Lupus
Anticoagulants/Phospholipid-Dependent-Antibodies
June 17, 2004
14:00 to 18:00
Cenacolo Room
Fondazione Giorgio Cini
Chairman:
Ph. G. De Groot, The Netherlands
Co-chairs: J. Arnout, Belgium; M. Galli, Italy; ; S. Machin, UK; V.
Pengo, Italy;
J.
Rand, USA; R. Roubey, USA
Number of attendees: more than 100
Introduction: The chairman opened the meeting and summarized
the goals of this meeting. Dr. Pengo distributed a
questionnaire with questions regarding antiphospholipid
testing in diagnostic laboratoria. The goal of this
questionnaire is to reach standardization in testing and the questionnaires
were collected at the end of the meeting.
Prospective
in LA diagnosis
Dr Pengo (Italy) discussed his recent
finding that by using different Ca2+-concentrations to activate
coagulation it is possible to discriminate between a b2-Glycoprotein I (b2GPI )-dependent lupus Anticoagulants (LAC) and a prothrombin
dependent LAC. A reduction in final calcium concentration, from
10mM to 5mM, increased
coagulation times in both dilute Russell Viper Venom Time (dRVVT)
and dilute Prothrombin Time (dPT)
when plasmas of patients with anti-b2GPI antibodies were used. Instead, all
LA-positive anti-b2GPI antibody negative patients
showed decreased coagulation times from mean These
results are confirmed by running dRVVT of normal
plasma spiked with affinity purified IgG anti-b2GPI antibodies. Therefore, when a
PL–dependent coagulation test is run twice, at different final calcium
concentrations, clinically important anti-b2GPI LA can be identified.
Dr. Arnout (Belgium) have used
his panel of monoclonal antibodies directed against b2GPI and prothrombin
that exert LAactivity to confirm the effects of
different Ca2+- concentrations on dRVVT and
dPT. He confirmed the original observations of dr. Pengo but showed that the
presence of a high level of a anti-prothrombin
antibody mediated LAC masks the presence of a b2GPI dependent LAC. Also the presence of an anti-factor VIII, LMWH. OAC etc
should be carefully evaluated before the test as proposed by dr.
Pengo could be introduced.
Dr. de Groot (the Netherlands) discussed the assay they use to
detect a b2GPI –dependent LAC. This assay is
based on the use of cardiolipin as confirming agent
in a PTT-LA coagulation assay. His group studied the clinical of the detection
of b2-glycoprotein I-dependent LAC in a cohort of
198 patients with autoimmune diseases. Presence of b2-glycoprotein I dependent LAC was strongly
associated with a history of thrombo-embolic
complications, an Odds Ratio of 42.3 (95% confidence interval 194.3 - 9.9) was
found. An increased frequency of thrombosis was not found in the 33 patients
with LAC independent of anti-b2-glycoprotein I antibodies (Odds Ratio 1.6, 95%
confidence interval 3.9 - 0.8). The use of a LAC assay with cardiolipin
as confirming agent strongly improves the detection of patients at risk for
thrombosis. This suggest that anti-b2-glycoprotein I antibodies with LAC activity
are antibodies responsible for the thrombo-embolic
complications in APS.
In a separate study he compared dr. Pengo’s assay with the assay
they developed in their own lab. In general a good correlation was found,
although some discrepancies were noticed., probably
due to the presence of mixtures of antibody populations.
A multicentre study
was launched in which the clinical relevance of a b2GPI-dependent LAC will be tested in
a cohort of at least 500 patients. This study will be performed in
collaboration with and supported by Stago R&D
(France)
Standardisation of the LAC
assay
Dr Arnout (Belgium) discussed the development of a new standard
for LAC testing. The standard should contain a mixture of patient pooled
plasmas, normal pooled plasma spiked with monoclonal antibodies and normal
pooled plasma. To make the standard, 10 L of patient plasma is needed. He
suggested to collect 100 mL
of plasma from 100 patients with high titer LAC. Dr Arnout
distributed a form with which participants could enlist donors for this pool. after a long discussion on ethics and selection of patients
a protocol and time table was accepted. The plan is to have the standard ready
for the next meeting in Sidney. A steering committee (Arnout, de Groot, Mackie, Jennings, Gray, Sie, de Moorloose & Pengo) will
guarantee the quality and progress of the procedure.
