1. Abstract and Introduction
Abstract
Purpose of review: Although originally described in the context of systemic lupus
erythematosus, antiphospholipid syndrome was then recognized as a primary
antiphospholipid syndrome without any underlying autoimmune disease in almost
half of the cases. However, cases of primary antiphospholipid syndrome were
reported to evolve into systemic lupus erythematosus over time suggesting that
these apparently different diseases areomehow related.
Recent findings: Peculiar biological systemic lupus erythematosus markers such as
an autoantibody response against chromatin antigens and complement activation
have been also described in patients with primary antiphospholipid syndrome.
Distinct polymorphisms of common genetic factors have been associated with
systemic lupus erythematosus and primary antiphospholipid syndrome supporting
the notion that these entities are indeed variants within a continuum of the same
disease.
Summary: A multiorgan involvement that cannot be explained by the thrombophilic
state per se and similar to the well known clinical manifestations in lupus is reported
in patients with primary antiphospholipid syndrome. Further studies, mainly genetic,
will better underline the proximity between primary antiphospholipid syndrome and
systemic lupus erythematosus.
Introduction
Antiphospholipid syndrome (APS) was first described in a group of patients with
systemic lupus erythematosus (SLE) who had increased levels of anticardiolipin
antibodies (aCL) and clinical features of recurrent thrombosis, thus defined as
secondary APS.[1] By 1985 it had become apparent that some of these patients
exhibited no features of underlying connective tissue disease, and the concept
emerged that APS could exist as a primary syndrome. At this stage, the term primary
APS (PAPS) was established, relating to a group of patients in whom recurrent
thrombosis, and or fetal losses, were associated with antiphospholipid antibodies
(aPL) [aCL, lupus anticoagulant, and anti-β2 glycoprotein I antibodies (anti-β2 GPI)]
and lacking any evidence of underlying connective tissue diseases.[2–4]
After 25 years 'in the air' and the reported cases of PAPS evolving into full-blown
SLE, the hypothesis that the syndrome represents wide spectrum of diseases rather
than discrete disease entities has been raised. In this respect, multiorgan
involvements and serological markers that could not be explained by APS per se
suggested that differences between SLE and APS, once very well defined, are at
least discussed now.[5] In this study we review in a systematic manner the findings
that support the proximity between PAPS and SLE.
2. The Proximity between Primary Antiphospholipid Syndrome
and Systemic Lupus Erythematosus: Common Clinical
Manifestations
The Following Clinical Manifestations are Commonly Present in Both
SLE and PAPS Skin Manifestations
Livedo reticularis is a skin vasculopathy that has been frequently described in
patients with both PASP and SLE. However, livedo reticularis is more frequently
detected in APS associated to SLE than in the primary form.[6,7] We also reported that
livedo reticularis was associated with thrombocytopenia, cardiac valves dysfunction,
epilepsy, and arthritis.[8] It is useful to speculate that skin involvement may identify
patients with a more systemic disorder and potentially closer to SLE-like disease (
Table 1 ).
Hematological Manifestations
In one of the reported cases in which PAPS evolved into SLE 12 years after the first
thrombotic episode, antinuclear antibodies (ANAs), thrombocytopenia, antidouble
stranded (ds)DNA antibodies and positive Coombs' test were all present.[9] The
association of autoimmune hemolytic anemia (AIHA) and cardiac valvular
vegetations/thickening (P < 0.0001), arterial thrombosis (P < 0.02), livedo reticularis
(P < 0.0001) and central nervous system (CNS) signs of epilepsy or chorea (P <
0.02 and P < 0.03, respectively) was reported in another study enrolling 308 PAPS
patients.[10]
Altogether these findings would suggest that hematological manifestations may
define a subgroup of patients with a significant risk for subsequent development of
SLE. However, definite conclusions cannot be drawn at the moment on the basis of
the literature data and specific prospective studies are needed.
