Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.

com

Postgrad Med J (1990) 66, 356 362 The Fellowship of Postgraduate Medicine, 1990

Antiphospholipid antibodies and Mycoplasma pneumoniae

infection
Neil Snowden, Philip B. Wilson, Maurice Longsonl* and Richard S.H.
Pumphrey
Regional Immunology Service, St Mary's Hospital Manchester, M13 OJH and The Virology Unit, Booth
Hall Children's Hospital, Manchester, UK.

Summary: Anticardiolipin antibody levels were measured in 57 patients with Mycoplasma

pneumoniae infection and 21 patients with other infections. Significantly more patients in the mycoplamsa
group had increased IgM and IgG anticardiolipin. Within the mycoplasma group significantly higher titres
were found in patients with severe infection (assessed by need for hospital admission) and in patients with
cold agglutinins. A tendency for particularly high titres to occur in patients with extra-pulmonary
complications was identified.

Introduction
Antibodies to negatively charged phospholipids, Immunology Service, large numbers of serum
such as cardiolipin, are associated with non- samples from normal controls and patients were
vasculitic thrombosis, recurrent spontaneous analysed. In addition to the expected positive
abortion, thrombocytopenia and neurological results from patients with SLE and low titre
disease.'2 This association was first noted in positives from other conditions, it was noted that in
systemic lupus erythematosus (SLE) where anti- a few patients, markedly increased titres of aPL
phospholipid antibodies (aPL) are very closely occurred unexpectedly. Particularly high titres
linked with3 (but are not identical to4) the so-called were noted in the serum of one patient with
'lupus anticoagulant'. Recently, patients have been Mycoplasma pneumoniae infection. This was inter-
described with high titre aPL, thrombosis and esting in view of the recognized extra pulmonary
recurrent abortion but with no evidence of connec- complications of mycoplasma infection, such as
tive tissue disease or other immunopathology. haemolysis, neurological disease'3 and stroke,'4"5
These patients may constitute a 'primary antiphos- some of which may have an immunological basis.
pholipid antibody syndrome'.56 Low titre aPL We therefore examined the hypothesis that
have been demonstrated in a number of other patients with Mycoplasma pneumoniae infection
conditions: acute infections (infectious mononucle- were particularly likely to develop high titres of
osis,7 human immunodeficiency virus infection8, aPL. A retrospective analysis of sera from patients
malignancy, ischaemic heart disease9 and in the with known M. pneumoniae infection was under-
healthy elderly."' The significance of this is unclear; taken. We attempted to correlate aPL levels with
in general it is felt to be representative of the low severity of illness and complications.
grade autoimmune responses of little functional We also had the opportunity to study serum
significance which accompany tissue damage and from one patient with an ischaemic stroke in
ageing" (in contrast with the aPL responses in SLE association with M. pneumoniae infection (not
and the primary aPL syndrome where there is some from the retrospective series).
evidence that these antibodies play a direct role in
the pathogenesis of the associated complications'2).
During development of an enzyme-linked immu-
nosorbent assay (ELISA) for aPL (using cardio- Materials and methods
lipin as antigen) at the North West Regional
Patients

Correspondence: N. Snowden M.R.C.P. Details of the patient with stroke in association

*Present address: Faculty of Medicine and Health with Mycoplasma pneumoniae infection have been
Sciences, Al Ain, United Arab Emirates. previously reported.'4 Briefly, a 31 year old
Accepted: 4 January 1990 previously fit male presented with a 3-week history
 Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

