Cold Haemagglutinin Disease: Serum Haemolysins & Clinical Significance

Clin. Lab. Haem.
2000, 22, 337±344
R.J. SOKOL*,
D.J. BOOKER*,
R. STAMPS*,
R. WALEWSKA
Cold haemagglutinin disease: clinical
signi®cance of serum haemolysins
*National Blood Service, Trent Centre, Shef®eld, Department of Haematology,
Leicester Royal In®rmary, Leicester, UK
Summary
Two hundred and twenty-one patients with cold haemagglutinins of thermal amplitude ³ 30 °C (considered to be a reasonable indicator of clinical signi®cance) were
classi®ed by in vitro haemolysin activity into three groups. Group 1 contained 116
individuals in whom haemolysins were never detected; the 74 patients in Group 2 had
monophasic haemolysins alone; whereas both monophasic and biphasic haemolysins
were detected in the 31 Group 3 patients. There was a signi®cantly higher proportion
of patients in Groups 2 and 3 with haptoglobin levels < 0.1 g/l compared with Groups
1 and 2, respectively (P < 0.005 and P < 0.001). Direct antiglobulin test results
showed that the autoimmune response became more complex and IgM predominant
through Groups 1±3, resulting in an increasing ability to activate complement which
was re¯ected in increasing haemolysin activity and number of patients with active
haemolysis. The 31 patients in Group 3 were mostly elderly (median age 71 years at
presentation) and the majority had chronic cold haemagglutinin disease (CHAD),
several in association with lymphoid neoplasms or carcinomas; only four had acute
CHAD.
The natural history of idiopathic chronic CHAD was of mild, well compensated haemolysis, punctuated by severe acute episodes necessitating intensive therapy. The
condition often remained active for long periods and did not appear to affect natural
lifespan. In some cases, no treatment (or just warmth) was needed; in others continuous or intermittent prednisolone and/or chlorambucil were effective; yet others
required a greater variety and more intense therapy, or treatment of associated
conditions. Blood transfusion support was frequently required when haemolysis was
severe.
Keywords
Anaemia, haemolytic, autoimmune, cold agglutinin disease, serum haemolysins
Introduction
Cold red cell autoantibody syndromes are usually divided
into two groups, cold haemagglutinin disease (CHAD) and
paroxysmal cold haemoglobinuria (PCH), based on their
clinical and serological features, including different patterns of in vitro complement mediated haemolysis. The
IgM class autoantibodies causing CHAD usually bind to
red cells and activate complement to produce haemolysis
in vitro at the same temperature, and as such are referred
Accepted for publication 3 July 2000
Correspondence: Dr R.J. Sokol, National Blood Service, Trent Centre,
Longley Lane, Shef®eld S5 7JN, UK. Fax: 0114 2034910; E-mail:
[email protected]
Ó 2000 Blackwell Science Limited
to as monophasic haemolysins. On the other hand, the IgG
class Donath-Landsteiner autoantibodies, considered pathognomonic of PCH, are usually described as biphasic
haemolysins since they bind to red cells at low temperatures but activate complement to produce in vitro haemolysis at 37 °C (Sokol et al., 1999). However, whether a
haemolytic autoantibody behaves biphasically or monophasically depends on its thermal range; for example, true
IgG Donath-Landsteiner antibodies occasionally have a
suf®ciently high thermal amplitude to cause haemolysis
monophasically (Ries et al., 1971; Boccardi et al., 1977;
Wolach et al., 1981; Lindgren et al., 1985) and conversely, the cold IgM autoantibodies of CHAD can cause
biphasic haemolysis if they have a relatively low thermal
337
Serum haemolysins in CHAD
range but a high complement ®xing ability (Dacie 1992).
Such cases are more common than PCH and it is
important that the two conditions are not confused,
though this is not usually a problem if full and proper
investigations are carried out. Likewise, they must not be
confused with individuals having both CHAD and PCH;
such mixed cases are exceedingly rare, but do exist (Sokol
et al., 1999).
Nowadays, serology reference laboratories rarely test for
the presence of serum haemolysins in patients with
suspected CHAD. In order to assess the signi®cance of
haemolysin test results and to evaluate whether they give
worthwhile information on the likely clinical signi®cance
of cold autoantibodies, the present study examines a series
of patients with wide thermal range cold autoagglutinins
who were classi®ed according to the pattern of in vitro
haemolysin activity, and discusses the clinico-pathological
features of the 31 individuals in whom the serum
haemolysins showed biphasic properties.