Renal Manifestations
Kidney involvement by aPL was originally associated with thrombotic events in
medium-large venous/arterial vessels or less frequently with thrombotic
microangiopathy.[4] More recently, nonthrombotic lesions could be found in kidneys of
patients with PAPS, establishing the term aPL-associated nephropathy to indicate
such kidney damage.[11] Moreover, association between the persistent presence of
aPL and the risk to develop an end-stage renal insufficiency was reported in lupus
patients with renal involvement. A chronic renal vasculopathy specifically mediated
by aPL was thought to be responsible.[12,13]
Apart from these findings, conditions like glomerulonephritis have been reported in
patients with 'classical' PAPS, suggesting that inflammatory processes could
possibly develop in some patients, so suggesting that a continuum between these
two diseases may exist also for kidney involvement.[14]
Central Nervous System Manifestations
3. Central nervous system involvement in SLE is common and results in different clinical
manifestations. Whereas ischemia secondary to vasculitis was reported to represent the
most important cause of the CNS manifestations in SLE, a noninflammatory
vasculopathy was associated with aPL.[15] In this regard, brain endothelial cells display a
higher and firmer expression of β2GPI on cell membranes that may offer a suitable target
for β2GPI-dependent aPL.[16] The antibody binding has been reported as a potential cause
of perturbation/damage of the cerebral microcirculation and eventually of the clinical
manifestations such as seizures, cognitive dysfunction or psychosis.[17] Alternatively anti-
β2GPI antibodies themselves may react with β2GPI expressed on the neuronal cell
membranes as recently suggested by some groups. The same antibodies may also affect
the endothelium of the blood–brain barrier making it much more permeable to
proinflammatory mediators or serum autoantibodies potentially active on neuronal
tissues. Hence, similar CNS manifestations may be supported by different mechanisms
all of which are present in both SLE and PAPS patients. In this regard it is very difficult
to state whether CNS involvement does represent a bridge between the two diseases or
simply a coincidence
The Slow March from Primary Antiphospholipid Syndrome to
Systemic Lupus Erythematosus
Follow-up studies reported cases in which PAPS developed into a complete picture
of SLE with an incidence ranging from 4 to 10%. The expression 'intermediate' APS
or lupus-like APS disease was coined to identify those patients displaying minor
features resembling SLE but in the absence of the full set of the ACR classification
criteria.[18,19]
A family history of lupus, the presence of Raynaud phenomenon, migraine, multiple
sclerosis-like features, hemolytic anemia, low C3 and C4, and Coombs' positivity,
were found in one study to confer a statistically significant risk for subsequent
development of SLE (P < 0.05).[20] Comparable finding was reported in another study
showing that PAPS may be a forerunner of a full-blown lupus.[21]
In order to test the inflammation and immune activation hypothesis in primary
thrombotic APS (PAPS) and to identify clinical and laboratory factors related to
inflammation and immune activation. PAPS (n = 41) patients were compared with
patients with inherited thrombophilia (n = 44) and controls (CTR, n = 39). IgGaCL,
IgG anti-β2-glycoprotein I (β2GPI), high-sensitivity CRP (hs-CRP), serum amyloid A
(SAA), CRP bound to oxidized low-density lipoprotein-β2GPI complex (CRP-oxLDL-
β2GPI) (as inflammatory markers) neopterin (NPT), soluble CD14 (sCD14) (as
immune activation markers) were assessed. PAPS showed higher plasma levels of
hs-CRP (P = 0.0004), SAA (P < 0.01), CRP-oxLDL-β2GPI (P = 0.0004), NPT (P <
0.0001), and sCD14 (P = 0.007) than inherited thrombophilia and CTR. Two
regression models were applied to the PAPS group: in the first, IgG aCL and IgG
β2GPI were included amongst the independent variables and in the second they
were excluded. In the first model, SAA (as the dependent variable) independently
related to the thrombosis number (P = 0.003); NPT (as the dependent variable)
independently related to thrombosis type (arterial, P = 0.03) and number (P = 0.04)
4. (as the dependent variable) independently related to IgG β2GPI (P = 0.0001), age
(0.001) and arterial thrombosis (P = 0.01); CRP-oxLDL-β2GPI (as the dependent
variable) independently related to IgG β2GPI (P = 0.0001). In the second model,
sCD14 and NPT independently related to each other (P = 0.002) (this was noted
also in the inherited thrombophilia group) (P < 0.0001) and CRP-oxLDL-β2GPI
independently predicted SAA (P = 0.002). Thus, low-grade inflammation and
immune activation occur in thrombotic PAPS and relate to clinical features and aPL
levels.[22••]
Altogether, the above studies point to the fact that the progression from PAPS to
SLE disease is possible. In many of these cases it takes place only after a long
period of time during the follow-up.
Lupus Serological Markers in Primary Antiphospholipid
Syndrome
Many Serological Markers that are Considered Specific in Lupus are
Found also in PAPS. Anti-nuclear Antibodies
The presence of antinuclear antibodies at titer higher than 1: 320 as well as anti-
dsDNA or antiextractable nuclear antigens (ENAs) have been initially regarded as
exclusion criteria for primary APS. Certainly, the frequency of these autoantibodies is
higher in patients with APS within SLE or lupus-like disease suggesting that these
antibodies are more linked to lupus itself than to APS.