ANTIPHOSPHOLIPID ANTIBODIES AND M. PNEUMONIAE INFECTION 357

of malaise, dyspnoea and cough and a 1-day Anticardiolipin ELISA

history of drowsiness and left sided weakness.
Chest X-ray revealed left sided pneumonic Antibodies to cardiolipin were detected by an
changes, a computed tomographic (CT) scan of the isotype specific ELISA, after the method of Harris
head showed a right cerebral hemisphere infarct and co-workers.'6"17
and serological testing revealed a four-fold increase Cardiolipin (Sigma) was prepared to a concen-
in complement fixation titres to M. pneumoniae. No tration of 50 fig/ml in ethanol; 100 pl of this
source for embolism or other risk factor for stroke solution was placed in each well of a 96 well
was identified. polystyrene microtitre plate. The ethanol was then
Serum samples were obtained from all patients evaporated by overnight storage at 4°C. The plates
with serological evidence of M. pneumoniae infec- were blocked for 60 minutes using 10% newborn
tion over a 2-year period (January 1986-January calf serum (Flow) in phosphate buffered saline
1988) detected at the North Manchester Regional (NCS/PBS). The blocking solution was then
Virus Laboratory. Names of patients were obtain- shaken from the plates. The serum samples were
ed from the laboratory weekly returns to the diluted 1/10 in NCS/PBS and 100 1l were added to
Communicable Disease Surveillance Centre. All the wells in duplicate. One row (8 wells) was not
sera had been stored at - 40°C and aPL levels were exposed to serum: NCS/PBS only was added to
measured within 48 hours of thawing. Where these wells. Each plate incorporated serial doubling
possible, paired serum samples were retrieved (one dilutions (from 1/10 to 1/640) of a standard serum
from the acute phase of the illness and the other from a patient with SLE and high titres of IgM and
taken 7-14 days later), but in most cases only the IgG anticardiolipin (aliquotted and stored at
second (convalescent) sample was available. Sero- - 40°C). This was used to derive a standard curve
logical evidence of M. pneumoniae infection was for each plate. Each plate also incorporated
defined as a four-fold increase in complement another known positive serum as an interplate
fixation (CF) titre between the first and second quality control. Serum samples from 19 normal
samples, or a single titre greater than or equal to controls (mean age 23; range 21-58) were also
1/320 in the context of an illness consistent with M. tested in order to provide further validation for the
pneumoniae infection. Sera were obtained from 57 normal ranges used. The plates were incubated for
patients in all (32 female, 25 male. Mean age 23. 60 minutes at room temperature and then washed
Range 21- 58; 19 patients aged 11 or under) with three times in PBS. One hundred 1tl of peroxidase
paired samples from 17. The CF assay was per- conjugated rabbit antibody to human IgG or IgM
formed according to a standard protocol using a (Dako) diluted 1/500 in NCS/PBS was added to
five -0.2 ml volume test. each well.
Sera were also obtained from 21 patients (mean Following a further 1 hour incubation the plates
age 14; range 1-46) with other infections present- were washed three times in PBS and 200 gl of
ing to the laboratory over the same time period I mM-ABTS (Sigma A-1888) containing 0.1 mM-
(4 hepatitis A, 2 Chlamydia psittaci, 4 parvovirus, hydrogen peroxide in citrate phosphate buffer
1 Herpes simplex encephalitis, 1 Coxiella burnettii, (pH 5.4) was added to each well. The reactions
1 Varicella zoster, 1 cytomegalovirus, 6 rubella and were then allowed to develop until the optical
1 influenza B. Infection was defined in terms of density of the 1/10 dilution of the standard serum
four-fold rise in antibody titre, organism-specific read approximately 1. The plates were then read
IgM or virus isolation) using a TitertekO Multiskan plate reader. The
A letter requesting further clinical details was standard curve (optical density versus dilution) was
sent to each of the consultants and general practi- plotted and optical densities from the test samples
tioners supervising the cases of mycoplasma infec- were read relative to this curve and expressed in
tion. Information was sought on: (1) the presenting arbitrary units derived from the mean and 3
illness (e.g. pneumonia, upper respiratory tract standard deviations of a large number of healthy
infection, (2) the severity of illness (e.g. treated at controls. For IgG anticardiolipin normal was
home, hospitalized, died), (3) acute extrapulmon- defined as A 10 U and for IgM anticardiolipin,
ary complications (e.g. haemolysis, neurological < 15 U. The use of a standard curve reduces
disease, skin rashes), (4) any illness (with particular interplate error due to variables such as time and
reference to vascular and neurological disease and temperature. Plates yielding more than 10% varia-
complication of pregnancy) occurring in the 3 tion from previous laboratory values for the
months following infection, (5) the presence or quality control sample were discarded. Samples
absence of cold agglutinins. These details were with anticardiolipin levels outside the most sen-
either recorded on a standard proforma by the sitive zone of the standard curve were retested at
clinician or extracted from the case notes by one of 1/100 dilution. The 19 controls gave means (±3
the investigators, in each case without knowledge standard deviation), for IgM and IgG isotypes, of
of the anticardiolipin titres. 3.3 ( ± 10.2) U and 2.8 ( ± 5.7) U respectively.
 Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