Materials and methods
The records of all patients referred to the Trent Centre
Immunohaematology Department between 13 December
1984 and 12 January 1996 were reviewed; cases with
cold red cell autoagglutinins having a thermal amplitude
of at least 30 °C by an albumin technique (Garratty et al.,
1977) were examined and those with a complete set of
haemolysin tests, but where full investigation had excluded a primary or concomitant diagnosis of PCH, were
selected for detailed study.
The investigations performed at this Centre have
previously been reported in full (Sokol, Hewitt & Stamps
1981; Sokol et al., 1988a, b, 1990a, b; Sokol et al., 1995).
They included direct antiglobulin tests using monospeci®c
anti-IgG, -IgM, -IgA, -C3c and -C3d by the standard
agglutination method and employing anti-IgG, -IgM and
-IgA in an enzyme-linked assay (ELISA). Serum collected
and separated at 37 °C was tested for autoantibodies using
saline, enzyme and indirect antiglobulin techniques;
autoantibody speci®city and thermal amplitude were
determined by titrating at 20 °C and 30 °C against adult
and cord cells using saline, albumin and enzyme methods
(Sokol et al., 1995). In addition, the warm separated
serum was tested for monophasic haemolysins using
acidi®ed serum (pH 6.8) at 20 °C and 37 °C (Sokol et al.,
1995); biphasic haemolysins were detected by the indirect
Donath-Landsteiner procedure (Sokol et al., 1995). Serum
haptoglobin levels were measured by their haemoglobin
binding capacity (Ratcliff & Hardwicke 1964), our normal
range being 0.4±2.0 g/l.
A method of population proportions was used to examine
the effects of haemolysin pattern on the likely presence or
absence of active haemolysis, de®ned by haptoglobin levels
of < 0.1 g/l and ³ 0.4 g/l, respectively. The difference
between proportions was tested for signi®cance by the Chisquare procedure (with Yates correction) at 1 degree of
freedom, and the 99% con®dence intervals were calculated
(Mack 1967; Gardner & Altman 1986). Because multiple
statistical comparisons were being made, the signi®cance
level for individual tests was set at P < 0.01, which
required a Chi-square value of ³ 6.63 (Sokol et al., 1990b).
Results
Two hundred and twenty-one patients with cold autoagglutinins active at 30 °C (or above) were classi®ed into
three groups on the pattern of in vitro haemolysin activity
(Tables 1±3). There were 116 in whom haemolysins were
never detected (Group 1); 74 patients had haemolysins but
these never showed biphasic properties (Group 2); and in
31 patients, both monophasic and biphasic haemolysins
(demonstrated by a positive Donath-Landsteiner test) were
present (Group 3). Direct antiglobulin tests were positive
in all patients (except one in Group 1) with increased
amounts of red cell bound C3d; in addition, C3c was
detected in 28, 45 and 55% in Groups 1±3, respectively.
Increased amounts of red cell bound immunoglobulins
were detected by ELISA in 68, 85 and 93% of patients in
Groups 1±3, respectively. IgM predominated, either alone
or in combination with IgG and/or IgA, and was found,
respectively, in 80, 90 and 100% of patients in Groups
1±3 who had a positive ELISA result.
There were signi®cantly higher proportions of patients
with haptoglobin levels < 0.1 g/l in Group 2 compared
with Group 1 (0.09 vs. 0.29; 99% con®dence interval for
Table 1. Group 1: 116 patients in whom in vitro haemolysins
were never detected
Males/females
53
Age at testing (years)
19±99 (median 66)
Speci®city of cold agglutinin (n)
I
i
`not I, i or Pr'
Pr
I/HI
82
14
12
7
1
Haptoglobin (g/l) (n)
< 0.1
0.1±0.3
³ 0.4
9
10
97
63
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
338
Table 2. Group 2: 74 patients in whom
in vitro haemolysins were detected but did
not show biphasic reactions
Males/females
30
Age at testing (years)
16±95 (median 71)
Speci®city of cold agglutinin (n)
I
i
`not I, i or Pr'
Pr
I/HI
64
1
3
3
3
Monophasic haemolysin pattern (n)
pH 6.8 (20 °C)
pH 6.8 (20 °C + 37 °C)
62
12
Haptoglobin (g/l) (n)
< 0.1
0.1±0.3
³ 0.4
18
11
45
Table 3. Group 3: 31 patients in whom biphasic haemolysins
were demonstrated
Males/females
17
Age at testing (years)
19±90 (median 79)
Speci®city of cold agglutinin (n)
I
i
`not I, i or Pr'
14
29
1
1
Monophasic haemolysin pattern (n)
pH 6.8 (20 °C) + biphasic haemolysins 19
pH 6.8 (20 °C + 37 °C) +
12
biphasic haemolysins
Haptoglobin (g/l) (n)*
< 0.1
0.1±0.3
³ 0.4
19
9
2
*Out of 30 tested.