Dr. Mackie (United Kingdom) assessed a new dilute prothrombin
time diagnostic assay, ACTICLOTÒ dPTtestä and ACTICLOT dPTconfirm (AdPT, American Diagnostica Inc, USA) for lupus anticoagulant (LA) detection. The
ACTICLOT dPTtestä is a lipidated
recombinant human tissue factor-based clotting assay designed to screen plasmas
for LA. The ACTICLOT dPTconfirm test utilizes a high
phospholipid-containing reagent to confirm the presence of LA. The clotting
times are compared using calculations similar to those employed for DRVVT
methods. AdPT was positive patients, 23 oral
anticoagulant and 17 new thrombophilia outpatients.
Samples were considered LA positive if they complied with International
criteria and were positive in at least 2/4 reference methods (APTT ratio using Actin FSL, two DRVVT methods employing either high phospholipid or platelet neutralization confirm reagents,
and StaClot LA). Two different lots of AdPT reagent were tested on an ACL300R analyzer and gave
similar results in normals and LA patients. The AdPT reagents were stable for at least 96hr and the within
run cv for normal and LA
positive samples was <5%. Cut-off values for LA were based on AdPT results from normal subjects. 19/23 APA group patients
had LA by AdPT (median and range 2.5 and 1.1-5.7 for
patient/normal ratio; 1.4 and 0.9-3.7 for test/confirm ratio). All 17 thrombophilia patients were LA negative and AdPT negative. Using the APA and thrombophilia
groups, and definite LA positive and negative samples, AdPT
showed 83% sensitivity and 100% specificity. When AdPT
was coupled with any of the other LA tests, sensitivity varied from 96-100% and
specificity 82-100%. When a sensitive APTT reagent was used for screening, with
AdPT and a DRVVT as specific tests, 96% sensitivity
and 100% specificity were achieved. If three tests were performed, using APTTr mainly as a screening test, specificity was 100%. In
oral anticoagulant patients, AdPT exhibited similar
problems to other reagents, suggesting that mixing tests should be performed.
The results suggest that AdPT is highly specific and
detects the majority of patients with LA, albeit a slightly different spectrum
of LA antibodies than seen with DRVVT or APTTr.
ACTICLOT dPT used in
combination with two recognized LA reagents provided the best diagnostic
sensitivity for LA in patient samples studied in 20 healthy normal subjects, 23
known antiphospholipid antibody (APA)
Dr. Jennings (United Kingdom) discussed the effect of pH and
buffering on DRVVT results with different commercial Kits. Following the
observation that the dRVVT results of normal pooled
plasma of NEQUAS-
distribution in England showed substantial variation he
found that the pH of the plasma strongly influenced the results of LAC testing.
When the pH of normal plasma is above 8.0, a dRVVT
becomes positive. As the pH of lyophilised plasma after reconstitution can be
above 8.0, he advised to use a Hepes buffer in dRVVT-testing.
Prospective
in ELISAs
Guido Reber (Switserland): At the Sapporo consensus meeting [6], anti-β2GPI antibodies (aβ2GPI) were
not included in the list of biological criteria defining the antiphospholipid syndrome (APS). At the SSC held in Boston July 2002, a new classification was
proposed, which included aβ2GPI either alone or in the presence
of lupus anticoagulant. One of the issues raised by adding this test to the
biological criteria was the standardisation of these assays. Since 1999, our
group has performed three collaborative studies on this topic. One study
addressed the performances of homemade assays and the last one looked for the
agreement between commercial tests (both were presented at the SSC meeting held
in Birmingham July 2003). In summary, the results
of our studies showed that the agreement was good for high positive samples,
fair to poor for medium positive and poor for low positive. For commercial
kits, there was agreement in sample classification (positive or negative)
between the ten commercial kits in 12/22 samples for IgG
isotype and in 5/22 for IgM
isotype. Analysis of the data prompted us to make
some proposals to improve the agreement between assays.
1) To perform duplicates
2) Control group for the determination
of the upper limit of the reference range (cut-off)
3) Calculation of the cut-off value
Because the distribution of the values is not
Gaussian, the cut-off has to be calculated by the method of percentiles and not
by adding standard deviations to the mean value. When analysing the
distribution of the values of the control group, it must be kept in mind that
anti-β2GPI antibodies are found in 2 to 5% of healthy
individuals, especially in individuals aged more than 60. Kits' manufacturers
are asked to indicate the units values corresponding
to the 95%, 97.5% and 99% of the distribution of the control group in addition
to the cut-off value they recommend. To fulfil regulation authorities
requirements, the proposed cut-off takes into account sensitivity and
specificity data obtained from clinical studies.
