAD requiring therapy should be included in prospective trials if available. Outside clinical trials, fludarabine-rituximab combination therapy should be considered in those who definitely need efficient treatment, especially if they have failed rituximab monotherapy, are not too young, are reasonably fit, and have not previously received cytotoxic chemotherapy. In patients failing to meet these criteria, single-agent rituximab should remain first-line treatment.

The Future

Favorable response to bendamustine-rituximab combination therapy has been reported in one patient, and a prospective study is ongoing.¹⁵  The targeted B-cell receptor pathway inhibitors ibrutinib and idelalisib have not been evaluated in CAD.

Given that hemolysis in CAD is entirely complement-dependent,^1,6  studies of complement modulators are highly interesting. Favorable effect of the C5 inhibitor eculizumab has been described in a prospective trial.⁸  Since the hemolysis is predominantly C3b-mediated in most patients, complement blockade at a more proximal, classical pathway level might, in theory, be more successful.^6,7  Preclinical studies with the anti-C1s monoclonal antibody TNT003 and its humanized counterpart TNT009 have shown favorable results.⁷  If sufficient clinical documentation for complement-directed therapies can be provided, such treatment will probably still not replace clonally directed therapies, which are more causal and do not need to be continued indefinitely. Complement-directed therapies seem very promising, however, in acute exacerbations, in patients undergoing surgery requiring hypothermia, and in those with severe CAD not responding to clonally directed immunochemotherapy.