difference 0.04±0.37; v2 10.42, P < 0.005) and in Group
3 compared with Group 2 (0.29 vs. 0.91, 99% con®dence
interval for difference 0.48±0.92; v2 22.04, P < 0.001).
The 31 patients in Group 3 were numbered according to
age at presentation (Table 4). Severity of disease was a
subjective attempt, after reviewing all available records, to
indicate how the condition affected the individual's dayto-day life ± `mild' representing very few or no problems,
`moderate' meaning that lifestyle had to be modi®ed to cope
with the disease, whereas `severe' meant serious impairment and the patient required frequent hospital treatment.
Minimum haemoglobin concentrations varied from 2.8/
12.8 g/dl, with 21 patients having values below 8.5 g/dl.
These data relate to a speci®c time during the course of a
mainly chronic condition (which sometimes lasted for more
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
44
than 20 years) when, for various reasons full investigations were carried out and biphasic haemolysins detected.
Discussion
The whole patient group was selected on the basis that
activity at ³ 30 °C (in albumin) was a reasonable
indicator of a cold autoantibody being clinically signi®cant
(Garratty et al., 1977; Garratty 1994). Our ®ndings
showed that this was not always the case; and previous
studies similarly demonstrated that although in vitro
assessment of autoagglutination characteristics tended to
correlate with the rate of haemolysis in vivo (Rosse &
Adams 1980), there were many exceptions (Abramson
1977; Dacie 1992). Patients with low titre, high thermal
amplitude cold agglutinins could be associated with severe
1 haemolysis (Schreiber et al., 1977) and conversely there
were cases with high titre, high thermal amplitude cold
agglutinins where there was never any evidence of
haemolysis (as in many Group 1 patients), even over a
period of years (Sniecinski et al., 1988).
Our results showed that the pattern of haemolysin
activity (Tables 1±3), taken in conjunction with thermal
amplitude studies, was a better indicator of the severity of
haemolysis than titration results alone. Less than 10% of
Group 1 patients (serum haemolysins not detected) had
haptoglobin levels of < 0.1 g/l, compared with approximately 30% of Group 2 patients (haemolysins present) and
with over 90% of Group 3 patients (where the haemolysins showed biphasic properties). Although a normal
