You are on page 1of 9


Diagnosis and treatment of cold agglutinin mediated autoimmune hemolytic anemia
Sigbjørn Berentsen
⁎, Geir E. Tjønnfjord
b, 1
Department of Medicine, Haugesund Hospital, Helse Fonna, P.O.Box 2170, NO-5504 Haugesund, Norway
Department of Haematology, Oslo University Hospital, Rikshospitalet and Institute of Clinical Medicine, University of Oslo, P.O.Box 4950, Nydalen, NO-0424 Oslo, Norway
a b s t r a c t a r t i c l e i n f o
Hemolytic anemia
Cold agglutinin
Cold agglutinin disease
Exact diagnosis of the subtype has essential therapeutic consequences in autoimmune hemolytic anemia.
Cold-antibody types include primary chronic cold agglutinin disease (CAD) and rare cases of cold agglutinin
syndrome (CAS) secondary to cancer or acute infection. Primary CAD is a clonal lymphoproliferative disorder.
Not all patients require pharmacological therapy, but treatment seems indicated more often than previously
thought. Corticosteroids should not be used to treat primary CAD. Half of the patients respond to rituximab
monotherapy; median response duration is 11 months. The most efficient treatment to date is fludarabine
and rituximab in combination, resulting in responses in 75%, complete responses in 20% and median response
duration of more than 66 months. Toxicity may be a concern, and an individualized approach is discussed.
Erythrocyte transfusions can be given provided specific precautions are undertaken. No evidence-based ther-
apy exists in secondary CAS, but optimal treatment of the underlying disorder is essential when feasible.
© 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Autoimmune hemolytic anemia (AIHA) is a group of uncommon
disorders characterized by hemolysis due to autoantibodies against
red blood cell surface antigens. The autoantibodies may be warm-
reactive with a temperature optimum at 37 °C or cold-reactive with
a temperature optimum way below the normal body temperature.
AIHA can be classified, accordingly, into warm and cold reactive anti-
body types and further subdivided based on the presence of underly-
ing or associated disorders. A widely accepted classification is shown
in Table 1.
Altogether, the cold-reactive types probably account for about 25%
of all AIHA.
1, 2
The involved autoantibodies are cold agglutinins (CA),
defined by their ability to agglutinate erythrocytes at an optimum
temperature of 0–4 °C (Fig. 1).
4, 5
Most CAs are of the immunoglobu-
lin(Ig)M class, although IgG or IgA CAs are occasionally found.
5, 6
pathogenesis and management of AIHA differ substantially depend-
ing of the characteristics of the autoantibody and, therefore, a correct
and precise diagnosis of the subtype has critical therapeutic conse-
quences. Particularly in primary cold agglutinin disease (CAD), con-
siderable progress has been made during the last 1–2 decades in the
knowledge of clinical features, humoral and cellular immunology
and bone marrow pathology.
4, 6–9
Therapy for primary CAD was
largely unsuccessful until 10 years ago, but efficient treatment op-
tions have now become available.
The term ‘cold (hem)agglutinin disease’ (CAD, CHAD) is some-
times used in a broad sense as a synonym for cold agglutinin syn-
drome (CAS), including all types of cold antibody AIHA.
3, 11–14
and others prefer to use the termCAD in a narrowsense, synonymous
with primary chronic CAD.
1, 10, 15
This particular, well-defined and
well-characterized clinicopathological entity should be called a dis-
ease, not syndrome. Although this reviewwill concentrate on primary
chronic CAD, we will also discuss the diagnosis and management of
acute and chronic secondary CAS. Mixed-type AIHA and paroxysmal
cold hemoglobinuria will not be addressed.
2. Pathogenicity of cold agglutinins
Cold hemagglutination was described in 1903 and found to occur
in humans in 1918.
16, 17
The association between cold hemagglutina-
tion and hemolysis was first reported in 1937.
CA can be deter-
mined semi-quantitatively by the titer, based on their ability to
agglutinate erythrocytes at 4 °C.
Screening for CA have shown that
a high proportion of the adult population has CA in serum without
any evidence of hemolysis or other disease.
5, 15
These normally occur-
ring CA are polyclonal and are found in low titers, usually below 64
and rarely exceeding 256.
On the contrary, in 172 consecutive indi-
viduals with monoclonal IgM in serum, significant CA activity was
found in 8.5% with titers between 512 and 65,500, and all individuals
with detectable CA had hemolysis.
Thus, monoclonal CA are gener-
ally far more pathogenic than polyclonal CA.
The thermal amplitude is defined as the highest temperature at
which the CA will react with the antigen.
4, 20
In general, the pathoge-
nicity of CA is more dependent on the thermal amplitude than on the
20, 21
The normally occurring CA have low thermal amplitudes. If
Blood Reviews 26 (2012) 107–115
⁎ Corresponding author. Tel.: +47 52732000; fax: +47 52770189.
E-mail addresses: (S. Berentsen), (G.E. Tjønnfjord).
Tel.: +47 23070713; fax: +47 23070470.
0268-960X/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
Contents lists available at SciVerse ScienceDirect
Blood Reviews
j our nal homepage: www. el sevi er . com/ l ocat e/ bl r e
the thermal amplitude exceeds 28–30 °C, erythrocytes will aggluti-
nate in the circulation in acral parts of the body even at mild ambient
temperatures and, often, complement fixation and complement-
mediated hemolysis will ensue. CA should not be confused with cryo-
globulins. Occasionally, however, patients have been reported in
whom the cryoprotein had both CA and cryoglobulin properties.
8, 22,
CA are most often directed against the Ii blood group system.
4, 24
About 90% of CA are anti-I specific while most of the remaining
ones show specificity for i.
3, 5
The I and i antigens are carbohydrate
macromolecules and the densities of these antigens on the erythro-
cyte surface are inversely proportional. Neonatal red blood cells al-
most exclusively express the i antigen, while the I antigen
predominates in individuals of 18 months of age and older.
CA with anti-I specificity are generally more pathogenic in children
and adults than those specific for the i antigen.
5, 25, 26
CA show specificity against the erythrocyte surface protein antigen
designated Pr and such CA can be highly pathogenic.
26, 27
other specificities have been reported but are probably very rare.
Cooling of blood during passage through acral parts of the circula-
tion allows CA to bind to erythrocytes and cause agglutination
(Fig. 1). Antigen-bound IgM-CA is more prone than IgG to bind com-
plement protein C1 and thereby initiate the classical complement
C1 esterase activates C4 and C2, generating C3 conver-
tase which leads to the formation of C3b. Upon returning to central
parts of the body with a temperature of 37 °C, IgM-CA detaches
from the cell surface, allowing agglutinated erythrocytes to separate
from each other, while C3b remains bound. A proportion of the C3b-
coated red cells are sequestered by cells of the reticulo-endothelial
system (RES), mainly in the liver. On the surface of the surviving
erythrocytes, C3b is cleaved, leaving high numbers of C3d molecules
on the cell surface. Complement activation may proceed beyond the
C3b formation step, resulting in C5 activation, formation of the mem-
brane attack complex and intravascular hemolysis. Due to surface-
bound regulatory proteins such as CD55 and CD59, however, the
complement activation is usually not sufficient to produce clinically
significant activation of the terminal complement pathway. The
major mechanism of hemolysis in stable disease, therefore, is the ex-
travascular destruction of C3b-coated erythrocytes by the RES.
29, 30,
32, 33
These mechanisms explain why the direct antiglobulin test (DAT)
is strongly positive for C3d in patients with CA mediated hemolysis
and, in a majority, negative for IgM and IgG. In up to 20% of patients
with primary CAD, however, DAT is also weakly positive for IgG,
which should not lead to a wrong diagnosis of mixed-type AIHA.