4) Humanized monoclonal antibodies as
reference calibrators
Because there are still no worldwide-accepted
anti-β2GPI calibrators, we propose to use humanized MAb dilutions as calibrators to allow comparisons between
assay systems. Our group is aware that a MAb, because
it is directed to one single epitope, cannot mirror
the diversity of the subsets of patients' autoimmune polyclonal antibodies with
different binding affinities, generally lower than MAb.
MAb may not detect micro heterogeneities in β2GPI preparations because they possibly do not
affect their binding, as would be the case for subsets of patients’ antibodies.
But MAb are not affected by batch-to-batch variation
because their affinity do not change notably with
time. This ensures that the stability of assay systems can be checked over many
years.
We hope that these proposals will improve the
agreement between assay results. Thirteen different commercial firms have
agreed to follow this proposal and to include the humanised monoclonal
antibodies as standard in their tests.
Does a panel of aPL tests improve diagnostics?
Monica Galli (Italy): A systematic review of the published articles on the antiphospholipid
syndrome showed lupus anticoagulants were a clear risk factor for thrombosis,
irrespective of the site and type of thrombosis, the presence of systemic lupus
erythematosus, and the methods used to detect them. Anticardiolipin and antib2-glycoprotein
I antibodies were possible risk factors of thrombosis, at least in some
selected situations, whereas the measurement of antiprothrombin
antibodies was not helpful to define the patient’s risk of thrombosis. In most
cases, these studies analyzed the relationship with thrombosis of each single antiphospholipid antibody, rather than the combination of
different antibodies with each other and/or with the coagulation tests commonly
used to detect lupus anticoagulants. Based on these premises, we retrospectively
studied a group of 103 well established lupus anticoagulant-positive patients
for whom laboratory data about anticardiolipin, antib2-glycoprotein I and antiprothrombin
antibodies were available, in order
to investigate whether laboratory patterns emerged for their association with
thrombosis in the antiphospholipid syndrome. First, anticardiolipin, antib2-glycoprotein I, and antiprothrombin antibodies were analyzed separately. We
observed the following statistically significant associations. No association
with arterial thrombosis reached significance.
Second, the antibodies
were combined in different patterns, which included from 2 to 5 laboratory
variables. Matching the variables for the antibody isotype,
60 laboratory patterns were generated. Their analysis with respect to three
different types of thrombosis made us available 180 different combinations.
Overall, 22 associations reached significance:
Type of thrombosis N. of significant
associations/Total N. of associations
___________________________________________________________________
- Any 14/60
- Venous 8/60
- Arterial 0/60
___________________________________________________________________
In all but 2 cases, anticardiolipin antibodies > 40 units (G or any isotype)
were present in the laboratory patterns that reached significance. Antib2-glycoprotein I antibodies (G or any isotype) were present in 11 significant patterns, and antiprothrombin antibodies (M or any isotype)
in 7 cases. KCT Ratio of 1:1 mixture > 1.5 was present in 7 patterns which
were significantly associated with thrombosis, dRVVT
Ratio of 1:1 mixture > 1.5 in other 7 cases, and both KCT and dRVVT Ratios > 1.5 in one case.
Increasing the number of variables of the laboratory patterns did not increase
the Odds Ratio towards thrombosis. Figures 1 and 2 show
the OR with 95% CI of the laboratory patterns which reached statistical
significance.
Conclusions
The analysis of the antibody patterns of a
large cohort of lupus anticoagulant-positive patients confirmed that the
presence of medium to high titres of IgG anticardiolipin antibodies, either alone or in various
combinations, increases their risk of thrombosis. Conversely, the role of the
other laboratory variables, irrespective of whether they were considered
separately or combined, is less clear.
The way the results of the various antiphospholipid antibodies are reported may partly explain
these findings. Unlike anticardiolipin antibodies,
antib2-glycoprotein I and antiprothrombin
antibodies were expressed in
a qualitative fashion, which does not allow establishing clinically relevant
cut offs. With respect to coagulation tests, dRVVT
and KCT ratios did not show any correlation with the patients’ history of
thrombosis. This is apparently in contrast with our previous observation that
the dRVVT profile rather than the KCT profile was
associated with thrombosis in lupus anticoagulant-positive patients. However,
we must underscore that the two studies had different designs, and that in the
present study we considered the ratio of the 1:1 coagulation times of the
mixture of patient’s with normal plasma, in order to rule possible interference
of oral anticoagulant treatment.