haptoglobin level does not necessarily exclude haemolysis
(Petz & Garratty 1980), it was felt that equating
haptoglobin levels of < 0.1 and ³ 0.4 g/l with the
presence and absence of active haemolysis was a reason-
339
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
R. J. Sokol et al.
M
M
M
M
F
F
19
55
58
60
60
61
62
64
65
65
66
66
67
68
69
71
75
78
79
79
79
79
79
79
80
NK
83
84
86
90
90
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
2 months
> 28 years
> 14 years
NK
> 5 years
> 11 years
diabetes mellitus
CLL
NHL
Ca lung
NK
Occasional blood transfusions
Blood transfusion
Nil
Blood transfusion
Prednisolone, cyclophosphamide, blood transfusion
Prednisolone, cyclophosphamide, blood transfusion
Chlorambucil (short course)
Warmth, monthly blood transfusion, chlorambucil, prednisolone,
splenectomy after 2 years (responded well)
Prednisolone, cyclophosphamide, azathioprine, blood transfusion
Antibiotics; oxygen; blood transfusion
Blood transfusion, chlorambucil
Nil
Prednisolone; blood transfusion
Chlorambucil (short course): blood transfusions
iv antibiotics, prednisolone, blood transfusion
Nil
Blood transfusion
Warmth, blood transfusions on three occasions
Warmth
Prednisolone, blood transfusion
Blood transfusion
Chlorambucil (short course ± no effect), blood transfusions near end of life
Avoidance of cold; prednisolone
NK
For AML only; nil for CHAD
Blood transfusion, prednisolone, chlorambucil, azathioprine,
cyclophosphamide, plasma exchange, splenectomy
Chlorambucil, prednisolone, blood transfusion
Treatment course
Blood transfusion, chlorambucil (no response to prednisolone)
Nil
rheumatoid arthritis Blood transfusion
mild/moderate
severe
Pneumonia
moderate/severe
NHL
mild
CLL
severe
Pemphigoid
moderate
severe
Pneumonia
mild/moderate
moderate/severe
mild/moderate
±
mild ± becoming
moderate with time
NK
NK
moderate
mild
NHL
died 2 weeks
after presentation
> 4 years
mild
> 3 years
mild
> 2 years
mild/moderate
> 12 years
mild/moderate
> 7 years
mild/moderate
> 13 years
severe
> 4 years
< 1 month
> 10 years
> 11 years
< 2 months
> 4 years
< 4/12
> 11 years
> 6 years
> 3 years
> 3 years
thyrotoxicosis
NK
AML
CLL
CRF
Associated
conditions
severe, gradually got
worse with time
severe
carcinomatosis
mild
moderate
mild±severe episodes
NK
mild
moderate/severe
Severity
NK
NK
chronic > 3 years
NK
died 6 days
after presentation
NK
died 2 months
severe
after presentation
chronic died 10 months
mild/moderate
after presentation
chronic died 6 months
severe
after presentation
chronic
acute
chronic
chronic
acute
chronic
acute
chronic
chronic
chronic
chronic
chronic
chronic
chronic
chronic
chronic
chronic
NK
chronic > 20 years
acute
chronic
chronic
NK
chronic
chronic
Acute/ Length of
chronic time affected
Serum haemolysins in CHAD
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
NK, not known (data not available); CRF, chronic renal failure; AML, acute myeloid leukaemia; CLL, chronic lymphocytic leukaemia; NHL, non-Hodgkin's lymphoma; Ca, carcinoma.
M
M
M
F
F
M
F
M
M
M
M
M
F
F
F
F
M
F
F
M
F
M
F
M
F
Sex
Patient Age at prenumber sentation
Table 4. Clinical features of Group 3 patients with biphasic haemolysins
340
able assumption for the purposes of statistical analysis; if
anything it made the tests more stringent by excluding
patients with marginally reduced levels (0.1±0.3 g/l).
The most common autoantibody speci®city in all patient
Groups was anti-I, but whether the apparent increasing
prevalence from Groups 1±3 (71%, 86% and 94%) was
signi®cant cannot be assessed as the number of patients in
Group 3 was too small. In other reports, the antibodies also
most commonly showed anti-I speci®city; examples of anti-i
were proportionally more frequent in CHAD occurring in
children and secondary to lymphoproliferative disorders
2 (Pruzanski & Shumak 1977; Crisp & Pruzanski 1982;
Nydegger et al., 1991; Englefriet et al., 1992; Garratty
1994). The ability to cause in vitro haemolysis was greater
with anti-I than with anti-i (Kirsch®nk et al., 1992).
As well as predicting the likelihood of active haemolysis,
classi®cation according to the pattern of in vitro haemolysin activity provided an insight into the autoimmune
response, with more complex and stronger reactions from
Group 1 through to Group 3 patients. Acid (monophasic)
haemolysins (pH 6.8) active at 20 °C were found in all
patients in Groups 2 and 3 (Tables 2 and 3); but titres of
³ 1 : 16 were found in 47% of Group 2 patients and in
94% of Group 3 patients (data not shown). Furthermore,
haemolysin activity at 37 °C (at pH 6.8) was detected in
16% of Group 2 patients compared with 39% in Group 3.
The biphasic haemolysins (the criterion for inclusion in
Group 3) were usually weakly reacting.
The increasingly complex pattern of direct antiglobulin
test results from Groups 1±3 showed that the immune
response in patients with cold autoagglutinins was more
varied (and complex) than commonly thought; as the
immune response became IgM predominant, the cold
autoantibodies had an increasing ability to activate
complement through Groups 1±3, which was re¯ected in
increasing haemolysin activity and proportion of patients
with active haemolysis, Group 3 patients, with biphasic
haemolysins, representing the severest form of CHAD. The
®ndings of increased amounts of IgG and IgA in many
patients were thought to re¯ect a more generalized
autoimmune response (Sokol et al., 1990a), multiple
immunoglobulin coating of red cells per se being known
to have an important effect on the degree of haemolysis
(Sokol et al., 1990a). These ®ndings were in contrast to
those of previous studies in patients with CHAD which
stressed the presence of large amounts of complement on
the red cells with only rare cases additionally showing
small amounts of immunoglobulin (Garratty et al., 1977;
3 Pruzanski & Shumak 1977; Schreiber et al., 1977; Crisp &
Pruzanski 1982; Nydegger et al., 1991; Garratty 1994;
Gilliland 1996; Berentsen et al., 1997).