358 N. SNOWDEN et al.

The assay described was developed, and the units 60 _

of measurement defined, before the report by A

Harris et al. 7 of the first attempt at inter-labor-

atory standardization of the anticardiolipin assay.
Subsequent calibration of our assay against 5 50 _-
standard sera obtained from St Thomas' Hospital, A

London suggests that, for IgM, 1 'Manchester unit'

is approximately equalivent to 0.6 'Harris units' 40 _
and for IgG, 1 'Manchester unit' is approximately
equivalent to 2.5 'Harris units'. Cl
IgM rheumatoid factor was detected by the
30 _
A

agglutination of dyed gel particles coated with -j

At

rabbit IgG (RAPA, Sera-Diagnostics). Other u

A
At
autoantibodies were detected by indirect A
A

immunofluorescence using as substrate (a) com- 20 _

A
AAAA
posite tissue blocks of rat stomach, liver and kidney A
A A

and (b) culture Hep 2 cells. Samples were screened A_*- A A AA

at a 1/20 dilution in PBS and diluted as appropri- AA

AA AA

ate. 10 _ ........ A
A-
A

Statistics
u.
4!
Comparison between test and control group IgM aCL IgG aCL IgM aCL IgG aCL
medians was made using the Mann Whitney U test. Controls M. pneumoniae
Comparisons between test and control groups and
the laboratory normal range were made by dividing Figure 1 Levels of IgM and IgG anticardiolipin (aCL) in
patients into 3 groups: those with aCL levels (a) 57 patients with Mycoplasma pneumoniae infection and
within the normal range, (b) above normal and (c) 21 patients with other infections. Hatched line indicates
above twice normal. The null hypothesis that the laboratory normal range.
distribution of test and control patients within
these groups was identical was tested using x2 (or Table I Number of patients with IgM and IgG
Fisher's exact test if numbers were small). Com- anticardiolipin (aCL) greater than normal range and
parison between anticardiolipin results and other greater than twice normal range.
laboratory data was made using Spearman's rank
order correlation coefficient. Two tailed tests were IgM aCL/U IgG aCL/U
used in all cases. 415 15-30 30 l10 10 -20 20
Mycoplasma 27 19 11 30 24 3
Results pneumoniae
(n = 27)
Controls 19 2 0 19 2 0
The patient with mycoplasma infection and a (n= 21)
stroke had IgM and IgG anticardiolipin (aCL)
levels of 46 and 7 respectively (sample taken 2 days For IgM aCL x2 = 12.26 (2 degrees of freedom)
after admission to hospital). P<0.01. For IgG aCL xI = 9.41, P<0.01 (I degree of
The anticardiolipin titres from the retrospective freedom, comparing values greater and less than normal
study are shown in Figure 1 (for patients with only. Numbers with values >20 too small for comparison
paired samples, only the values from the second by x2.
sample are shown). The median absolute values
were significantly higher in the mycoplasma groups both occasions had positive 1gM. The median IgM
(Mann-Whitney U test, P <0.001 for both IgM and IgG aCL levels from the second of the paired
and IgG aCL). Significantly more patients in the samples were not significantly different from those
mycoplasma group had IgM and IgG anticardio- from patients with only a single (convalescent)
lipin titres outside the normal range and, for IgM samples (Mann-Whitney U test P>0.1).
aCL, greater than twice the normal range (Table I). Clinical details were obtained on 40/57 (70%)
Analysis of the 17 paired samples is shown in patients. Twenty seven patients were hospitalized
Table II. Seven of the 8 samples positive for IgG and 13 were treated at home. Twenty six of the
anticardiolipin on both occasions, or becoming hospitalized patients had clinical and/or radio-
positive on the second sample had positive IgM graphic evidence of pneumonia. One developed
aCL. Four of 8 patients negative for IgG aCL on bilateral facial nerve palsies (lower motor neu-
 Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