6, 34
3. Primary chronic cold agglutinin disease
3.1. Epidemiology
Primary CAD accounts for about 15% of all cases of AIHA.
1, 2, 35
prevalence in Norway has been estimated to 16 per million inhabi-
tants and the incidence rate to 1 per million inhabitants per year.
The median age of patients with CAD is 76 years (range, 51–96)
with a median age at onset of symptoms of 67 years (range, 30–92).
3.2. Clinical, immunological and pathological characteristics
By definition, all patients with CAD have hemolysis, but occasional
patients are not anemic because the hemolysis is fully compensated.
Most patients, however, have manifest hemolytic anemia. Of 16 pa-
tients described in an early publication, five had hemoglobin (Hgb)
levels below 7.0 g/dL and one had levels below 5.0 g/dL.
Hgb levels
ranged from 4.5 g/dL to normal in a more recent population-based
descriptive study of 86 Norwegian patients.
In the same study, the
median Hgb level was 8.9 g/dL and the lower tertile was 8.0 g/dL.
Fifty per cent of the patients had been considered transfusion depen-
dent for shorter or longer periods during the course of the disease,
and 70% had received drug therapy. Although the term ‘cold’ refers
to the biological properties of the CA, not the clinical features, approx-
imately 90% of the patients experienced cold-induced acrocyanosis
and/or Raynaud phenomena.
These symptoms ranged from slight
to disabling. Characteristic seasonal variations in the severity of he-
molytic anemia have been well documented.
In at least two-thirds
of the patients, exacerbation of hemolytic anemia is also triggered
by febrile infections or major trauma.
6, 38, 39
The explanation for
this paradoxical exacerbation is that during steady-state CAD, most
patients are complement-depleted with low levels of C3 and, in par-
ticular, C4. During acute phase reactions, C3 and C4 are repleted and
complement-induced hemolysis increases.
4, 39
It has been known for decades that the CA in patients with prima-
ry CAD are usually monoclonal IgMantibodies, most often with κ light
chain restriction.
40, 41
Our study of 86 unselected patients showed
that this was the case in more than 90% of the patients, whereas
monoclonal IgG, IgA or λ light chain restriction were rare findings.
In 6% of the patients, monoclonal Ig could not be detected despite
otherwise characteristic primary CAD. This is probably a matter of
sensitivity. Anti-I CA in patients with primary CAD show restriction
to the IGHV4-34 gene segment.
During the last 15 years it has become clear that in a majority of pa-
tients, clonality at the B-cell level can also be demonstrated by flowcy-
tometry and/or immunohistochemistry.
6, 8
We found a clonal, CD20
B-lymphocyte population and a cellular κ/λ ratio of more than 3.5 in
90% of bone marrow aspirates,
while the sensitivity of flow cytometry
was lowif performed in peripheral blood.
A clonal lymphoproliferative
bone marrow disorder, usually discrete, was confirmed by immunohis-
tochemistry in 75% of the patients.
The lymphoproliferation was most
frequently classified as lymphoplasmacytic lymphoma (LPL; 50% of the
Table 1
Autoimmune hemolytic anemia.
Warm-antibody type
Cold-antibody type
Primary chronic cold agglutinin disease
Secondary cold agglutinin syndrome
Associated with malignant disease
Acute, infection-associated
Paroxysmal cold hemoglobinuria
Mixed cold- and warm-antibody type
Fig. 1. Peripheral blood smear from a patient with primary CAD.
108 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
total patient cohort) or marginal zone lymphoma (MZL; 8% of the pa-
6, 42
The histopathology findings are listed in Table 2.
Since LPL is a frequent finding and most if not all patients have
monoclonal IgM, a considerable overlap exists between primary
CAD and Waldenström's macroglobulinemia (WM).
In most pa-
tients who do not fulfill the criteria for LPL/WM or MZL, primary
CAD can be classified as an IgM-related disorder (IgM-RD). IgM-RD
is defined as a clinical condition characterized by specific properties
of monoclonal IgM proteins, but without lymphoma.
In a clinical
context, however, CAD with definitive or merely detectable clonal
lymphoproliferation should be regarded a continuum, not distinct en-
tities. Unlike warm-antibody AIHA, primary CAD does not appear to
be associated with other autoimmune diseases, probably reflecting a
competent regulation of the immune system.
4, 6
3.3. Diagnosis
Primary CAD should be suspected in elderly patients with chronic
hemolytic anemia and/or the cold-induced circulatory symptoms
mentioned above. Although not specific, the observation of aggluti-
nated erythrocytes in a peripheral blood smear (Fig. 1) will increase
the suspicion of CA-mediated phenomena. When hemolysis has
been confirmed by biochemical tests and, often, elevated absolute re-
ticulocyte counts, polyspecific DAT is required to detect autoimmune
pathogenesis. Monospecific DAT is also mandatory and will be
strongly positive for C3d and negative (or weakly positive in 20%
of the patients) for IgG.
6, 34
CA titers should be determined using
serial two-fold dilutions of serum before adding a suspension of
adult 0RhD positive erythrocytes and incubating at 4 °C.
A titer of
64 or more is required for diagnosis, but most patients have sub-
stantially higher titers.
Determination of the thermal amplitude
may be useful in some situations, but is usually not necessary in
order to establish the diagnosis. The concentrations of complement
proteins C3 and C4 should be determined, but normal levels do not
exclude CAD.
4, 6, 31, 39
Confirming the presence of a clonal lymphoproliferative disorder
has potential therapeutic consequences, even though negative find-
ings may be a matter of sensitivity and do not exclude primary
6, 31
Capillary electrophoresis or agarose electrophoresis with
immunofixation should always be performed. If no monoclonal
band can be detected on electrophoresis, immunofixation should
still be done. A trephine biopsy should be examined by an experi-
enced lymphoma pathologist, and we also recommend flow cyto-
metric immunophenotyping of bone marrow lymphocytes.
8, 9
History and clinical examination, supplemented by radiological imag-
ing as required, will usually be sufficient to exclude cases of second-
ary chronic CAS described below. The diagnostic criteria for primary
CAD are summarized in Table 3.
6, 31
Fig. 2 shows a diagnostic
Importantly, in order to achieve sufficient sensitivity, serum for
immunoglobulin analyses and CA titration must be obtained from
blood specimens kept at 37–38 °C from sampling until serum has
been removed from the clot.
4. Chronic cold agglutinin syndrome secondary to malignant
4.1. Definition, frequency and underlying disorders
After primary CAD was shown to be a clonal lymphoproliferative
disease, there has been some confusion in the literature regarding
Table 2
Bone marrow histopathology in primary chronic cold agglutinin disease.
Immunohistochemistry findings n %
Normal or reactive lymphocytosis 7 11
Irregular lymphoid hyperplasia 9 13
Non-Hodgkin's B cell lymphoma 50 76
Lymphoplasmacytic lymphoma 33 50
Marginal zone lymphoma 5 8
Small lymphocytic B-cell lymphoma 4 6
Clonal lymphocytosis/other small B-cell lymphoma 8 12
Total 66 100
Table 3
Diagnostic criteria for primary chronic cold agglutinin disease.