Vittorio Pengo (Italy) performed a comarable
study. Over a 6-year period, 618 consecutive patients (55% of
whom with previous documented thromboembolic events)
were referred to their clinic for Antiphospholipid
antibody detection. LA was detected
according to internationally accepted
recommendations. Enzyme-Linked-Immunosorbent Assays
(ELISA) detect aCL and
Anti-human b2GPI
antibodies. Patients’ records were reviewed for the presence of previous thromboembolic events or obstetrical complications and each
patient received a physical examination. LA (Odds Ratio 2.9, Confidence
Interval 1.1-7.6) and ab2GPI (Odds Ratio 2.3, Confidence Interval
0.9-5.8) but not aCL (Odds Ratio 0.7, Confidence
Interval 0.3-1.6) as individual test positively are significantly associated to
thromboembolic events. The rates of thromboembolic events or obstetric complications were 94%
(34/36) in patients with a triple positivity, 58%
(18/31) in those aCL-and ab2GPI- positive but LA- negative, 50%
(4/8) in patients in which ab2GPI antibodies was the single positive assay,
35% (7/20) in patients in which medium-high titre aCL
was the single positive test, while no one of 5 patients in which Lupus
Anticoagulant was the single positive test has had previous thromboembolic
events. When results were tested in a logistic regression analysis, only the
complete positive antiphospholipid profile was
independently associated with thromboembolic events
(OR 14.8, CI 3.5-62.2). ACL and ab2GPI antibody positive but LA negative profile
showed an association with thromboembolic events
without reaching statistic significance. The mean level of IgG
Anti-human b2-glycoprotein
I antibodies was statistically higher in triple positive profile and might
account for positive Lupus Anticoagulant. These results show that LA and IgG ab2GPI antibodies
as individual positive test but specially the combined positivity
of all three tests in a single patient may be considerable for identifying
patients with antiphospholipid syndrome (APS).
Dr. Machin (United Kingdom) commented on these observations
and showed that in the patient population of his hospital a substantially
number of APS patients had only a positive anti cardiolipin.
After an intense discussion agreement on
headlines became apparent but no agreement could be reached on details. It was
proposed that the chairman and cochairs should make a
proposal on aPl testing and
that this proposal should be proposed at the next meeting in Sidney. Hopefully, in Sidnet
an agreement of new criteria on aPL
testing could be reached.
Relevance of
other antibodies
Dr. Rand (USA): gave an overview on annexin A5 and its possible role in APS. He discussed an
assay he developed in his lab that was based on the inhibition of a clotting
test by annexin A5. In the presence of aPL, the inhibitory effect of annexin A5 becomes less. The ‘shortening’ of the Annexin A5 mediated prolongation of the clotting test
correlated with the presence of aPl,
although there is an overlap between controls and patients. More studies are
needed before this assay can be introduced as a aPL specific test.
Dr. de Groot (the Netherlands) investigated together with dr. Mackie the prevalence of IgG
or IgM annexin A5 (A5)
antibodies. An ELISA was performed with purified recombinant A5 coated onto the
plate in two different laboratories on patients with various autoimmune
diseases, mainly the antiphospholipid syndrome (APS).
In laboratory A, 2/36 patients were positive for
anti-A5 IgG, both patients had APS and had
complications during pregnancy. 6/36 patients had anti-A5 IgM
antibodies, 4/6 patients had a history of pregnancy loss. There was no
association between thrombosis and A5 antibodies. In laboratory B, 26/198
patients were positive for IgM A5 antibodies with an
odds ratio of 2.7 (significant) for thrombosis and 1.3 (non-significant) for
pregnancy loss. 53 patients were positive for IgG
antibodies with an odds ratio of 1.5 (non-significant) for thrombosis and 2.2
(non-significant) for pregnancy loss. In this laboratory the A5 antigen levels
were also measured and were non-significantly lower in patients with IgM and/or IgG anti-A5 (7.9 ± 0.7
ng/ml plasma) than in patients without anti-A5 (9.8 ±
0.9 ng/ml plasma).
In conclusion, the measurement of annexin A5
antibodies seems to have no additional value for the detection of patient at risk
for pregnancy loss. The relation with a history of thrombosis needs further
studies.
Dr. Mackie (United Kingdom) investigated together with dr. de root the
prevalence of IgG or IgM antiprotein S antibodies. A prototype commercial ELISA
method for protein S antibodies was assessed in two different laboratories. The
assay used microplates coated with
highly purified human protein S, with bovine serum albumin in the blocking and
wash buffers. Peroxidase conjugated monoclonal
anti-human IgG or IgM were
used for detection and humanised murine monoclonal
anti-protein S for calibration. Good agreement was obtained between duplicate
wells and replicate calibrant dilution curves on the
same plate. The optical density of calibrant varied
between plates and was therefore used to normalise the results. Each laboratory
established separate normal ranges using 36-40 healthy normal subjects each.