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
Patients in Group 3 were predominantly elderly and,
apart from one individual aged 19 years, all the others
were 55 or over, with a median age of 71 years at
presentation (Table 4). Similarly, in other published series,
the majority of patients with CHAD (irrespective of their
haemolysin status) were over 60 (Crisp & Pruzanski 1982;
Hadnagy1993; Nydegger et al., 1991; Dacie 1992;
Berentsen et al., 1997); and one report noted that only
8% of patients were under 40 years (Crisp & Pruzanski
1982). CHAD does occur in children, albeit rarely, and our
own studies (Sokol et al., 1992) showed that it was at least
as common as warm type autoimmune haemolytic
anaemia. The slight male bias (ratio 1 : 0.8) was of
interest, and an excess of males with CHAD was also noted
previously (Nydegger et al., 1991).
The majority of patients in this and in other studies;
(Rosse & Adams 1980; Crisp & Pruzanski 1982; Nydegger
et al., 1991; Berentsen et al., 1997) had the chronic
condition; we could not classify a few cases as death
occurred within a short time of diagnosis. Only four
patients had acute CHAD (Table 4). In one (number 1), this
was a brief episode at the time when he was rejecting a
renal transplant and was possibly secondary to a viral
infection; haemolysis was mild. In two of the three others
(numbers 16 and 21), severe haemolysis was associated
with pneumonia (though not due to Mycoplasma pneumoniae (Nydegger et al., 1991; Dacie 1992; Rordorf et al.,
1994)); in the remaining patient (number 19 ± who also
had pemphigoid), there was no obvious precipitating
incident; all three individuals were very ill, but eventually made good recoveries without recurrence.
In view of the known associations between autoimmune haemolytic anaemia and malignancy (Sokol et al.,
1992; Sokol et al., 1994) and between malignancy and
age, it was not surprising that several of the patients were
also suffering from lymphoid neoplasms or carcinoma
(Table 4). The relationship between cold autoagglutinins
and lymphoproliferative disorders is complex. Some
authorities regard idiopathic CHAD as nonmalignant
(although with a slight but de®nite increase in marrow
lymphoid cells) (Nydegger et al., 1991); others feel it is a
premalignant or low grade lymphoproliferative disease
4 (Berentsen et al., 1997); while yet others think of
conditions with persistent cold agglutinins as a spectrum
that varies from benign autoimmune chronic CHAD to
malignant lymphoma (Crisp & Pruzanski 1982).
The natural history of our Group 3 patients with
idiopathic chronic CHAD was commonly of mild, well
compensated haemolysis that required no treatment for
long periods, punctuated with acute episodes of severe
haemolysis (67% of patients had haemoglobin levels of less
341
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
R. J. Sokol et al.
Serum haemolysins in CHAD
than 8.5 g/dl and values below 6.0 g/dl were not uncommon) necessitating intensive therapy. This was very much
an individual picture and in some patients the treatment
requirements became more frequent or greater as time
went by; in others they did not. The severity of disease
column (Table 4) attempts to show this variation. The
condition could remain active for long periods (Table 4),
over 28 years in patient 2, which taken with the high age
at presentation, suggests that CHAD does not seriously
affect the natural lifespan. Those patients who died within
a year of presentation either had disseminated malignancy
(numbers 8 and 28), or congestive cardiac failure (number
29), or had already reached their 90th birthday (numbers
30 and 31). These ®ndings are in keeping with those in
the literature, where most patients with idiopathic CHAD
had chronic low grade haemolysis, with occasional acute
exacerbations and with very much an individual course to
the disease, the patients usually remaining in good general
condition (Petz & Garratty 1980; Rosse & Adams 1980;
Nydegger et al., 1991; Dacie 1992; Hillen & Bakker 1994;
Berentsen et al., 1997). In some patients, there was a
progressive worsening of symptoms over the years
throughout autumn and winter (Hillen & Bakker 1994).