Level Criteria Procedures and comments
Required for
Chronic hemolysis
Polyspecific DAT positive
Monospecific DAT strongly
positive for C3d
DAT is usually negative for IgG,
but occasionally weakly positive
Cold agglutinin (CA) titer
≥64 at 4 °C
Blood specimen must be kept at
37–38 °C from sampling until
serum is removed from the clot
No overt malignant disease Clinical assessment. Radiology as
but not
required for
Monoclonal IgM κ in serum
(or, rarely, IgG, IgA or λ
Serum must be obtained as for CA
titer. Immunofixation should be
done even if no band is visible on
Cellular κ/ λ ratio >3.5 (or,
rarely, b0.9) in B-lymphocyte
Flow cytometry in bone marrow
Clonal lymphoproliferative
bone marrow disorder by
Trephine biopsy
not AIHA
Polyspecific DAT
Monospecific DAT
for C3d
for C3d
> 64
Assess for infection or overt malignancy
Primary CAD
Onset of
anemia before
Secondary CAS
clonality if
(Table 3)
Fig. 2. Diagnostic algorithm for cold-antibody mediated AIHA. Abbreviations: AIHA, au-
toimmune hemolytic anemia; CA, cold agglutinin; CAD, cold agglutinin disease; CAS,
cold agglutinin syndrome; w-, warm-antibody type.
109 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
the terms ‘secondary’ versus ‘primary’. Patients with chronic CAD
recognized by us and others as having a clonal B-cell disorder, most
often non-progressive and clinically non-malignant, undoubtedly
represent the same majority that has traditionally been diagnosed
with primary or idiopathic CAD.
1, 6, 8, 36
In these patients, the disease
should still be called primary CAD. The term ‘secondary’ chronic CAS
should be reserved for those patients in whom the cold-antibody
mediated hemolytic anemia complicates an overt and well-defined
malignant disease different from LPL and MZL.
1, 6, 31, 42, 47, 48
Among 295 consecutive individuals with AIHA described by
Dacie, 7 patients (2.4%) were classified as having CAS secondary to
malignant disease.
In the very large series of AIHA by Sokol's
group, the frequency seemed higher.
CAS has been described in pa-
tients diagnosed with diffuse large B-cell lymphoma, Hodgkin's
lymphoma, carcinomas, sarcomas, metastatic melanoma and chron-
ic myeloproliferative disorders.
1, 2, 12, 13, 47–50
For the following rea-
sons, however, both the reported frequencies and some of the
assumed associations should be regarded uncertain. First, particu-
larly in case reports, patients may simply have suffered from two
independent diseases; cancer and primary CAD. Second, sufficient
details have often not been provided to critically review the diagno-
sis of the co-existing or underlying malignancy. In a comprehensive
series from a regional transfusion center, for example, advanced se-
rological characterization of the AIHA was done at the referral cen-
ter, whereas the reported frequencies of malignant disorders were
based on more or less poorly specified information from the refer-
ring hospitals.
Third, althoughrarely, we have observedtransformation
to diffuse large B-cell lymphoma of the low-grade lymphoproliferative
disorder characteristic for primary CAD.
Fourth, changing and more
standardized tumor classifications should justify a re-interpretation of
early reports. This issue is highlighted by a description from1978 of “sar-
coma” in two CAD patients who would probably be classified according
to the 2008 version of the World Health Organization (WHO) classifica-
tion as having LPL or splenic MZL.
42, 51
The re-classification into aggres-
sive non-Hodgkin's lymphoma of certain tumors previously perceived
as lymphocyte depleted Hodgkin's lymphoma is also well-known.
In our experience, true secondary CAS is far more uncommon than
primary CAD. The best documentation for a clearly malignant disease
resulting in CAS seems to have been provided in non-Hodgkin's
12, 13, 47, 48
4.2. Clinical and pathological features and diagnostic criteria
Besides the autoimmune hemolysis, the clinical and pathological
features of secondary CAS depend on the underlying malignancy.
The diagnosis can sometimes be based on the occurrence of CA medi-
ated AIHA in a patient already diagnosed with an aggressive lympho-
ma. In other cases, the diagnostic pathway shown in Fig. 2 will be
relevant. The DAT features and occurrence of CA in serum do usually
not differ substantially from the findings in primary CAD.
In con-
trast to the κ light chain phenotype found in almost all patients
with primary CAD, however, the light chain restriction can be λ as
well as κ.
47, 48
5. Acute cold agglutinin syndrome associated with infection
5.1. Mycoplasma pneumoniae
An association between CA and respiratory disease was already
observed in 1918.
More precisely, the occurrence of high-titer CA
in primary atypical pneumonia was described in 1943 and soon
thereafter identified as a cause of hemolytic anemia in such pa-
52, 53
Probably as part of the physiological immune response,
most patients with M. pneumoniae infections produce CA. In the ma-
jority, these CA do not give rise to significant hemolysis; and before
specific tests became widely available, demonstration of CA activity
was used as a diagnostic tool in Mycoplasma infections. In some pa-
tients, however, production of high-titer, high-thermal amplitude
CA may result in AIHA which may occasionally be severe.
295 patients with AIHA, Mycoplasma or primary atypical pneumonia
was identified as the probable cause in 23 (8%).
Conversely, the fre-
quency of clinically significant hemolysis in patients with M. pneumo-
niae infection is unknown. Six (24%) of 25 patients admitted to a
referral center with this type of pneumonia had hemolysis; severe
in two patients and mild to moderate in four.
In general hospitals
and in the community, however, the frequency of hemolytic compli-
cations is probably much lower.
The autoantibodies are polyclonal, usually anti-I specific and al-
most invariably of the IgM class.
CA titers typically range between
512 and 32000. DAT is always positive for C3d.
3, 56
Most reported pa-
tients have been adults, and AIHA typically occurs during the second
or third week after the febrile illness has started.
In most published
cases the onset has been sudden with pallor, jaundice and, some-
times, prostration. Intravascular hemolysis, as evidenced by hemoglo-
binuria, has been reported in several patients. In general, the
prognosis is good and the hemolytic complication is self-remitting
within 4–6 weeks, although a lethal course has been described in
one patient.
55, 56
5.2. Epstein–Barr virus infection
A number of case reports have been published on CA mediated
AIHA in infectious mononucleosis with confirmed Epstein–Barr
virus (EBV) etiology.
As compared to M. pneumoniae pneumonia,
however, infectious mononucleosis is an infrequent cause of AIHA, ac-
counting for approximately 1% of the cases.
1, 2
Conversely, the fre-
quency of clinically significant hemolysis in EBV infections is
unknown but probably low. Hospital-based data have indicated that
hematological complications, being generally mild and including sev-
eral manifestations other than AIHA, occur in 25–50% of patients with
EBV infection.
Since patient selection will influence such figures and
most individuals with infectious mononucleosis are not hospitalized,
the frequency is probably much lower among patients with EBV-
infection in the community.
CA found in EBV-infections are polyclonal and almost invariably
specific for the i-antigen.
57, 60, 61
The immunoglobulin class may be
either IgM or IgG.
Rheumatoid factor-like IgM-IgG complexes have
also been reported to act as CA in single cases.
In most published re-
ports, DAT has been positive for C3d only. Anti-i titers are usually
modest, typically at 256–512, and the hemolytic anemia is transient
and generally mild.
5.3. Other infections
Several authors have reported on transient CAS mediated by anti-i
autoantibodies following cytomegalovirus (CMV) infection.
60, 61
Rarely, CA-mediated hemolysis has been described in adenovirus in-
fections, influenza A, varicella, rubella, Legionella pneumophilica pneu-
monia, listeriosis and pneumonia caused by Chlamydia species.
11, 60,
We observed a slight, transient CAS in an otherwise healthy
23 year old man two weeks after a Chlamydia pneumoniae pneumo-
nia. Severe CAS with a prolonged course and cryoglobulin activity of
the CA has been reported following Escherichia coli lung infection.
Autoantibody specificities in these rare cases have included anti-I,
anti-i and anti-Pr.