ANTIPHOSPHOLIPID ANTIBODIES AND M. PNEUMONIAE INFECTION 359

Table II Paired serum samples: numbers of patients with acute (sample 1) and
convalescent (sample 2) IgM and IgG aCL above and below the normal
range.
Both Sample I negative Sample 1 positive Both
positive Sample 2 positive Sample 2 negative negative
IgM aCL 5 4 3 5
n= 17
IgG aCL 3 5 1 8
n= 17

rone), one developed erythema multiforme (the Table III Comparison of aCL titres in home vs hos-
patient without pneumonia) and one (Down syn- pitalized groups
drome) had repeated hospital admissions. No
patient died. Of the patients treated at home, 7 had IgM aCLU IgG aCLU
evidence of pneumonia and 6 had upper respira- l15 15 -30 >30 <10 >10
tory tract infections or influenza-like illnesses. Five
patients (1 hospitalized, 4 managed at home) had Hospitalized 9 9 9 10 17
prolonged (>2 months) post-infectious respira- n = 27
tory symptoms and malaise. No other complica- Treated at home 9 4 0 9 4
n= 13
tions were recorded. In the control infection group
9 patients were hospitalized and 12 treated at home For IgM aCL, comparing patients above and below
(not significantly different from mycoplasma group 301L P = 0.03. For IgG aCL, above and below 10 U
x2= 3.5, P>0.05). P = 0.12 (Fisher's exact test).
The distribution of positive anticardiolipin titres
in the home treated and hospitalized groups were U0
n-r
A
shown in Table III. Significantly more patients in
the hospitalized group had IgM anticardiolipin
titres > 30 units. The 3 patients with a complicated 50 -

course had titres of IgM and IgG anticardiolipin of A

47 U and 20 U (facial nerve palsies), 35 and 31
(Down syndrome, repeated admissions) and 33 U 40
and 14 U (erythema multiforme).
Cold agglutinins were measured in 12 patients cn A
and were positive in 5 (all greater than 1:512). IgM
anticardiolipin (but not IgG) titres were signifi- 30 -

-J A
cantly higher in the group with positive cold C-)
C.)
agglutinins (Figure 2). There was, however, a A
significant positive correlation in the mycoplasma 20
A
group between IgM and IgG anticardiolipin levels A A
A A A
(Spearman's rank correlation coefficient r, = 0.45 A A A
(n = 58): P<0.01). The hypothesis was therefore A
examined that high titres of IgM anticardiolipin 10 k- A A
might be due to IgM rheumatoid factor (IgM RF) A't
A A A
binding to low titre IgG anticardiolipin. IgM RF A
was measured in 23 patients; the 16 patients with U
the highest IgM anticardiolipin titres and 7 patients +ga
with titres in the normal range. Positive results IgMaCL IgGaCL
(mainly low titre) were found in 8/16 patients with
high levels and 1/7 with low (see Figure 3). There Figure 2 Levels of IgM and IgG aCL in patients with
was, however, no simple relation between IgM RF ( + ) and without ( - ) cold agglutinins. For IgM aCL the
and anticardiolipin titre and the three patients with levels are significantly higher in the ( + ) group (P = 0.01
the highest anticardiolipin titres had no detectable Mann Whitney U test).
IgM RF in this assay.
The presence of other autoantibodies was also the facial nerve palsies and erythema multiforme).
assessed in these 23 patients. Two patients had low Two patients had smooth muscle antibodies at 1/20
titre (1/20) antinuclear antibodies, with a speckled (anticardiolipin titres IgM 59 U and 41 U and IgG
pattern (interestingly, these were the patients with 5 U and 14 U respectively).
 Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