Survival times of greater than 5 years were quoted for
idiopathic CHAD and for CHAD associated with chronic
lymphocytic leukaemia, but of less than 3 years for
patients where the disease was secondary to other
lymphoproliferative disorders (Crisp & Pruzanski 1982).
Not surprisingly, in view of their high median age, at least
18 of our patients are now known to be deceased, but as
far as could be seen from the available hospital case notes,
not one died of intractable haemolysis. Fulminating
haemolysis causing death is rare in CHAD and only a
few individual cases have been reported (Seldon et al.,
1980; Rousey & Smith 1990; Mandigers et al., 1996).
Treatment varied (Table 4) and was tailored to an
individual; in some cases none was needed, or avoidance of
cold and/or keeping warm were suf®cient. In other
patients, either continuous or intermittent treatment with
prednisolone and/or chlorambucil were used with good
effect. Yet others (e.g. number 6) required a greater variety
and more intense therapy. Antibiotics (and oxygen in one
case) were important in the patients with pneumonia.
Blood transfusion support was frequently required at times
when haemolysis was severe. Splenectomy was carried out
in two individuals and was particularly successful in one
(number 14). The treatment of associated conditions
(Table 4), for example lymphoproliferative malignancies,
was often also effective against the CHAD.
These treatment modalities were similar to those in the
published literature, which also recognized that many
cases required no active intervention; otherwise, warmth
(or avoidance of cold) supported by blood transfusion on
occasions formed the basis of management (Ulvestad
1998; Petz & Garratty 1980; Rosse & Adams 1980; Rosse
5 1990; Nydegger et al., 1991; Englefriet et al., 1992;
Rordorf et al., 1994; Berentsen et al., 1997). Chlorambucil
was probably ®rst choice if drug treatment was needed
(Ulvestad 1998; Chaplin 1982; Englefriet et al., 1992;
Rordorf et al., 1994; Berentsen et al., 1997) (though there
was a suggestion in two cases that it might have induced
the development of lymphoma (Chaplin 1982)). Cyclophosphamide, vincristine, azathioprine, ¯udarabine and
monoclonal antibody directed against CD20, with or
without plasma exchange and steroids, were also prescribed in various combinations in more severe (or
desperate) cases, with varying degrees of success (Seldon
et al., 1980; Rousey & Smith 1990; Nydegger et al., 1991;
Jacobs 1996; Mandigers et al., 1996; Berentsen et al.,
1997; Lee & Kueck 1998). Although steroids (prednisolone) were generally disappointing, exceptions have
been reported, particularly in the subgroup of CHAD
characterized by low titre, high thermal amplitude cold
agglutinins (Lahav et al., 1989; Schreiber et al., 1977;
Englefriet et al., 1992; Gilliland 1996); also, the synthetic
steroid danazol was successful in 4 patients (Geffray &
Najman 1992). There were several reports of the use of
a-interferon-2b, some successful (O'Connor et al., 1989;
Fest et al., 1994; Rordorf et al., 1994), others not (Hillen &
Bakker 1994). Similarly, splenectomy was thought to be
of limited value, though exceptions (as our patient 14
exempli®es) were reported (Rordorf et al., 1994; Gilliland
1996). Some authorities have concluded that there is no
really satisfactory treatment for CHAD, a view with which
we would concur, and it is fortunate that active intervention is not often needed (Rosse 1990; Petz & Garratty
1980).
Until the present work, CHAD had never been considered from the perspective of the autoantibodies' ability to
cause in vitro haemolysis. Our study showed that haemolysin test results provided valuable information on the
complexity of the cold autoimmune response, were
signi®cantly related to the likelihood of an autoantibody
causing haemolysis in vivo and identi®ed a group of
patients, characterized by the presence of biphasic haemolysins, with the severest form of CHAD.
Acknowledgements
We thank our clinical colleagues for allowing access to
case records of patients under their care and Mrs Carol
Mitchell for secretarial assistance.
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
342
References
Abramson N. (1977) Cold agglutinins. New England Journal of
Medicine 297, 727±728.
Berentsen S., Bù K., Shammas F.V., Myking A.O. & Ulvestad E.
(1997) Chronic cold agglutinin disease of the `idiopathic' type
is a premalignant or low-grade malignant lymphoproliferative
disease. APMIS 105, 354±362.
Boccardi V., D'Annibali S., Di Natale G., Girelli G. & Summonti D.