5.4. Diagnosis of infection-associated cold agglutinin syndrome
Cold-antibody AIHA with infectious etiology typically involves a
young adult or adolescent with M. pneumoniae pneumonia or infec-
tious mononucleosis. Anemia or hemoglobinuria in such patients
should immediately raise the suspicion of secondary AIHA. In addition
110 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
to the work-up required to diagnose the infection, assessment of hemo-
lysis, polyspecific as well as monospecific DAT and CAtitration will usu-
ally establish the diagnosis. No monoclonal spike is found on
electrophoresis. In some cases, paroxysmal cold hemoglobinuria and
infection-induced exacerbation of primary CAD will have to be ruled
out as differential diagnoses.
6. Other secondary cold agglutinin syndromes
A few case reports have described chronic CA-mediated hemolysis
in patients with systemic lupus erythematosus (SLE). In one of these
publications, the presence of a clonal disorder was considered but
could not be confirmed.
These very rare cases of SLE-associated
CAS should not be confused with primary CAD. Several authors have
reported the development of CA-mediated hemolysis after allogenic
stem cell transplantation. In some of these patients, the AIHA seemed
related to the transplantation per se; in other cases it was associated
with virus infection.
64, 67
CAS has also been described during preg-
nancy in one single patient.
7. Treatment
7.1. Primary chronic cold agglutinin disease
7.1.1. Non-pharmacological management
Until a decade ago, pharmacological therapy for primary CAD was
largely ineffective.
6, 69
Partly based on this fact and partly because the
severity of the clinical features have not been appreciated, counseling
has been regarded the mainstay of management.
3, 6, 36
However, doc-
umentation of efficacy is mainly anecdotal.
15, 70
Still, in our clinical
experience, staying warm seems to alleviate the symptoms and can
probably prevent severe exacerbations of hemolytic anemia. In par-
ticular, the head, face and extremities should be protected against
cold exposure.
36, 69, 71
Some patients experience increased Hgb levels
and alleviation of circulatory symptoms after relocating to warmer
regions during the cold season, but severely symptomatic CAD does
exist even in the subtropics. Infusion of cold liquids should be
avoided. Surgery under hypothermia requires specific precautions,
e.g. preoperative plasmapheresis.
72, 73
Erythrocyte transfusions can safely be given provided appropriate
precautions are undertaken.
31, 69
In contrast to the compatibility
problems characteristic for warm-antibody AIHA, it is usually easy
to find compatible donor erythrocytes, and screening tests for irregu-
lar blood group antibodies are most often negative. Antibody screen-
ing and, if required, compatibility tests should be performed at 37 °C.
The patient and, in particular, the extremity chosen for infusion
should be kept warm, and the use of an in-line blood warmer is
Failure to observe required precautions has resulted
in dismal or, very rarely, even fatal outcome.
72, 74
Because comple-
ment proteins can exacerbate hemolysis, transfusion of blood prod-
ucts with a high plasma content should probably be avoided.
In a population-based retrospective series on primary CAD we iden-
tified three splenectomised patients, none of whom had responded to
the splenectomy.
This observation is not surprising, since clearance
of C3b-opsonized erythrocytes primarily occurs in the liver. Improve-
ment after splenectomy has been occasionally reported among the
rare patients with CAD mediated by IgG instead of IgM autoanti-
Probably because nearly all IgMis intravascular, plasmaphere-
sis efficiently induces clinical improvement in acute situations or before
surgery requiring hypothermia.
These remissions are short-lived,
7.1.2. Unspecific immunosuppressive and supportive therapies
Although patients with CAD often have received corticosteroids,
this practice has never been supported by systematic studies.
Among 38 consecutive patients seen at the Hammersmith Hospital
in London, only occasional patients responded to therapy with ste-
Similar clinical experience has been obtained by others.
71, 76
Studied retrospectively, 43% of unselected Norwegian patients
with CAD had been treated with corticosteroids for shorter or longer
periods. Responses had been observed in only 14% of those treated,
and the few patients who did respond usually required high doses
in order to maintain the remission.
The requirement for unaccept-
ably high maintenance doses in the occasional responders has also
been observed by others.
Monotherapy with chlorambucil or cyclo-
phosphamide has shown some beneficial effect on laboratory param-
eters, and clinical improvement has been described.
76, 78
The clinical
response rates, however, are in the same low order of magnitude as
for corticosteroids.
A few patients treated with azathioprine have
been reported in the literature, none of whom responded.
6, 34
two small series of therapy with interferon-α or low-dose cladribine,
respectively, these drugs failed to induce clinical remission, although
some conflicting data have been published with interferon-α.
Symptomatic therapy with erythropoietin or its analogues seems
widely used in the USA, but not so often in Western and Northern
Europe (S. Berentsen, unpublished observation). Folic acid supple-
mentation is rather commonly prescribed.
None of these support-
ive measures have been systematically studied. In exacerbation of
hemolysis triggered by febrile illness, immediate treatment of any
bacterial infection is indicated.
4, 31, 39
7.1.3. Therapies directed at the pathogenic B-cell clone
The first major advance in treatment of primary CAD was the
achievement of remission following monotherapy with the human-
ized, chimeric monoclonal anti-CD20 antibody rituximab. Several
case reports on rituximab therapy have been published since
and we reported in 2001 promising results of a small, pro-
spective trial.
Two larger, prospective, uncontrolled trials of 37 and
20 courses of therapy, respectively, were published in 2004 and
87, 88
The dosage of rituximab was 375 mg/m
weekly for four
weeks in both studies; and the baseline data, response definitions
and response data were similar. The response criteria used in our
trial are listed in Table 4.
We found an overall response rate of
54%. With the exception of complete response (CR) in one patient,
all remissions were partial responses (PR). Ten patients were treated
for relapse after previously having received rituximab therapy, and
six of them responded to a second course. The responders achieved
a median increase in Hgb levels of 4.0 g/dL. The median time to re-
sponse was 1.5 months (range, 0.5–4.0 months) and the median ob-
served response duration was 11 months (range, 2–42 months). The
treatment was well tolerated.
Retrospectively, the results of rituximab monotherapy were also
studied in the population-based descriptive study of 86 Norwegian
Forty patients were reported to have received rituximab
Table 4
Primary chronic cold agglutinin disease: response criteria used in clinical trials.
Absence of anemia
No signs of hemolysis
Disappearance of clinical symptoms of CAD
No monoclonal serum protein
No signs of clonal lymphoproliferation as assessed by bone
marrow histology, immunohistochemistry and flow cytometry
A stable increase in hemoglobin levels by at least 2.0 g/dL or to
the normal range
A reduction of serum IgM levels by at least 50% of the initial level
or to the normal range
Improvement of clinical symptoms
Transfusion independence
No response
Any outcome not meeting the criteria for CR or PR
111 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
monotherapy. As far as permitted by the retrospective design, the
previously published response criteria (Table 4) were used for the
data analysis. Twenty-three responders (58%) were identified; two
(5%) achieved CR and 21 (53%) achieved PR. Responses were ob-
served following a second and even a third course of rituximab in pa-
tients who had relapsed after previous therapy.
These findings
confirmthe essential results of the prospective studies; rituximab sin-
gle agent therapy is an efficient and well tolerated treatment for pri-
mary CAD. CR is uncommon, however; the median response duration
is relatively short; and 40-50% of the patients do not respond.
We wanted to improve on the results achieved by rituximab
monotherapy. The purine nucleoside analogues, e.g. fludarabine, are
powerful therapeutic agents in several lymphoproliferative diseases
and remission of CAD has been observed in at least two patients
after monotherapy with fludarabine.
6, 89
Furthermore, combined
treatment with fludarabine and rituximab has resulted in high re-
sponse rates and sustained remissions in WM and other low-grade
non-Hodgkin's lymphoma.
90, 91
We published in 2010 a prospective, uncontrolled trial of fludar-
abine and rituximab in combination in 29 patients with primary
CAD requiring treatment.