360 N. SNOWDEN et al.

factor activity, although some amplifying effect of

160 rheumatoid factor cannot be ruled out in the cases
in which it was present.
Two contrasting models (not entirely mutually
o 80 exclusive) can be developed for the relationship
between antiphospholipid antibodies and M.
40 pneumoniae infection. Firstly, the antibodies may
be representative of the low grade immune respon-
ses (principally IgM, often low affinity and highly
20 A AAAA AA cross-reactive) to intracellular constituents seen in
<20 A A A.
A A . A A A tissue damage due to many causes.'0,20 The
10 20 30 40 50 60 specificity for mycoplasma infection, the tendency
lgM aCL units to develop IgG antibodies and the low levels of
other autoantibodies would tend to argue against
Figure 3 Relation between IgM aCL and IgM this. Secondly, the antibody production may reflect
rheumatoid factor. some particular tendency of M. pneumoniae to
stimulate autoantibody production (whether by
antigenic cross-reactivity on some other
Discussion immunomodulatory effect). The correlation with
cold agglutinin production (usually identified with
The series of patients with mycoplasma infection IgM anti-I21) is interesting. It is perhaps worthy of
presented here is comparable in clinical spectrum note that the 4 patients in which other autoanti-
to a recently published retrospective review,'8 bodies were detected by immunofluorescence were
although our rate of complications is somewhat all cases with high anticardiolipin binding and/or
lower (neurological problems 1/40 versus 2/47, extrapulmonary complications.
erythema multiforme 1/40 versus 2/47. No cases of The results presented here suggest that eleva-
arthritis, hepatitis or pericarditis were detected). tions in anticardiolipin titre are more marked in
The susceptibility of patients with Down syndrome patients with severe infection (assessed in terms of
was also noted in the retrospective series. the need for hospital admission). There is a possible
We have confirmed the presence of antiphospho- correlation of antiphospholipid titre with cold
lipid antibodies in acute infection. Our results agglutinin production, but the numbers are small
suggest that these antibodies occur in more than and warrant further study.
50% of patients with mycoplasma infection and The possibility that high levels of anticardiolipin
that these patients may develop higher levels of binding may be predictive (and even possibly
antibody when compared to individuals with other pathogenetic) of extrapulmonary complications in
infections (although this should be interpreted M. pneumoniae infection deserves further consider-
cautiously, as the range of infections in the control ation. The numbers presented here do not justify
group was low and was limited to those infections such a conclusion but it is interesting to note that
serologically investigated in a hospital virology the two patients with the facial nerve palsies and
laboratory. There is also a suggestion, although not the stroke had the second and third highest IgM
statistically significant, that the control group had anticardiolipin titres.
less severe disease). It has often been speculated (but rarely substan-
These results contrast with those of Vaarala et tiated, except in the case of anti-I21) that the
al.'9 who found no selective increase among extrapulmonary features of M. pneumoniae infec-
patients with mycoplasma infection. However, the tion may have an autoimmune basis. In one of the
anticardiolipin assay in that study used Tween 20 (a few studies to address this question directly24 a
detergent) as a washing and blocking agent. In our patient with mycoplasma infection and transverse
experience this removes most of the phospholipid myelitis was found, by indirect immunofluo-
from the microtitre plates. rescence, to have antibodies directed against cent-
Examination of the IgG anticardiolipin respon- ral nervous system neurones. The patient improved
ses particularly in the paired sera, suggest that the with plasma exchange. Intriguingly, this patient
IgM response is associated with subsequent also had a circulating inhibitor of in vitro coagula-
development of IgG in around 50-70% of cases. tion, with features similar to the lupus anticoagu-
The results from the paired sera also suggest that lant. As indicated above, this property is very
the IgM response is sustained to at least 7-14 days closely linked to the possession of antiphospholipid
in the majority of cases. We feel this validates the antibodies. There are also suggestions that anti-
use of principally convalescent sera in this series. phospholipid antibodies from patients with SLE
It is unlikely that the high IgM responses seen in may bind neuronal membrane lipids.25 It is unfor-
this study can be explained in terms of rheumatoid tunate that, because of the retrospective nature of
 Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