Nevertheless and contrary to old beliefs, many studies reported on the occurrence of
antinuclear antibodies in APS without a full-blown systemic disease supporting the
idea that an antinuclear autoimmune signature may be present in the syndrome.[5,23]
Antinucleosome antibodies (anti-NCS) are reported to be highly sensitive and
specific for SLE and to correlate with the disease activity. They may appear in the
early stages of the disease, in many cases even before anti-dsDNA antibodies
themselves, being a marker to identify potential patients who may develop SLE.
Studies with small series of patients reported a high prevalence of anti-NCS
antibodies in PAPS. A recent multicenter study[24••] confirmed these findings showing
that up to 50% of PAPS displayed medium/high titer of anti-NCS autoantibodies.
In spite of a follow-up of at least 2 years in the last study, the number of patients
evolving into SLE was very small and comparable to that previously reported. Hence
we do not have a sound demonstration that anti-NCS autoantibodies may predict the
evolution of the syndrome.
Complement Activation in Primary Antiphospholipid Syndrome
Complement consumption by immune complexes is widely accepted as a key
pathogenic mechanism in SLE tissue damage. There is sound evidence that
complement activation is required both in experimental models of aPL-mediated fetal
loss and thrombosis.[25,26••] The presence of aPL ('first hit') increases the risk for
thrombophilia. Then, innate immunity (i.e. toll-like receptors and the activation of
5. complement) by sensing microbial agents might synergize ('second hit') and
contribute to the development of clotting events.[27] At the present time there are few
clues regarding the involvement of complement in the pathogenesis of APS.
The prevalence of hypocomplementemia was assessed in a large number of patients
diagnosed either with SLE or with PAPS in association with the main clinical,
hematological and immunological features of these diseases.[28] Patients with PAPS
displayed low complement values in 33/57 (47%), and those were associated with
higher prevalence of livedo reticularis (P = 0.02), thrombocytopenia (P = 0.004), and
positive anti-dsDNA antibodies (P = 0.04), pointing to the fact that these patients are
viewed as a subset of patients with SLE-related markers. On the contrary, a more
recent study confirmed significantly lower complement levels in primary APS patients
compared with patients with non-SLE connective tissue diseases (C3, 81.07 ± 17.86
vs. 109.8 ± 22.76 mg/dl, P = 0.000005; C4, 13.04 ± 8.49 vs. 21.70 ± 6.96 mg/dl, P =
0.0001; CH50, 31.32 ± 8.76 vs. 41.40 ± 7.70 mg/dl, P = 0.000004) or healthy
volunteers. Most primary APS patients showed elevated serum C3a and C4a, and
PAPS patients with low serum C3 or C4 had significantly higher levels of C3a or C4a
compared with healthy controls. No patients had low serum complement regulatory
factors suggesting complement activation rather than a deficiency. Although no
relationship was found with the clinical manifestations of the syndrome,
hypocomplementemia was significantly more frequent in patients with lupus
anticoagulant.[29•]
In any case, complement involvement appears to be a pathogenic mechanism
common to both APS and SLE. Again, it is very difficult to state whether such a fact
represents a bridge between the two diseases or simply a coincidence.
Does a Genetic Background Explain the Proximity or the
Difference between Primary Antiphospholipid Syndrome and
Systemic Lupus Erythematosus?
The evidence for a genetic background in SLE has been known for at least 20 years
based on the studies on twins, on familiar clustering, on ethnic associations as well
as on murine in-vivo models. Linkage analysis and association studies have
identified several candidate genes for lupus susceptibility.[30] It is widely accepted that
SLE susceptibility requires multiples genes, each of them displaying only a small or
moderate effect individually. A 'threshold liability' hypothesis has been suggested:
the individual probability to develop the disease would depend on the presence of a
certain number of susceptibility alleles, but the disease would be the result of the
eventual interaction between the genome and the environment.