The median age was 73 years (range,
39–87 years). Eligible patients received rituximab 375 mg/m
days 1, 29, 57 and 85; and fludarabine orally, 40 mg/m
on days
1–5, 29–34, 57–61 and 85–89. Growth factors, co-trimoxazole or
antiviral agents were not used routinely. Table 4 shows the re-
sponse criteria. Responses were observed in 22 patients (76%); six
(21%) achieved CR and 16 (55%) achieved PR. Ten patients had pre-
viously been non-responsive to rituximab monotherapy. In this
subgroup, CR was observed in one patient and PR in six. Median in-
crease in Hgb level was 3.1 g/dL in the responders and 4.0 g/dL
among those who achieved CR. Median time to response was
4.0 months, and estimated median response duration was more
than 66 months.
Although comparison of non-randomized trials
must be interpreted with caution, the much higher response rates,
promising frequency of CR and very long response duration ob-
served after the combination therapy compare favorably with the
results achieved by rituximab monotherapy.
Toxicity, however, occurred more frequently with the fludarabine-
rituximab combination.
Hematological toxicity grades 3–4 were ob-
served in 12 patients (41%), including grade 4 neutropenia in four
(14%). Seventeen patients (59%) experienced grade 1–3 infection.
All infections were successfully treated except for one old, frail non-
responder who died of pneumonia after nine months. Three patients
(10%) suffered herpes zoster reactivation, but Pneumocystis jirovecii
pneumonia or infection grade 4 was not observed. Fludarabine-
induced warm-antibody AIHA did not occur, but three patients
(10%) experienced a transient, mild exacerbation of CAD precipitated
by infection.
39, 92
The study was not designed to address the risk of
myelodysplasia or late-occurring hematological malignancies. Al-
though not specific to nucleoside analogues, such late events have
been reported after fludarabine-based therapy for WM.
This con-
cern should not be prohibitive to the use of the combination therapy,
but lead to a balanced, individualized consideration of risk versus
7.1.4. Recommendations
There seems to be a discrepancy between the restrictive attitude
to pharmacological therapy for CAD often found in the literature
and the real requirement for therapy.
Recommendations to avoid
medications may simply reflect the fact that in the past, treatment
was ineffective. Underestimation of the severity of anemia and clini-
cal symptoms in this particular patient population may also have
influenced the attitudes. In selected patients, actually, the circulatory
symptoms may be sufficiently disabling to justify therapy even if the
hemolysis is fully compensated.
10, 92
Some patients, however, do
have a mild disease in which the anemia is slight and the circulatory
symptoms modest or absent. In consequence, CAD should not be
regarded an indication for therapy in every patient, and the decision
to treat should be based on an individualized assessment. Reasonable
criteria for starting drug therapy are symptomatic anemia, transfu-
sion dependence, and/or disabling circulatory symptoms.
10, 87, 92
Corticosteroids should not be used to treat primary CAD.
6, 15, 31, 69
Outside clinical trials, the fludarabine and rituximab combination
should be regarded the most efficient treatment to date and should
be considered in elderly patients requiring therapy if they are other-
wise reasonably fit and have no relevant co-morbidity. The combina-
tion has proved useful even in patients non-responsive to monotherapy
Given the toxicity, however, a balanced assessment of
risk versus benefit should be undertaken in every case. In the occasional
young patients as well as the very old and co-morbid ones, rituximab
monotherapy should be considered first-line treatment. Those who re-
lapse after having responded to rituximab as single agent therapy may,
depending on an individualized assessment, receive another course of
rituximab or proceed to combination therapy.
7.1.5. Future perspectives
Although significant advances have been made during the last ten
years in treating primary CAD, future studies should aim to achieve
still higher response rates, increased numbers of CR and even more
prolonged response duration. Less toxic regimens and efficient
second-line therapies should also be regarded realistic achievements.
For example, it has been proposed to explore the safety and efficacy of
fludarabine and rituximab in combination using reduced doses of flu-
92, 93
It may also be worthwhile investigating combinations
of monoclonal antibodies with newer chemotherapeutic agents. The
relationship between primary CAD and WMshould encourage studies
of several, more or less targeted therapies shown to be feasible and
efficient in WM.
In primary CAD, improvement has been observed
in two patients following bortezomib monotherapy
; and high re-
sponse rates have been achieved in WM following treatment with a
bortezomib-based combination regimen.
The monoclonal anti-C5 antibody, eculizumab, is a potent comple-
ment inhibitor shown to be an efficient therapeutic agent in paroxys-
mal nocturnal hemoglobinuria (PNH).
In steady-state CAD, on the
other hand, most of the hemolysis is not thought to be intravascular
and C5-mediated.
30, 31
Furthermore, the administration of eculizumab
in PNHhas been shown to unmask the low-grade, C3b-mediated extra-
vascular hemolysis assumed to predominate in CAD.
Infusions of ecu-
lizumab have been reported, however, to result in rapid improvement
in a patient with primary CAD
; and unpublished observations may in-
dicate a marked and sustained suppression of hemolysis during contin-
ued therapy (A. Röth, personal communication). These observations
should be further explored for two reasons. First, this therapeutic ap-
proach might prove useful in subgroups, e.g. in acute situations (infec-
tions or surgery with exacerbation of hemolysis) or in severely
hemolytic patients not responding to therapy directed against the path-
ogenic B-cell clone. Second, if efficacy is confirmed, such results may
challenge our present understanding of hemolysis in CAD, theoretically
leading to a re-consideration of which hemolytic mechanism is most
In order to further improve on current treatment options in pri-
mary CAD, patients requiring therapy should be considered for inclu-
sion in prospective trials if available.
7.2. Management of secondary cold agglutinin syndromes
No evidence-based therapy exists for the CAS per se in cold-
antibody mediated AIHA secondary to clearly malignant or infectious
diseases. Prospective trials or well-designed retrospective series of con-
secutive patients have not been published, and all recommendations
found in the literature are based on case reports, clinical experience
and theoretical considerations.
112 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
For obvious reasons, however, optimal treatment of the underlying
disease is important whenever feasible.
15, 69
Particularly in curable ma-
lignancies such as aggressive lymphomas, achieving complete remis-
sion is usually accompanied by resolution of the hemolysis. M.
pneumoniae pneumonia should be treated according to appropriate
guidelines; although due to the time of onset of hemolysis, antibiotic
therapy will often have been completed before the hemolytic anemia
manifests itself.
In some virus infections, causal therapy is not possi-
ble; in others, such as CMV infection in immunocompetent individuals,
antiviral therapy is not necessary because the CAS is slight and self-
The efficacy of corticosteroid therapy has not been systematically
investigated, and opinions promoted in textbooks and review articles
differ widely.
3, 14, 15, 69
We treated a 45 year old man admitted with
Mycoplasma pneumonia and an initial Hgb level of 3.5 g/dL due to
an anti-I mediated severe CAS. He received erythromycin and high-
dose prednisolone followed by rapid tapering of the corticosteroid
dose. His condition improved rapidly and he became transfusion in-
dependent within days. Similar case reports have been published in
the literature.
100, 101
Since spontaneous resolution eventually occurs
in all patients, however, guidelines can hardly be built upon case ob-
servations. Therefore, there is no documentation for using corticoste-
roids routinely in CAS secondary to infection. Until more data are
provided, corticosteroid therapy may be considered if the hemolysis
is severe and spontaneous improvement does not occur within
some days. Plasmapheresis may be helpful in selected, extreme
72, 102
If indicated, erythrocyte transfusions can safely be given
provided the same precautions are undertaken as in primary CAD.
8. Practice points
• Diagnosing the subtype of AIHA precisely is essential for the choice
of therapy.