(1977) Mycoplasma pneumoniae infection complicated by
paroxysmal cold hemoglobinuria with anti-P speci®city of
biphasic hemolysin. Blut 34, 211±214.
Chaplin H. (1982) Lymphoma in primary cold hemagglutinin
disease treated with chlorambucil. Archives of Internal Medicine
142, 2119±2123.
Crisp D. & Pruzanski W. (1982) B-cell neoplasms with homogeneous cold-reacting antibodies (cold agglutinins). American
Journal of Medicine 72, 915±922.
Dacie J. (1992) The Haemolytic Anaemias, Vol. 3, the AutoImmune Haemolytic Anaemias. 3rd edn. Churchill Livingstone. Edinburgh.
Englefriet C.P., von Overbeeke M.A.M., dem Borne A.E.G. & Kr.
(1992) Autoimmune hemolytic anemia. Seminars in Hematology 29, 3±12.
Fest T., de WazieÁres B., Lamy B., Maskani M., Vuitton D. &
Dupond J.L. (1994) Successful response to alpha-interferon 2b
in a refractory IgM autoagglutinin-mediated hemolytic anemia. Annals of Hematology 69, 147±149.
Gardner M.J. & Altman D.G. (1986) Con®dence intervals rather
than P values: estimation rather than hypothesis testing.
British Medical Journal 292, 746±750.
Garratty G. (1994) Autoimmune hemolytic anemia. In: Immunobiology of Transfusion Medicine. (Ed. Garratty G.). Marcel
Dekker, New York, 493±521.
Garratty G., Petz L.D. & Hoops J.K. (1977) The correlation of cold
agglutinin titrations in saline and albumin with haemolytic
anaemia. British Journal of Haematology 35, 587±595.
Geffray E. & Najman A. (1992) Ef®cacite du danazol dans
l'aneÂmie heÂmolytique auto-immune avec agglutinines froides:
4 observations. Presse Medicale 21, 1472±1475.
Gilliland B.C. (1996) Autoimmune hemolytic anemia. In: Principles of Transfusion Medicine, 2nd edn. (Ed. Rossi E.C., Simon
T.L., Moss G.S. & Gould S.A.) Williams & Wilkins, Baltimore,
101±120.
Hadnagy C. (1993) Agewise distribution of idiopathic cold
agglutinin disease. Zeitschrift Fur Gerontologie 26, 199±201.
Hillen H.F.P. & Bakker S.J.L. (1994) Failure of interferon-a-2b
therapy in chronic cold agglutinin disease. European Journal of
Haematology 53, 242±243.
Jacobs A. (1996) Cold agglutinin hemolysis responding to
¯udarabine therapy. American Journal of Hematology 53,
279±280.
Kirsch®nk M., Fritze H. & Roelcke D. (1992) Complement activation by cold agglutinins. Vox Sanguinis 63, 220±226.
Lahav M., Rosenberg I. & Wysenbeek A.J. (1989) Steroidresponsive idiopathic cold agglutinin disease: a case report.
Acta Haematologica 81, 166±168.
Lee E.J. & Kueck B. (1998) Rituxan in the treatment of cold
agglutinin disease. Blood 92, 3490±3491.
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
Lindgren S., Zimmerman S., Gibbs F. & Garratty G. (1985) An
unusual Donath-Landsteiner antibody detectable at 37°C by
the antiglobulin test. Transfusion 25, 142±144.
Mack C. (1967) Essentials for Statistics for Scientists and Technologists. Plenum Press. New York.
Mandigers C.M.P.W., Keuning J.J. & Booij A.C. (1996) A patient
with fatal cold haemagglutinins. Netherlands Journal of Medicine
49, 209±211.
Nydegger U.E., Kazatchkine M.D. & Miescher P.A. (1991)
Immunopathologic and clinical features of hemolytic anemia
due to cold agglutinins. Seminars in Hematology 28, 66±77.
O'Connor B.M., Clifford J.S., Lawrence W.D. & Logue G.L. (1989)
Alpha-interferon for severe cold agglutinin disease. Annals of
Internal Medicine 111, 255±256.
Petz L.D. & Garratty G. (1980)Acquired Immune Hemolytic Anemias. Churchill Livingstone. New York.
Pruzanski W. & Shumak K.H. (1977) Biologic activity of coldreacting autoantibodies (®rst of two parts). New England Journal
of Medicine 297, 538±542.