In a similar way, proofs of the genetic background of APS lie on familiar clustering of
cases, on murine models and on the association with human leukocyte antigens
(HLAs) (Table 2).[31–34,35•,36••] The ability to mount a response against phospholipid-
binding proteins has been associated with some HLA alleles (HLA-DR4, HLA-DR7,
HLA-DRw53 and HLA-DQB1*0302). Such an association was found both in PAPS
and in APS secondary to SLE, whereas a completely different pattern of HLA allele
association was reported in lupus patients in comparison with PAPS. This finding
6. would suggest that a common genetic background between SLE and PAPS may be
related just to the ability to produce or not aPL.[31]
Regarding the other candidate genes for lupus susceptibility, there are studies only
on the association between aPL and FcγR and PDCD1 polymorphisms. In a large
international meta-analysis study,[32] FcγRIIA-R/H131 polymorphism displays different
effects on APS and SLE. In fact, although the RR genotype was enriched in the
entire group of APS cases, this was driven mostly by patients with secondary APS.
In another study[33] on the role of an intronic polymorphism in the PDCD1 gene with
the risk of sporadic SLE, PDCD1 polymorphism is significantly associated with
protection against the occurrence of aPL both in SLE cases and in the general
population.
In another study,[34] the association of the mannose-binding lectin (MBL) pathway of
complement activation with SLE disease activity and/or the production of SLE-
related autoantibodies was investigated. MBL pathway activity was found to be
reduced in patients carrying MBL variant alleles. In respect to this, anticardiolipin and
anti-C1q autoantibodies were observed more frequently in patients with MBL variant
alleles. Thus, in patients with SLE, a reduced functional activity of the MBL pathway
of complement, in relation to expression of MBL variant alleles, is associated with
increased levels of aCL.
In a most recent study,[35•] it was shown that activity of the mammalian target of
rapamycin (mTOR), which is a sensor of the mitochondrial transmembrane potential,
is increased in lupus T cells, and that activation of mTOR causes the loss of TCR in
lupus T ells through HRES-1/Rab4-dependent lysosomal degradation. The HRES-1
human endogenous retroviral sequence is centrally located at the 1q42
chromosomal region relative to microsatellites previously associated with SLE.
Interestingly, linkage disequilibrium between HRES-1 single-nucleotide
polymorphisms (SNPs) at bases 653 and 1259 was reduced in patients with SLE (P
= 0.04). The HRES-1 653C/1259C-harboring alleles were associated with the
presence of renal disease (P = 0.002), whereas the 956A allele was associated with
the APS in APS patients with SLE (P = 0.003).[36••]
As a whole these findings do suggest that the ability to produce aPL may be
genetically determined. However, full-blown SLE seems to be the result of the
coordinated action of several genes that are not always detectable in APS. Further
studies may eventually demonstrate whether the difference between SLE and APS is
the result of the lack of specific genes able to drive the autoimmune response
towards a more systemic disease.
Conclusion
Although the reviewed literature is not homogeneous regarding the inclusion criteria
used, the final message is that PAPS may be systemic in some cases, involving
CNS, kidneys, and the skin. In addition, the presence of other autoantibodies apart
7. from aPL is not very rare and both these aspects should be therefore periodically
analyzed. Patients with PAPS, in whom systemic manifestations and anti-
dsDNA/nucleosomes antibodies co-exist, should be considered a subset that is
located between PAPS and SLE (Fig. 1). It is useful to speculate on the fact that a
gradual accumulation of autoantibodies has been reported in clinical lupus. Hence,
the appearance of autoantibodies other than aPL may be the expression of a similar
phenomenon. Still open is the question why a complete evolution towards lupus is
relatively so rare and developing in a long period of time. As mentioned before, it is
possible that the lack of the appropriate susceptibility gene panel is responsible for
such a block.
Figure 1.
Systemic clinical manifestations such as CNS, kidney and skin involvement,
together with laboratory markers such as antinuclear antibodies and persistent
(Enlarge Image) thrombocytopenia may define a subgroup of PAPS closer to full-blown SLE
Clinical markers such as glomerulonephritis, livedo reticularis and CNS involvement,
together with laboratory markers such as anti-dsDNA antibodies and persistent
attenuation of platelets (thrombocytopenia), may define a subgroup of PAPS patients that
are of higher potential to develop full-blown SLE. CNS, central nervous system; PAPS,
primary antiphospholipid syndrome; SLE, systemic lupus erythematosus.
The beneficial effect of hydroxychloroquine, and other immunomodulatory therapies
in reducing aCL titers and preventing thrombosis in PAPS patients, has been
recently suggested as additional therapeutic options.[37] Theoretically the
administration of compounds able to affect the immune response may be beneficial
not only in preventing thrombosis but also in inhibiting the development of full-blown
SLE in some patients.