• Primary CAD is a well-defined disease characterized by a clonal B-
lymphocyte proliferation.
• Not all patients with primary CAD require pharmacological therapy,
but treatment seems indicated more often than previously
• Corticosteroids should not be used to treat primary CAD.
• In primary CAD, the most efficient therapy documented to date is
the administration of fludarabine and rituximab in combination.
Toxicity, however, may be a concern.
• Rituximab monotherapy is well documented and should be consid-
ered in some patients.
• No evidence-based recommendations can be given for the manage-
ment of secondary CAS, but optimal treatment of the underlying
disorder is essential when feasible.
• Erythrocyte transfusions can safely be given in cold-antibody AIHA
provided appropriate precautions are observed.
9. Research agenda
• The molecular, immunological and immunohistochemical charac-
teristics of the clonal lymphoproliferation in primary CAD should
be further studied.
• Further therapeutic trials are warranted in primary CAD, aiming at
still higher rates of complete remissions, even more prolonged re-
sponse duration, less toxic regimens and efficient therapeutic op-
tions in the second-line setting.
• Patients with CAD requiring therapy should be considered included
in prospective trials if available.
• In the rare and diverse cases of secondary CAS, therapeutic trials are
probably unrealistic. International, well-designed, descriptive series
of consecutive patients, however, might add valuable scientific
Conflict of interest
The authors declare no financial or other conflicts of interest.
[1] Dacie J. The auto-immune haemolytic anaemias: introduction. In: Dacie J, editor.
The Haemolytic Anaemias, Vol. 3. London: Churchill Livingstone; 1992. p. 1–5.
[2] 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)
[3] Gehrs BC, Friedberg RC. Autoimmune hemolytic anemia. Am J Hematol
[4] Ulvestad E, Berentsen S, Bo K, Shammas FV. Clinical immunology of chronic cold
agglutinin disease. Eur J Haematol 1999;63(4):259–66.
[5] Dacie J. Auto-immune haemolytic anaemia (AIHA): cold-antibody syndromes II:
immunochemistry and specificity of the antibodies; serum complement in auto-
immune haemolytic anaemia. In: Dacie J, editor. The Haemolytic Anaemias, Vol.
3. London: Churchill Livingstone; 1992. p. 240–95.
[6] Berentsen S, Ulvestad E, Langholm R, Beiske K, Hjorth-Hansen H, Ghanima W,
et al. Primary chronic cold agglutinin disease: a population based clinical study
of 86 patients. Haematologica 2006;91(4):460–6.
[7] Pascual V, Victor K, Spellerberg M, Hamblin TJ, Stevenson FK, Capra JD. VH re-
striction among human cold agglutinins. The VH4-21 gene segment is required
to encode anti-I and anti-i specificities. J Immunol 1992;149(7):2337–44.
[8] Berentsen S, Bo K, Shammas FV, Myking AO, Ulvestad E. Chronic cold agglutinin
disease of the “idiopathic” type is a premalignant or low-grade malignant lym-
phoproliferative disease. APMIS 1997;105(5):354–62.
[9] Berentsen S, Ulvestad E, Tjonnfjord GE. B-lymphocytes as targets for therapy in
chronic cold agglutinin disease. Cardiovasc Hematol Disord Drug Targets
[10] Berentsen S. How I manage cold agglutinin disease. Br J Haematol 2011;153(3):
[11] King JW, May JS. Cold agglutinin disease in a patient with Legionnaires' disease.
Arch Intern Med 1980;140(11):1537–9.
[12] Niainle F, Hamnvik OP, Gulmann C, Bermingham C, Kelly J, Mc EP, et al. Diffuse
large B-cell lymphoma with isolated bone marrow involvement presenting
with secondary cold agglutinin disease. Int J Lab Hematol 2008;30(5):444–5.
[13] Eskazan AE, Akmurad H, Ongoren S, Ozer O, Ferhanoglu B. Primary gastrointesti-
nal diffuse large B cell lymphoma presenting with cold agglutinin disease. Case
Rep Gastroenterol 2011;5(2):262–6.
[14] Powers A, Silberstein LE. Autoimmune hemolytic anemia. In: Hoffman R, Benz EJ,
Shattil SJ, Furie B, Silberstein LE, McGlave P, et al, editors. Hematology: basic
principles and practice. Philadelphia: Churchill Livingstone Elsvier; 2008.
p. 645–57.
[15] Gertz MA. Management of cold haemolytic syndrome. Br J Haematol 2007;138(4):
[16] Landsteiner K. Über Beziehungen zwischen dem Blutserum und den Körperzel-
len. Münch Med Wochenschr 1903;50:1812–4.
[17] Clough MC, Richter IM. A study of an auto-agglutinin occurring in human serum.
Johns Hopkins Hosp Bull 1918;29:86–93.
[18] Rosenthal F, Corten M. Über das Phänomen der Autohämagglutination und über
die Eigenscaften der Kältehämagglutinine. Folia Haematol (Leipzig) 1937;58:
[19] Stone MJ, McElroy YG, Pestronk A, Reynolds JL, Newman JT, Tong AW. Human
monoclonal macroglobulins with antibody activity. Semin Oncol 2003;30(2):
[20] Rosse WF, Adams JP. The variability of hemolysis in the cold agglutinin syn-
drome. Blood 1980;56(3):409–16.
[21] Garratty G, Petz LD, Hoops JK. The correlation of cold agglutinin titrations in sa-
line and albumin with haemolytic anaemia. Br J Haematol 1977;35(4):587–95.
[22] Poldre P, Pruzanski W, Chiu HM, Dotten DA. Fulminant gangrene in transient
cold agglutinemia associated with Escherichia coli infection. Can Med Assoc J
[23] Bhattacharyya J, Mihara K, Takihara Y, Kimura A. Successful treatment of IgM-
monoclonal gammopathy of undetermined significance associated with cryoglo-
bulinemia and cold agglutinin disease with immunochemotherapy with rituxi-
mab, fludarabine, and cyclophosphamide. Ann Hematol Sep. 8 2011 [E-pub
ahaead of print].
[24] Issitt PD. I blood group system and its relationship to disease. J Med Lab Technol
[25] Marsh WL. Anti-i: a cold antibody defining the Ii relationship in human red cells.
Br J Haematol 1961;7:200–9.
[26] Roelcke D. Cold agglutination. Transfus Med Rev 1989;3(2):140–66.
[27] Brain MC, Ruether B, Valentine K, Brown C, ter Keurs H. Life-threatening hemo-
lytic anemia due to an autoanti-Pr cold agglutinin: evidence that glycophorin A
antibodies may induce lipid bilayer exposure and cation permeability indepen-
dent of agglutination. Transfusion 2010;50(2):292–301.
[28] Jonsen J, Kass E, Harboe M. Complement and complement components in ac-
quired hemolytic anemia with high titer cold antibodies. Acta Med Scand
[29] Jaffe CJ, Atkinson JP, Frank MM. The role of complement in the clearance of cold
agglutinin-sensitized erythrocytes in man. J Clin Invest 1976;58(4):942–9.
[30] Zilow G, Kirschfink M, Roelcke D. Red cell destruction in cold agglutinin disease.
Infusionsther Transfusionsmed 1994;21(6):410–5.
113 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
[31] Berentsen S, Beiske K, Tjonnfjord GE. Primary chronic cold agglutinin disease: an
update on pathogenesis, clinical features and therapy. Hematology 2007;12(5):
[32] Lewis SM, Szur L, Dacie JV. The pattern of erythrocyte destruction in haemolytic
anaemia, as studied with radioactive chromium. Br J Haematol 1960;6:122–39.