Ratcliff A.P. & Hardwicke J. (1964) Estimation of serum haemoglobin-binding capacity (haptoglobin) on Sephadex G.100.
Journal of Clinical Pathology 17, 676±679.
Ries C.A., Garratty G., Petz L.D. & Fudenberg H.H. (1971) Paroxysmal cold hemoglobinuria: report of a case with an exceptionally high thermal range Donath-Landsteiner antibody.
Blood 38, 491±499.
Rordorf R., Barth A., Nydegger U. & Tobler A. (1994) Behandlung einer schweren idiopathischen KaÈlteagglutinin-krankheit
mit Interferon-a2b. Schweizerische Medizinische Wochenschrift
124, 56±61.
Rosse W.F. (1990) Clinical Immunohematology: Basic Concepts
and Clinical Applications. Blackwell Scienti®c Publications.
Boston.
Rosse W.F. & Adams J.P. (1980) The variability of hemolysis in
the cold agglutinnin syndrome. Blood 56, 409±416.
Rousey S.R. & Smith R.E. (1990) A fatal case of low titer anti-PR
cold agglutinin disease. American Journal of Hematology 35,
286±287.
Schreiber A.D., Herskovitz B.S. & Goldwein M. (1977) Low-titer
cold-hemagglutinin disease: mechanism of hemolysis and
response to corticosteroids. New England Journal of Medicine
296, 1490±1494.
Seldon M., Isbister J.P., Raik E. & Biggs J.C. (1980) A fatal case of
cold autoimmune hemolytic anemia. American Journal of Clinical Pathology 73, 716±717.
Sniecinski I., Margolin K., Shulman I., Oien L., Meyer E. & Branch
D.R. (1988) High-titer, high-thermal-amplitude cold autoagglutinin not associated with hemolytic anemia. Vox Sanguinis
55, 26±29.
Sokol R.J., Booker D.J. & Stamps R. (1992) The Pathology of
autoimmune haemolytic anaemia. Journal of Clinical Pathology
45, 1047±1052.
Sokol R.J., Booker D.J. & Stamps R. (1994) Erythrocyte autoantibodies, autoimmune haemolysis, and carcinoma. Journal of
Clinical Pathology 47, 340±343.
Sokol R.J., Booker D.J. & Stamps R. (1995) Investigation of
patients with autoimmune haemolytic anaemia and provision
of blood for transfusion. Journal of Clinical Pathology 48,
602±610.
343
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
R. J. Sokol et al.
Serum haemolysins in CHAD
Sokol R.J., Booker D.J. & Stamps R. (1999) Paroxysmal cold
haemoglobinuria: a clinico-pathological study of patients with
a positive Donath-Landsteiner test. Hematology 4, 137±164.
Sokol R.J., Hewitt S., Booker D.J. & Bailey A. (1990a) Red cell
autoantibodies, multiple immunoglobulin classes, and autoimmune hemolysis. Transfusion 30, 714±717.
Sokol R.J., Hewitt S., Booker D.J. & Bailey A. (1990b) Erythrocyte
autoantibodies, subclasses of IgG and autoimmune haemolysis.
Autoimmunity 6, 99±104.
Sokol R.J., Hewitt S., Booker D.J. & Morris B.M. (1988a) Patients
with red cell autoantibodies: selection of blood for transfusion.
Clinical and Laboratory Haematology 10, 257±264.
Sokol R.J., Hewitt S., Booker D.J., Stamps R. & Booth J.R. (1988b)
An enzyme-linked direct antiglobulin test for assessing erythrocyte bound immunoglobulins. Journal of Immunological
Methods 106, 31±35.
Sokol R.J., Hewitt S. & Stamps B.K. (1981) Autoimmune haemolysis: an 18-year study of 865 cases referred to a regional
transfusion centre. British Medical Journal 282, 2023±2027.
Ulvestad E. (1998) Paradoxical haemolysis in a patient with cold
agglutinin disease. European Journal of Haematology 60, 93±100.
Wolach B., Heddle N., Barr R.D., Zipursky A., Pai K.R.M. &
Blajchman M.A. (1981) Transient Donath-Landsteiner haemolytic anaemia. British Journal of Haematology 48, 425±434.
Ó 2000 Blackwell Science Ltd., Clin. Lab. Haem., 22, 337±344
13652257, 2000, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1046/j.1365-2257.2000.00320.x by Universidad De Chile, Wiley Online Library on [25/05/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
344