ANTIPHOSPHOLIPID ANTIBODIES AND M. PNEUMONIAE INFECTION 361

our study, we were not able to assess our patients antiphospholipid antibodies is syphilis. However,
for lupus anticoagulant activity. In patients with the phospholipid binding found here has a different
SLE this seems to be a more specific predictor of antigenic specificity22 (for neutral and positively
vascular and other complications.26 charged lipids) from that found in SLE, and is
A recent retrospective study has identified infec- poorly detected by solid phase anticardiolipin
tion as an important independent risk factor in assays such as the ELISA described here (reference
patients with ischaemic stroke aged under 50 23 and unpublished results from our laboratory on
(relative risk 14.5 in the month following infec- 4 high titre VDRL sera, kindly supplied by Mr D.
tion).27 The explanation for this is not clear and Ellis of the Central Serology Laboratory, Withing-
although changes in haematological variables ton Hospital). It therefore seems likely that the
(such as fibrinogen concentration and platelet antiphospholipid antibodies in mycoplasma infec-
count) occurring as acute phase responses could be tion have a specificity for negatively charged lipids
implicated, the role of antiphospholipid antibodies but insufficient data are presented here to draw
would seem to be worthy of further investigation further comparisons with SLE.
(no cases or controls in this study, from Helsinki,27 Finally it should be noted that there are some
had mycoplasma infection, suggesting the inci- important clinical differences between the extra-
dence was low at the time of study). It would be pulmonary features of mycoplasma infection and
interesting to know whether other infections (such the lupus related antiphospholipid antibody syn-
as the principally bacterial infection recorded in the drome. Neurological disease is relatively common
Finnish27 study) are capable of producing anti- and thrombotic events relatively rare in myco-
phospholipid responses of a degree comparable plasma infection'8 whereas the reverse is true in
with those seen in mycoplasma infection - or SLE and in the phospholipid antibody syndrome.'
whether certain individuals produce a marked The thrombocytopenia and fetal loss noted in SLE
increase in aPL titre regardless of the infecting have not been reported in mycoplasma disease.
organism. It would thus seem unreasonable to draw too
To what extent are the antiphospholipid anti- close a parallel between these two groups of
bodies seen in M. pneumoniae infection comparable disorders, and their relation to antiphospholipid
to those seen in SLE and the primary antiphos- antibodies. However, it would be precipitate to
pholipid syndrome? This question needs to be consider antiphospholipid antibody responses in
answered in both quantitative and qualitative M. pneumoniae infection (and possibly other infec-
terms. Quantitatively the positive responses in our tions) as an entirely irrelevant epiphenomenon.
patients fall into the 'medium positive' range
defined by Harris (IgM = 10-80 'Manchester
units' IgG = 6-30 'Manchester units'). Values Acknowledgement
within this range are associated with thrombosis
and other SLE complications, although less so than We wish to thank Dr W.D.W. Rees for permission to
values in the 'high positive' range.30 However, our study the patient with stroke and mycoplasma infection,
patients tended to have principally IgM anti- Dr M.R. Haeney for the use of this patient's sera, Miss J.
cardiolipin, whereas complications in lupus and Ditchfield for the St Thomas' reference sera, Mrs E.
lupus-like patients are more strongly associated Crosdale for help with sample location, Dr M.E. Ellis for
with IgG28 (but have been reported in patients with initially drawing our attention to these antibodies in
mycoplasma infection and the many physicians in the
only IgM anticardiolipin29). North West who allowed us to study their patients and
Qualitatively, antiphospholipid antibodies are provided clinical details. We are also grateful to Dr S.J.
undoubtedly heterogeneous in terms of antigen Richmond for her permission to use her diagnostic index
binding. An infection strongly associated with of patients.