[33] Kirschfink M, Knoblauch K, Roelcke D. Activation of complement by cold agglu-
tinins. Infusionsther Transfusionsmed 1994;21(6):405–9.
[34] Chandesris MO, Schleinitz N, Ferrera V, Bernit E, Mazodier K, Gayet S, et al. Cold
agglutinins, clinical presentation and significance: retrospective analysis of 58
patients. Rev Med Interne 2004;25(12):856–65.
[35] Genty I, Michel M, Hermine O, Schaeffer A, Godeau B, Rochant H. Characteristics
of autoimmune hemolytic anemia in adults: retrospective analysis of 83 cases.
Rev Med Interne 2002;23(11):901–9.
[36] Schubothe H. The cold hemagglutinin disease. Semin Hematol 1966;3(1):27–47.
[37] Lyckholm LJ, Edmond MB. Images in clinical medicine. Seasonal hemolysis due to
cold-agglutinin syndrome. N Engl J Med 1996;334(7):437.
[38] Ulvestad E. Paradoxical haemolysis in a patient with cold agglutinin disease. Eur
J Haematol 1998;60(2):93–100.
[39] Ulvestad E, Berentsen S, Mollnes TE. Acute phase haemolysis in chronic cold ag-
glutinin disease. Scand J Immunol 2001;54(1–2):239–42.
[40] Harboe M, Deverill J. Immunochemical properties of cold haemagglutinins.
Scand J Haematol 1964;61:223–37.
[41] Harboe M, van Furth R, Schubothe H, Lind K, Evans RS. Exclusive occurrence of K
chains in isolated cold haemagglutinins. Scand J Haematol 1965;2(3):259–66.
[42] WHO classification of tumours of haematopoietic and lympoid tissues. 4 ed.
Lyon: International Agency for Resarch on Cancer; 2008.
[43] Berentsen S. Cold agglutinin-mediated autoimmune hemolytic anemia in Wal-
denstrom's macroglobulinemia. Clin Lymphoma Myeloma 2009;9(1):110–2.
[44] Owen RG, Treon SP, Al Katib A, Fonseca R, Greipp PR, McMaster ML, et al. Clini-
copathological definition of Waldenstrom's macroglobulinemia: consensus
panel recommendations from the Second International Workshop on Walden-
strom's Macroglobulinemia. Semin Oncol 2003;30(2):110–5.
[45] Stone MJ, Merlini G, Pascual V. Autoantibody activity in Waldenstrom's macro-
globulinemia. Clin Lymphoma 2005;5(4):225–9.
[46] Cesana C, Barbarano L, Miqueleiz S, Lucchesini C, Ricci F, Varettoni M, et al. Clin-
ical characteristics and outcome of immunoglobulin M related disorders. Clin
Lymphoma 2005;5(4):261–4.
[47] Crisp D, Pruzanski W. B-cell neoplasms with homogeneous cold-reacting anti-
bodies (cold agglutinins). Am J Med 1982;72(6):915–22.
[48] Dacie J. Haemolytic anaemias associated with malignant lymphomas other than
Hodgkin's disease and chronic lymphocytic leukaemia (CLL). In: Dacie J, editor.
The Haemolytic Anaemias, Vol. 4. London: Churchill Livingstone; 1995. p. 27–40.
[49] Al Matham K, Alabed I, Zaidi SZ, Qushmaq KA. Cold agglutinin disease in fibrola-
mellar hepatocellular carcinoma: a rare association with a rare cancer variant.
Ann Saudi Med 2011;31(2):197–200.
[50] Carlsson J, Batge R, Rahlf G, Tebbe U. Cold agglutinin disease and metastatic mel-
anoma. Med Klin (Munich) 1994;89(6):343–5.
[51] Diebold J, Reynes M, Tricot G, Zafrani E, Weill B, Dao C, et al. Splenic lympho-
plasmocytic sarcomas discovered by exploratory laparotomy in 2 cases of cold
agglutinin disease. Sem Hop 1978;54(43–44):1325–30.
[52] Peterson OL, Ham TH, Finland M. Cold agglutinins (autohemagglutinins) in pri-
mary atypical pneumonias. Science 1943;97(2511):167.
[53] Ginsberg HS. Acute hemolytic anemia in primary atypical pneumonia associated
with high titer of cold agglutinins; report of a case. N Engl J Med 1946;234:
[54] Linz DH, Tolle SW, Elliot DL. Mycoplasma pneumoniae pneumonia. Experience at
a referral center. West J Med 1984;140(6):895–900.
[55] Schubothe H, Merz KP, Weber S, Dahm K, Schmitz W, Altmann HW. Akute auto-
immunhämolytische Anemie vom Kältenantikörpertype mit tödlichem Ausgang.
Acta Haematol (Basel) 1970;44(1):111–23.
[56] Dacie J. Auto-immune haemolytic anaemia (AIHA): cold-antibody syndromes
III: haemolytic anaemia following mycoplasma pneumonia. In: Dacie J, editor.
The Haemolytic Anaemias, Vol. 3. London: Churchill Livingstone; 1992.
p. 296–312.
[57] Jenkins WJ, Koster HG, Marsh WL, Carter RL. Infectious mononucleosis: an
unsuspected source of anti-i. Br J Haematol 1965;11:480–3.
[58] Wentworth P, Bate LR. Acute hemolytic anemia secondary to infectious mononu-
cleosis. Can Med Assoc J 1980;123(6):482–6.
[59] Luzuriaga K, Sullivan JL. Infectious mononucleosis. N Engl J Med 2010;362(21):
[60] Dacie J. Auto-immune haemolytic anaemia (AIHA): cold-antibody syndromes IV:
haemolytic anaemia following infectious mononucleosis and other viral infec-
tions. In: Dacie J, editor. The Haemolytic Anaemias, Vol. 3. London: Churchill Liv-
ingstone; 1992. p. 313–28.
[61] Berlin BS, Chandler R, Green D. Anti-“i” antibody and hemolytic anemia associat-
ed with spontaneous cytomegalovirus mononucleosis. Am J Clin Pathol
[62] Vanherweghem JL, de Jonghe M, Van Geel P. Acute hemolytic syndrome caused
by cold agglutinins: unusual manifestation of ornithosis. Nouv Presse Med
[63] Terada K, Tanaka H, Mori R, Kataoka N, Uchikawa M. Hemolytic anemia associat-
ed with cold agglutinin during chickenpox and a review of the literature. J
Pediatr Hematol Oncol 1998;20(2):149–51.
[64] Mori T, Yamada Y, Aisa Y, Uemura T, Ishida A, Ikeda Y, et al. Cold agglutinin dis-
ease associated with adenovirus infection after allogeneic bone marrow trans-
plantation. Bone Marrow Transplant 2005;36(3):263–4.
[65] Schoindre Y, Bollee G, Dumont MD, Lesavre P, Servais A. Cold agglutinin syn-
drome associated with a 2009 influenza A H1N1 infection. Am J Med
[66] Srinivasan N, Oswal A, Garg S, Nahar J, Gosmonova A, Nahar R. Cold agglutinin
induced hemolysis in a newly diagnosed systemic lupus erythematosus. Am J
Med Sci 2010;339(3):270–3.
[67] Azuma E, Nishihara H, Hanada M, Nagai M, Hiratake S, Komada Y, et al. Recurrent
cold hemagglutinin disease following allogeneic bone marrow transplantation
successfully treated with plasmapheresis, corticosteroid and cyclophosphamide.
Bone Marrow Transplant 1996;18(1):243–6.
[68] Batalias L, Trakakis E, Loghis C, Salabasis C, Simeonidis G, Karanikolopoulos P,
et al. Autoimmune hemolytic anemia caused by cold agglutinins in a young preg-
nant woman. J Matern Fetal Neonatal Med 2006;19(4):251–3.