Laboratory A studied 14 serial samples from a 20-year-old female with thrombotic necrosis of the extremities. Six of these samples
were IgG anti-protein S (aPS)
positive and the results correlated with protein S activity, antigen and
neutralising activity. aPS
disappeared with treatment and clinical improvement. Each laboratory studied
separate samples from patients with SLE, PAPS, lupus-like disease, pregnancy
complications, acquired inhibitors and controls. Patients were judged APA
positive if they had positive results for lupus anticoagulant, anti-cardiolipin or anti-B2GPI assays. Laboratory A
studied 118 such samples, of which 18/86 APA positive samples and 1/32 APA
negative samples had aPS (14 IgG,
5 IgM). Laboratory B studied 198 samples, of which
64/128 APA positive samples had aPS (54 IgG, 3 IgM, 7 dual positive); and
17/70 APA negative samples had aPS (14 IgG, 1 IgM, 2 dual positive).
Overall, 38% of APA positive patients had aPS and 18%
of APA negative patients; the presence of aPS thus
having a high association with other APA (FET, p=0.0002). There were no
significant associations of aPS with pregnancy loss
or thrombosis, although further, prospective studies are required.
Treatment of
APS patients
Guido Finazzi (Italy) discussed the optimal intensity of
oral anticoagulation for the prevention of recurrent thrombosis in patients
with the antiphospholipid antibody syndrome is
uncertain. Retrospective studies have suggested that only doses of warfarin adjusted to achieve an international normalized
ratio (INR) of more than 3.0 are effective, whereas a recent randomised
clinical trial comparing high (INR range 3.0 to 4.0) vs. moderate (INR 2.0 to
3.0) intensities of anticoagulation failed to confirm this assumption. They
conducted a randomized trial in which patients with persistent lupus
anticoagulant and/or moderate to high levels of anticardiolipin
antibodies and previous thrombosis were given either high-intensity (INR range
3.0-4.5, target 3.5) or standard intensity (INR range 2.0-3.0, target 2.5) warfarin therapy. We sought to
determine whether intensive anticoagulation was superior to standard treatment
in preventing symptomatic recurrent thromboembolism without increasing the rate
of bleeding complications. The WAPS trial was launched in the framework of the
SSC Subcommittee on antiphospholipid antibodies/lupus anticoagulants and
started in July 1997. A total of 109 patients were
recruited in the trial and followed for a median of 3.6 years. Mean INR during follow-up was 3.2 (SD 0.6) and 2.5 (SD 0.3)
(p<0.0001) in the high- and standard-intensity groups respectively.
Recurrent thrombosis was observed in 6 of 54 patients (11.1 percent) assigned
to receive high-intensity warfarin and in 3 of 55 (5.5%) assigned to receive
conventional treatment (hazard ratio for the high intensity group, 1.97; 95
percent confidence interval 0.49-7.89). Major and minor bleeding occurred in 15
patients (2 major) (27.8%) assigned to receive high-intensity warfarin and 8 (3
major) (14.6%) assigned to receive conventional treatment (hazard ratio 2.18;
95 percent confidence interval 0.92-5.15). High-intensity warfarin was not superior to
standard treatment in preventing recurrent thrombosis in patients with the antiphospholipid syndrome and was associated with an apour results with those of a similar trial recently
published support the recommendation that conventional thromboprophylaxis
with warfarin targeted at INR 2.5 is usually
appropriate for these patients.
The
results of this meeting in Venice are:
1. A collaborative study will be started on the value
of testing b2Glycoprotein I specific LAC (de Groot / Woodhams). Results will be presented in Sidney.
2. A standard will be developed for LAC testing (Arnout) The first results will be
presented in Sidney.
3. A proposal for better standardisation for the b2GPI ELISA has been discussed (Reber). 13 different commercial firms have already agreed
to follow these rules and to include the humanised monoclonal antibodies as
standard in their tests. The proposal will be published.
4. A consensus is
in reach for new criteria for aPL
testing. A proposal will be formulated and presented in Sidney
(chair and co-chairs).
5. The European
WAPS study is finished (Finazzi). It is concluded
that it is not necessary to treat patients with APS with high intensity warfarin. These results will be published. It is recommendated to treat patients with warfarin
targeted at an INR of 2.5. Further analysis of the observery
arm of the study will follow.