References
1. Asherson, R.A & Harris, E.N. Anticardiolipin antibodies - 5. Harris, E.N., Asherson, R.A. & Hughes, G.R.V. Antiphos-
clinical associations. Postgrad Med J 1987, 62: 1081- 1087. pholipid antibodies - autoantibodies with a difference. Ann
2. Oppenheimer, S. & Hoffbrand, B.I. Optic neuritis and Rev Med 1986, 39: 261-271.
myelopathy in systemic lupus erythematosus. Can J Neurol 6. Hughes, G.R.V., Harris, E.N. & Gharavi, A.E. The anticar-
Sci 1986, 13: 129-132. diolipin syndrome. J Rheumatol 1986, 13: 486-489.
3. Harris, E.N. Gharavi, A.E., Tincani, A. et al. Affinity purified 7. Misra, R., Venables, P.J.W., Watkins, R.P.F. & Maini, R.N.
anticardiolipin and anti-DNA antibodies. J Clin Lab Autoimmunity to cardiolipin in infectious mononucleosis.
Immunol 1985, 17: 155-162. Lancet 1987, ii: 629.
4. Derksen, R.H.W.M., Beisma, D., Bauma, B.N. et al. Discor- 8. Di Prima, M.A., Sorice, M., Vullo, V., Mastroianni, C.M.,
dant affects of prednisone on anticardiolipin antibodies and Amendolea, M.A. & Masala, C. Anticardiolipin antibody in
the lupus anticoagulant. Arthritis Rheum 1986, 29: the acquired immunodeficiency syndrome: a marker of
1295-1296. Pneumocystis carinii infection? J Infection 1989, 18: 100 -1 0 1.
 Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