[69] Dacie J. Treatment and prognosis of cold-antibody AIHA. In: Dacie J, editor. The
Haemolytic Anaemias, Vol. 3. London: Churchill Livingstone; 1992. p. 502–8.
[70] Bartholomew JR, Bell WR, Shirey RS. Cold agglutinin hemolytic anemia: manage-
ment with an environmental suit. Ann Intern Med 1987;106(2):243–4.
[71] Nydegger UE, Kazatchkine MD, Miescher PA. Immunopathologic and clinical fea-
tures of hemolytic anemia due to cold agglutinins. Semin Hematol 1991;28(1):
[72] Rosenfield RE, Jagathambal. Transfusion therapy for autoimmune hemolytic ane-
mia. Semin Hematol 1976;13(4):311–21.
[73] Zoppi M, Oppliger R, Althaus U, Nydegger U. Reduction of plasma cold agglutinin
titers by means of plasmapheresis to prepare a patient for coronary bypass sur-
gery. Infusionsther Transfusionsmed 1993;20(1–2):19–22.
[74] Lodi G, Resca D, Reverberi R. Fatal cold agglutinin-induced haemolytic anaemia:
a case report. J Med Case Reports 2010;4:252.
[75] Silberstein LE, Berkman EM, Schreiber AD. Cold hemagglutinin disease associat-
ed with IgG cold-reactive antibody. Ann Intern Med 1987;106(2):238–42.
[76] Worlledge SM, Brain MC, Cooper AC, Hobbs JR, Dacie J. Immmunosuppressive
drugs in the treatment of autoimmune haemolytic anaemia. Proc R Soc Med
[77] Schreiber AD, Herskovitz BS, Goldwein M. Low-titer cold-hemagglutinin disease.
Mechanism of hemolysis and response to corticosteroids. N Engl J Med
[78] Hippe E, Jensen KB, Olesen H, Lind K, Thomsen PE. Chlorambucil treatment of pa-
tients with cold agglutinin syndrome. Blood 1970;35(1):68–72.
[79] Hillen HF, Bakker SJ. Failure of interferon-alpha-2b therapy in chronic cold ag-
glutinin disease. Eur J Haematol 1994;53(4):242–3.
[80] O'Connor BM, Clifford JS, Lawrence WD, Logue GL. Alpha-interferon for severe
cold agglutinin disease. Ann Intern Med 1989;111(3):255–6.
[81] Rordorf R, Barth A, Nydegger U, Tobler A. Treatment of severe idiopathic cold-
agglutinin diseases using interferon-alpha 2b. Schweiz Med Wochenschr
[82] Berentsen S, Tjonnfjord GE, Shammas FV, Bergheim J, Hammerstrom J, Langholm
R, et al. No response to cladribine in five patients with chronic cold agglutinin
disease. Eur J Haematol 2000;65(1):88–90.
[83] Lee EJ, Kueck B. Rituxan in the treatment of cold agglutinin disease. Blood
[84] Cohen Y, Polliack A, Zelig O, Goldfarb A. Monotherapy with rituximab induces rapid
remission of recurrent cold agglutinin-mediated hemolytic anemia in a patient
with indolent lympho-plasmacytic lymphoma. Leuk Lymphoma 2001;42(6):
[85] Engelhardt M, Jakob A, Ruter B, Trepel M, Hirsch F, Lubbert M. Severe cold hemag-
glutinin disease (CHD) successfully treated with rituximab. Blood 2002;100(5):
[86] Berentsen S, Tjonnfjord GE, Brudevold R, Gjertsen BT, LangholmR, Lokkevik E, et al.
Favourable response to therapy withthe anti-CD20 monoclonal antibody rituximab
in primary chronic cold agglutinin disease. Br J Haematol 2001;115(1):79–83.
[87] Berentsen S, Ulvestad E, Gjertsen BT, Hjorth-Hansen H, Langholm R, Knutsen H,
et al. Rituximab for primary chronic cold agglutinin disease: a prospective
study of 37 courses of therapy in 27 patients. Blood 2004;103(8):2925–8.
[88] Schollkopf C, KjeldsenL, BjerrumOW, Mourits-AndersenHT, NielsenJL, Christensen
BE, et al. Rituximab in chronic cold agglutinin disease: a prospective study of 20 pa-
tients. Leuk Lymphoma 2006;47(2):253–60.
[89] Jacobs A. Cold agglutinin hemolysis responding to fludarabine therapy. Am J
Hematol 1996;53(4):279–80.
[90] Treon SP, Branagan AR, Ioakimidis L, Soumerai JD, Patterson CJ, Turnbull B, et al.
Long-term outcomes to fludarabine and rituximab in Waldenstrom macroglobu-
linemia. Blood 2009;113(16):3673–8.
[91] Czuczman MS, Koryzna A, Mohr A, Stewart C, Donohue K, Blumenson L, et al.
Rituximab in combination with fludarabine chemotherapy in low-grade or follic-
ular lymphoma. J Clin Oncol 2005;23(4):694–704.
[92] Berentsen S, Randen U, Vagan AM, Hjorth-Hansen H, Vik A, Dalgaard J, et al. High
response rate and durable remissions following fludarabine and rituximab com-
bination therapy for chronic cold agglutinin disease. Blood 2010;116(17):
[93] Stone MJ. Heating up cold agglutinins. Blood 2010;116(17):3119–20.
[94] Treon SP. How I treat Waldenstrom macroglobulinemia. Blood 2009;114(12):
[95] Carson KR, Beckwith LG, Mehta J. Successful treatment of IgM-mediated autoim-
mune hemolytic anemia with bortezomib. Blood 2010;115(4):915.
[96] Treon SP, Loakimidis L, Soumerai JD, Patterson CJ, Hunter ZR, Feiner A, et al. Primary
therapy of Waldenstrom's macroglobulinemia with bortezomib, dexamethasone
and rituximab: Results of the WMCTG clinical trial 05–180. ASCO Annual Meeting
2008: Abstract 8519. J Clin Oncol 2008;26(May 20 suppl).
114 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115
[97] Hillmen P, Young NS, Schubert J, Brodsky RA, Socie G, Muus P, et al. The comple-
ment inhibitor eculizumab in paroxysmal nocturnal hemoglobinuria. N Engl J
Med 2006;355(12):1233–43.
[98] Hill A, Rother RP, Arnold L, Kelly R, Cullen MJ, Richards SJ, et al. Eculizumab pre-
vents intravascular hemolysis in patients with paroxysmal nocturnal hemoglo-
binuria and unmasks low-level extravascular hemolysis occurring through C3
opsonization. Haematologica 2010;95(4):567–73.
[99] Roth A, HuttmannA, Rother RP, DuhrsenU, Philipp T. Long-termefficacy of the com-
plement inhibitor eculizumab in cold agglutinin disease. Blood 2009;113(16):
[100] Chu CS, Braun SR, Yarbro JW, Hayden MR. Corticosteroid treatment of hemolytic
anemia associated with Mycoplasma pneumoniae pneumonia. South Med J
[101] Tsuruta R, Kawamura Y, Inoue T, Kasaoka S, Sadamitsu D, Maekawa T. Corticoste-
roid therapy for hemolytic anemia and respiratory failure due to Mycoplasma
pneumoniae pneumonia. Intern Med 2002;41(3):229–32.
[102] Geurs F, Ritter K, Mast A, Van M V. Successful plasmapheresis in corticosteroid-
resistant hemolysis in infectious mononucleosis: role of autoantibodies against
triosephosphate isomerase. Acta Haematol 1992;88(2–3):142–6.
115 S. Berentsen, G.E. Tjønnfjord / Blood Reviews 26 (2012) 107–115