362 N. SNOWDEN et al.

9. Klemp, P., Cooper, R.C., Strauss, F.J., Jordaan, E.R., 20. Isenberg, D. & Shoenfeld, Y. The origin and significance of
Przybojewski, J.Z. & Nel, N. Anticardiolipin antibodies in anti-DNA antibodies. Immunol Today 1987, 8: 279-282.
ischaemic heart disease. Clin Exp Immunol 1988, 74: 21. Mollison, P.L., Engelfriet, C.P. & Contreras, M. Blood
254-257. Transfusion in Clinical Medicine. Blackwell Scientific Publica-
10. Manoussakis, N.M., Tzioufas, A.G., Sillis, M.P., Pange, tions, Oxford, pp. 419-421.
P.J.E., Goudevenos, J. & Mountsopoulos, H.M. High 22. Colaco, C.B. & Male, D.K. Antiphospholipid antibodies in
prevalence of anticardiolipin and other autoantibodies in a syphilis and thrombotic subset of SLE: distinct profiles of
healthy elderly population. Clin Exp Immunol 1987, 69: epitope specificity. Clin Exp Immunol 1985, 59: 449-457.
557-565. 23. Harris, E.N., Gharavi, A.E., Loizou, S. et al. Cross reactivity
11. Isenberg, D.A., Maddison, P., Swana, G. et al. Profile of of antiphospholipid antibodies. J Clin Lab Immunol 1985, 16:
autoantibodies in the serum of patients with tuberculosis, 1-6.
klebsiella and other gram-negative infections. Clin Exp 24. Cotter, F.E., Bainbridge, D. & Newland, A.C. Neurological
Immunol 1987, 67: 516-523. deficit associated with Mycoplasma pneumoniae reversed by
12. Carreras, L.O., Defreyn, G., Machin, S.J. et al. Arterial plasma exchange. Br Med J 1983, 286: 22.
thrombosis, intrauterine death and the lupus anticoagulant: 25. Harris, E.N., Englert, H., Derve, E. et al. Antiphospholipid
detection of immunoglobulin interfering with prostacylin antibodies in acute Guillain-Barre syndrome. Lancet 1983,
formation. Lancet 1981, i: 244-246. ii: 11361-11362.
13. Twomey, J.A. & Espir, M.L.E Neurological manifestations 26. Derksen, R.H.W.M., Hasselaar, P., Blokzijl, L., Gmelig
and Mycoplasma pneumoniae infection. Br Med J 1979, 2: Meyling, F.H.J. & De Groot, P.G. Coagulation screen is
832-833. more specific than the anticardiolipin antibody ELISA in
14. Dowd, A.B., Grace, R. & Rees, W.D.E. Cerebral infarction defining a thrombotic subset of lupus patients. Ann Rheum
associated with Mycoplasma pneumoniae infection. Lancet Dis 1988, 47: 364-377.
1987, ii: 567. 27. Syrjanen, J., Valtonen, V.V., Ilvanainen, M., Kaste, M. &
15. Mulder, L.J.M.M. & Spriengs, E.L.H. Stroke due to DIC Huttunen, J.K. Preceding infection as an important risk
Mycoplasma pnewnoniae infection. Lancet 1987, ii: factor for ischaemic brain infarction in young and middle
1152-1153. aged patients. Br Med J 1988, 296: 1156- 1160.
16. Loizou, S., McRea, J.D., Rudge, A.C., Reynolds, R., Boyle, 28. Gharavi, A.C., Harris, E.N., Asherson, R.A. & Hughes,
C.C. & Harris, E.N. Measurement of anticardiolipin G.R.V. Anticardiolipin antibodies: isotype distribution and
antibodies by ELISA: standardization and quantitation of phospholipid specificity. Ann Rheum Dis 1987, 46: 1-6.
results. Clin Exp Immunol 1985, 62: 738-745. 29. Drew, P., Asherson, R.A., Zuk, R.J., Goodwin, F.J. &
17. Harris, E.N., Gharavi, A.E., Patel, S.P. & Hughes, G.R.V. Hughes, G.R.V. Aortic occlusion in SLE associated with
Evaluation of the anticardiolipin antibody test: report on an antiphospholipid antibodies. Ann Rheum Dis 1987, 46:
international workshop held 4 April 1986. Clin Exp Immunol 612-616.
1987, 68: 215-222. 30. McHugh, N.J., Maymo, J., Skinner, R.P., James, 1. &
18. Ali, N.J., Sillis, M., Andrews, B.E., Jenkins, P.F. & Harrison, Maddison, P.J. Anticardiolipin antibodies, livedo reticularis
B.D.W. The clinical spectrum and diagnosis of Mycoplasma and major cerebrovascular and renal disease in SLE. Ann
pneumoniae infection. Q J Med 1986, 58: 241-251. Rheum Dis 1988, 47: 110-115.
19. Vaarala, O., Palosuo, T., Kleemola, M. & Aho, K. Anticar-
diolipin responses in acute infections. Clin Immunol
Immunopathol 1986, 41: 8-15.
Downloaded from http://pmj.bmj.com/ on October 5, 2015 - Published by group.bmj.com

Antiphospholipid antibodies
and Mycoplasma
pneumoniae infection.
N. Snowden, P. B. Wilson, M. Longson and R.
S. Pumphrey

Postgrad Med J 1990 66: 356-362

doi: 10.1136/pgmj.66.775.356

Updated information and services can be found

at:
http://pmj.bmj.com/content/66/775/356

These include:

Email alerting Receive free email alerts when new articles cite
service this article. Sign up in the box at the top right
corner of the online article.

Notes

To request permissions go to:

http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to:

http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to:

http://group.bmj.com/subscribe/